dq6

See attachment titled dq6

Week 6 – Discussion

No unread replies.No replies.

Your initial discussion thread is due on Day 3 (Thursday) and you have until Day 7 (Monday) to respond to your classmates. Your grade will reflect both the quality of your initial post and the depth of your responses. Refer to the Discussion Forum Grading Rubric under the Settings icon above for guidance on how your discussion will be evaluated.

Promoting Maintenance and Generalization Save Time On Research and Writing Hire a Pro to Write You a 100% Plagiarism-Free Paper. Get My Paper

Incorporating objectives focused specifically on maintenance and generalization into IEP goals is best practice. Goals for maintenance and generalization focus on long term outcomes for students with mild to moderate disabilities. The readings and Instructor Guidance this week provided you with EBP strategies that promote maintenance and generalization of IEP goals.

Initial Post: Post an initial response that addresses the following areas using the discussion board forum:

Select only one of the following bullets to focus your discussion with your peers. Be sure to include both scholarly sources and personal experiences; also consider posing questions related to your discussion to your colleagues within your initial post to extend the conversation through their guided response:

· Based on the readings, what are the long-term effects of maintenance and generalization instructional practices on the outcomes of students with mild to moderate disabilities? How can educators promote positive outcomes through goal development or EBP strategies?

Or

· How can you identify and evaluate EBP strategies to promote maintenance and generalization of IEP goals and objectives across skills for students with mild to moderate disabilities? How can these concepts be incorporated directly into IEP goals and objectives?

ESE668: EVIDENCE-BASED INSTRUCTIONAL METHODS FOR STUDENTS WITH MILD TO MODERATE DISABILITIES

Instructor Guidance

Week 6

Congratulations! Welcome to the final week of ESE 668: Evidence-Based Instructional Methods for Students with Mild to Moderate Disabilities. Please be sure to review the Week Six homepage for this course to see:

· The specific learning outcomes for the week.

· The schedule overview.

· The required and recommended resources.

· The introduction to the week.

· A listing of the assessments.

Next, be sure to read this entire Instructor Guidance page.

Overview

This week, we will discuss one final component needed in EBP methods for students with mild to moderate disabilities: Maintenance and generalization of skills. Now is the time to reflect upon all you have learned and experienced in this course. In Week Six, you will prepare your Final Project consisting of the requirements in the guidance. 

Intellectual Elaboration

Stages of Skill Development and Learning

When a student is presented with a new skill, there are four stages of learning they will go through (Young, West, Howard, & Whitney, 1986):

1. Acquisition.

2. Fluency.

3. Maintenance.

4. Generalization.

As educators, our initial aim is acquisition. Goals and objectives must be written using S.M.A.R.T. for this express purpose of skill acquisition. What about the other stages? Is it enough for a student to be able to perform a skill in the classroom? What about their efficiency in doing so? Fluency involves accuracy and speed. If a student can add numbers when provided a problem on paper, is it enough that she is successful in doing so in 30 minutes per problem? Probably not. We also need to consider writing goals to ensure fluency of skills.

Maintenance of Skills

Now what about maintenance? Will skills be helpful for our students if they are able to fluently complete the skill in the classroom when it is taught, but a week later cannot perform the skill as a prerequisite for the next stage of the task? What about if a student was able to complete math facts fluently in 5th grade, but did not maintain the skill after secondary school? Goals must be explicitly written to consider maintenance of skills. One way to do this is to align goals and objectives with CCSS as we discussed in Week Four of this course. When we have long-term goals that build upon the previously mastered skills, we promote maintenance of acquired skills over time.

Technology of Generalization in Special Education

“Generalization refers to

the transfer of what is learned in one setting or situation

to another setting or situation

without explicit teaching or programming in the second transfer setting.”

–Autism Ontario, 2011, pg. 35

One of your required readings this week is a seminal article discussing the technology of generalization, which is often a missed step in the education process. We cannot assume that all students with passively generalize the skills they have learned in the classroom to other people, environments, situations, or stimuli. We as educators must specifically implement strategies that teach generalization and incorporate generalization technology into our S.M.A.R.T. goals and objectives (Stokes & Baer, 1977):

Putting It All Together: Evidence-based instructional methods for students with mild to moderate disabilities

Well, you did it! We have explored six weeks of evidence-based instructional methods for students with mild to moderate disabilities from many facets of development and implementation of instruction. It is now time to look toward demonstrating what you have learned by compiling and composing your Final Assignment: Huang’s IEP.

Assessment Guidance

This section includes additional specific assistance for excelling in the discussions for Week Six beyond what is given with the instructions for the assessments. If you have questions about what is expected on any assessment for Week Six, contact your instructor using the Ask Your Instructor discussion before the due date. Both the Discussion and Final Project are opportunities for you to further demonstrate mastery with the four course learning outcomes, which are noted on the course Syllabus.

Discussion: Promoting Maintenance & Generalization

In this final discussion, you will choose only one of two optional topics on maintenance and generalization of skills to reflect upon and discuss with your classmates. Use what you have learned about scholarly searches in Week Five to guide you in finding additional information to support your discussion. 

Final Project: The Individualized Education Program (IEP)

For this final assignment, you will be assuming the role of a special educator and will create an IEP for Huang, a student in Mr. Franklin’s class. Using the IEP rubric provided complete the IEP Template and reflection for Huang. You may use components have personally developed throughout this class (i.e. PLAAFPs, Goals & Objectives, Accommodations & Modifications, S.M.A.R.T. goals) to complete the corresponding sections of the IEP template. Review all feedback that was provided and make any needed changes before including them in your IEP document. Be sure to complete all sections of the rubric. As well, in a second document you will provide a narrative detailing the EBP strategies you would recommend for implementing Huang’s IEP, as well as a reflection on the relevancy of this IEP process to your current or future professional goals.

References

Autism Ontario. (2011). 

Ideas for promoting generalization of social skills (Links to an external site.)

. Social Matters. Toronto, ON: Author. Retrieved from https://autismontario.novosolutions.net/default.asp?id=108
Burns, M. K., Egan, A. M., Kunkel, A. K., McComas, J., Peterson, M. M., Rahn, N. L., & Wilson, J. (2013). Training for Generalization and Maintenance in RTI Implementation: Front-Loading for Sustainability. Learning Disabilities Research & Practice (Wiley-Blackwell), 28(2), 81-88. doi:10.1111/ldrp.12009.
Osnes, P. G., & Lieblein, T. (2003). An explicit technology of generalization. The Behavior Analyst Today, 3(4), 364-374. doi:10.1037/h0099994
Stokes, T. F., & Baer, D. M. (1977). An implicit technology of generalization. Journal of Applied Behavior Analysis, 10(2), 349-367. doi:10.1901/jaba.1977.10-349
Young, K.R., West, R., Howard, V., & Whitney, R. (1986). 

Acquisition, fluency training, generalization, and maintenance of dressing skills of two developmentally disabled children (Links to an external site.)

. Education and Treatment of Children, 9 (1), 16-29. Retrieved from http://www.jstor.org/stable/42898943

Required Resources

Article

s

Burns, M. K., Egan, A. M., Kunkel, A. K., McComas, J., Peterson, M. M., Rahn, N. L., & Wilson, J. (2013). 

Training for generalization and maintenance in RTI implementation: Front-loading for sustainability

. Learning Disabilities Research & Practice, 28(2), 81-88. doi:10.1111/ldrp.12009

· The full-text version of this article is available through the EBSCOhost database in the Ashford University Library. This article discusses ways in which programming for generalization and maintenance can be incorporated into interventions such as RTI. This article will support your Promotion Maintenance and Generalization discussion this week.

Osnes, P. G., & Lieblein, T. (2003). 

An explicit technology of generalization

. The Behavior Analyst Today, 3(4), 364-374. doi:10.1037/h0099994

· The full-text version of this article is available through the EBSCOhost database in the Ashford University Library. This article builds upon the seminal article from Stokes and Baer (1977) to extend the strategies that can be used for targeting generalization of skills. This article will support your Promotion Maintenance and Generalization discussion this week.

Stokes, T. F., & Baer, D. M. (1977). 

An implicit technology of generalization (Links to an external site.)

. Journal of Applied Behavior Analysis, 10(2), 349-367. doi:10.1901/jaba.1977.10-349

· The full-text version of this article is available through the National Center for Biotechnology Information website at https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1311194/pdf/jaba00113-0179 . This seminal article on generalization provides seven strategies for promoting generalization of skills. This article will support your Promotion Maintenance and Generalization discussion this week.

Recommended Resource

Article

Young, K.R., West, R., Howard, V., & Whitney, R. (1986). 

Acquisition, fluency training, generalization, and maintenance of dressing skills of two developmentally disabled children (Links to an external site.)

. Education and Treatment of Children, 9(1), 16-29. Retrieved from http://www.jstor.org/stable/42898943

· This article on stages of learning provides a background on acquisition, fluency, generalization, and maintenance of skills. This article will support your Promotion Maintenance and Generalization discussion this week.

Learning Disabilities Research & Practice, 28(2), 81–88
C© 2013 The Division for Learning Disabilities of the Council for Exceptional Children

Training for Generalization and Maintenance in RtI Implementation:
Front-Loading for Sustainability

Matthew K. Burns, Andrea M. Egan, Amy K. Kunkel, Jennifer McComas, Meredith M. Peterson,
Naomi L. Rahn, and Jennifer Wilson

University of Minnesota

Response to Intervention (RtI) is being implemented as a new initiative in PK-12 schools
with increasing frequency. However, the model must be sustained at the school level, which
is potentially difficult due to a number of challenges brought about by systems change. This
article applied the Stokes and Baer (1977) framework for programming for generalization and
maintenance of behavior change to suggest specific activities in which schools could engage
to better ensure RtI sustainability. We specifically discussed ways to (1) introduce to natural
maintaining contingencies, (2) train with sufficient exemplars, (3) train loosely, (4) program
common stimuli, (5) mediate generalization, and (6) train to generalize. Directions for future
research are included.

Response to Intervention (RtI) and other multitiered inter-
vention systems are being adopted nationwide with increas-
ing frequency (Berkeley, Bender, Peaster & Saunders, 2009)
to increase student achievement for all students, reduce re-
ferrals to special education, and close existing achievement
gaps (Fuchs, Fuchs & Stecker, 2010). RtI has the potential to
positively affect both systemic and student outcomes (Burns,
Appleton & Stehouwer, 2005), but, some question whether
the RtI movement will sustain over time (Burns, 2007; Ys-
seldyke, 2005). RtI initiatives must ultimately be sustained
at the school level, and organizations adopting a system of
RtI are faced with a multitude of challenges brought about
by systems change (Grimes, Kurns & Tilly, 2006).

Previous research has found that implementation integrity
could be a serious threat to the validity of RtI models (Gansle
& Noell, 2007). For example, school personnel consistently
assessed fidelity of implementation for interventions that oc-
curred at tier 2, but did not assess fidelity at tier 1, and the
alignment between tiers was not explicit (Hill, King, Lemons
& Partanen, 2012). Moreover, implementation integrity of
problem-solving teams (PSTs) was low to the point of po-
tentially affecting student outcomes (Burns & Symington,
2002). Some of the challenges regarding implementation in-
tegrity can be avoided by building on the existing knowledge
of the school personnel, streamlining processes, and using a
clear system of communication between interventionist and
teacher (Johnson, Pool & Carter, 2012). However, implemen-
tation integrity can still wane as the implementation moves
further from the initial supports (Burns & Symington, 2002;
Kovaleski, Gickling, Morrow & Swank, 1999), which further
highlights the need to focus on sustainability.

Requests for reprints should be sent to Matthew K. Burns, University of
Minnesota. Electronic inquiries should be sent to burns258@umn.edu.

Sustainability is best obtained by changing the system
in which the initiative is implemented (Hargreaves & Fink,
2000). Systems change is an “intentional process designed
to alter the status quo by shifting and realigning the form
and function of a targeted system” (Foster-Fishman, Nowell
& Yang, 2007, p. 197), and is multifaceted with theoretical,
ethical, and pragmatic implications (Noell & Gansle, 2009).
Prior to implementation, theoretical and ethical dimensions
of systems change are considered, including issues of what to
change, why to change it, and how that change will take place.
Promoting adoption and implementation of RtI in schools re-
quires that stakeholders see the value of its implementation
in their schools (McIntosh, Filter, Bennett, Ryan & Sugai,
2010) and that teacher “buy-in” is high. Moreover, educators
are more likely to implement interventions or practices in
which they have experience, support, and belief in overall
effectiveness (Fixsen, Blase, Naoom & Wallace, 2009). Al-
though RtI implementation research has found that collabo-
ration is important for teacher acceptance, teacher buy-in can
be difficult to gain (Mahdavi & Beebe-Franenberger, 2009).

Alberto and Troutman (2009) suggest the school and
teacher environment should be examined to uncover what
teachers value and invoke strategies or interventions that
may have existing support. After schools have resolved the
theoretical and ethical dimensions of systems change, the
pragmatic aspects of implementation, including generaliza-
tion and sustainability, can be addressed (Noell & Gansle,
2009). Behavior change among all organizational partici-
pants (i.e., teachers and administrators) is crucial (Sarason,
1996), and must be accompanied by sustained environmen-
tal supports that are responsive and adapted to inevitable
challenges (Grimes et al., 2006). Moreover, in order for the
long-term goal of sustained change to be realized, the ex-
pected behaviors exhibited by key stakeholders must gen-
eralize across situations and maintain over time (Sarason,
1990; 1996).

82 BURNS ET AL.: SUSTAINABILITY OF RTI IMPLEMENTATION

There is an inexorable link between generalization and
sustainability within educational reform (Hargreaves & Fink,
2000). Generalization occurs when a learned behavior con-
tinues to occur across time, setting, and target in the absence
of the conditions that promoted its acquisition (Stokes &
Baer, 1977). Thus, generalization is at least a prerequisite
for sustainability, but it could be argued that promoting gen-
eralization over time could provide a framework to address
sustainability because sustainability is the continued behav-
ior over time after the conditions in which it was required are
removed or changed.

Stokes and Baer (1977) introduced a framework for as-
sessing and programming for generalization and mainte-
nance of behavior change. Prior to that publication, the most
frequent method of considering generalization in behavior
change programs was to “train and hope” (p. 351). In other
words, new behaviors were trained and any generalization
across settings, time, or responses were not actively planned;
rather, it was hoped that generalization would occur. A re-
cent survey of special education directors found that the most
common support for RtI implementation provided by state
departments of education was short-term trainings and pro-
fessional development (Werts, Lambert & Carpenter, 2009).
Accordingly, RtI implementation could attempt to be gen-
eralized through train and hope, but it will likely not be
successful without sustained environmental support. How-
ever, deliberate programming for generalization and mainte-
nance of expected behaviors in a system of RtI could result
in successful outcomes and sustained actions. The process
of programming for generalization of RtI implementation
includes the following techniques discussed by Stokes and
Baer in 1977: (1) introduce to natural maintaining contin-
gencies, (2) train sufficient exemplars, (3) train loosely, (4)
program common stimuli, (5) mediate generalization, and (6)
train to generalize.

The purpose of this article is to discuss each of these strate-
gies within the context of implementation and sustainability
of RtI and implications for practice in schools. The goal will
be to describe specific actions that schools can take to pro-
mote generalization and maintenance of practices in order for
RtI implementation to be sustained over time. In other words,
we will discuss ways that school personnel can frontload im-
plementation efforts to better assure sustainability. Table 1
provides a succinct summary of the generalization strategies
and related practices for RtI sustainability. We will also pro-
vide suggestions for future research, which will likely be the
primary outcome associated with these suggested strategies
and practices.

INTRODUCE TO NATURAL MAINTAINING
CONTINGENCIES

Stokes and Baer (1977) stated that introducing naturally
maintaining contingencies is the most dependable way to ob-
tain generalization, even though this strategy may not always
be feasible. To generalize in this manner is to transfer the
behavioral control to the natural contingencies that operate
in the environment where the practice will occur. Apply-
ing naturally maintaining contingencies in training involves

teaching behaviors or practices and bringing them into con-
tact with naturally existing contingencies for reinforcement.

Using naturally occurring contingencies to promote gen-
eralization can take many varied forms. For example, imagine
that during a professional development session, teachers are
taught to examine student data and make instructional pro-
gramming decisions based on the data. Now imagine that a
month later when they examine their students’ progress, the
data indicate that the most struggling students have made sub-
stantial gains. Seeing those substantial gains may naturally
reinforce the practice of making instructional programming
decisions based on student data. In this example, the instruc-
tional leader might select specific examples of the data-based
instructional programming that resulted in the substantial
academic gains for a sample of the students and could discuss
the types of instructional programming decisions that would
be more and less likely to produce future academic gains.

The first step in planning for a sustainable RtI model at
the individual school level might be to have school person-
nel implement the model within their daily practice (Fixsen
et al., 2009), rather than having district or university person-
nel handle the initial implementation. Despite the temptation
to provide significant support during the initial implementa-
tion of RtI, if school personnel are the ones who implement
the RtI-related practices, then they are likely to directly ex-
perience the natural successes that result. Similarly, teachers
should be included in all aspects of planning and implement-
ing RtI, including making intervention decisions, and doing
so resulted in improved student outcomes (Lembke, Garman,
Deno & Stecker, 2010). For example, teachers could create
the list of instructional practices and interventions for spec-
ified skill deficits available for use at each tier of service
in their building. The list could take the form of a menu
of evidence-based options and include evidence-based in-
structional practices and interventions that teachers in the
building have used and found successful. A teacher may be
more likely to implement an instructional practice or inter-
vention that s/he has found successful in the past because
s/he has witnessed the effect it had on producing academic
growth and therefore has contacted the natural consequence,
student success, which was produced by implementing the
practice or intervention.

Implementation of RtI should be considered within the
context of what components are already in place and what
components need to be established. If numerous components
need to be added, a format for establishing the model that al-
lows individual teachers to make a relatively small number of
changes to their practice at a time is advisable (Grimes et al.,
2006). This approach allows teachers to come into contact
with naturally maintaining contingencies, whereas if they are
forced to change numerous aspects of their practice at once,
they are less likely to contact the reinforcing consequences of
any one of the practices (McIntosh et al., 2010). When sys-
tems change is time-consuming and requires implementation
of numerous novel practices, competition with existing, less
effective practices presents a considerable challenge (Noell
& Gansle, 2009). Alternately, schools might consider invit-
ing teachers to be involved in or responsible for particular
components of RtI (e.g., screening, interventions) according
to their interest (Johnson et al., 2012). Preference is related

LEARNING DISABILITIES RESEARCH 83

TABLE 1
Strategies for Generalization from Stokes and Baer (1977) and Accompanying Activities to Build Sustainability in Response to Intervention (RtI)

Implementation

Strategy Description Activities

Natural maintaining Teach the skill to be reinforced by naturally • Involve school personnel in implementation decisions
contingencies existing contingencies • School personnel implement interventions and assessments

• Use efficient data collection procedures
Train sufficient exemplars Use numerous examples during training • Provide ongoing professional development in the core components/skill

sets of RtI
• Use a broad range of examples of forms that RtI core components can

take (e.g., collecting progress monitoring data for a variety of academic
skills)

• Train personnel to implement multiple aspects of the grade-level and
problem-solving team processes

Train loosely Expose learners to a diverse array of the
contexts or situations in which skill set is

• Train using a variety of contexts and situations in which the same set of
skills are required (e.g., monitor progress in multiple areas)

to be used • Use a broad range of examples (e.g., what teams are called, which data
collection tools are selected)

Program common stimuli Incorporate into training stimuli that are
common across contexts or situations

• Use grade-level teams as professional learning communities to make
decisions at various tiers

• Configure teams (e.g., grade-level teams) of consistent members who
will address a variety of contexts and situations together

Mediate generalization Incorporate tools or strategies that the • Use implementation fidelity protocols and checklists
learner can readily use across contexts or
situations

• Provide continuous feedback to school personnel (e.g., team processes,
intervention fidelity, assessment procedures)

Train to generalize Raise awareness of need for generalization
during training and suggest use of trained
skill sets across contexts and situations

• Discuss how existing RtI practices contextualize into other areas of
practice

to quality of reinforcement and if teachers are encouraged
to participate in aspects of RtI that fit their preferences, they
may experience relatively higher-quality reinforcement for
their participation.

Finally, response effort impacts the effects of naturally
maintaining contingencies of reinforcement. For example,
the amount of data collected in an RtI model should not
be exorbitant, but rather focused on useful information
that can be collected efficiently (Horner, Sugai & Todd,
2001). Assessment procedures should be quick and easy,
and yet result in sufficiently reliable data and valid decisions
(e.g., curriculum-based measurement). When teachers are
involved in collecting their own data, it allows them to see
the effects of their practice through a direct link to student
outcomes (McIntosh et al., 2010). Moreover, previous RtI im-
plementation efforts emphasized the importance of stream-
lining data collection and giving the teachers responsibility
for collecting the data (Johnson et al., 2012). However, teach-
ers must view data collection and analysis as an investment
(Horner et al., 2001), and the payoff of positive outcome data
presents natural reinforcement for teachers. If data reveal an
absence of positive outcomes for certain students, teachers
are provided an efficient and effective means by which to in-
form further instruction, and will see the benefits of program
modification for students as interventions are intensified and
data collection continues.

TRAIN SUFFICIENT EXEMPLARS

Training sufficient exemplars is described as one of the most
valuable techniques for programming generalization (Stokes

& Baer, 1977). Teaching only a single exemplar limits the
effectiveness of the lesson to the teaching situation, whereas
providing additional exemplars across a variety of situations
is crucial for generalization of the skill set to occur across
a variety of situations. To illustrate, in a special education
program, an instructor might teach how to use a vending
machine. Such teaching necessitates some careful planning
because there are a wide variety of vending machines, many
of which require different approaches. Some vending ma-
chines require pushing the button that depicts the product,
others require finding the code for the product and enter-
ing into a keypad. Depositing money can take the form of
coins, bills, or a combination, and the coin slot is sometimes
vertical and sometimes horizontal. By exposing students to
these variations during training, they are more likely to ex-
perience success when they use a vending machine when
they are not with their teacher. Within the context of pro-
fessional development, which is a crucial aspect of effective
RtI implementation (Kratochwill et al., 2007), it is wise to
provide educators with several examples of potential imple-
mentation models including structures for delivering quality
core instruction for all students, a variety of screening and
progress monitoring tools, evidence-based interventions for
tiered intervention delivery, and teaming strategies for data-
based decision making. How these individual components
are implemented within a school can vary depending on the
school’s model. Providing educators with a variety of ex-
amples of these core components, as well as examples of
successful RtI models in other schools or districts, can allow
educators to adapt and adopt an ideal model for the situations
their setting presents, leading to a much greater probability
of sustaining RtI within a given school.

84 BURNS ET AL.: SUSTAINABILITY OF RTI IMPLEMENTATION

Although schools are limited by the standardized assess-
ments they are required to use with students with and without
disabilities, the screening and progress monitoring tools are
typically open to teacher discretion. Providing teachers with
training and materials to monitor progress across content ar-
eas (i.e., reading, writing, and math) and with various tools
(e.g., oral reading fluency, timed mathematics probes) im-
parts additional examples of efficient assessments of student
outcomes, thus creating multiple ways from which teachers
can choose to monitor progress. Furthermore, varied exam-
ples of monitoring frequency (e.g., weekly or bimonthly) and
data collection personnel (e.g., paraprofessionals, volunteers,
or classroom teachers) provides teachers with additional op-
tions. For example, screening or monitoring data might be
completed by a small cadre of individuals across grades and
days to limit disruptions to the classroom and instruction.
Conversely, such data might also be collected in a unified
approach involving many individuals completing all class-
rooms within a shorter time span.

Ensuring that teachers are knowledgeable of multiple
evidence-based interventions will allow them to make in-
formed intervention decisions for struggling students to re-
ceive targeted interventions in appropriate groups. However,
providing training in too many intervention approaches can
be counterproductive to maintaining naturally occurring con-
tingencies and may overwhelm the teachers. Thus, it might
be beneficial to train teachers and school personnel in in-
tervention implementation within content areas, age groups,
and academic needs, which may help promote a foundation
for a successful model and its generalization.

Training educators on various teaming strategies may be
one of the most important considerations for RtI sustain-
ability, particularly with frequent changes in staffing models,
movement of administrators within a district, and teacher
turnover. Many schools implementing RtI form grade-level
teams, which are strongly related to the quality of later prepa-
rations for sustainability (Perkins et al., 2011). Although the
function of grade-level teams can differ among schools, de-
pending on the RtI model in place, they are often involved in
examining screening data for all students, analyzing progress
monitoring data, making informed intervention decisions re-
garding struggling students, and discussing adaptations and
modifications to the model at each tier of instruction. Pro-
viding schools with examples of successful grade-level team
models and professional development of effective teaming
strategies will allow them to choose and adapt the best model
fitting their resources, increasing the sustainability of RtI over
time. Unfortunately, inconsistent implementation of school-
based teams is well documented and a potential threat to RtI
implementation (Burns, Vanderwood & Ruby, 2005), which
reinforces the need for schools to train personnel with posi-
tive examples before implementation begins.

TRAIN LOOSELY

Whereas training sufficient exemplars involves teaching in
such a way that individuals make appropriate adaptations and
adjustments in their behavior (e.g., how to indicate a selection
with any vending machine) given the specific requirements

of the context or situation, training loosely (Stokes & Baer,
1977) refers to teaching a behavior or skill set such that it oc-
curs in the presence of a variety of contexts and situations. To
train loosely, an approach must be taken that exposes learners
to a diverse array of situations in which the same response
might be expected. For example, in a classroom, a teacher
might say, “Have a seat,” “Take a seat,” “Find your place,”
or gesture toward a circle of chairs; in all cases, the expected
behavior is for the student to sit down. In the previous sec-
tion, we mentioned grade-level teams, which often function
to provide a forum and structure, as well as accountability
for analyzing progress monitoring data, making informed in-
tervention decisions, and discussing necessary adjustments
to the model at each tier of instruction. However, PSTs can
also play an important role in RtI models, especially within
tier 3. PSTs go by a wide variety of names across the coun-
try, including but not limited to Instructional Support Team,
Instructional Leadership Team, Academic Leadership Team,
Child Study Team, and Teacher Support Team. By inter-
changeably using a variety of names for teams but pointing
out their unifying function, the notion of problem-solving
instruction and interventions within a team is trained loosely.
The purpose of training loosely is to allow for responding
in a singularly appropriate manner in a variety of situations
that differ superficially but are functionally equivalent. Thus,
transfer of the targeted behavior to new situations is facil-
itated by exposure to the many contextual dimensions that
may vary.

The concept of training loosely can inform multiple as-
pects important to the sustainability of RtI, including data
collection, intervention delivery, and teaming strategies for
effective decision making. School personnel must pay at-
tention to the fit between the conceptual framework of a
school-wide program and the local, contextual variables of
a given school (McIntosh et al., 2010). While adherence to
the conceptual framework of RtI is necessary to increase the
efficacy of the practice, acknowledgement of contextual fit
is important to its sustainability within a given school en-
vironment (Goldenberg, 2003). For example, a school may
strongly embrace a strengths-based approach to instructional
planning. In this case, the term “PST” would likely be less
acceptable than the term “Instructional Leadership Team.”
Increased flexibility of RtI implementation combined with
an emphasis on local control may create the potential for
RtI to sustain in a manner that is both building-based and
consistent with the general concept.

The function and makeup of PSTs might also allow for
flexibility regarding how often the team meets, who is re-
sponsible for leading the meetings, and the relationship be-
tween team discussions and professional development. For
example, some schools may use a designated leader who
organizes and leads meetings, whereas flexibility and sus-
tainability may be enhanced by having several individuals
within a school able to lead meetings at different times. Sim-
ilarly, the data analysis completed by PSTs may serve as a
springboard for related professional development, or could
support already-implemented school-wide initiatives. This
relationship is often reciprocal and can buttress the sustain-
ability of similar data-driven practices such as School-wide
Positive Behavior Interventions and Supports.

LEARNING DISABILITIES RESEARCH 85

PROGRAM COMMON STIMULI

Programming common stimuli, another technique used to
train for generalization, involves incorporating in training
stimuli that are essential features and therefore will likely be
present in a variety of generalized situations (Stokes & Baer,
1977). One example in school settings with students has been
the use of peers as common stimuli to promote generalization
of desired social interactions across settings (Stokes & Baer,
1976). Incorporating peers as common stimuli to train for
generalization can be applied to professional development
related to RtI.

The literature on professional development, particularly
the use of professional learning communities (PLCs), pro-
vides an opportunity to utilize peers as common stimuli to
build sustainability of RtI. It is increasingly clear that high-
quality professional development in schools represents an
essential link between teacher performance and student out-
comes (Kratochwill et al., 2007). In a PLC, teachers work
together in small groups on a particular topic to analyze and
improve school practices to enhance student learning. PLCs
are composed of three “Big Ideas” (DuFour, 2004): (1) ensur-
ing that students learn, (2) building a culture of collaboration,
and (3) a focus on results. The mission of professional devel-
opment in a PLC framework is not simply that students are
taught, but rather that they learn. When learning does not oc-
cur for all students, a PLC will focus on improving teaching
practices to enhance student learning. To build a culture of
collaboration, PLCs provide an ideal occasion to use peers as
common stimuli to build sustainability. With a focus on re-
sults, improving student achievement through collaboration
between teachers becomes routine work for everyone in the
school. PLCs allow schools to “create a multi-tiered, coordi-
nated, and collective response to support students” (DuFour,
2011, p. 61).

Within RtI, teachers in grade-level teams comprise the
PLCs. Teachers on a grade-level teamwork together to adopt
specific aspects of RtI. As new aspects of RtI are adopted, the
likelihood that a teacher will successfully implement new RtI
components is increased if it is done in the context and with
the support of the other teachers on the grade-level team. The
presence of peer teachers can facilitate generalization by sim-
ulating the environment—the grade-level team—in which
successful adoption of the initial RtI components occurred.
Research regarding professional development related to RtI
found that isolated training was not sufficient (Kratochwill
et al., 2007). This seems particularly relevant to the sus-
tainability of RtI, and the use of peers to program common
stimuli in professional development practices in schools can
ensure generalization and maintenance of the model.

MEDIATE GENERALIZATION AND TRAIN TO
GENERALIZE

Two final ways to increase sustainability and generalization
of RtI are to: (1) build procedures into the RtI process that
will increase the likelihood of generalization of desired be-
haviors, and (2) directly discuss and ask for generalization.
In mediated generalization, a response that is likely to be

used in new situations is established to promote generaliza-
tion (Stokes & Baer, 1977). For example, to multiply poly-
nomials, we are taught “FOIL”—first, outside, inside, last,
which is the order in which the products are to be computed
(Crawford, 1980). Thus, whenever one is confronted with a
polynomial, use of FOIL will facilitate successful multiplica-
tion of the polynomial in any situation. Within an RtI context,
tools for individual teacher and program self-evaluation may
play a mediating role in generalization of RtI components.
These tools include checklists for fidelity of implementation
of specific evidence-based practices (e.g., a reading interven-
tion), and for implementation of various aspects of the RtI
process more generally. Fidelity of implementation at both
the teacher and school levels should be evaluated to ensure
the effectiveness of RtI (Riley-Tillman & Burns, 2009).

Teacher-level implementation across RtI components and
settings can be measured through observations of imple-
mentation fidelity of specific evidence-based practices in
the classroom using checklists developed by researchers or
school districts. For example, the St. Croix River Education
District in Minnesota and Heartland Area Education Agency
in Iowa have developed checklists for assessing fidelity of
implementation of specific instructional or intervention pro-
grams (see Table 2). School-level evaluation tools are also
necessary for measuring generalization and maintenance of
RtI over time. For example, the School-wide Evaluation Tool
(SET; Sugai, Lewis-Palmer, Todd & Horner, 2001), was de-
signed to evaluate implementation fidelity of School-wide
Positive Behavior Support. Data from the SET are reviewed
by school teams and state-level teams to guide sustainabil-
ity efforts at both levels (McIntosh, Filter, Bennett, Ryan &
Sugai, 2010).

In addition to considering mediators, Stokes and Baer
(1977) advise directly discussing generalization and suggest-
ing that individuals generalize the desired behaviors or skills
sets to other contexts or situations. Training to generalize
involves explicitly suggesting or reminding the implementer
(e.g., teacher) to implement the RtI components in novel sit-
uations. Within an RtI framework, professional development
efforts should include discussions with staff of how exist-
ing RtI skill sets, such as universal screening and data-based
decision making, could be generalized to other areas of prac-
tice. For example, in a school already implementing RtI in
reading, school leaders might initiate discussions of how the
RtI model could be expanded to include math or behavior.
As generalization occurs, staff efforts should be reinforced.
Reinforcing generalization when it happens results in quick
wins for teachers and other RtI team members.

DIRECTIONS FOR FUTURE RESEARCH

Although the recommendations made by Stokes and Baer
(1977) are well grounded in research, the application to sus-
taining RtI requires additional research. Schools are com-
plex systems with several considerations when implement-
ing change initiatives (Fixsen et al., 2009). Thus, researchers
could examine a method to best identify potential applica-
tion (e.g., quality core instruction, screening and progress
monitoring tools, evidence-based interventions for tiered

86 BURNS ET AL.: SUSTAINABILITY OF RTI IMPLEMENTATION

TABLE 2
Response to Intervention (RtI) Implementation Checklists

Resource Tool(s)

Evidence-Based Intervention Network http://ebi.missouri.edu/ Intervention protocols for reading, math, writing, and
behavior

Heartland Area Education Agency http://www.aea11.k12.ia.us/idm/ Observation and permanent product
checkists.html treatment integrity checklists for academic interventions

National Center on Response to Intervention http://www. RtI Integrity Rubric and Worksheet
rti4success.org/categorycontents/continuously_improving/page

Pennsylvania Department of Education http://www.pattan.net/category/ • Response to Instruction & Intervention
Educational%20Initiatives/Response%20to%20Instruction%20and% (RtII) Readiness and Implementation
20Intervention%20%28RtII%29 (Elementary): Self-Assessment Tool

• Secondary RtII Framework: Self-Assessment Tool
Path to Reading Excellence in School Sites http://www. • Reading intervention protocols for all three tiers

cehd.umn.edu/reading/PRESS/default.html • Intervention implementation checklists
• Professional development materials

RtI Action Network http://www.rtinetwork.org/ Self-Assessment of Problem Solving Implementation
getstarted/checklists-and-forms

Scientifically based research http://gosbr.net/ • Reading and math intervention protocols
• Assessment tools

St. Croix River Education District http://www.scred. • Integrity checklists for reading interventions
k12.mn.us/School/Index.cfm/go:site.Page/Page:3/index.html

Technical Assistance Center on Positive Behavioral Interventions and PBIS evaluation checklists including:
Supports (PBIS) http://www.pbis.org/evaluation/evaluation_ • School-Wide Evaluation Tool (SET)

tools.aspx • Early Childhood System-wide Evaluation Tool: Program
Wide (EC SET-PW)

• Benchmarks for Advanced Tiers (BAT)

intervention delivery, and data-based decision-making mod-
els) that matches the needs and circumstances of each
unique system. Second, research has found that school-based
problem-solving teams were effective, but implementation
integrity of the process may have substantially reduced team
effectiveness (Burns & Symington, 2002), and implementa-
tion integrity was rarely assessed in tier 1 (Hill et al., 2012).
Moreover, unanticipated staffing changes can occur within
schools, which may result in a change in problem-solving and
leadership teams. When team members leave, responsibilities
have the potential to shift or be forgotten. Ensuring neces-
sary components of the RtI process are in place throughout
change is crucial to sustainability. Thus, additional research
is needed to examine issues such as the essential attributes of
an effective team and how to best measure integrity of core
instruction.

Implementing multiple changes, such as training suffi-
cient exemplars, training loosely, and programming common
stimuli, comes with additional difficulties that could provide
targets for additional research. Moreover, future researchers
could examine the recommendations made here to determine
both effectiveness and a potential heuristic to prioritize the
strategies given characteristics of the schools.

CONCLUSION

Education has a long history of fads in which, as Ellis (2005)
elegantly stated, “today’s flagship is often tomorrow’s aban-
doned shipwreck” (p. 200). RtI has the potential to be the
next in a long line of innovations about which school person-

nel are initially enthusiastic and result in immediate gains
in student learning, but then implementation wanes as the
initial enthusiasm fades. Educational change is a slow and
difficult process, but it can result in lasting reform if school
personnel consider long-term implications during the initial
phases. Applying the framework for generalization during
initial RtI implementation could potentially frontload sus-
tainability efforts and provide a roadmap to sustainability.
The goal of this article was to suggest potential methods to
apply generalization strategies to RtI implementation efforts,
primarily to provide directions for future research. Some of
the strategies mentioned above would be easily implemented
and some would require extensive research. However, given
the increased frequency of RtI implementation, the research
seems warranted.

Acknowledgments

This publication was made possible in part by Grant Number
H325D090012 from the United States Department of Edu-
cation Office of Special Education Programs. Its contents are
solely the responsibility of the authors and do not necessarily
represent the official views of the USDE OSEP.

REFERENCES

Alberto, P. A., & Troutman, A. C. (2009). Applied behavior analysis for
teachers (8th ed). Upper Saddle River, NJ: Pearson.

Berkeley, S., Bender, W. N., Peaster, L. G., & Saunders, L. (2009). Imple-
mentation of response to intervention: A snapshot of progress. Journal
of Learning Disabilities, 42, 85–95.

LEARNING DISABILITIES RESEARCH 87

Burns, M. K., & Symington, T. (2002). A meta-analysis of prereferral in-
tervention teams: Systemic and student outcomes. Journal of School
Psychology, 40, 437–447.

Burns, M. K. (2007). RTI WILL fail, unless . . . .Communiqué, 35(5), 38–40.
Burns, M. K., Appleton, J. J., & Stehouwer, J. D. (2005). Meta-analytic

review of responsiveness-to-intervention research: Examining field-
based and research-implemented models. Journal of Psychoeduca-
tional Assessment, 23, 381–394.

Burns, M. K., Vanderwood, M., & Ruby, S. (2005). Evaluating the readiness
of prereferral intervention teams for use in a problem-solving model:
Review of three levels of research. School Psychology Quarterly, 20,
89–105.

Crawford, C. G. (1980). Math without fear. New York: New View-
points/Vision Books.

DuFour, R. (2004). What is a “professional learning community?” Educa-
tional Leadership, 61(8), 6–11.

DuFour, R. (2011). Work together, but only if you want to. Phi Delta Kappan,
92(5), 57–61.

Ellis, A. K. (2005). Research on educational innovations (4th ed.). Larch-
mont, NY: Eye on Education.

Fixsen, D. L., Blase, K. A., Naoom, S. F., & Wallace, F. (2009). Core
implementation components. Research on Social Work Practice, 19,
531–540.

Foster-Fishman, P. G., Nowell, B., & Yang, H. (2007). Putting the sys-
tem back into systems change: A framework for understanding and
changing organizational and community systems. American Journal of
Community Psychology, 39, 197–215.

Fuchs, D., Fuchs, L. S., & Stecker, P. M. (2010). The “blurring” of special
education in a new continuum of general education placements and
services. Exceptional Children, 76, 301–323.

Gansle, K. A., & Noell, G. H. (2007). The fundamental role of intervention
implementation in assessing resistance to intervention. In S. Jimerson,
M. K. Burns, & A. M. VanDerHeyden (Eds.), Handbook of response to
intervention: The science and practice of assessment and intervention
(pp. 244–254). New York: Springer.

Goldenberg, C. (2003). Settings for school improvement. International Jour-
nal of Disability, Development and Education, 50, 7–16.

Grimes, J., Kurns, S., & Tilly, W. D. III. (2006). Sustainability: An enduring
commitment to success. School Psychology Review, 35, 224–244.

Hargreaves, A. & Fink, D. (2000). The three dimensions of education reform.
Educational Leadership, 57(7), 30–34.

Hill, D. R., King, S. A., Lemons, C. J., & Partanen, J. N. (2012). Fidelity of
implementation and instructional alignment in Response to Interven-
tion research. Learning Disabilities Research & Practice, 27, 116–124.

Horner, R. H., Sugai, G., & Todd, A. W. (2001). “Data” need not be a
four-letter word: Using data to improve schoolwide discipline. Beyond
Behavior, 11, 20–22.

Johnson, E. S., Pool, J. L., & Carter, D. R. (2012). Lessons learned from a
tiered service delivery implementation project. Intervention In School
& Clinic, 47, 139–143.

Kovaleski, J. F., Gickling, E. E., Morrow, H., & Swank, P. (1999). High
versus low implementation of instructional support teams: A case for
maintaining program fidelity. Remedial & Special Education, 20, 170–
183.

Kratochwill, T. R., Volpiansky, P., Clements, M., & Ball, C. (2007). Profes-
sional development in implementing and sustaining multitier preven-
tion models: Implications for response to intervention. School Psychol-
ogy Review, 36, 618–631.

Lembke, E. S., Garman, C., Deno, S. L., & Stecker, P. M. (2010). One ele-
mentary school’s implementation of Response to Intervention. Reading
& Writing Quarterly, 26, 361–373.

Mahdavi, J. N., & Beebe-Frankenberger, M. E. (2009). Pioneering RTI
systems that work. Teaching Exceptional Children, 42(2), 64–72.

McIntosh, K., Filter, K. J., Bennett, J. L., Ryan, C., & Sugai, G. (2010).
Principles of sustainable prevention: Designing scale-up of school-wide
positive behavior support to promote durable systems. Psychology in
the Schools, 47, 5–21.

Noell, G. H., & Gansle, K. A. (2009). Moving from good ideas in educa-
tional systems change to sustainable program implementation: Com-
ing to terms with some of the realities. Psychology in the Schools, 46,
78–88.

Perkins, D. F., Feinberg, M. E., Greenberg, M. T., Johnson, L. E., Chilenski,
S. M., Mincemoyer, C. C., et al. (2011). Team factors that predict to
sustainability indicators for community-based prevention teams. Eval-
uation and Program Planning, 34, 283–291.

Riley-Tillman, T. C. & Burns, M. K. (2009). Single case design for measur-
ing response toeducational intervention. New York: Guilford.

Sarason, S. (1990). The predictable failure of educational reform. San Fran-
cisco, CA: Jossey-Bass.

Sarason, S. (1996). Revisiting the culture of school and the problem of
change. New York: Teachers College Press.

Stokes, T. F., & Baer, D. M. (1976). Preschool peers as mutual generalization-
facilitating agents. Behavior Therapy, 7, 549–556.

Stokes, T. F., & Baer, D. M. (1977). An implicit technology of generalization.
Journal of Applied Behavior Analysis, 10, 349–367.

Sugai, G., Lewis-Palmer, T., Todd, A. W., & Horner, R. H. (2001). School-
Wide Evaluation Tool (SET). Eugene, OR: Educational and Community
Supports.

Werts, M. G., Lambert, M., & Carpenter, E. (2009). What special ed-
ucation directors say about RTI. Learning Disability Quarterly, 32,
245–254.

Ysseldyke, J. (2005). Assessment and decision making for students with
learning disabilities: What if this is as good as it gets. Learning Dis-
ability Quarterly, 28, 125–128.

About the Authors

Matthew K. Burns is a Professor of Educational Psychology, Coordinator of the School Psychology program, and Co-Director
of the Minnesota Center for Reading Research at the University of Minnesota. His research interests include response to
intervention, using curriculum-based assessment for instructional design to determine academic interventions, and facilitating
problem-solving teams.

Andrea M. Egan is a doctoral student in special education at the University of Minnesota. Her research interests include
assessment and intervention strategies for students with co-occurring academic and behavioral problems and methods to
address these within a response to intervention framework.

Amy K. Kunkel is a graduate research assistant in Educational Psychology at the University of Minnesota. Her research
interests include computer-assisted instruction and response to intervention.

Jennifer McComas, Ph.D., is a Professor with the Special Education Program in the Department of Educational Psychology
at the University of Minnesota. Her current research interests include functional analysis and treatment for problem behavior
and academic skill deficits, the influence of the principles of behavior on learning, and the influence of social context on severe
problem behavior.

88 BURNS ET AL.: SUSTAINABILITY OF RTI IMPLEMENTATION

Meredith M. Peterson is a doctoral student in Educational Psychology at the University of Minnesota. Her research inter-
ests include assessment and intervention strategies for students with behavioral problems within a response to intervention
framework.

Naomi L. Rahn is a doctoral candidate in Educational Psychology, Special Education at the University of Minnesota. She
has over 15 years of experience in early childhood special education. Her research interests include naturalistic language
interventions, response to intervention, and teacher preparation.

Jennifer Wilson is a doctoral candidate in Educational Psychology, Special Education at the University of Minnesota. She
holds a Director of Special Education license and has over 10 years of experience in the field. Her research interests include
response to intervention and teacher preparation.

Copyright of Learning Disabilities Research & Practice (Blackwell Publishing Limited) is the property of

Wiley-Blackwell and its content may not be copied or emailed to multiple sites or posted to a listserv without

the copyright holder’s express written permission. However, users may print, download, or email articles for

individual use.

1977, 10, 349-367

AN IMPLICIT TECHNOLOGY OF GENERALIZATION’
TREVOR F. STOKES AND DONALD M. BAER

THE UNIVERSITY OF MANITOBA AND THE UNIVERSITY OF KANSAS

Traditionally, discrimination has been understood as an active process, and a technology
of its procedures has been developed and practiced extensively. Generalization, by con-
trast, has been considered the natural result of failing to practice a discrimination
technology adequately, and thus has remained a passive concept almost devoid of a
technology. But, generalization is equally deserving of an active conceptualization and
technology. This review summarizes the structure of the generalization literature and
its implicit embryonic technology, categorizing studies designed to assess or program
generalization according to nine general headings: Train and Hope; Sequential Modifi-
cation; Introduce to Natural Maintaining Contingencies; Train Sufficient Exemplars;
Train Loosely; Use Indiscriminable Contingencies; Program Common Stimuli; Mediate
Generalization; and Train “To Generalize”.
DESCRIPTORS: generalization, treatment-gain durability, followup measures, main-

tenance, postcheck methodology

Traditionally, many theorists have considered
generalization to be a passive phenomenon. Gen-
eralization was not seen as an operant response
that could be programmed, but as a description
of a “natural” outcome of any behavior-change
process. That is, a teaching operation repeated
over time and trials inevitably involves varying
samples of stimuli, rather than the same set
every time; in the same way, it inevitably evokes
and reinforces varying samples of behavior,
rather than the same set every time. As a conse-
quence, it is predictable that newly taught re-
sponses would be controlled not only by the
stimuli of the teaching program, but by others
somewhat resembling those stimuli (Skinner,
1953, p. 107ff.). Similarly, responses resembling
those established directly, yet not themselves ac-
tually touched by the teaching procedures, would
appear as a result of the teaching (Keller and

‘Preparation of this paper was supported in part by
PHS Training Grant 00183, Program Project Grant
HD 00870, and Research Grant MH 11739. Reprints
may be obtained either from T. F. Stokes, Department
of Psychology, University of Manitoba, Winnipeg,
Manitoba, Canada R3T 2N2, or D. M. Baer, Depart-
ment of Human Development, University of Kansas,
Lawrence, Kansas 66045.

Schoenfeld, 1950, p. 168ff.). Thus, generaliza-
tion was something that happened, not some-
thing produced by procedures specific to it.

If generalization seemed absent or insignifi-
cant, it was simply to be assumed that the teach-
ing process had managed to maintain unusually
tight control of the stimuli and responses in-
volved, allowing little sampling of their varie-
ties. This assumption was strongly supported by
the well-known techniques of discrimination: by
differential reinforcement (in general, by any
differential teaching) of certain stimuli relative
to others, and/or certain responses relative to
others, generalization could be programmatically
restricted and diminished to a very small range.
Thus, it was discrimination that was understood
as an active process, and a technology of its pro-
cedures was developed and practiced extensively.
But generalization was considered the natural
result of failing to practice discrimination’s tech-
nology adequately, and thus remained a passive
concept almost devoid of a technology. Never-
theless, in educational practice, and in the devel-
opment of theories aimed at serving both practice
and a better understanding of human function-
ing, generalization is equally as important as dis-

349

NUMBER 2 (SUMMER) 1977JOURNAL OF APPLIED BEHAVIOR ANALYSIS

TREVOR F. STOKES and DONALD M. BAER

crimination, and equally deserving of an active
conceptualization.

Generalization has been and doubtless will
remain a fundamental concern of applied behav-
ior analysis. A therapeutic behavioral change,
to be effective, often (not always) must occur
over time, persons, and settings, and the effects
of the change sometimes should spread to a
variety of related behaviors. Even though the
literature shows many instances of generaliza-
tion, it is still frequently observed that when a
change in behavior has been accomplished
through experimental contingencies, then that
change is manifest where and when those contin-
gencies operate, and is often seen in only transi-
tory forms in other places and at other times.
The frequent need for generalization of thera-

peutic behavior change is widely accepted, but it
is not always realized that generalization does
not automatically occur simply because a behav-
ior change is accomplished. Thus, the need ac-
tively to program generalization, rather than
passively to expect it as an outcome of certain
training procedures, is a point requiring both
emphasis and effective techniques (Baer, Wolf,
and Risley, 1968). That such exhortations have
often been made has not always ensured that
researchers in the field have taken serious note
of and, therefore, proceeded to analyze ade-
quately the generalization issues of vital concern
to their programs. The emphasis, refinement, and
elaboration of the principles and procedures that
are meant to explain and produce generalization
when it does not occur “naturally” is an impor-
tant area of unfinished business for applied be-
havior analysis.

The notion of generalization developed here
is an essentially pragmatic one; it does not
closely follow the traditional conceptualizations
(Keller and Schoenfeld, 1950; Skinner, 1953).
In many ways, this discussion will sidestep much
of the controversy concerning terminology. Gen-
eralization will be considered to be the occur-
rence of relevant behavior under different, non-
training conditions (i.e., across subjects, settings,
people, behaviors, and/or time) without the

scheduling of the same events in those conditions
as had been scheduled in the training conditions.
Thus, generalization may be claimed when no
extratraining manipulations are needed for extra-
training changes; or may be claimed when some
extra manipulations are necessary, but their cost
or extent is clearly less than that of the direct
intervention. Generalization will not be claimed
when similar events are necessary for similar ef-
fects across conditions.
A technology of generalization programming

is almost a reality, despite the fact that until re-
cently, it had hardly been recognized as a prob-
lem in its own right. Within common teaching
practice, there is an informal germ of a technol-
ogy for generalization. Furthermore, within the
practice of applied behavior analysis (especially
within the past 5 yr or so), there has appeared
a budding area of “generalization-promotion”
techniques. The purpose of this review is to sum-
marize the structure of that generalization litera-
ture and its implicit embryonic technology. Some
270 applied behavior analysis studies relevant to
generalization in that discipline were reviewed.2
A central core of that literature, consisting of
some 120 studies, contributes directly to a tech-
nology of generalization. In general, techniques
designed to assess or to program generalization
can be loosely categorized according to nine
general headings:

1. Train and Hope
2. Sequential Modification
3. Introduce to Natural Maintaining Contin-

gencies
4. Train Sufficient Exemplars

5. Train Loosely

6. Use Indiscriminable Contingencies

7. Program Common Stimuli

2Ninety per cent of the literature reviewed was
from five journals: Behaviour Research and Therapy;
Behavior Therapy; Journal of Applied Behavior Anal-
ysis; Journal of Behavior Therapy and Experimental
Psychiatry; and Journal of Experimental Child Psy-
chology. Seventy-seven per cent of the literature re-
viewed has been published since 1970.

350

AN IMPLICIT TECHNOLOGY OF GENERALIZATION

8. Mediate Generalization

9. Train “To Generalize”.

This review characterizes each category, and
describes some examples of research that illus-
trate the generalization analyses or program-
ming involved in each category. Obviously, all
the relevant references cannot be discussed in this
review.3 The nine categories listed above were
induced from the literature; they are not a priori
categories. Consequently, studies do not always
fit neatly into these categories. It should also be
noted that not all studies reviewed were thorough
experimental analyses of generalization. Often
inferences were necessary to categorize the re-
search. However, the following discussion still
may provide a useful organization and concep-
tualization of generalization and its program-
ming.

1. Train and Hope
In applied behavior analysis research, the most

frequent method of examining generalization, so
far, may be labelled Train and Hope. After a
behavior change is effected through manipula-
tion of some response consequences, any existent
generalization across responses, settings, experi-
menters, and time, is concurrently and/or sub-
sequently documented or noted, but not actively
pursued. It is usually hoped that some generali-
zation may occur, which will be welcomed yet
not explicitly programmed. These hopeful probes
for stimulus and response generalization char-
acterize almost half of the applied literature on
generalization. The studies have considerable
importance, for they begin to document the ex-
tent and limits of generalization of particular
operant intervention techniques. While not being

3Complete reference lists and detailed tables de-
scribing subjects, procedures, and generalization of
all studies reviewed are deposited with the National
Auxiliary Publications Service (NAPS). For copies,
order NAPS Document #02873. Order from ASIS/
NAPS Co., C/O Microfiche Publications, 305 East
46th Street, New York, New York 10017. Remit
with order for each copy $3.00 for microfiche or
$19.50 for photocopies. Make checks payable to
Microfiche Publications.

examples of the programming of generalization,
they are a sound first step in any serious analysis
of generalization. When generalization is desired,
but is shown to be absent or deficient, program-
ming procedures can then be instituted.

For example, useful generalization across set-
tings was documented by Kifer, Lewis, Green,
and Phillips (1974). In an experimental class-
room setting, parent-child pairs were taught to
negotiate in conflict situations. During simulated
role-playing, instructions, practice, and feedback
were used to teach the negotiation behaviors of
fully stating one’s position, identifying the issues
of conflict, and suggesting options to resolve the
conflict. The data showed increased use of nego-
tiation behaviors and the reaching of agreements
in actual parent-child conflict situations at home.
An assessment of generalization across experi-

menters was described by Redd and Birnbrauer
(1969), who demonstrated that control over the
cooperative play of retarded children did not
generalize from an adult who dispensed contin-
gent edible reinforcement to five other adults
who had not participated in training.

Studies that are examples of Train and Hope
across time are those in which there was a change
from the intervention procedures, either to a less
intensive but procedurally different program, or
to no program or no specifically defined pro-
gram. Data or anecdotal observations were re-
ported concerning the maintenance of the origi-
nal behavior change over the specified time
intervening between the termination of the
formal program and the postchecks. An example
of a followup evaluation was the study by Azrin,
Sneed, and Foxx (1973). An intensive training
program involving reinforcement of correct toi-
leting and positive practice procedures promptly
decreased bedwetting by 12 retarded persons.
The reduced rate of accidents was maintained
during a three-month followup assessment.

Perhaps there are many more studies in the
Train and Hope category than would have been
expected (about 135, of which 65 % are across
Time). However, despite its obvious value, this
research is frequently characterized by a lack of

351

TREVOR F. STOKES and DONALD M. BAER

comprehensiveness and depth of the generaliza-
tion analysis. Even though generalized behavior
change was frequently reported, extensive, wide-
ranging, and practical generalization was not
often noted or even sought. The continued de-
velopment of behavior analysis almost surely
will demand more extensive collection of gener-
alization data than is presently the fashion. The
extent and limits of applied behavioral interven-
tions may be well documented and understood if
measurement is extended over longer periods of
time, over more than one circumscribed part of
the day, with more than one related response,
and with more than a restricted part of the social
and physical environment. It is as important for
the field to formalize the conditions of the non-
occurrence of generalization as it is to document
the conditions associated with the display of un-
programmed generalization.

Most of the Train-and-Hope research described
successful generalization-approximately 90%
of Train-and-Hope studies. By definition, there
was no further need to program generalization
in those studies where generalization had been
exhibited within the Train-and-Hope para-
digm-presuming, of course, that the generali-
zation exhibited was considered sufficient to meet
the therapeutic goals of the various modification
programs (not necessarily a valid presumption
in the Train-and-Hope research). This prepon-
derance of positive data may simply reflect the
tendency of some researchers not to report their
generalization data if measurement procedures
were instituted to probe for any generalized be-
havior changes, but generalization was shown to
be absent. Some researchers may view nongen-
eralization as reflecting a deficiency or ineffective-
ness of their procedures to develop a desirable
generalized performance. Behavior analysts,
nevertheless, should encourage their fellow re-
searchers to document and to analyze experimen-
tally their apparent failures, rather than allowing
them to slide into oblivion. A detailed and sys-
tematic understanding of generalization and its
programming could result. Alternatively, re-
searchers might view their generalization base-

lines as being essentially independent of the mod-
ified baseline; thus, to report nongeneralization
would serve no useful purpose, for its nonoccur-
rence would be expected. Again, any such docu-
mentation contributes to our understanding of
the extent and limits of generalization, as well as
serving as an indication of the frequent necessity
of generalization-programming techniques.

There is another reason for the predominance
of positive results in this section: if nongeneral-
ization was clearly evident, and the modification
of this state was important, then a form of lim-
ited programming was frequently instituted. Ex-
amples of this research will be discussed in the
next category, “Sequential Modification”.

2. Sequential Modification
These studies exemplify a more systematic

approach to generalization than the Train-and-
Hope research. Again, a particular behavior
change is effected, and generalization is assessed.
But then, if generalization is absent or deficient,
procedures are initiated to accomplish the desired
changes by systematic sequential modification in
every nongeneralized condition, i.e., across re-
sponses, subjects, settings, or experimenters. The
possibility of unprogrammed generalization typi-
cally was not examined in these sequential modi-
fication studies, because after the initial demon-
stration of nongeneralization, all other baselines
were exhausted. That is, after changes had been
produced directly in all baselines, generalization
to nonrecorded responses, subjects, settings, and
experimenters may have occurred, but could not
be examined.

For example, Meichenbaum, Bowers, and Ross
(1968) reported an absence of generalization of
behavior changes from an afternoon intervention
period to the morning period in a classroom for
institutionalized female adolescent offenders.
Money dispensed contingent on on-task behav-
iors effected desired behavior changes during the
afternoon, but generalization to the morning
period required that the same manipulations be
applied there as well (sequential modification
across settings). Similarly, generalization across

352

AN IMPLICIT TECHNOLOGY OF GENERALIZATION

settings of the disruptive and oppositional behav-
ior of two children was investigated by Wahler
(1969). He demonstrated control of these behav-
iors in the home by using differential attention
and timeout operations. When generalization to
the children’s school behavior was not evidenced,
similar contingency operations were employed to
accomplish changes in that setting as well.
The category of Sequential Modification char-

acterizes much of the actual practice of many
behavior analysts. Sequential modification is
merely a systematized experimental procedure
that formalizes and allows evaluation of these
typical therapeutic endeavors. The tactic of
scheduling behavior-change programs in every
condition to which generalization is desired is
frequently employed. The rationale for these
procedures is as follows. If a desired generaliza-
tion is not likely to be exhibited after changing a
behavior in a particular condition, or a number
of conditions, e.g., settings, then the researcher or
practitioner works to effect changes across con-
ditions as a matter of course, rather than as an
outcome of the display or nondisplay of general-
ization. Thus, a behavior analyst is likely to ad-
vise the scheduling of consequences in every
relevant condition in preference to the dispens-
ing of consequences in only one or a few condi-
tions, while hoping for generalization, but likely
not seeing it.

3. Introduce to Natural Maintaining
Contingencies

Perhaps the most dependable of all general-
ization programming mechanisms is one that
hardly deserves the name: the transfer of behav-
ioral control from the teacher-experimenter to
stable, natural contingencies that can be trusted
to operate in the environment to which the sub-
ject will return, or already occupies. To a con-
siderable extent, this goal is accomplished by
choosing behaviors to teach that normally will
meet maintaining reinforcement after the teach-
ing (Ayllon and Azrin, 1968).

Baer and Wolf (1970) reported a study by
Ingram that illustrated the mechanism of “trap-

ping”, where a preschool child was taught an
entry response that exposed the child to the
natural contingencies of peers in the preschool
environment. Preschool teachers modified the
low rate of skillful interaction of the child by
priming others to interact with the subject and
reinforcing appropriate interactions. The data
showed that over time the teachers lost control
of the interaction behavior, which remained
high; it was assumed that the group’s natural
consequences for interaction had taken control
of the subject’s behavior. Thus, to program gen-
eralization, the child perhaps needed only to be
introduced adequately to the natural reinforcers
inherent in active preschool play and interaction.
Some early analyses of preschool children’s be-
havior have stressed that if the child can be so
introduced (through the operation of differential
attention from teachers) to a reinforcing pre-
school natural environment, then the behaviors
eventually do not need to be maintained by con-
tinued contrived modification of the environ-
ment. For example, Hall and Broden (1967)
modified the manipulative play, climbing, and
social interaction of three subjects through social
reinforcement operations. Behavior changes were
shown to be durable and successful followup
data at three months were described.

Buell, Stoddard, Harris, and Baer (1968) dem-
onstrated the collateral development of appro-
priate social behavior (e.g., touching, verbalizing,
and playing with other children) accompanying
the reinforcement of increased use of outdoor
play equipment by a 3-yr-old girl. This entry re-
sponse to the natural reinforcement community
was tactically sound because the child’s motor
behavior was modified in a setting where the
resulting behavior would tend automatically to
increase social contact with other children, and
this natural social environment could maintain
the child’s new skills, but indeed may also be
expected to sharpen and refine them, and add
entirely new ones as well.
Most of the research concerning natural main-

taining contingencies has involved children, per-
haps because such techniques seem particularly

353

TREVOR F. STOKES and DONALD M. BAER

suitable, especially to their social behavior. Re-
search would profit by determining what natural
reinforcement communities exist for various be-
haviors and subjects, and what economical means
may be employed to ensure entry to these behav-
ioral traps.

Unfortunately, in some instances there may be
no natural reinforcement operating to develop
and maintain skills. For example, in the case of
retarded and institutionalized persons whose de-
pendency has become a stable fact in the lives of
their caretakers, some re-arrangement of the
natural environment may be necessary. A few
studies have introduced subjects to semicontrived
or redesigned “natural” reinforcement communi-
ties. A simple but meaningful example was pro-
vided by Horner (1971), who taught a 5-yr-old
institutionalized retarded boy to walk on crutches
in an experimental setting. The child was then
prompted to generalize the new walking skill
to other settings and activities to which he previ-
ously had been taken in a wheelchair by solici-
tous caretakers, by enlisting those caretakers to
refrain from offering this help. Within 15 days
after treatment was concluded, the child walked
on crutches to all those activities and settings,
eventually extending his ambulation skills to any
part of his world. Stolz and Wolf (1969) trained
a 16-yr-old, “blind” retarded male to discriminate
visual stimuli. Then, the environment was so
structured that assistance was not given in situa-
tions where it had previously been given as a
matter of course. When the boy was required to
use visual cues to help himself in a cafeteria line,
he soon emitted the necessary behaviors. How-
ever, these studies did not establish the function-
ality of their procedures in the maintenance of
behavior changes.

Another significant example was provided by
Seymour and Stokes (1976). In their study, insti-
tutionalized delinquent girls were taught to so-
licit reinforcement (cf. Graubard, Rosenberg,
and Miller, 1971) from their natural community,
the staff of their residential institution. In their
case, the staff had rarely displayed any systematic
attempts at reinforcing desirable behavior shown

by the girls, perhaps on the presumption that the
girls were “bad” and not reinforcible in any case.
However, the experimenters were able to teach
the girls that when their work was objectively
good, and when staff persons were nearby, a
simple skill of calling these adults’ attention to
their good work would result in fairly consistent
reinforcement. Thus, this was a case in which
experimental reinforcement was used to develop
a response in the subjects that would tap and
cultivate the available but dormant natural com-
munity. In theory, this new skill should have
obviated the need for further experimental re-
inforcement, for the praise evoked should have
functioned to maintain both the girls’ work and
cueing, and the cueing, in turn, should have func-
tioned to maintain staff praise. The Seymour and
Stokes’ study could not be continued long enough
to establish whether this would happen, and so
it remains a logically appealing but still unex-
plored method of enhancing generalization:
teaching the subject a means of recruiting a nat-
ural community of reinforcement to maintain
that generalization. Perhaps an even greater ad-
vantage of such procedures is a change in the
locus of control: the subjects can become more
prominent agents of their own behavior change,
rather than being hapless pawns of more-or-less
random environmental contingencies.

Restructuring the environment thus becomes a
target of research aimed at extending the gener-
alization of newly taught skills; even though, at
a technical level, this operation may not be con-
sidered generalization, but rather transfer of
control from one reinforcement contingency to
another. In any event, it is a much neglected
topic of experimental research, although widely
recognized as a desirable, and even essential
characteristic of any rehabilitative effort.

Some natural contingencies are inevitably at
work contributing to the maintenance of inap-
propriate behavior. For example, peer-group
control of inappropriate behavior has often been
suspected and sometimes documented (Buehler,
Patterson, and Furniss, 1966; Gelfand, Gelfand,
and Dobson, 1967; Solomon and Wahler, 1973).

354

AN IMPLICIT TECHNOLOGY OF GENERALIZATION

It would seem reasonable, then, that if the pat-
tern of reinforcement of inappropriate behavior
is modified, the observed outcome may errone-
ously, but happily be attributed to generalization.
For example, Bolstad and Johnson (1972) pre-
sented data that showed that both experimental
and control subjects in the same classroom were
all affected (although not to the same extent) by
experimental manipulation of the reinforcement
contingency for the experimental subjects,
whereas control subjects in a different classroom
were not so affected. The authors presented data
that may account for these differences. The con-
trol subjects in the experimental classroom, who
were also disruptive students, had fewer disrup-
tive interactions with the experimental subjects
during the treatment phases than during base-
line. This possible generalization effect may be
due to the disruption of the natural contingencies
operating in that environment. That is, other
disruptive students previously supported some of
the disruptive behavior of the control subjects,
but during treatment these experimental subjects
did not support the disruptive behavior of their
peers and, thus, a “generalized” decrease in dis-
ruptive behavior by the control subjects resulted.

4. Train Sufficient Exemplars
If the result of teaching one exemplar of a

generalizable lesson is merely the mastery of the
exemplar taught, with no generalization beyond
it, then the obvious route to generalization is to
teach another exemplar of the same generaliza-
tion lesson, and then another, and then another,
and so on until the induction is formed (i.e., until
generalization occurs sufficiently to satisfy the
problem posed). Examples of such programming
techniques will be described in this category of
training sufficient exemplars, perhaps one of the
most valuable areas of programming. Certainly
it is the generalization-programming area most
prominent and extensive in the present literature.

In the research discussed previously under
the categories of Train and Hope and Sequen-
tial Modification, the typical analysis of gener-
alization concerned the measurement of gener-

alization to only a few (and often only one)
extraexperimental responses, subjects, settings,
experimenters, or times. When the absence of
generalization was noted, sometimes it was ac-
complished by further direct intervention in
every nongeneralized condition (i.e., Sequential
Modification). Having completed such modifica-
tions, the possibility of more extensive general-
ized effects (i.e., beyond the two or three modified
baselines) was not examined. In the training of
sufficient exemplars, generalization to untrained
stimulus conditions and to untrained responses
is programmed by the training of sufficient ex-
emplars (rather than all) of these stimulus con-
ditions or responses.
A systematic demonstration of programmed

generalization and measurement of generalized
effects beyond intervention conditions was re-
ported by Stokes, Baer, and Jackson (1974). They
established that training and maintenance of re-
tarded childrens’ greeting responses by one ex-
perimenter was not usually sufficient for the
generalization of the response across experi-
menters. However, high levels of generalization
to over 20 members of the institution staff (and
newcomers as well) who had not participated
in the training of the response were recorded,
after a second experimenter trained and main-
tained the response in conjunction with the first
experimenter. Thus, when generalization did not
prevail after the training of one stimulus exem-
plar, it was programmed by training a greater
diversity of stimulus (trainer) conditions. Simi-
larly, Garcia (1974) taught a conversational
speech form to two retarded children, and, upon
discovering a lack of stable generalization across
experimenters after one training input, pro-
grammed generalization across experimenters by
having a second experimenter teach the same
responses.
A sufficient-stimulus-exemplars demonstration

of programmed generalization across settings has
been described by Allen (1973). Allen modified
the bizarre verbalizations of an 8-yr-old boy by
differential attention procedures. Ignoring bi-
zarre verbalizations and praise for appropriate

355

TREVOR F. STOKES and DONALD M. BAER

interaction reduced bizarre verbalizations during
evening camp activities. However, there was no
generalization to three other camp settings. After
additional training in a second setting, some
generalization to the unmanipulated settings was
noted. This generalization was further enhanced
by intervention in the third setting. Unfortu-
nately, the experimental procedures did not allow
sufficient time to document the full extent of
generalization after training in two settings, but
generalization after training in two settings was
clearly evident. Griffiths and Craighead (1972)
similarly programmed generalization across set-
tings. A 30-yr-old retarded woman received
praise and tokens for correct articulation in
speech therapy. Generalization to a residential
cottage was not observed until the same proce-
dures were instituted there. Following training
in these two stimulus exemplars, generalization
to a third nontraining setting (a classroom) was
observed.

Very little research concerned with generaliza-
tion programming has dealt with the training of
sufficient stimulus exemplars. The infrequent
research that has been published is characterized
largely by programming across experimenters.
This work has been promising, for after a modest
number of training inputs, generalization appar-
ently will occur with persons not involved in
training-unquestionably a valuable and inex-
pensive outcome. However, the present implica-
tion of these studies is limited because of the
restricted nature of the type of subjects and
responses analyzed. Further work is also needed
to give direction to the optimal conditions
whereby the most extensive generalization will
be achieved with a minimal training expendi-
ture. Nevertheless, it is optimistic to note how
frequently a sufficient number of exemplars is
a small number of exemplars. Frequently, it is
no more than two. In particular, there may well
be reason to suspect that the use of two trainers
will yield excellent results in terms of generaliza-
tion. This possibility, obviously an economical
one, certainly merits systematic study of its po-
tential and limits.

Although very little research has been re-
ported, the analysis of generalization program-
ming by training in a number of settings is a
virtually untapped area of far-reaching value.
However, consistent optimism should follow ex-
amination of the studies showing generalization
after training in only a few settings. Unfortu-
nately, behavior analysts seem too often satisfied
with the modification of a single, well-defined
behavior in one setting, e.g., a laboratory pre-
school. Discriminated programs are often accept-
able, and sometimes even desirable. When gener-
alization is a valid concern, but researchers and
practitioners do not act as if this were so, the
discriminated behavior of researchers is most
probably inhibitory to the development of an
effective generalization technology.

Over the past 10 yr, there has developed an
extensive literature discussing the programmed
generalization of responses through the training
of sufficient response exemplars. A response class
has been operationally defined to describe the
fact that some responses are organized such that
operations applied to a subset of responses in the
class affect the other members of that class in the
same manner. For example, Baer, Peterson, and
Sherman (1967) reinforced various motor imita-
tions by retarded children. They found that as
long as reinforcement followed some imitative
responses, other imitations continued to be per-
formed without training or reinforcement.
A topographical analysis of generalized imita-

tion has been made by Garcia, Baer, and Fire-
stone ( 1971). Four retarded children were trained
to imitate three different topographical types of
response: small motor, large motor, and short
vocal. These subjects were also probed for their
imitation of other unreinforced responses: short
motor, long motor, short vocal, and long vocal.
Generalized imitation was observed with each
subject, but this generalization reflected the par-
ticular dimensions of the topographical response
currently being trained or having previously re-
ceived training. Thus, generalization may occur
within well-defined classes and may not gener-
alize to other classes unless some special training

356

AN IMPLICIT TECHNOLOGY OF GENERALIZATION

(generalization programming) occurs within that
class as well. These data depict one possible limi-
tation of the generality of generalized imitation,
as well as pointing to the need to train response
exemplars that will adequately reflect the di-
versity of the generalization being programmed.

Children’s grammatical development has been
another prominent area of research dealing with
generalized behavior. The concept of response
class is again pivotal in these studies, which
conceptualize the rules of morphological gram-
mar as equivalent to response class phenomena.
For example, Guess, Sailor, Rutherford, and
Baer (1968) developed the generative correct
use of plurals by a retarded girl. After teaching
a number of exemplars of the correct plural
response, the girl appropriately labelled new
objects in the singular or plural without further
direct training relevant to those objects. Plural
usage had become a generalized response class;
the morphological rule had been established.
Schumaker and Sherman (1970) rewarded three
retarded children for the correct production of
past- and present-tense forms of verbs. As past-
and present-tense forms of verbs within an in-
flectional class were modified, there occurred a
generalized usage of untrained verbs to similar
tense forms.

There has been considerable research to estab-
lish the importance of the training of sufficient
response exemplars. A survey of these (approxi-
mately 60) studies shows that the number of
exemplars found to be “sufficient” for a desirable
level and durability of generalization varies
widely, probably determined primarily by the
nature of the task and the subject’s prior skills
relevant to it. Most of this research was con-
cerned with the development of motor and vocal
imitations, and the beginning development of
grammar and syntax. The development of ques-
tion-asking and instruction-following is also well
represented.

In conclusion, examination of the sufficient
exemplar research points to a significant (and
long-familiar) generalization-programming pro-
cedure: a number of stimulus and/or response

exemplars should undergo training. That is, to
program the generalized performance of certain
responses across various setting conditions or
persons, training should occur across a (suffi-
cient) number of setting conditions and/or with
various persons. In a similar manner, generaliza-
tion across responses can be programmed reliably
by the training of a number of responses. Diver-
sity of exemplars seems to be the rule to follow
in pursuit of the maximum generalization. Suffi-
cient diversity to reflect the dimensions of the
desired generalization is a useful tactic. However,
diversity may also be our greatest enemy: too
much diversity of exemplars and not enough
(sufficient) exemplars of similar responses may
make potential gains disproportional to the in-
vestment of training effort. The optimal combi-
nation of sufficient exemplars and sufficient di-
versity to yield the most valuable generalization
is critically in need of analysis. Is the best pro-
cedure to train many exemplars with little diver-
sity at the outset, and then expand the diversity
to include dimensions of the desired generaliza-
tion? Or is it a more productive endeavor to
train fewer exemplars that represent a greater
diversity, and persist in the training until gen-
eralization emerges’?

5. Train Loosely

One relatively simple technique can be con-
ceptualized as merely the negation of discrimi-
nation technique. That is, teaching is conducted
with relatively little control over the stimuli
presented and the correct responses allowed, so
as to maximize sampling of relevant dimensions
for transfer to other situations and other forms
of the behavior. A formal example of this most
often informal technique was provided by
Schroeder and Baer (1972), who taught vocal
imitation skills to retarded children in both of
two ways, one emphasizing tight restriction of
the vocal skills being learned at the moment
(serial training of vocal imitations), and the
other allowing much greater range of stimuli
within the current problem (concurrent training

357

TREVOR F. STOKES and DONALD M. BAER

of imitations). The latter method was charac-
terized repeatedly by greater generalization to
as-yet-untaught vocal imitation problems, thus
affirming “loose” teaching techniques as a con-
tributor to wider generalization.

It will be appreciated that the literature of the
field contains very few examples of this type.
Researchers always have attempted to maintain
thorough control and careful restriction and
standardization of their teaching procedures,
primarily to allow easy subsequent interpretation
of the nature of their (successful) teaching tech-
niques. Yet the import of this technique is that
careful management of teaching techniques to a
precisely repetitive handful of stimuli or formats
may, in fact, correspondingly restrict generaliza-
tion of the lessons being learned. The ultimate
force of this recommendation remains to be seen.
What seems required is programmatic research
aimed at assessing the generalization character-
istics of lessons taught under careful, restricted
conditions, relative to similar lessons taught
under looser, more variable conditions.

6. Use Indiscriminable Contingencies
Intermittent schedules of reinforcement have

been shown repeatedly to be particularly resistant
to extinction, relative to continuous schedules
(Ferster and Skinner, 1957). Resistance to extinc-
tion may be regarded as a form of generaliza-
tion-generalization across time subsequent to
learning. The essential feature of intermittent
schedules may be their unpredictability-the
impossibility of discriminating reinforcement oc-
casions from nonreinforcement occasions until
after the fact. Thus, if contingencies of reinforce-
ment or punishment, or the setting events that
mark the presence or absence of those contingen-
cies, are made indiscriminable, then generaliza-
tion may well be observed.

In generalization, behavior occurs in settings in
which it will not be reinforced, just as it does in
settings in which it will be reinforced. Then, the
analogue to an intermittent schedule, extended
to settings, is a condition in which the subject

cannot discriminate in which settings a response
will be reinforced or not reinforced. A potential
approximation to such a condition was presented
in a study by Schwarz and Hawkins (1970).
In that experiment, the behavior of a sixth-grade
child was videotaped during math and spelling
classes. Later, after each school day had ended,
the child was shown the tape of the math class
and awarded reinforcers according to how often
good posture, absence of face-touching, and ap-
propriate voice-loudness were evident on that
tape. Although reinforcers were awarded only
on the basis of behaviors displayed during the
math class, desirable improvements were ob-
served during the spelling class as well. In that
reinforcement was delayed, this technique must
have made it difficult for the child to discriminate
in which class the behaviors were critical for
earning reinforcement. In other words, the gen-
eralized success of the study may well be at-
tributable to the partly indiscriminable nature
of the reinforcement contingency.

In general, it may be suspected that delayed
reinforcement often will have the advantage of
making the times and places in which the con-
tingency actually operates indiscriminable to the
subject. However, this advantage is an advantage,
by hypothesis, primarily for the goal of general-
ization. Otherwise, delayed reinforcement would
often be considered an inefficient technique, most
especially so for the initial development of a new
skill. Indeed, it may be exactly in the realm of
disadvantaged persons such as retarded children
that the usual inefficiency of delayed reinforce-
ment may seem the most severe handicap to its
use. However, its potential for fostering general-
ization suggests strongly that further research be
invested in this procedure (and any others that
make reinforcement contingencies properly in-
discriminable), to develop methods of applying
it perhaps only after the initial development of a
new skill, in the interests of promoting gen-
eralization.

Less than a dozen studies of generalization
interpretable as cases of indiscriminable rein-
forcement contingencies can be found in the

358

AN IMPLICIT TECHNOLOGY OF GENERALIZATION

literature. Kazdin (1973), for example, showed
that teacher attention to one retarded child was
responded to by another child as if it were rein-
forcement for on-task behavior. Indeed, the
onlooker reacted with increased on-task behavior,
even when the teacher attended to the target
child’s off-task behavior. Possibly, prior experi-
ence with reinforcement contingent on the peers’
on-task behavior was sufficient to make all future
praise (contingent or not) discriminative for on-
task behavior. In other words, with sufficient
prior experience, the onlooker may have stopped
observing the contingency in which the rein-
forcement operated and responded only to the
reinforcing stimulus’ presence, making the con-
tingency functionally indiscriminable.

Generalization across subjects has similarly
been reported by Broden, Bruce, Mitchell, Carter,
and Hall (1970) in a classroom of culturally
disadvantaged children. When positive teacher
attention was given for one child’s attention to
academic work, the attending of a peer also in-
creased. This generalization was also a probable
function of the cueing properties of teacher rein-
forcement. However, the generalization observed
may also have been due to the manipulation of
natural social consequences received by the non-
target child through peer attention, or may have
been caused by a slight increase in the amount
of teacher attention to the nontarget child. These
effects deserve further systematic evaluation be-
cause of their relevance to the classroom prac-
tices of many teachers who strive to instruct
effectively but are unable to devote extensive
time to individual children.

Pendergrass (1972) showed that timeout could
be employed to decrease the destructive behavior
of two retarded children. With one subject, de-
creased rates were also observed with another
response (self-biting) which was sometimes
chained to the destructive behavior, but not
itself subjected to contingent timeout. However,
with the second subject, generalization to a sec-
ond response (autistic jerking movement) was
not observed. Analysis of the data revealed that
the two behaviors occurred simultaneously more

frequently with the subject with whom general-
ization was evidenced. Thus, with this subject,
punishment of the generalization response oc-
curred more frequently when destructive behav-
ior was punished. Unfortunately, it was not
determined how often the self-biting occurred
at times not simultaneous with the destructive
behavior. Therefore, the schedule of punishment
for self-biting was not established, i.e., whether
biting occurred only when destructive behavior
occurred and, therefore, always met the timeout
contingency. In this example (which was not
intended to be a careful analysis of the indis-
criminable reinforcement concept), not only was
the reinforcement contingency somewhat diffi-
cult to discriminate, but the two behaviors (de-
structive and self-destructive responses) also may
well have been only somewhat differentiated by
the subject.

Thus, preventing the ready discrimination of
contingencies is a generalization-programming
technique worthy of application and research.
Perhaps a random or haphazard delivery of re-
inforcement will (if luck or good judgement
prevails) function to modify targetted behavior
as well as behavior occurring in proximal time
or space. Even noncontingent reinforcement,
delivered at the outset of an intervention pro-
gram, may retard initial effects, but may work
to later advantage in generalization outcomes.

Finally, Kazdin and Polster (1973) showed
once again the usefulness of intermittent sched-
ules to delay subsequent extinction, relative to
continuous schedules of reinforcement. Social
interaction by two retardates was reinforced with
tokens. After establishing social interaction, one
subject received continuous reinforcement and
the other, intermittent reinforcement. During
extinction, only the subject who received inter-
mittent reinforcement continued to interact so-
cially with peers. However, these results may
simply reflect different extinction rates by two
subjects. The research was essentially a group
study where N 1. Adequate single-subject ex-
perimental control was lacking. Therefore, repli-
cation of these procedures would be desirable.

359

TREVOR F. STOKES and DONALD M. BAER
7. Program Common Stimuli

The passive approach to generalization de-
scribed earlier need not be a completely imprac-
tical one. If it is supposed that generalization
will occur, if only there are sufficient stimulus
components occurring in common in both the
training and generalization settings, then a rea-
sonably practical technique is to guarantee that
common and salient stimuli will be present in
both. One predictor of the salience of a stimulus
to be chosen for this role is its already established
function for other important behaviors of the
subject.

Children’s peers may represent peculiarly
suitable candidates for a stimulus common to
both training and generalization settings. An
example has been provided by Stokes and Baer
(1976). In their study, two children exhibiting
serious learning disabilities were recruited to
learn several word-recognition skills. One child
was taught these skills and concurrently shown
how to teach them to the other child, thus acting
as a peer-tutor. It was found that both children
reliably learned the skills, but that neither gen-
eralized them reliably or stably to somewhat dif-
ferent settings in which the other child usually
was absent. However, when the peer-tutor was
brought into those settings, then each child simi-
larly showed greatly increased and stabilized
generalization, even though there were never
any consequences for generalization. Similar
demonstrations have been provided by Johnston
and Johnston (1972) for the skill of speech
articulation. In that study, peers were rewarded
for correct monitoring of the subjects’ articula-
tion. Generalization of correct articulation oc-
curred only when the “monitoring” peer was
present. Unfortunately, it was not determined
clearly whether generalization was evidenced be-
cause of the discriminative properties of the
peers’ presence in both settings, or whether the
peers actively continued their monitoring in the
generalization setting.

Rincover and Koegel (1975) have also incor-
porated functional training stimuli into the gen-

eralization setting. Autistic children were re-
warded for imitation and instruction-following
in a training setting. Four of their 10 subjects
then did not exhibit generalization to a different
setting. Therefore, to program for this general-
ization, various aspects of the training procedures
(e.g., hand movement by therapist) or physical
training environment (e.g., table and chairs)
were systematically introduced to the generaliza-
tion setting to control generalization. Making
the experimental setting more closely resemble
the regular classroom (generalization setting)
was the programming procedure employed by
Koegel and Rincover (1974). They decreased
the teacher-to-student ratio in the experimental
setting from 1-to-i to 1-to-8. After these special
programming conditions were instituted, there
was increased performance on previously learned
and new behaviors learned in the classroom.
Walker and Buckley (1972) programmed gener-
alization of the effects of remedial training of
social and academic classroom behavior by estab-
lishing common stimuli between the experimen-
tal remedial classroom and the childrens’ regular
classroom by using the same academic materials
in both classrooms.

The literature of this field shows only a hand-
ful of studies deliberately making use of a com-
mon stimulus in both training and generalization
settings. Obviously, this is a technological dimen-
sion urgently in need of thorough development.
The use of peers as the common stimulus has
much to recommend it as a practical and natural
technique. To what extent peers need to partici-
pate in the training setting has not yet been
determined, although the absence of generaliza-
tion sometimes shown when peers are present
in nontraining settings, suggests that peers not
involved in a training setting will not likely
acquire sufficient discriminative function to con-
trol generalized responding. The use of common
physical stimuli is in even greater need of sys-
tematic research. A common stimulus approach
to generalization would encourage the incorpora-
tion into training settings of (naturally occur-
ring) physical stimuli that are frequently promi-

360

AN IMPLICIT TECHNOLOGY OF GENERALIZATION

nent or functional in nontraining environments.
If these stimuli are well chosen, and can be made
functional and salient in the training procedures,
then generalization may thereby be programmed.

8. Mediate Generalization

Mediated generalization is well known as a
theoretical mechanism explaining generalization
of highly symbolic learnings (Cofer and Foley,
1942). In essence, it requires establishing a re-
sponse as part of the new learning that is likely
to be utilized in other problems as well, and will
constitute sufficient commonality between the
original learning and the new problem to result
in generalization. The most commonly used
mediator is language, apparently. However, the
deliberate application of language to accomplish
generalization is rare in the literature reviewed,
and correspondingly little is known about what
aspects of a language response make for best
mediation.
A sophisticated analysis of mediated general-

ization was conducted by Risley and Hart ( 1968),
who taught preschool children to report at the
end of play on their play-material choices. Men-
tion of a given choice was reinforced with snacks,
which produced increased mentioning of that
choice, but no change in the children’s actual
use of that play-material. When reinforcement
was restricted to true reports of play-material
choices, however, the children then changed their
play behavior (the next day) so that when
queried about that play, they could truthfully
report on their use of the specified play material
and earn reinforcement. Control over any choice
of play materials proved possible with this tech-
nique, which placed teaching contingencies not
on the play, but on a potential mediator (verbal
report) of that play behavior. That the reports
were only potential mediators was apparent in
the early stages of the study, when the children
readily reported (untruly) their use of play ma-
terials with no corresponding actual behavior
with those materials; at that stage, they earned
reinforcement even so. When the reinforcement

was restricted to true reports, the reports then
became mediators of play behavior. The lesson
generalized, such that after several sequential
experiences with these procedures, the children
then used reports about play as mediators, even
without reinforcement being restricted to only
true reports. Israel and O’Leary (1973) used
essentially the same paradigm to compare the
effects of having children report first what they
would play with later, in contrast to having them
report after play what they had done (the Risley
and Hart method); they found that reinforcing
postreports (when they were true) produced more
actual behavior (the next day) than reinforcing
the actual behavior when it agreed with the
earlier promise to perform it. This technique has
been extended subsequently to the case of social
skills, specifically sharing and praising between
young children (Rogers-Warren and Baer, 1976).
In that case, modelling was added, such that the
young children would have a thorough chance
to learn the nature of the relatively complex
responses at issue.

Obviously, verbal mediation can easily fail,
most especially in those situations in which the
verbal mediators have little meaning (i.e., tightly
restricted discriminative value) for the subjects.
It is commonplace to find children agreeing to a
query (e.g., about whether they praised or shared)
without any knowledge of what that must entail
in actual behavior. In the case of retarded chil-
dren, it might be particularly true that the ability
to use verbal responses as mediators would lag
behind that of normal children using the same
language responses. It may be reasonable to
suggest that in the development of language-
training programs, systematic attention be given
to the training of language skills sufficiently well
elaborated to function as mediators of nonverbal
behavior. Language is a response, of course; it is
also, equally obviously, a stimulus to the speaker
as well as to the listener. Thus, it meets perfectly
the logic of a salient common stimulus, to be
carried from any training setting to any general-
ization setting that the child may ever enter.
It also perfectly exemplifies the essence of the

361

TREVOR F. STOKES and DONALD M. BAER

active generalization approach recommended
earlier.
The mediation of generalization is also exem-

plified in the behavior analysis research of self-
control and self-management procedures. That
is, self-control procedures such as self-recording,
taught as part of an intervention program, may
function to promote generalization: such tech-
niques are easy to transport and may be em-
ployed readily to facilitate responding under
generalization conditions. Some research that has
employed any or all of the various tactics of self-
assessment, self-recording, self-determination of
reinforcement, and/or self-administration of re-
inforcement (Glynn, Thomas, and Shee, 1973),
has also displayed maintenance and generaliza-
tion of behavior change; however, the correla-
tion is not perfect.

Broden, Hall, and Mitts (1971) reported that
after an eighth-grade girl experienced self-
recording of study behavior and teacher praise
for improved study, her study behavior main-
tained at a high level for a recorded three weeks.
Although the individual effects of the self-record-
ing and praise were not determined, it is possible
that the self-recording procedures contributed
significantly to this generalization.

Drabman, Spitalnik, and O’Leary (1973)
taught disruptive children to match their teach-
er’s evaluations of their appropriate classroom
behavior. Tokens were dispensed for appropriate
classroom behavior and accurate matching. Dis-
ruptive classroom behavior decreased and was
maintained at low levels during a 12-day phase
when tokens were not dispensed for self-record-
ing accuracy. Generalized behavior improvement
was also evident during a 15-min no-token
period within the experimental hour. These
changes were possibly a function of the close
temporal proximity of the token periods, which
frequently immediately preceded or followed
the generalization period.

The role of self-control procedures in medi-
ating generalization has often been proposed.
Research would do well to examine the contri-
bution of self-control tactics in generalization

and maintenance, especially when formal inter-
vention manipulations have ceased to operate.
The effects of accompanying procedures should
be experimentally separated from self-control
effects, and the role of each of the various self-
control tactics (Glynn et al., 1973) should be
individually analyzed. The potential of self-
mediated generalization is apparent, but its im-
plications and practical utility still remain to be
assessed.

9. Train “To Generalize”
If generalization is considered as a response

itself, then a reinforcement contingency may be
placed on it, the same as with any other operant.
Informally, teachers often do this when they urge
a student who has been taught one example of
a general principle to “see” another example
as “the same thing”. (In principle, they are also
attempting to make use of language as a medi-
ator of generalization, relying on the supposed
characteristics of words like “same” to accom-
plish the generalization.) Common observation
suggests that the method often fails, and that
when it does succeed, little extrinsic reinforce-
ment is offered as a consequence. A more formal
example of the technique was seen in a study
by Goetz and Baer (1973), in which three pre-
school children were taught to generalize the
response of making block forms (in blockbuild-
ing play). Descriptive social reinforcement was
offered only for every different form the child
made, i.e., contingent on every first appearance
of any blockbuilding form within a session, but
not for any subsequent appearances of that form.
Thus, the child was rewarded for moving along
the generalization gradient underlying block-
form inventions, and never for staying at any
one point. In general, the technique succeeded, in
that the children steadily invented new block
forms while this contingency was in use. Thus,
there exists the possibility of programming rein-
forcement specifically, perhaps only, for move-
ment along the generalization gradient desired.

In largely unspecified ways, perhaps two other
studies exemplify this logic. Herbert and Baer

362

AN IMPLICIT TECHNOLOGY OF GENERALIZATION

(1972), for example, taught two mothers of
deviant children to give social reinforcement
only to their children’s appropriate behaviors,
but taught the mothers from the outset to judge
all behavior according to criteria they helped
to develop, rather than attack only a few speci-
fied child responses. These mothers learned a
generalized skill because they applied correct
social contingencies to categories that included
virtually all appropriate child behavior likely to
occur. Behavior changes were maintained at 20
and 24 weeks after completion of formal train-
ing. Similarly, Parsonson, Baer, and Baer (1974)
taught two teachers of retarded children to apply
generalized correct social contingencies to all
likely appropriate and inappropriate behaviors
of preschool retarded children. These effects were
also durable over several months. Apparently
generalized changes were produced in these stud-
ies by Herbert and Baer and Parsonson et al.,
but the extent and quality of that generalization
was not quantified as such.

Very few studies of this type are found in the
literature of applied behavior analysis, probably
because of the preference of behaviorists to con-
sider generalization as an outcome of behavioral
change, rather than as a behavior itself. Ulti-
mately, this behavioristic stance may well prove
durable and consistent. Meanwhile, it is worth
hypothesizing that “to generalize” may be treated
as if it were an operant response, and reinforced
as such, simply to see what useful results occur.

Consequently, one other technique deserves
discussion: the systematic use of instructions to
facilitate generalization. Thus, if a behavior is
taught and generalization is not displayed, the
least expensive of all techniques is to tell the
subject about the possibility of generalization
and then ask for it. If that generalization then
occurs, it may well be referred to as “instructed
generalization”. If the effects of that instruction
are themselves to become generalized (yielding
a “generalized generalizer”?), then reinforcement
of the generalized behavior, on a suitable sched-
ule, might well be prudent, at least at first. Per-
haps it is simply a very elaborate version of this

technique that is being practiced when a client
is taught to relax in a somewhat anxiety-arousing
situation, and reinforced (socially) for doing so;
and then is instructed to relax in a somewhat
more powerful anxiety-arousing situation, etc.
That is, systematic desensitization to a heirarchy
of stimuli may be analyzed as reinforcing not just
relaxation, but also generalization along an al-
ready constructed generalization gradient (cf.
Yates, 1970, p. 64ff.).

CONCLUSION

The structure of the generalization literature
and its implicit embryonic technology has been
summarized. The most frequent treatments of
generalization are also the least analytical-those
described as Train and Hope and Sequential
Modification. Included in the category of Train
and Hope were those studies where the potential
for generalization had been recognized, its pres-
ence or absence noted, but no particular effort
was expended to accomplish generalization. By
contrast, some limited programming was imple-
mented in the Sequential Modification research.
In these studies, given an absence of reliable
generalization, procedures to effect changes were
instituted directly in every nongeneralized condi-
tion. Although contributing significantly to our
understanding of the generalization of behavior-
change programs, these studies are not examples
of the programming of generalization.

Seven categories were discussed that directly
relate to a technology of generalization. First,
the potential role of Natural Maintaining Con-
tingencies was discussed. According to this tactic,
generalization may be programmed by suitable
trapping manipulations, where responses are in-
troduced to natural reinforcement communities
that refine and maintain those skills without
further therapeutic intervention. The Training
of Sufficient Exemplars is numerically the most
extensive area of programming: generalization
to untrained stimulus conditions and to un-
trained responses is programmed by the training
of sufficient exemplars of those stimulus condi-

363

TREVOR F. STOKES and DONALD M. BAER

tions or responses. Train Loosely is a program-
ming technique in which training is conducted
with relatively little control over the stimuli and
responses involved, and generalization is thereby
enhanced. To invoke the tactic of Indiscrimi-
nable Contingencies, the contingencies of re-
inforcement or punishment, or the setting events
marking the presence or absence of those con-
tingencies, are deliberately made less predictable,
so that it becomes difficult to discriminate rein-
forcement occasions from nonreinforcement oc-
casions. Common Stimuli may be employed in
generalization programming by incorporating
into training settings social and physical stimuli
that are salient in generalization settings, and
that can be made to assume functional or obvious
roles in the training setting. Mediated General-
ization requires establishing a response as part
of new learning that is likely to be utilized in
other problems as well, and thus result in gener-
alization. The final technique, Train “To Gener-
alize”, involves reinforcing generalization itself
as if it were an explicit behavior. These program-
ming techniques should be researched further
and usefully applied in programs in which gen-
eralization is relevant.

This list of generalized tactics conceals within
itself a much smaller list of specific tactics. These
specific tactics can be presented as a small pic-
ture of the generalization technology in its pres-
ent most pragmatic form, not only to offer a set
of what-to-do possibilities, but also to emphasize
how very small the current technology is and
how much development it requires:

1. Look for a response that enters a natural
community; in particular, teach subjects to
cue their potential natural communities to
reinforce their desirable behaviors.

2. Keep training more exemplars; in particu-
lar, diversify them.

3. Loosen experimental control over the stim-
uli and responses involved in training; in
particular, train different examples concur-
rently, and vary instructions, SDs, social
reinforcers, and backup reinforcers.

4. Make unclear the limits of training contin-
gencies; in particular, conceal, when pos-
sible, the point at which those contingen-
cies stop operating, possibly by delayed
reinforcement.

5. Use stimuli that are likely to be found in
generalization settings in training settings
as well; in particular, use peers as tutors.

6. Reinforce accurate self-reports of desirable
behavior; apply self-recording and self-
reinforcement techniques whenever possi-
ble.

7. When generalizations occur, reinforce at
least some of them at least sometimes, as
if “to generalize” were an operant response
class.

There are many examples of generalization
and nongeneralization of behavior changes. The
fact that apparently unprogrammed generaliza-
tion has been demonstrated (particularly across
time) is valuable. It heralds a practicality de-
sirable in any technology of behavior: that every
one of a subjects’ responses, in every setting,
with every experimenter, and at every conceiv-
able time does not need to meet specific treat-
ment consequences for that program to accom-
plish and maintain important behavior changes.
Alternatively, the fact that generalization is not
always observed and durability is not inevitable
means that there is hope for behavior modifica-
tion: behavior can always be modified and
changes are not necessarily irreversible. That is,
once behavior has been modified, there is still
the possibility of reconditioning if changes are
undesirable or inappropriate, or if new inappro-
priate behaviors develop. If both appropriate
and inappropriate behavior changes were to per-
sist and prove irreversible, it would presage the
demise of any technology of behavioral inter-
vention. This occurrence of nongeneralization
also underlines the need to develop a technology
of generalization, so that programming will be
a fundamental component of any procedures
when durability and generalization of behavior
changes are desirable.

364

AN IMPLICIT TECHNOLOGY OF GENERALIZATION 365

A most important question is prompted by an
examination of the previous research: does gen-
eralization ever occur without programming?
In the above research, generalization was not al-
ways evident. In fact, the highly discriminated
effects of some operant programs were some-
times documented. We have seen that the behav-
ior analysis literature describes various programs
that have shown that generalization may be pro-
moted or programmed by particular intervention
techniques. It seems reasonable to suggest, then,
that many of the successful Train-and-Hope
examples cited above may be undiagnosed in-
stances of informal or inadvertent programming
techniques, rather than an absence of program-
ming techniques. It cannot be discounted, and is
indeed possible, that these generalization exam-
ples may simply depict successful programmed
generalization, and neither the authors of those
papers, nor the present authors have recognized
or hypothesized the programming technique.

Perhaps the most pragmatic orientation for
behavior analysts is to assume that generalization
does not occur except through some form of pro-
gramming. Thus, the best course of action seems
to be that of systematic measurement and analy-
sis of variables that may have been functional in
any apparently unprogrammed generalization.
These analyses should be included as part of all
research where “unprogrammed” generalized be-
havior changes are evidenced, for discriminated
behavior changes may well be the rule if gen-
eralization is not specifically programmed. Such
analyses, if successful, will contribute to a tech-
nology of generalization by further developing
the understanding of critical variables that func-
tion to produce generalization, and would further
emphasize the need always to be concerned not
only with generalization issues, but with the vari-
ous techniques that accomplish generalization.

In other words, behavioral research and prac-
tice should act as if there were no such animal
as “free” generalization-as if generalization
never occurs “naturally”, but always requires
programming. Then, “programmed generaliza-
tion” is essentially a redundant term, and snould

be descriptive only of the active regard of re-
searchers and practitioners.

REFERENCES
Allen, G. J. Case study: Implementation of behav-

ior modification techniques in summer camp set-
tings. Behavior Therapy, 1973, 4, 570-575.

Ayllon, T. and Azrin, N. H. The token economy.
New York: Appleton-Century-Crofts, 1968.

Azrin, N. H., Sneed, T. J., and Fox, R. M. Dry-bed:
A rapid method of eliminating bedwetting (enu-
resis) of the retarded. Behaviour Research and
Therapy, 1973, 11, 427-434.

Baer, D. M., Peterson, R. F., and Sherman, J. A. The
development of imitation by reinforcing behav-
ioral similarity to a model. journal of the Experi-
mental Analysis of Behavior, 1967, 10, 405-416.

Baer, D. M. and Wolf, M. M. The entry into natu-
ral communities of reinforcement. In R. Ulrich,
T. Stachnik, and J. Mabry (Eds), Control of hu-
man behavior: Volume II. Glenview, Illinois:
Scott, Foresman, 1970. Pp. 319-324.

Baer, D. M., Wolf, M. M., and Risley, T. R. Some
current dimensions of applied behavior analysis.
Journal of Applied Behavior Analysis, 1968, 1,
91-97.

Bolstad, 0. D. and Johnson, S. M. Self-regulation in
the modification of disruptive behavior. Journal
of Applied Behavior Analysis, 1972, 5, 443-454.

Broden, M., Hall, R. V., and Mitts, B. The effect of
self-recording on the classroom behavior of two
eighth-grade students. Journal of Applied Behav-
ior Analysis, 1971, 4, 191-199.

Broden, M., Bruce, C., Mitchell, M. A., Carter, U.,
and Hall, R. V. Effects of teacher attention on
attending behavior of two boys at adjacent desks.
Journal of Applied Behavior Analysis, 1970, 3,
199-203.

Buehler, R. E., Patterson, G. R., and Furniss, J. M.
The reinforcement of behavior in institutional set-
tings. Behaviour Research and Therapy, 1966, 4,
157-167.

Buell, J., Stoddard, P., Harris, F. R., and Baer, D. M.
Collateral social development accompanying rein-
forcement of outdoor play in a preschool child.
Journal of Applied Behavior Analysis, 1968, 1,
167-173.

Cofer, C. N. and Foley, J. P. Mediated generaliza-
tion and the interpretation of verbal behavior: I.
Prolegomena. Psychological Review, 1942, 49,
513-540.

Drabman, R. S., Spitalnik, R., and O’Leary, K. D.
Teaching self control to disruptive children.
Journal of Abnormal Psychology, 1973, 82, 10-
16.

Ferster, C. B. and Skinner, B. F. Schedules of rein-
forcement. New York: Appleton-Century-Crofts,
1957.

366 TREVOR F. STOKES and DONALD M. BAER

Garcia, E. The training and generalization of a
conversational speech form in nonverbal retardates.
Journal of Applied Behavior Analysis, 1974, 7,
137-149.

Garcia, E., Baer, D. M., and Firestone, I. The devel-
opment of generalized imitation within topo-
graphically determined boundaries. Journal of
Applied Behavior Analysis, 1971, 4, 101-112.

Gelfand, D. M., Gelfand, S., and Dobson, W. R.
Unprogrammed reinforcement of patients behav-
ior in a mental hospital. Behaviour Research and
Therapy, 1967, 5, 201-207.

Glynn, E. L., Thomas, J. D., and Shee, S. M. Behav-
ioral self-control of on-task behavior in an ele-
mentary classroom. Journal of Applied Behavior
Analysis, 1973, 6, 105-113.

Goetz, E. M. and Baer, D. M. Social control of form
diversity and the emergence of new forms in chil-
dren’s blockbuilding. Journal of Applied Behav-
ior Analysis, 1973, 6, 105-113.

Graubard, P. S., Rosenberg, H., and Miller, M. B.
Student applications of behavior modification to
teachers and environments or ecological ap-
proaches to social deviancy. In E. A. Ramp and
B. L. Hopkins (Eds), A new direction for educa-
tion: behavior analysis. Lawrence, Kansas: Sup-
port and Development Center for Follow
Through, 1971. Pp. 80-101.

Griffiths, H. and Craighead, W. E. Generalization
in operant speech therapy for misarticulation.
Journal of Speech and Hearing Disorders, 1972,
37, 485-494.

Guess, D., Sailor, W., Rutherford, G., and Baer, D. M.
An experimental analysis of linguistic develop-
ment: the productive use of the plural morpheme.
Journal of Applied Behavior Analysis, 1968, 1,
297-306.

Hall, R. V. and Broden, M. Behavior changes in
brain-injured children through social reinforce-
ment. Journal of Experimental Child Psychology,
1967, 5, 463-479.

Herbert, E. W. and Baer, D. M. Training parents
as behavior modifiers: self-recording of contin-
gent attention. Journal of Applied Behavior Anal-
ysis, 1972, 5, 139-149.

Horner, R. D. Establishing use of crutches by a
mentally retarded spina bifida child. Journal of
Applied Behavior Analysis, 1971, 4, 183-189.

Israel, A. C. and O’Leary, K. D. Developing corre-
spondence between children’s words and deeds.
Child Development, 1973, 44, 575-581.

Johnston, J. M. and Johnston, G. T. Modification of
consonant speech-sound articulation in young
children. Journal of Applied Behavior Analysis,
1972, 5, 233-246.

Kazdin, A. E. The effect of vicarious reinforcement
on attentive behavior in the classroom. Journal of
Applied Behavior Analysis, 1973, 6, 71-78.

Kazdin, A. E. and Polster, R. Intermittent token re-

inforcement and response maintenance in extinc-
tion. Behavior Therapy, 1973, 4, 386-391.

Keller, F. S. and Schoenfeld, W. N. Principles of
psychology. New York: Appleton-Century-Crofts,
1950.

Kifer, R. E., Lewis, M. A., Green, D. R., and Phillips,
E. L. Training predelinquent youths and their
parents to negotiate conflict situations. Journal of
Applied Behavior Analysis, 1974, 7, 357-364.

Koegel, R. L. and Rincover, A. Treatment of psy-
chotic children in a classroom environment: I.
Learning in a large group. Journal of Applied
Behavior Analysis, 1974, 7, 45-59.

Meichenbaum, D. H., Bowers, K. S., and Ross, R. R.
Modification of classroom behavior of institu-
tionalized female adolescent offenders. Behaviour
Research and Therapy, 1968, 6, 343-353.

Parsonson, B. S., Baer, A. M., and Baer, D. M. The
application of generalized correct social contin-
gencies by institutional staff: an evaluation of
the effectiveness and durability of a training pro-
gram. Journal of Applied Behavior Analysis,
1974, 7, 427-437.

Pendergrass, V. E. Timeout from positive reinforce-
ment following persistent high-rate behavior in
retardates. Journal of Applied Behavior Analysis,
1972, 5, 85-91.

Redd, W. H. and Birnbrauer, J. S. Adults as dis-
criminative stimuli for different reinforcement
contingencies with retarded children. Journal of
Experimental Child Psychology, 1969, 7, 440-
447.

Rincover, A. and Koegel, R. L. Setting generality
and stimulus control in autistic children. Journal
of Applied Behavior Analysis, 1975, 8, 235-246.

Risley, T. R. and Hart, B. M. Developing corre-
spondence between the nonverbal and verbal be-
havior of preschool children. Journal of Applied
Behavior Analysis, 1968, 1, 267-281.

Rogers-Warren, A. and Baer, D. M. Correspondence
between saying and doing: teaching children to
share and praise. Journal of Applied Behavior
Analysis, 1976, 9, 335-354.

Schroeder, G. L. and Baer, D. M. Effects of concur-
rent and serial training on generalized vocal imi-
tation in retarded children. Developmental Psy-
chology, 1972, 6, 293-301.

Schumaker, J. and Sherman, J. A. Training genera-
tive verb usage by imitation and reinforcement
procedures. Journal of Applied Behavior Analysis,
1970, 3, 273-287.

Schwarz, M. L. and Hawkins, R. P. Application of
delayed reinforcement procedures to the behavior
of an elementary school child. Journal of Applied
Behavior Analysis, 1970, 3, 85-96.

Seymour, F. W. and Stokes, T. F. Self-recording in
training girls to increase work and evoke staff
praise in an institution for offenders. Journal of
Applied Behavior Analysis, 1976, 9, 41-54.

AN IMPLICIT TECHNOLOGY OF GENERALIZATION 367

Skinner, B. F. Science and human behavior. New
York: Macmillan, 1953.

Solomon, R. W. and Wahler, R. G. Peer reinforce-
ment control of classroom problem behavior.
Journal of Applied Behavior Analysis, 1973, 6,
49-56.

Stokes, T. F. and Baer, D. M. Preschool peers as
mutual generalization-facilitating agents. Behavior
Therapy, 1976, 7, 549-556.

Stokes, T. F., Baer, D. M., and Jackson, R. L. Pro-
gramming the generalization of a greeting re-
sponse in four retarded children. Journal of Ap-
plied Behavior Analysis, 1974, 7, 599-610.

Stolz, S. B. and Wolf, M. M. Visually discriminated

behavior in a “blind” adolescent retardate. Journal
of Applied Behavior Analysis, 1969, 2, 65-77.

Wahler, R. G. Setting generality: some specific and
general effects of child behavior therapy. Journal
of Applied Behavior Analysis, 1969, 2, 239-246.

Walker, H. M. and Buckley, N. K. The use of posi-
tive reinforcement in conditioning attending be-
havior. Journal of Applied Behavior Analysis,
1968, 1, 245-250.

Yates, A. J. Behavior therapy. New York: John
Wiley and Sons, 1970.

Received 22 December 1975.
(Final acceptance 3 June 1976.)

1977, 10, 349-367

AN IMPLICIT TECHNOLOGY OF GENERALIZATION’
TREVOR F. STOKES AND DONALD M. BAER

THE UNIVERSITY OF MANITOBA AND THE UNIVERSITY OF KANSAS

Traditionally, discrimination has been understood as an active process, and a technology
of its procedures has been developed and practiced extensively. Generalization, by con-
trast, has been considered the natural result of failing to practice a discrimination
technology adequately, and thus has remained a passive concept almost devoid of a
technology. But, generalization is equally deserving of an active conceptualization and
technology. This review summarizes the structure of the generalization literature and
its implicit embryonic technology, categorizing studies designed to assess or program
generalization according to nine general headings: Train and Hope; Sequential Modifi-
cation; Introduce to Natural Maintaining Contingencies; Train Sufficient Exemplars;
Train Loosely; Use Indiscriminable Contingencies; Program Common Stimuli; Mediate
Generalization; and Train “To Generalize”.
DESCRIPTORS: generalization, treatment-gain durability, followup measures, main-

tenance, postcheck methodology

Traditionally, many theorists have considered
generalization to be a passive phenomenon. Gen-
eralization was not seen as an operant response
that could be programmed, but as a description
of a “natural” outcome of any behavior-change
process. That is, a teaching operation repeated
over time and trials inevitably involves varying
samples of stimuli, rather than the same set
every time; in the same way, it inevitably evokes
and reinforces varying samples of behavior,
rather than the same set every time. As a conse-
quence, it is predictable that newly taught re-
sponses would be controlled not only by the
stimuli of the teaching program, but by others
somewhat resembling those stimuli (Skinner,
1953, p. 107ff.). Similarly, responses resembling
those established directly, yet not themselves ac-
tually touched by the teaching procedures, would
appear as a result of the teaching (Keller and

‘Preparation of this paper was supported in part by
PHS Training Grant 00183, Program Project Grant
HD 00870, and Research Grant MH 11739. Reprints
may be obtained either from T. F. Stokes, Department
of Psychology, University of Manitoba, Winnipeg,
Manitoba, Canada R3T 2N2, or D. M. Baer, Depart-
ment of Human Development, University of Kansas,
Lawrence, Kansas 66045.

Schoenfeld, 1950, p. 168ff.). Thus, generaliza-
tion was something that happened, not some-
thing produced by procedures specific to it.

If generalization seemed absent or insignifi-
cant, it was simply to be assumed that the teach-
ing process had managed to maintain unusually
tight control of the stimuli and responses in-
volved, allowing little sampling of their varie-
ties. This assumption was strongly supported by
the well-known techniques of discrimination: by
differential reinforcement (in general, by any
differential teaching) of certain stimuli relative
to others, and/or certain responses relative to
others, generalization could be programmatically
restricted and diminished to a very small range.
Thus, it was discrimination that was understood
as an active process, and a technology of its pro-
cedures was developed and practiced extensively.
But generalization was considered the natural
result of failing to practice discrimination’s tech-
nology adequately, and thus remained a passive
concept almost devoid of a technology. Never-
theless, in educational practice, and in the devel-
opment of theories aimed at serving both practice
and a better understanding of human function-
ing, generalization is equally as important as dis-

349

NUMBER 2 (SUMMER) 1977JOURNAL OF APPLIED BEHAVIOR ANALYSIS

TREVOR F. STOKES and DONALD M. BAER

crimination, and equally deserving of an active
conceptualization.

Generalization has been and doubtless will
remain a fundamental concern of applied behav-
ior analysis. A therapeutic behavioral change,
to be effective, often (not always) must occur
over time, persons, and settings, and the effects
of the change sometimes should spread to a
variety of related behaviors. Even though the
literature shows many instances of generaliza-
tion, it is still frequently observed that when a
change in behavior has been accomplished
through experimental contingencies, then that
change is manifest where and when those contin-
gencies operate, and is often seen in only transi-
tory forms in other places and at other times.
The frequent need for generalization of thera-

peutic behavior change is widely accepted, but it
is not always realized that generalization does
not automatically occur simply because a behav-
ior change is accomplished. Thus, the need ac-
tively to program generalization, rather than
passively to expect it as an outcome of certain
training procedures, is a point requiring both
emphasis and effective techniques (Baer, Wolf,
and Risley, 1968). That such exhortations have
often been made has not always ensured that
researchers in the field have taken serious note
of and, therefore, proceeded to analyze ade-
quately the generalization issues of vital concern
to their programs. The emphasis, refinement, and
elaboration of the principles and procedures that
are meant to explain and produce generalization
when it does not occur “naturally” is an impor-
tant area of unfinished business for applied be-
havior analysis.

The notion of generalization developed here
is an essentially pragmatic one; it does not
closely follow the traditional conceptualizations
(Keller and Schoenfeld, 1950; Skinner, 1953).
In many ways, this discussion will sidestep much
of the controversy concerning terminology. Gen-
eralization will be considered to be the occur-
rence of relevant behavior under different, non-
training conditions (i.e., across subjects, settings,
people, behaviors, and/or time) without the

scheduling of the same events in those conditions
as had been scheduled in the training conditions.
Thus, generalization may be claimed when no
extratraining manipulations are needed for extra-
training changes; or may be claimed when some
extra manipulations are necessary, but their cost
or extent is clearly less than that of the direct
intervention. Generalization will not be claimed
when similar events are necessary for similar ef-
fects across conditions.
A technology of generalization programming

is almost a reality, despite the fact that until re-
cently, it had hardly been recognized as a prob-
lem in its own right. Within common teaching
practice, there is an informal germ of a technol-
ogy for generalization. Furthermore, within the
practice of applied behavior analysis (especially
within the past 5 yr or so), there has appeared
a budding area of “generalization-promotion”
techniques. The purpose of this review is to sum-
marize the structure of that generalization litera-
ture and its implicit embryonic technology. Some
270 applied behavior analysis studies relevant to
generalization in that discipline were reviewed.2
A central core of that literature, consisting of
some 120 studies, contributes directly to a tech-
nology of generalization. In general, techniques
designed to assess or to program generalization
can be loosely categorized according to nine
general headings:

1. Train and Hope
2. Sequential Modification
3. Introduce to Natural Maintaining Contin-

gencies
4. Train Sufficient Exemplars

5. Train Loosely

6. Use Indiscriminable Contingencies

7. Program Common Stimuli

2Ninety per cent of the literature reviewed was
from five journals: Behaviour Research and Therapy;
Behavior Therapy; Journal of Applied Behavior Anal-
ysis; Journal of Behavior Therapy and Experimental
Psychiatry; and Journal of Experimental Child Psy-
chology. Seventy-seven per cent of the literature re-
viewed has been published since 1970.

350

AN IMPLICIT TECHNOLOGY OF GENERALIZATION

8. Mediate Generalization

9. Train “To Generalize”.

This review characterizes each category, and
describes some examples of research that illus-
trate the generalization analyses or program-
ming involved in each category. Obviously, all
the relevant references cannot be discussed in this
review.3 The nine categories listed above were
induced from the literature; they are not a priori
categories. Consequently, studies do not always
fit neatly into these categories. It should also be
noted that not all studies reviewed were thorough
experimental analyses of generalization. Often
inferences were necessary to categorize the re-
search. However, the following discussion still
may provide a useful organization and concep-
tualization of generalization and its program-
ming.

1. Train and Hope
In applied behavior analysis research, the most

frequent method of examining generalization, so
far, may be labelled Train and Hope. After a
behavior change is effected through manipula-
tion of some response consequences, any existent
generalization across responses, settings, experi-
menters, and time, is concurrently and/or sub-
sequently documented or noted, but not actively
pursued. It is usually hoped that some generali-
zation may occur, which will be welcomed yet
not explicitly programmed. These hopeful probes
for stimulus and response generalization char-
acterize almost half of the applied literature on
generalization. The studies have considerable
importance, for they begin to document the ex-
tent and limits of generalization of particular
operant intervention techniques. While not being

3Complete reference lists and detailed tables de-
scribing subjects, procedures, and generalization of
all studies reviewed are deposited with the National
Auxiliary Publications Service (NAPS). For copies,
order NAPS Document #02873. Order from ASIS/
NAPS Co., C/O Microfiche Publications, 305 East
46th Street, New York, New York 10017. Remit
with order for each copy $3.00 for microfiche or
$19.50 for photocopies. Make checks payable to
Microfiche Publications.

examples of the programming of generalization,
they are a sound first step in any serious analysis
of generalization. When generalization is desired,
but is shown to be absent or deficient, program-
ming procedures can then be instituted.

For example, useful generalization across set-
tings was documented by Kifer, Lewis, Green,
and Phillips (1974). In an experimental class-
room setting, parent-child pairs were taught to
negotiate in conflict situations. During simulated
role-playing, instructions, practice, and feedback
were used to teach the negotiation behaviors of
fully stating one’s position, identifying the issues
of conflict, and suggesting options to resolve the
conflict. The data showed increased use of nego-
tiation behaviors and the reaching of agreements
in actual parent-child conflict situations at home.
An assessment of generalization across experi-

menters was described by Redd and Birnbrauer
(1969), who demonstrated that control over the
cooperative play of retarded children did not
generalize from an adult who dispensed contin-
gent edible reinforcement to five other adults
who had not participated in training.

Studies that are examples of Train and Hope
across time are those in which there was a change
from the intervention procedures, either to a less
intensive but procedurally different program, or
to no program or no specifically defined pro-
gram. Data or anecdotal observations were re-
ported concerning the maintenance of the origi-
nal behavior change over the specified time
intervening between the termination of the
formal program and the postchecks. An example
of a followup evaluation was the study by Azrin,
Sneed, and Foxx (1973). An intensive training
program involving reinforcement of correct toi-
leting and positive practice procedures promptly
decreased bedwetting by 12 retarded persons.
The reduced rate of accidents was maintained
during a three-month followup assessment.

Perhaps there are many more studies in the
Train and Hope category than would have been
expected (about 135, of which 65 % are across
Time). However, despite its obvious value, this
research is frequently characterized by a lack of

351

TREVOR F. STOKES and DONALD M. BAER

comprehensiveness and depth of the generaliza-
tion analysis. Even though generalized behavior
change was frequently reported, extensive, wide-
ranging, and practical generalization was not
often noted or even sought. The continued de-
velopment of behavior analysis almost surely
will demand more extensive collection of gener-
alization data than is presently the fashion. The
extent and limits of applied behavioral interven-
tions may be well documented and understood if
measurement is extended over longer periods of
time, over more than one circumscribed part of
the day, with more than one related response,
and with more than a restricted part of the social
and physical environment. It is as important for
the field to formalize the conditions of the non-
occurrence of generalization as it is to document
the conditions associated with the display of un-
programmed generalization.

Most of the Train-and-Hope research described
successful generalization-approximately 90%
of Train-and-Hope studies. By definition, there
was no further need to program generalization
in those studies where generalization had been
exhibited within the Train-and-Hope para-
digm-presuming, of course, that the generali-
zation exhibited was considered sufficient to meet
the therapeutic goals of the various modification
programs (not necessarily a valid presumption
in the Train-and-Hope research). This prepon-
derance of positive data may simply reflect the
tendency of some researchers not to report their
generalization data if measurement procedures
were instituted to probe for any generalized be-
havior changes, but generalization was shown to
be absent. Some researchers may view nongen-
eralization as reflecting a deficiency or ineffective-
ness of their procedures to develop a desirable
generalized performance. Behavior analysts,
nevertheless, should encourage their fellow re-
searchers to document and to analyze experimen-
tally their apparent failures, rather than allowing
them to slide into oblivion. A detailed and sys-
tematic understanding of generalization and its
programming could result. Alternatively, re-
searchers might view their generalization base-

lines as being essentially independent of the mod-
ified baseline; thus, to report nongeneralization
would serve no useful purpose, for its nonoccur-
rence would be expected. Again, any such docu-
mentation contributes to our understanding of
the extent and limits of generalization, as well as
serving as an indication of the frequent necessity
of generalization-programming techniques.

There is another reason for the predominance
of positive results in this section: if nongeneral-
ization was clearly evident, and the modification
of this state was important, then a form of lim-
ited programming was frequently instituted. Ex-
amples of this research will be discussed in the
next category, “Sequential Modification”.

2. Sequential Modification
These studies exemplify a more systematic

approach to generalization than the Train-and-
Hope research. Again, a particular behavior
change is effected, and generalization is assessed.
But then, if generalization is absent or deficient,
procedures are initiated to accomplish the desired
changes by systematic sequential modification in
every nongeneralized condition, i.e., across re-
sponses, subjects, settings, or experimenters. The
possibility of unprogrammed generalization typi-
cally was not examined in these sequential modi-
fication studies, because after the initial demon-
stration of nongeneralization, all other baselines
were exhausted. That is, after changes had been
produced directly in all baselines, generalization
to nonrecorded responses, subjects, settings, and
experimenters may have occurred, but could not
be examined.

For example, Meichenbaum, Bowers, and Ross
(1968) reported an absence of generalization of
behavior changes from an afternoon intervention
period to the morning period in a classroom for
institutionalized female adolescent offenders.
Money dispensed contingent on on-task behav-
iors effected desired behavior changes during the
afternoon, but generalization to the morning
period required that the same manipulations be
applied there as well (sequential modification
across settings). Similarly, generalization across

352

AN IMPLICIT TECHNOLOGY OF GENERALIZATION

settings of the disruptive and oppositional behav-
ior of two children was investigated by Wahler
(1969). He demonstrated control of these behav-
iors in the home by using differential attention
and timeout operations. When generalization to
the children’s school behavior was not evidenced,
similar contingency operations were employed to
accomplish changes in that setting as well.
The category of Sequential Modification char-

acterizes much of the actual practice of many
behavior analysts. Sequential modification is
merely a systematized experimental procedure
that formalizes and allows evaluation of these
typical therapeutic endeavors. The tactic of
scheduling behavior-change programs in every
condition to which generalization is desired is
frequently employed. The rationale for these
procedures is as follows. If a desired generaliza-
tion is not likely to be exhibited after changing a
behavior in a particular condition, or a number
of conditions, e.g., settings, then the researcher or
practitioner works to effect changes across con-
ditions as a matter of course, rather than as an
outcome of the display or nondisplay of general-
ization. Thus, a behavior analyst is likely to ad-
vise the scheduling of consequences in every
relevant condition in preference to the dispens-
ing of consequences in only one or a few condi-
tions, while hoping for generalization, but likely
not seeing it.

3. Introduce to Natural Maintaining
Contingencies

Perhaps the most dependable of all general-
ization programming mechanisms is one that
hardly deserves the name: the transfer of behav-
ioral control from the teacher-experimenter to
stable, natural contingencies that can be trusted
to operate in the environment to which the sub-
ject will return, or already occupies. To a con-
siderable extent, this goal is accomplished by
choosing behaviors to teach that normally will
meet maintaining reinforcement after the teach-
ing (Ayllon and Azrin, 1968).

Baer and Wolf (1970) reported a study by
Ingram that illustrated the mechanism of “trap-

ping”, where a preschool child was taught an
entry response that exposed the child to the
natural contingencies of peers in the preschool
environment. Preschool teachers modified the
low rate of skillful interaction of the child by
priming others to interact with the subject and
reinforcing appropriate interactions. The data
showed that over time the teachers lost control
of the interaction behavior, which remained
high; it was assumed that the group’s natural
consequences for interaction had taken control
of the subject’s behavior. Thus, to program gen-
eralization, the child perhaps needed only to be
introduced adequately to the natural reinforcers
inherent in active preschool play and interaction.
Some early analyses of preschool children’s be-
havior have stressed that if the child can be so
introduced (through the operation of differential
attention from teachers) to a reinforcing pre-
school natural environment, then the behaviors
eventually do not need to be maintained by con-
tinued contrived modification of the environ-
ment. For example, Hall and Broden (1967)
modified the manipulative play, climbing, and
social interaction of three subjects through social
reinforcement operations. Behavior changes were
shown to be durable and successful followup
data at three months were described.

Buell, Stoddard, Harris, and Baer (1968) dem-
onstrated the collateral development of appro-
priate social behavior (e.g., touching, verbalizing,
and playing with other children) accompanying
the reinforcement of increased use of outdoor
play equipment by a 3-yr-old girl. This entry re-
sponse to the natural reinforcement community
was tactically sound because the child’s motor
behavior was modified in a setting where the
resulting behavior would tend automatically to
increase social contact with other children, and
this natural social environment could maintain
the child’s new skills, but indeed may also be
expected to sharpen and refine them, and add
entirely new ones as well.
Most of the research concerning natural main-

taining contingencies has involved children, per-
haps because such techniques seem particularly

353

TREVOR F. STOKES and DONALD M. BAER

suitable, especially to their social behavior. Re-
search would profit by determining what natural
reinforcement communities exist for various be-
haviors and subjects, and what economical means
may be employed to ensure entry to these behav-
ioral traps.

Unfortunately, in some instances there may be
no natural reinforcement operating to develop
and maintain skills. For example, in the case of
retarded and institutionalized persons whose de-
pendency has become a stable fact in the lives of
their caretakers, some re-arrangement of the
natural environment may be necessary. A few
studies have introduced subjects to semicontrived
or redesigned “natural” reinforcement communi-
ties. A simple but meaningful example was pro-
vided by Horner (1971), who taught a 5-yr-old
institutionalized retarded boy to walk on crutches
in an experimental setting. The child was then
prompted to generalize the new walking skill
to other settings and activities to which he previ-
ously had been taken in a wheelchair by solici-
tous caretakers, by enlisting those caretakers to
refrain from offering this help. Within 15 days
after treatment was concluded, the child walked
on crutches to all those activities and settings,
eventually extending his ambulation skills to any
part of his world. Stolz and Wolf (1969) trained
a 16-yr-old, “blind” retarded male to discriminate
visual stimuli. Then, the environment was so
structured that assistance was not given in situa-
tions where it had previously been given as a
matter of course. When the boy was required to
use visual cues to help himself in a cafeteria line,
he soon emitted the necessary behaviors. How-
ever, these studies did not establish the function-
ality of their procedures in the maintenance of
behavior changes.

Another significant example was provided by
Seymour and Stokes (1976). In their study, insti-
tutionalized delinquent girls were taught to so-
licit reinforcement (cf. Graubard, Rosenberg,
and Miller, 1971) from their natural community,
the staff of their residential institution. In their
case, the staff had rarely displayed any systematic
attempts at reinforcing desirable behavior shown

by the girls, perhaps on the presumption that the
girls were “bad” and not reinforcible in any case.
However, the experimenters were able to teach
the girls that when their work was objectively
good, and when staff persons were nearby, a
simple skill of calling these adults’ attention to
their good work would result in fairly consistent
reinforcement. Thus, this was a case in which
experimental reinforcement was used to develop
a response in the subjects that would tap and
cultivate the available but dormant natural com-
munity. In theory, this new skill should have
obviated the need for further experimental re-
inforcement, for the praise evoked should have
functioned to maintain both the girls’ work and
cueing, and the cueing, in turn, should have func-
tioned to maintain staff praise. The Seymour and
Stokes’ study could not be continued long enough
to establish whether this would happen, and so
it remains a logically appealing but still unex-
plored method of enhancing generalization:
teaching the subject a means of recruiting a nat-
ural community of reinforcement to maintain
that generalization. Perhaps an even greater ad-
vantage of such procedures is a change in the
locus of control: the subjects can become more
prominent agents of their own behavior change,
rather than being hapless pawns of more-or-less
random environmental contingencies.

Restructuring the environment thus becomes a
target of research aimed at extending the gener-
alization of newly taught skills; even though, at
a technical level, this operation may not be con-
sidered generalization, but rather transfer of
control from one reinforcement contingency to
another. In any event, it is a much neglected
topic of experimental research, although widely
recognized as a desirable, and even essential
characteristic of any rehabilitative effort.

Some natural contingencies are inevitably at
work contributing to the maintenance of inap-
propriate behavior. For example, peer-group
control of inappropriate behavior has often been
suspected and sometimes documented (Buehler,
Patterson, and Furniss, 1966; Gelfand, Gelfand,
and Dobson, 1967; Solomon and Wahler, 1973).

354

AN IMPLICIT TECHNOLOGY OF GENERALIZATION

It would seem reasonable, then, that if the pat-
tern of reinforcement of inappropriate behavior
is modified, the observed outcome may errone-
ously, but happily be attributed to generalization.
For example, Bolstad and Johnson (1972) pre-
sented data that showed that both experimental
and control subjects in the same classroom were
all affected (although not to the same extent) by
experimental manipulation of the reinforcement
contingency for the experimental subjects,
whereas control subjects in a different classroom
were not so affected. The authors presented data
that may account for these differences. The con-
trol subjects in the experimental classroom, who
were also disruptive students, had fewer disrup-
tive interactions with the experimental subjects
during the treatment phases than during base-
line. This possible generalization effect may be
due to the disruption of the natural contingencies
operating in that environment. That is, other
disruptive students previously supported some of
the disruptive behavior of the control subjects,
but during treatment these experimental subjects
did not support the disruptive behavior of their
peers and, thus, a “generalized” decrease in dis-
ruptive behavior by the control subjects resulted.

4. Train Sufficient Exemplars
If the result of teaching one exemplar of a

generalizable lesson is merely the mastery of the
exemplar taught, with no generalization beyond
it, then the obvious route to generalization is to
teach another exemplar of the same generaliza-
tion lesson, and then another, and then another,
and so on until the induction is formed (i.e., until
generalization occurs sufficiently to satisfy the
problem posed). Examples of such programming
techniques will be described in this category of
training sufficient exemplars, perhaps one of the
most valuable areas of programming. Certainly
it is the generalization-programming area most
prominent and extensive in the present literature.

In the research discussed previously under
the categories of Train and Hope and Sequen-
tial Modification, the typical analysis of gener-
alization concerned the measurement of gener-

alization to only a few (and often only one)
extraexperimental responses, subjects, settings,
experimenters, or times. When the absence of
generalization was noted, sometimes it was ac-
complished by further direct intervention in
every nongeneralized condition (i.e., Sequential
Modification). Having completed such modifica-
tions, the possibility of more extensive general-
ized effects (i.e., beyond the two or three modified
baselines) was not examined. In the training of
sufficient exemplars, generalization to untrained
stimulus conditions and to untrained responses
is programmed by the training of sufficient ex-
emplars (rather than all) of these stimulus con-
ditions or responses.
A systematic demonstration of programmed

generalization and measurement of generalized
effects beyond intervention conditions was re-
ported by Stokes, Baer, and Jackson (1974). They
established that training and maintenance of re-
tarded childrens’ greeting responses by one ex-
perimenter was not usually sufficient for the
generalization of the response across experi-
menters. However, high levels of generalization
to over 20 members of the institution staff (and
newcomers as well) who had not participated
in the training of the response were recorded,
after a second experimenter trained and main-
tained the response in conjunction with the first
experimenter. Thus, when generalization did not
prevail after the training of one stimulus exem-
plar, it was programmed by training a greater
diversity of stimulus (trainer) conditions. Simi-
larly, Garcia (1974) taught a conversational
speech form to two retarded children, and, upon
discovering a lack of stable generalization across
experimenters after one training input, pro-
grammed generalization across experimenters by
having a second experimenter teach the same
responses.
A sufficient-stimulus-exemplars demonstration

of programmed generalization across settings has
been described by Allen (1973). Allen modified
the bizarre verbalizations of an 8-yr-old boy by
differential attention procedures. Ignoring bi-
zarre verbalizations and praise for appropriate

355

TREVOR F. STOKES and DONALD M. BAER

interaction reduced bizarre verbalizations during
evening camp activities. However, there was no
generalization to three other camp settings. After
additional training in a second setting, some
generalization to the unmanipulated settings was
noted. This generalization was further enhanced
by intervention in the third setting. Unfortu-
nately, the experimental procedures did not allow
sufficient time to document the full extent of
generalization after training in two settings, but
generalization after training in two settings was
clearly evident. Griffiths and Craighead (1972)
similarly programmed generalization across set-
tings. A 30-yr-old retarded woman received
praise and tokens for correct articulation in
speech therapy. Generalization to a residential
cottage was not observed until the same proce-
dures were instituted there. Following training
in these two stimulus exemplars, generalization
to a third nontraining setting (a classroom) was
observed.

Very little research concerned with generaliza-
tion programming has dealt with the training of
sufficient stimulus exemplars. The infrequent
research that has been published is characterized
largely by programming across experimenters.
This work has been promising, for after a modest
number of training inputs, generalization appar-
ently will occur with persons not involved in
training-unquestionably a valuable and inex-
pensive outcome. However, the present implica-
tion of these studies is limited because of the
restricted nature of the type of subjects and
responses analyzed. Further work is also needed
to give direction to the optimal conditions
whereby the most extensive generalization will
be achieved with a minimal training expendi-
ture. Nevertheless, it is optimistic to note how
frequently a sufficient number of exemplars is
a small number of exemplars. Frequently, it is
no more than two. In particular, there may well
be reason to suspect that the use of two trainers
will yield excellent results in terms of generaliza-
tion. This possibility, obviously an economical
one, certainly merits systematic study of its po-
tential and limits.

Although very little research has been re-
ported, the analysis of generalization program-
ming by training in a number of settings is a
virtually untapped area of far-reaching value.
However, consistent optimism should follow ex-
amination of the studies showing generalization
after training in only a few settings. Unfortu-
nately, behavior analysts seem too often satisfied
with the modification of a single, well-defined
behavior in one setting, e.g., a laboratory pre-
school. Discriminated programs are often accept-
able, and sometimes even desirable. When gener-
alization is a valid concern, but researchers and
practitioners do not act as if this were so, the
discriminated behavior of researchers is most
probably inhibitory to the development of an
effective generalization technology.

Over the past 10 yr, there has developed an
extensive literature discussing the programmed
generalization of responses through the training
of sufficient response exemplars. A response class
has been operationally defined to describe the
fact that some responses are organized such that
operations applied to a subset of responses in the
class affect the other members of that class in the
same manner. For example, Baer, Peterson, and
Sherman (1967) reinforced various motor imita-
tions by retarded children. They found that as
long as reinforcement followed some imitative
responses, other imitations continued to be per-
formed without training or reinforcement.
A topographical analysis of generalized imita-

tion has been made by Garcia, Baer, and Fire-
stone ( 1971). Four retarded children were trained
to imitate three different topographical types of
response: small motor, large motor, and short
vocal. These subjects were also probed for their
imitation of other unreinforced responses: short
motor, long motor, short vocal, and long vocal.
Generalized imitation was observed with each
subject, but this generalization reflected the par-
ticular dimensions of the topographical response
currently being trained or having previously re-
ceived training. Thus, generalization may occur
within well-defined classes and may not gener-
alize to other classes unless some special training

356

AN IMPLICIT TECHNOLOGY OF GENERALIZATION

(generalization programming) occurs within that
class as well. These data depict one possible limi-
tation of the generality of generalized imitation,
as well as pointing to the need to train response
exemplars that will adequately reflect the di-
versity of the generalization being programmed.

Children’s grammatical development has been
another prominent area of research dealing with
generalized behavior. The concept of response
class is again pivotal in these studies, which
conceptualize the rules of morphological gram-
mar as equivalent to response class phenomena.
For example, Guess, Sailor, Rutherford, and
Baer (1968) developed the generative correct
use of plurals by a retarded girl. After teaching
a number of exemplars of the correct plural
response, the girl appropriately labelled new
objects in the singular or plural without further
direct training relevant to those objects. Plural
usage had become a generalized response class;
the morphological rule had been established.
Schumaker and Sherman (1970) rewarded three
retarded children for the correct production of
past- and present-tense forms of verbs. As past-
and present-tense forms of verbs within an in-
flectional class were modified, there occurred a
generalized usage of untrained verbs to similar
tense forms.

There has been considerable research to estab-
lish the importance of the training of sufficient
response exemplars. A survey of these (approxi-
mately 60) studies shows that the number of
exemplars found to be “sufficient” for a desirable
level and durability of generalization varies
widely, probably determined primarily by the
nature of the task and the subject’s prior skills
relevant to it. Most of this research was con-
cerned with the development of motor and vocal
imitations, and the beginning development of
grammar and syntax. The development of ques-
tion-asking and instruction-following is also well
represented.

In conclusion, examination of the sufficient
exemplar research points to a significant (and
long-familiar) generalization-programming pro-
cedure: a number of stimulus and/or response

exemplars should undergo training. That is, to
program the generalized performance of certain
responses across various setting conditions or
persons, training should occur across a (suffi-
cient) number of setting conditions and/or with
various persons. In a similar manner, generaliza-
tion across responses can be programmed reliably
by the training of a number of responses. Diver-
sity of exemplars seems to be the rule to follow
in pursuit of the maximum generalization. Suffi-
cient diversity to reflect the dimensions of the
desired generalization is a useful tactic. However,
diversity may also be our greatest enemy: too
much diversity of exemplars and not enough
(sufficient) exemplars of similar responses may
make potential gains disproportional to the in-
vestment of training effort. The optimal combi-
nation of sufficient exemplars and sufficient di-
versity to yield the most valuable generalization
is critically in need of analysis. Is the best pro-
cedure to train many exemplars with little diver-
sity at the outset, and then expand the diversity
to include dimensions of the desired generaliza-
tion? Or is it a more productive endeavor to
train fewer exemplars that represent a greater
diversity, and persist in the training until gen-
eralization emerges’?

5. Train Loosely

One relatively simple technique can be con-
ceptualized as merely the negation of discrimi-
nation technique. That is, teaching is conducted
with relatively little control over the stimuli
presented and the correct responses allowed, so
as to maximize sampling of relevant dimensions
for transfer to other situations and other forms
of the behavior. A formal example of this most
often informal technique was provided by
Schroeder and Baer (1972), who taught vocal
imitation skills to retarded children in both of
two ways, one emphasizing tight restriction of
the vocal skills being learned at the moment
(serial training of vocal imitations), and the
other allowing much greater range of stimuli
within the current problem (concurrent training

357

TREVOR F. STOKES and DONALD M. BAER

of imitations). The latter method was charac-
terized repeatedly by greater generalization to
as-yet-untaught vocal imitation problems, thus
affirming “loose” teaching techniques as a con-
tributor to wider generalization.

It will be appreciated that the literature of the
field contains very few examples of this type.
Researchers always have attempted to maintain
thorough control and careful restriction and
standardization of their teaching procedures,
primarily to allow easy subsequent interpretation
of the nature of their (successful) teaching tech-
niques. Yet the import of this technique is that
careful management of teaching techniques to a
precisely repetitive handful of stimuli or formats
may, in fact, correspondingly restrict generaliza-
tion of the lessons being learned. The ultimate
force of this recommendation remains to be seen.
What seems required is programmatic research
aimed at assessing the generalization character-
istics of lessons taught under careful, restricted
conditions, relative to similar lessons taught
under looser, more variable conditions.

6. Use Indiscriminable Contingencies
Intermittent schedules of reinforcement have

been shown repeatedly to be particularly resistant
to extinction, relative to continuous schedules
(Ferster and Skinner, 1957). Resistance to extinc-
tion may be regarded as a form of generaliza-
tion-generalization across time subsequent to
learning. The essential feature of intermittent
schedules may be their unpredictability-the
impossibility of discriminating reinforcement oc-
casions from nonreinforcement occasions until
after the fact. Thus, if contingencies of reinforce-
ment or punishment, or the setting events that
mark the presence or absence of those contingen-
cies, are made indiscriminable, then generaliza-
tion may well be observed.

In generalization, behavior occurs in settings in
which it will not be reinforced, just as it does in
settings in which it will be reinforced. Then, the
analogue to an intermittent schedule, extended
to settings, is a condition in which the subject

cannot discriminate in which settings a response
will be reinforced or not reinforced. A potential
approximation to such a condition was presented
in a study by Schwarz and Hawkins (1970).
In that experiment, the behavior of a sixth-grade
child was videotaped during math and spelling
classes. Later, after each school day had ended,
the child was shown the tape of the math class
and awarded reinforcers according to how often
good posture, absence of face-touching, and ap-
propriate voice-loudness were evident on that
tape. Although reinforcers were awarded only
on the basis of behaviors displayed during the
math class, desirable improvements were ob-
served during the spelling class as well. In that
reinforcement was delayed, this technique must
have made it difficult for the child to discriminate
in which class the behaviors were critical for
earning reinforcement. In other words, the gen-
eralized success of the study may well be at-
tributable to the partly indiscriminable nature
of the reinforcement contingency.

In general, it may be suspected that delayed
reinforcement often will have the advantage of
making the times and places in which the con-
tingency actually operates indiscriminable to the
subject. However, this advantage is an advantage,
by hypothesis, primarily for the goal of general-
ization. Otherwise, delayed reinforcement would
often be considered an inefficient technique, most
especially so for the initial development of a new
skill. Indeed, it may be exactly in the realm of
disadvantaged persons such as retarded children
that the usual inefficiency of delayed reinforce-
ment may seem the most severe handicap to its
use. However, its potential for fostering general-
ization suggests strongly that further research be
invested in this procedure (and any others that
make reinforcement contingencies properly in-
discriminable), to develop methods of applying
it perhaps only after the initial development of a
new skill, in the interests of promoting gen-
eralization.

Less than a dozen studies of generalization
interpretable as cases of indiscriminable rein-
forcement contingencies can be found in the

358

AN IMPLICIT TECHNOLOGY OF GENERALIZATION

literature. Kazdin (1973), for example, showed
that teacher attention to one retarded child was
responded to by another child as if it were rein-
forcement for on-task behavior. Indeed, the
onlooker reacted with increased on-task behavior,
even when the teacher attended to the target
child’s off-task behavior. Possibly, prior experi-
ence with reinforcement contingent on the peers’
on-task behavior was sufficient to make all future
praise (contingent or not) discriminative for on-
task behavior. In other words, with sufficient
prior experience, the onlooker may have stopped
observing the contingency in which the rein-
forcement operated and responded only to the
reinforcing stimulus’ presence, making the con-
tingency functionally indiscriminable.

Generalization across subjects has similarly
been reported by Broden, Bruce, Mitchell, Carter,
and Hall (1970) in a classroom of culturally
disadvantaged children. When positive teacher
attention was given for one child’s attention to
academic work, the attending of a peer also in-
creased. This generalization was also a probable
function of the cueing properties of teacher rein-
forcement. However, the generalization observed
may also have been due to the manipulation of
natural social consequences received by the non-
target child through peer attention, or may have
been caused by a slight increase in the amount
of teacher attention to the nontarget child. These
effects deserve further systematic evaluation be-
cause of their relevance to the classroom prac-
tices of many teachers who strive to instruct
effectively but are unable to devote extensive
time to individual children.

Pendergrass (1972) showed that timeout could
be employed to decrease the destructive behavior
of two retarded children. With one subject, de-
creased rates were also observed with another
response (self-biting) which was sometimes
chained to the destructive behavior, but not
itself subjected to contingent timeout. However,
with the second subject, generalization to a sec-
ond response (autistic jerking movement) was
not observed. Analysis of the data revealed that
the two behaviors occurred simultaneously more

frequently with the subject with whom general-
ization was evidenced. Thus, with this subject,
punishment of the generalization response oc-
curred more frequently when destructive behav-
ior was punished. Unfortunately, it was not
determined how often the self-biting occurred
at times not simultaneous with the destructive
behavior. Therefore, the schedule of punishment
for self-biting was not established, i.e., whether
biting occurred only when destructive behavior
occurred and, therefore, always met the timeout
contingency. In this example (which was not
intended to be a careful analysis of the indis-
criminable reinforcement concept), not only was
the reinforcement contingency somewhat diffi-
cult to discriminate, but the two behaviors (de-
structive and self-destructive responses) also may
well have been only somewhat differentiated by
the subject.

Thus, preventing the ready discrimination of
contingencies is a generalization-programming
technique worthy of application and research.
Perhaps a random or haphazard delivery of re-
inforcement will (if luck or good judgement
prevails) function to modify targetted behavior
as well as behavior occurring in proximal time
or space. Even noncontingent reinforcement,
delivered at the outset of an intervention pro-
gram, may retard initial effects, but may work
to later advantage in generalization outcomes.

Finally, Kazdin and Polster (1973) showed
once again the usefulness of intermittent sched-
ules to delay subsequent extinction, relative to
continuous schedules of reinforcement. Social
interaction by two retardates was reinforced with
tokens. After establishing social interaction, one
subject received continuous reinforcement and
the other, intermittent reinforcement. During
extinction, only the subject who received inter-
mittent reinforcement continued to interact so-
cially with peers. However, these results may
simply reflect different extinction rates by two
subjects. The research was essentially a group
study where N 1. Adequate single-subject ex-
perimental control was lacking. Therefore, repli-
cation of these procedures would be desirable.

359

TREVOR F. STOKES and DONALD M. BAER
7. Program Common Stimuli

The passive approach to generalization de-
scribed earlier need not be a completely imprac-
tical one. If it is supposed that generalization
will occur, if only there are sufficient stimulus
components occurring in common in both the
training and generalization settings, then a rea-
sonably practical technique is to guarantee that
common and salient stimuli will be present in
both. One predictor of the salience of a stimulus
to be chosen for this role is its already established
function for other important behaviors of the
subject.

Children’s peers may represent peculiarly
suitable candidates for a stimulus common to
both training and generalization settings. An
example has been provided by Stokes and Baer
(1976). In their study, two children exhibiting
serious learning disabilities were recruited to
learn several word-recognition skills. One child
was taught these skills and concurrently shown
how to teach them to the other child, thus acting
as a peer-tutor. It was found that both children
reliably learned the skills, but that neither gen-
eralized them reliably or stably to somewhat dif-
ferent settings in which the other child usually
was absent. However, when the peer-tutor was
brought into those settings, then each child simi-
larly showed greatly increased and stabilized
generalization, even though there were never
any consequences for generalization. Similar
demonstrations have been provided by Johnston
and Johnston (1972) for the skill of speech
articulation. In that study, peers were rewarded
for correct monitoring of the subjects’ articula-
tion. Generalization of correct articulation oc-
curred only when the “monitoring” peer was
present. Unfortunately, it was not determined
clearly whether generalization was evidenced be-
cause of the discriminative properties of the
peers’ presence in both settings, or whether the
peers actively continued their monitoring in the
generalization setting.

Rincover and Koegel (1975) have also incor-
porated functional training stimuli into the gen-

eralization setting. Autistic children were re-
warded for imitation and instruction-following
in a training setting. Four of their 10 subjects
then did not exhibit generalization to a different
setting. Therefore, to program for this general-
ization, various aspects of the training procedures
(e.g., hand movement by therapist) or physical
training environment (e.g., table and chairs)
were systematically introduced to the generaliza-
tion setting to control generalization. Making
the experimental setting more closely resemble
the regular classroom (generalization setting)
was the programming procedure employed by
Koegel and Rincover (1974). They decreased
the teacher-to-student ratio in the experimental
setting from 1-to-i to 1-to-8. After these special
programming conditions were instituted, there
was increased performance on previously learned
and new behaviors learned in the classroom.
Walker and Buckley (1972) programmed gener-
alization of the effects of remedial training of
social and academic classroom behavior by estab-
lishing common stimuli between the experimen-
tal remedial classroom and the childrens’ regular
classroom by using the same academic materials
in both classrooms.

The literature of this field shows only a hand-
ful of studies deliberately making use of a com-
mon stimulus in both training and generalization
settings. Obviously, this is a technological dimen-
sion urgently in need of thorough development.
The use of peers as the common stimulus has
much to recommend it as a practical and natural
technique. To what extent peers need to partici-
pate in the training setting has not yet been
determined, although the absence of generaliza-
tion sometimes shown when peers are present
in nontraining settings, suggests that peers not
involved in a training setting will not likely
acquire sufficient discriminative function to con-
trol generalized responding. The use of common
physical stimuli is in even greater need of sys-
tematic research. A common stimulus approach
to generalization would encourage the incorpora-
tion into training settings of (naturally occur-
ring) physical stimuli that are frequently promi-

360

AN IMPLICIT TECHNOLOGY OF GENERALIZATION

nent or functional in nontraining environments.
If these stimuli are well chosen, and can be made
functional and salient in the training procedures,
then generalization may thereby be programmed.

8. Mediate Generalization

Mediated generalization is well known as a
theoretical mechanism explaining generalization
of highly symbolic learnings (Cofer and Foley,
1942). In essence, it requires establishing a re-
sponse as part of the new learning that is likely
to be utilized in other problems as well, and will
constitute sufficient commonality between the
original learning and the new problem to result
in generalization. The most commonly used
mediator is language, apparently. However, the
deliberate application of language to accomplish
generalization is rare in the literature reviewed,
and correspondingly little is known about what
aspects of a language response make for best
mediation.
A sophisticated analysis of mediated general-

ization was conducted by Risley and Hart ( 1968),
who taught preschool children to report at the
end of play on their play-material choices. Men-
tion of a given choice was reinforced with snacks,
which produced increased mentioning of that
choice, but no change in the children’s actual
use of that play-material. When reinforcement
was restricted to true reports of play-material
choices, however, the children then changed their
play behavior (the next day) so that when
queried about that play, they could truthfully
report on their use of the specified play material
and earn reinforcement. Control over any choice
of play materials proved possible with this tech-
nique, which placed teaching contingencies not
on the play, but on a potential mediator (verbal
report) of that play behavior. That the reports
were only potential mediators was apparent in
the early stages of the study, when the children
readily reported (untruly) their use of play ma-
terials with no corresponding actual behavior
with those materials; at that stage, they earned
reinforcement even so. When the reinforcement

was restricted to true reports, the reports then
became mediators of play behavior. The lesson
generalized, such that after several sequential
experiences with these procedures, the children
then used reports about play as mediators, even
without reinforcement being restricted to only
true reports. Israel and O’Leary (1973) used
essentially the same paradigm to compare the
effects of having children report first what they
would play with later, in contrast to having them
report after play what they had done (the Risley
and Hart method); they found that reinforcing
postreports (when they were true) produced more
actual behavior (the next day) than reinforcing
the actual behavior when it agreed with the
earlier promise to perform it. This technique has
been extended subsequently to the case of social
skills, specifically sharing and praising between
young children (Rogers-Warren and Baer, 1976).
In that case, modelling was added, such that the
young children would have a thorough chance
to learn the nature of the relatively complex
responses at issue.

Obviously, verbal mediation can easily fail,
most especially in those situations in which the
verbal mediators have little meaning (i.e., tightly
restricted discriminative value) for the subjects.
It is commonplace to find children agreeing to a
query (e.g., about whether they praised or shared)
without any knowledge of what that must entail
in actual behavior. In the case of retarded chil-
dren, it might be particularly true that the ability
to use verbal responses as mediators would lag
behind that of normal children using the same
language responses. It may be reasonable to
suggest that in the development of language-
training programs, systematic attention be given
to the training of language skills sufficiently well
elaborated to function as mediators of nonverbal
behavior. Language is a response, of course; it is
also, equally obviously, a stimulus to the speaker
as well as to the listener. Thus, it meets perfectly
the logic of a salient common stimulus, to be
carried from any training setting to any general-
ization setting that the child may ever enter.
It also perfectly exemplifies the essence of the

361

TREVOR F. STOKES and DONALD M. BAER

active generalization approach recommended
earlier.
The mediation of generalization is also exem-

plified in the behavior analysis research of self-
control and self-management procedures. That
is, self-control procedures such as self-recording,
taught as part of an intervention program, may
function to promote generalization: such tech-
niques are easy to transport and may be em-
ployed readily to facilitate responding under
generalization conditions. Some research that has
employed any or all of the various tactics of self-
assessment, self-recording, self-determination of
reinforcement, and/or self-administration of re-
inforcement (Glynn, Thomas, and Shee, 1973),
has also displayed maintenance and generaliza-
tion of behavior change; however, the correla-
tion is not perfect.

Broden, Hall, and Mitts (1971) reported that
after an eighth-grade girl experienced self-
recording of study behavior and teacher praise
for improved study, her study behavior main-
tained at a high level for a recorded three weeks.
Although the individual effects of the self-record-
ing and praise were not determined, it is possible
that the self-recording procedures contributed
significantly to this generalization.

Drabman, Spitalnik, and O’Leary (1973)
taught disruptive children to match their teach-
er’s evaluations of their appropriate classroom
behavior. Tokens were dispensed for appropriate
classroom behavior and accurate matching. Dis-
ruptive classroom behavior decreased and was
maintained at low levels during a 12-day phase
when tokens were not dispensed for self-record-
ing accuracy. Generalized behavior improvement
was also evident during a 15-min no-token
period within the experimental hour. These
changes were possibly a function of the close
temporal proximity of the token periods, which
frequently immediately preceded or followed
the generalization period.

The role of self-control procedures in medi-
ating generalization has often been proposed.
Research would do well to examine the contri-
bution of self-control tactics in generalization

and maintenance, especially when formal inter-
vention manipulations have ceased to operate.
The effects of accompanying procedures should
be experimentally separated from self-control
effects, and the role of each of the various self-
control tactics (Glynn et al., 1973) should be
individually analyzed. The potential of self-
mediated generalization is apparent, but its im-
plications and practical utility still remain to be
assessed.

9. Train “To Generalize”
If generalization is considered as a response

itself, then a reinforcement contingency may be
placed on it, the same as with any other operant.
Informally, teachers often do this when they urge
a student who has been taught one example of
a general principle to “see” another example
as “the same thing”. (In principle, they are also
attempting to make use of language as a medi-
ator of generalization, relying on the supposed
characteristics of words like “same” to accom-
plish the generalization.) Common observation
suggests that the method often fails, and that
when it does succeed, little extrinsic reinforce-
ment is offered as a consequence. A more formal
example of the technique was seen in a study
by Goetz and Baer (1973), in which three pre-
school children were taught to generalize the
response of making block forms (in blockbuild-
ing play). Descriptive social reinforcement was
offered only for every different form the child
made, i.e., contingent on every first appearance
of any blockbuilding form within a session, but
not for any subsequent appearances of that form.
Thus, the child was rewarded for moving along
the generalization gradient underlying block-
form inventions, and never for staying at any
one point. In general, the technique succeeded, in
that the children steadily invented new block
forms while this contingency was in use. Thus,
there exists the possibility of programming rein-
forcement specifically, perhaps only, for move-
ment along the generalization gradient desired.

In largely unspecified ways, perhaps two other
studies exemplify this logic. Herbert and Baer

362

AN IMPLICIT TECHNOLOGY OF GENERALIZATION

(1972), for example, taught two mothers of
deviant children to give social reinforcement
only to their children’s appropriate behaviors,
but taught the mothers from the outset to judge
all behavior according to criteria they helped
to develop, rather than attack only a few speci-
fied child responses. These mothers learned a
generalized skill because they applied correct
social contingencies to categories that included
virtually all appropriate child behavior likely to
occur. Behavior changes were maintained at 20
and 24 weeks after completion of formal train-
ing. Similarly, Parsonson, Baer, and Baer (1974)
taught two teachers of retarded children to apply
generalized correct social contingencies to all
likely appropriate and inappropriate behaviors
of preschool retarded children. These effects were
also durable over several months. Apparently
generalized changes were produced in these stud-
ies by Herbert and Baer and Parsonson et al.,
but the extent and quality of that generalization
was not quantified as such.

Very few studies of this type are found in the
literature of applied behavior analysis, probably
because of the preference of behaviorists to con-
sider generalization as an outcome of behavioral
change, rather than as a behavior itself. Ulti-
mately, this behavioristic stance may well prove
durable and consistent. Meanwhile, it is worth
hypothesizing that “to generalize” may be treated
as if it were an operant response, and reinforced
as such, simply to see what useful results occur.

Consequently, one other technique deserves
discussion: the systematic use of instructions to
facilitate generalization. Thus, if a behavior is
taught and generalization is not displayed, the
least expensive of all techniques is to tell the
subject about the possibility of generalization
and then ask for it. If that generalization then
occurs, it may well be referred to as “instructed
generalization”. If the effects of that instruction
are themselves to become generalized (yielding
a “generalized generalizer”?), then reinforcement
of the generalized behavior, on a suitable sched-
ule, might well be prudent, at least at first. Per-
haps it is simply a very elaborate version of this

technique that is being practiced when a client
is taught to relax in a somewhat anxiety-arousing
situation, and reinforced (socially) for doing so;
and then is instructed to relax in a somewhat
more powerful anxiety-arousing situation, etc.
That is, systematic desensitization to a heirarchy
of stimuli may be analyzed as reinforcing not just
relaxation, but also generalization along an al-
ready constructed generalization gradient (cf.
Yates, 1970, p. 64ff.).

CONCLUSION

The structure of the generalization literature
and its implicit embryonic technology has been
summarized. The most frequent treatments of
generalization are also the least analytical-those
described as Train and Hope and Sequential
Modification. Included in the category of Train
and Hope were those studies where the potential
for generalization had been recognized, its pres-
ence or absence noted, but no particular effort
was expended to accomplish generalization. By
contrast, some limited programming was imple-
mented in the Sequential Modification research.
In these studies, given an absence of reliable
generalization, procedures to effect changes were
instituted directly in every nongeneralized condi-
tion. Although contributing significantly to our
understanding of the generalization of behavior-
change programs, these studies are not examples
of the programming of generalization.

Seven categories were discussed that directly
relate to a technology of generalization. First,
the potential role of Natural Maintaining Con-
tingencies was discussed. According to this tactic,
generalization may be programmed by suitable
trapping manipulations, where responses are in-
troduced to natural reinforcement communities
that refine and maintain those skills without
further therapeutic intervention. The Training
of Sufficient Exemplars is numerically the most
extensive area of programming: generalization
to untrained stimulus conditions and to un-
trained responses is programmed by the training
of sufficient exemplars of those stimulus condi-

363

TREVOR F. STOKES and DONALD M. BAER

tions or responses. Train Loosely is a program-
ming technique in which training is conducted
with relatively little control over the stimuli and
responses involved, and generalization is thereby
enhanced. To invoke the tactic of Indiscrimi-
nable Contingencies, the contingencies of re-
inforcement or punishment, or the setting events
marking the presence or absence of those con-
tingencies, are deliberately made less predictable,
so that it becomes difficult to discriminate rein-
forcement occasions from nonreinforcement oc-
casions. Common Stimuli may be employed in
generalization programming by incorporating
into training settings social and physical stimuli
that are salient in generalization settings, and
that can be made to assume functional or obvious
roles in the training setting. Mediated General-
ization requires establishing a response as part
of new learning that is likely to be utilized in
other problems as well, and thus result in gener-
alization. The final technique, Train “To Gener-
alize”, involves reinforcing generalization itself
as if it were an explicit behavior. These program-
ming techniques should be researched further
and usefully applied in programs in which gen-
eralization is relevant.

This list of generalized tactics conceals within
itself a much smaller list of specific tactics. These
specific tactics can be presented as a small pic-
ture of the generalization technology in its pres-
ent most pragmatic form, not only to offer a set
of what-to-do possibilities, but also to emphasize
how very small the current technology is and
how much development it requires:

1. Look for a response that enters a natural
community; in particular, teach subjects to
cue their potential natural communities to
reinforce their desirable behaviors.

2. Keep training more exemplars; in particu-
lar, diversify them.

3. Loosen experimental control over the stim-
uli and responses involved in training; in
particular, train different examples concur-
rently, and vary instructions, SDs, social
reinforcers, and backup reinforcers.

4. Make unclear the limits of training contin-
gencies; in particular, conceal, when pos-
sible, the point at which those contingen-
cies stop operating, possibly by delayed
reinforcement.

5. Use stimuli that are likely to be found in
generalization settings in training settings
as well; in particular, use peers as tutors.

6. Reinforce accurate self-reports of desirable
behavior; apply self-recording and self-
reinforcement techniques whenever possi-
ble.

7. When generalizations occur, reinforce at
least some of them at least sometimes, as
if “to generalize” were an operant response
class.

There are many examples of generalization
and nongeneralization of behavior changes. The
fact that apparently unprogrammed generaliza-
tion has been demonstrated (particularly across
time) is valuable. It heralds a practicality de-
sirable in any technology of behavior: that every
one of a subjects’ responses, in every setting,
with every experimenter, and at every conceiv-
able time does not need to meet specific treat-
ment consequences for that program to accom-
plish and maintain important behavior changes.
Alternatively, the fact that generalization is not
always observed and durability is not inevitable
means that there is hope for behavior modifica-
tion: behavior can always be modified and
changes are not necessarily irreversible. That is,
once behavior has been modified, there is still
the possibility of reconditioning if changes are
undesirable or inappropriate, or if new inappro-
priate behaviors develop. If both appropriate
and inappropriate behavior changes were to per-
sist and prove irreversible, it would presage the
demise of any technology of behavioral inter-
vention. This occurrence of nongeneralization
also underlines the need to develop a technology
of generalization, so that programming will be
a fundamental component of any procedures
when durability and generalization of behavior
changes are desirable.

364

AN IMPLICIT TECHNOLOGY OF GENERALIZATION 365

A most important question is prompted by an
examination of the previous research: does gen-
eralization ever occur without programming?
In the above research, generalization was not al-
ways evident. In fact, the highly discriminated
effects of some operant programs were some-
times documented. We have seen that the behav-
ior analysis literature describes various programs
that have shown that generalization may be pro-
moted or programmed by particular intervention
techniques. It seems reasonable to suggest, then,
that many of the successful Train-and-Hope
examples cited above may be undiagnosed in-
stances of informal or inadvertent programming
techniques, rather than an absence of program-
ming techniques. It cannot be discounted, and is
indeed possible, that these generalization exam-
ples may simply depict successful programmed
generalization, and neither the authors of those
papers, nor the present authors have recognized
or hypothesized the programming technique.

Perhaps the most pragmatic orientation for
behavior analysts is to assume that generalization
does not occur except through some form of pro-
gramming. Thus, the best course of action seems
to be that of systematic measurement and analy-
sis of variables that may have been functional in
any apparently unprogrammed generalization.
These analyses should be included as part of all
research where “unprogrammed” generalized be-
havior changes are evidenced, for discriminated
behavior changes may well be the rule if gen-
eralization is not specifically programmed. Such
analyses, if successful, will contribute to a tech-
nology of generalization by further developing
the understanding of critical variables that func-
tion to produce generalization, and would further
emphasize the need always to be concerned not
only with generalization issues, but with the vari-
ous techniques that accomplish generalization.

In other words, behavioral research and prac-
tice should act as if there were no such animal
as “free” generalization-as if generalization
never occurs “naturally”, but always requires
programming. Then, “programmed generaliza-
tion” is essentially a redundant term, and snould

be descriptive only of the active regard of re-
searchers and practitioners.

REFERENCES
Allen, G. J. Case study: Implementation of behav-

ior modification techniques in summer camp set-
tings. Behavior Therapy, 1973, 4, 570-575.

Ayllon, T. and Azrin, N. H. The token economy.
New York: Appleton-Century-Crofts, 1968.

Azrin, N. H., Sneed, T. J., and Fox, R. M. Dry-bed:
A rapid method of eliminating bedwetting (enu-
resis) of the retarded. Behaviour Research and
Therapy, 1973, 11, 427-434.

Baer, D. M., Peterson, R. F., and Sherman, J. A. The
development of imitation by reinforcing behav-
ioral similarity to a model. journal of the Experi-
mental Analysis of Behavior, 1967, 10, 405-416.

Baer, D. M. and Wolf, M. M. The entry into natu-
ral communities of reinforcement. In R. Ulrich,
T. Stachnik, and J. Mabry (Eds), Control of hu-
man behavior: Volume II. Glenview, Illinois:
Scott, Foresman, 1970. Pp. 319-324.

Baer, D. M., Wolf, M. M., and Risley, T. R. Some
current dimensions of applied behavior analysis.
Journal of Applied Behavior Analysis, 1968, 1,
91-97.

Bolstad, 0. D. and Johnson, S. M. Self-regulation in
the modification of disruptive behavior. Journal
of Applied Behavior Analysis, 1972, 5, 443-454.

Broden, M., Hall, R. V., and Mitts, B. The effect of
self-recording on the classroom behavior of two
eighth-grade students. Journal of Applied Behav-
ior Analysis, 1971, 4, 191-199.

Broden, M., Bruce, C., Mitchell, M. A., Carter, U.,
and Hall, R. V. Effects of teacher attention on
attending behavior of two boys at adjacent desks.
Journal of Applied Behavior Analysis, 1970, 3,
199-203.

Buehler, R. E., Patterson, G. R., and Furniss, J. M.
The reinforcement of behavior in institutional set-
tings. Behaviour Research and Therapy, 1966, 4,
157-167.

Buell, J., Stoddard, P., Harris, F. R., and Baer, D. M.
Collateral social development accompanying rein-
forcement of outdoor play in a preschool child.
Journal of Applied Behavior Analysis, 1968, 1,
167-173.

Cofer, C. N. and Foley, J. P. Mediated generaliza-
tion and the interpretation of verbal behavior: I.
Prolegomena. Psychological Review, 1942, 49,
513-540.

Drabman, R. S., Spitalnik, R., and O’Leary, K. D.
Teaching self control to disruptive children.
Journal of Abnormal Psychology, 1973, 82, 10-
16.

Ferster, C. B. and Skinner, B. F. Schedules of rein-
forcement. New York: Appleton-Century-Crofts,
1957.

366 TREVOR F. STOKES and DONALD M. BAER

Garcia, E. The training and generalization of a
conversational speech form in nonverbal retardates.
Journal of Applied Behavior Analysis, 1974, 7,
137-149.

Garcia, E., Baer, D. M., and Firestone, I. The devel-
opment of generalized imitation within topo-
graphically determined boundaries. Journal of
Applied Behavior Analysis, 1971, 4, 101-112.

Gelfand, D. M., Gelfand, S., and Dobson, W. R.
Unprogrammed reinforcement of patients behav-
ior in a mental hospital. Behaviour Research and
Therapy, 1967, 5, 201-207.

Glynn, E. L., Thomas, J. D., and Shee, S. M. Behav-
ioral self-control of on-task behavior in an ele-
mentary classroom. Journal of Applied Behavior
Analysis, 1973, 6, 105-113.

Goetz, E. M. and Baer, D. M. Social control of form
diversity and the emergence of new forms in chil-
dren’s blockbuilding. Journal of Applied Behav-
ior Analysis, 1973, 6, 105-113.

Graubard, P. S., Rosenberg, H., and Miller, M. B.
Student applications of behavior modification to
teachers and environments or ecological ap-
proaches to social deviancy. In E. A. Ramp and
B. L. Hopkins (Eds), A new direction for educa-
tion: behavior analysis. Lawrence, Kansas: Sup-
port and Development Center for Follow
Through, 1971. Pp. 80-101.

Griffiths, H. and Craighead, W. E. Generalization
in operant speech therapy for misarticulation.
Journal of Speech and Hearing Disorders, 1972,
37, 485-494.

Guess, D., Sailor, W., Rutherford, G., and Baer, D. M.
An experimental analysis of linguistic develop-
ment: the productive use of the plural morpheme.
Journal of Applied Behavior Analysis, 1968, 1,
297-306.

Hall, R. V. and Broden, M. Behavior changes in
brain-injured children through social reinforce-
ment. Journal of Experimental Child Psychology,
1967, 5, 463-479.

Herbert, E. W. and Baer, D. M. Training parents
as behavior modifiers: self-recording of contin-
gent attention. Journal of Applied Behavior Anal-
ysis, 1972, 5, 139-149.

Horner, R. D. Establishing use of crutches by a
mentally retarded spina bifida child. Journal of
Applied Behavior Analysis, 1971, 4, 183-189.

Israel, A. C. and O’Leary, K. D. Developing corre-
spondence between children’s words and deeds.
Child Development, 1973, 44, 575-581.

Johnston, J. M. and Johnston, G. T. Modification of
consonant speech-sound articulation in young
children. Journal of Applied Behavior Analysis,
1972, 5, 233-246.

Kazdin, A. E. The effect of vicarious reinforcement
on attentive behavior in the classroom. Journal of
Applied Behavior Analysis, 1973, 6, 71-78.

Kazdin, A. E. and Polster, R. Intermittent token re-

inforcement and response maintenance in extinc-
tion. Behavior Therapy, 1973, 4, 386-391.

Keller, F. S. and Schoenfeld, W. N. Principles of
psychology. New York: Appleton-Century-Crofts,
1950.

Kifer, R. E., Lewis, M. A., Green, D. R., and Phillips,
E. L. Training predelinquent youths and their
parents to negotiate conflict situations. Journal of
Applied Behavior Analysis, 1974, 7, 357-364.

Koegel, R. L. and Rincover, A. Treatment of psy-
chotic children in a classroom environment: I.
Learning in a large group. Journal of Applied
Behavior Analysis, 1974, 7, 45-59.

Meichenbaum, D. H., Bowers, K. S., and Ross, R. R.
Modification of classroom behavior of institu-
tionalized female adolescent offenders. Behaviour
Research and Therapy, 1968, 6, 343-353.

Parsonson, B. S., Baer, A. M., and Baer, D. M. The
application of generalized correct social contin-
gencies by institutional staff: an evaluation of
the effectiveness and durability of a training pro-
gram. Journal of Applied Behavior Analysis,
1974, 7, 427-437.

Pendergrass, V. E. Timeout from positive reinforce-
ment following persistent high-rate behavior in
retardates. Journal of Applied Behavior Analysis,
1972, 5, 85-91.

Redd, W. H. and Birnbrauer, J. S. Adults as dis-
criminative stimuli for different reinforcement
contingencies with retarded children. Journal of
Experimental Child Psychology, 1969, 7, 440-
447.

Rincover, A. and Koegel, R. L. Setting generality
and stimulus control in autistic children. Journal
of Applied Behavior Analysis, 1975, 8, 235-246.

Risley, T. R. and Hart, B. M. Developing corre-
spondence between the nonverbal and verbal be-
havior of preschool children. Journal of Applied
Behavior Analysis, 1968, 1, 267-281.

Rogers-Warren, A. and Baer, D. M. Correspondence
between saying and doing: teaching children to
share and praise. Journal of Applied Behavior
Analysis, 1976, 9, 335-354.

Schroeder, G. L. and Baer, D. M. Effects of concur-
rent and serial training on generalized vocal imi-
tation in retarded children. Developmental Psy-
chology, 1972, 6, 293-301.

Schumaker, J. and Sherman, J. A. Training genera-
tive verb usage by imitation and reinforcement
procedures. Journal of Applied Behavior Analysis,
1970, 3, 273-287.

Schwarz, M. L. and Hawkins, R. P. Application of
delayed reinforcement procedures to the behavior
of an elementary school child. Journal of Applied
Behavior Analysis, 1970, 3, 85-96.

Seymour, F. W. and Stokes, T. F. Self-recording in
training girls to increase work and evoke staff
praise in an institution for offenders. Journal of
Applied Behavior Analysis, 1976, 9, 41-54.

AN IMPLICIT TECHNOLOGY OF GENERALIZATION 367

Skinner, B. F. Science and human behavior. New
York: Macmillan, 1953.

Solomon, R. W. and Wahler, R. G. Peer reinforce-
ment control of classroom problem behavior.
Journal of Applied Behavior Analysis, 1973, 6,
49-56.

Stokes, T. F. and Baer, D. M. Preschool peers as
mutual generalization-facilitating agents. Behavior
Therapy, 1976, 7, 549-556.

Stokes, T. F., Baer, D. M., and Jackson, R. L. Pro-
gramming the generalization of a greeting re-
sponse in four retarded children. Journal of Ap-
plied Behavior Analysis, 1974, 7, 599-610.

Stolz, S. B. and Wolf, M. M. Visually discriminated

behavior in a “blind” adolescent retardate. Journal
of Applied Behavior Analysis, 1969, 2, 65-77.

Wahler, R. G. Setting generality: some specific and
general effects of child behavior therapy. Journal
of Applied Behavior Analysis, 1969, 2, 239-246.

Walker, H. M. and Buckley, N. K. The use of posi-
tive reinforcement in conditioning attending be-
havior. Journal of Applied Behavior Analysis,
1968, 1, 245-250.

Yates, A. J. Behavior therapy. New York: John
Wiley and Sons, 1970.

Received 22 December 1975.
(Final acceptance 3 June 1976.)