Infants

explain what the paper was about.  What was the research question?  What hypotheses did the authors test?  How did they test their hypothesis (that is, what did they do?) and what were the results?  What do they think their results mean?  Be sure to focus only on the essential information—you can assume that I have read the paper and know the details. 

Did the authors convince you that their research question is important and that their hypotheses are meaningfully related to their research question? Why or why not?

What do you think about the interpretations of the data and the associated conclusions the authors draw? Are there additional experiments that you think would be relevant and interesting for future investigations?

Article evaluations should be no longer than 1 page, single-spaced

Toddlers Help a Peer

Robert Hepach
Leipzig University and Max Planck Institute for

Evolutionary Anthropology

Nadine Kante
Max Planck Institute for Evolutionary Anthropology

Michael Tomasello
Duke University and Max Planck Institute for Evolutionary Anthropology

Toddlers are remarkably prosocial toward adults, yet little is known about their helping behavior toward
peers. In the present study with 18- and 30-month-old toddlers (n = 192, 48 dyads per age group), one child
needed help reaching an object to continue a task that was engaging for both children. The object was within
reach of the second child who helped signi!cantly more often compared to a no-need control condition. The
helper also ful!lled the peer’s need when the task was engaging only for the child needing help. These !nd-
ings suggest that toddlers’ skills and motivations of helping do not depend on having a competent and help-
ful recipient, such as an adult, but rather they are much more “exible and general.

During the 2nd year of life young children help
adults in a variety of ways. They readily share a
valued resource with adults (Brownell, Svetlova, &
Nichols, 2009; Svetlova, Nichols, & Brownell, 2010)
and from as early as 12 months provide adults with
helpful information through means of pointing
(Liszkowski, Carpenter, Striano, & Tomasello,
2006). Children between the ages of 1 and 2 com-
fort adults in distress (Bischof-K!ohler, 1991; Zahn-
Waxler, Radke-Yarrow, Wagner, & Chapman,
1992), as well as help adults complete their instru-
mental goals (Warneken & Tomasello, 2007). This
motivation is intrinsic (Warneken & Tomasello,
2008) and is driven by a motive to see the person
in need receive help, thereby suggesting that young
children have a genuine concern for others’ well-
being (Hepach, Vaish, & Tomasello, 2012, 2013).

However, a challenge for this notion of a general
prosocial motive stems from the fact that young
children’s helping behavior has almost exclusively

been studied with an adult recipient. The issue is
that the history of adult–child interactions is gener-
ally cooperative, given that adults share with, com-
fort, and instrumentally help children on a regular
basis (Deynoot-Schaub & Riksen-Walraven, 2006).
Adults scaffold social interactions with toddlers in
ways that peers do not. This scaffolding can range
from providing cues (Svetlova et al., 2010) accord-
ing to the adult’s need, to creating joint commit-
ments (Warneken, Chen, & Tomasello, 2006), and
to encouraging and rewarding prosocial behavior
(Dahl, 2015; Waugh, Brownell, & Pollock, 2015).
Therefore, adults are competent, helpful partners
when it comes to ful!lling children’s needs. On
the other hand, and in contrast to their experience
with adults, toddlers have very little experience
with being helped by their peers or cooperating
with their peers (Brownell, Ramani, & Zerwas,
2006). When toddlers seek access to a resource,
such as an attractive toy, con”icts may arise
because both children demand access to the toy at
the same time (Hay, 2006; Hay & Ross, 1982), with
such disputes typically requiring the intervention
of an adult for resolution (Ross, Tesla, Kenyon, &
Lollis, 1990; Williams, Mastergeorge, & Ontai,
2010). This raises the possibility that toddlers are
not generally motivated to help others, but rather

We thank all the families who participated in the studies. The
following people helped with data collection and we are grateful
for their assistance: Philipp Berger, Hella Franke, Sara Rossi, and
Susanne Mauritz. In addition, Nora Drewes, Hille Nike St!uhring,
Julia J!ager, Ken Krempler, and Mira Weihe helped with tran-
scribing and the reliability coding of the data. We thank Roger
Mundry for his help with statistics as well as Felix Warneken
and Sarah Peoples for their helpful comments on an earlier ver-
sion of the article.

Correspondence concerning this article should be addressed to
Robert Hepach, Department of Research Methods in Early Child
Development, Faculty of Education, Leipzig University, 04109
Leipzig. Electronic mail may be sent to robert.hepach@uni-leipzig.
de.

© 2016 The Authors
Child Development © 2016 Society for Research in Child Development, Inc.
All rights reserved. 0009-3920/2017/8805-0022
DOI: 10.1111/cdev.12686

Child Development, September/October 2017, Volume 88, Number 5, Pages 1642–1652

their prosocial behavior is restricted to individuals
who are competent and helpful to them, that is,
an adult partner (see also Brownell, Nichols, &
Svetlova, 2005).

The relevant age range to study toddler peer
helping is during and after the 2nd year of life
given that a number of studies demonstrate devel-
opmental changes in children’s social behavior
around that age. Before 2 years of age toddlers are
very poor at coordinating with peers (Brownell &
Brown, 1992; Brownell & Carriger, 1990; Brownell
et al., 2006; Eckerman, Davis, & Didow, 1989; Hay,
1979). For example, in one study, one child had to
push a lever so that a toy was retrievable for a sec-
ond child. Children could only do this reliably after
their second birthdays (Brownell & Carriger, 1990).
In another study children had to pull simultane-
ously on separate handles to play with an appara-
tus, and again the same age of accomplishment was
found (Brownell et al., 2006). In addition, 30-
month-old compared to 18-month-old toddlers
needed less communicative support to provide an
out-of-reach object for a recipient (Svetlova et al.,
2010). With regard to cooperating with siblings, 2-
year-old children’s cooperative behavior is more
frequent than 18-month-olds’, but this is contingent
on the sibling’s previous cooperative behavior
(Dunn & Munn, 1986). This relative incompetence
with a peer partner is particularly striking given
that 18-month-old toddlers readily coordinate with
adult partners in similar task situations (e.g., War-
neken et al., 2006).

Although observational work has shown that
children convey bene!ts to peers most notably in
the context of sharing (Bar-Tal, Raviv, & Goldberg,
1982; Eckerman, Whatley, & Kutz, 1975; Hay, Cas-
tle, Davies, Demetriou, & Stimson, 1999), children’s
motives in these free-play situations are not entirely
clear. They could be initiating a rewarding play
interaction, trying to ward off an aggressive peer,
or initiating any number of other types of interac-
tions. To determine whether children’s previously
documented helping behavior in play observations
is indeed helping (i.e., a response to a peer’s unful-
!lled need), an experimental task is needed. Such
an approach will also help to determine whether
toddlers’ instrumental helping shows a develop-
mental trajectory either more similar to peer collab-
oration, emerging only after 2 years of age, or more
similar to helping adults, which emerges at least
10 months earlier.

In the present study, therefore, we studied 18-
and 30-month-old children who were confronted
with a situation in which they had access to a toy

that a peer needed to operate a game. In one situ-
ation, the game was engaging for both children,
such that helping the partner led to an entertaining
result for both. In another condition that more clo-
sely paralleled previous studies with adult part-
ners, the child needing help was engaged in the
task alone. The crucial question was whether the
peer who had access to the needed toy would help
their partner to obtain the out-of-reach object.
Importantly, we included a control condition in
which children did not need the out-of-reach
objects because they already had access to other
similar objects.

Method

Participants

Participants were dyads of 18-month-old children
(n = 96, 48 girls, age range = 1 year 5 months
15 days to 1 year 6 months 27 days; M = 18 months
and 6 days, SD = 9 days) and 30-month-old children
(n = 96, 48 girls, age range = 2 years 5 months
16 days to 2 years 6 months and 27 days; M = 3

0

months and 3 days, SD = 9 days). Each child was
paired with an unfamiliar same-sex peer (48 dyads
for each age group). Children were recruited from a
database and parents gave informed consent before
their child participated in the study. Participants
were recruited from a midsized German city (popu-
lation approximately 500,000; median household
monthly income approximately 1,400!). Families vis-
ited the research institute and all children (predomi-
nately White Caucasian) received a small toy after
participation.

For the 30-month-old participants, six additional
children were tested but excluded due to parental
interference (n = 2), because the child did not want
to participate (n = 3), or due to an experimenter
error (n = 1). For an additional 10 children of the
same age group, at least one test trial had to be
excluded from further analyses due to parental
interference (n = 7), because one child did not
want to participate (n = 1), or due to an experi-
menter error (n = 2). For the 18-month-old partici-
pants, 21 additional children were tested but
excluded due to parental interference (n = 5), the
child not being able to walk on her own without
assistance (n = 7), not wanting to participate (n = 4),
an experimenter error (n = 2), or due to equipment
failure (n = 3). For an additional nine children of the
same age group, data for at least one test trial had to
be excluded due to parental interference (n = 3), the
child not wanting to participate (n = 4), or due to

Toddler Peer Helping 1643

equipment failure (n = 2). In all the above cases data
of one trial or both trials were excluded for the dyad.
Data were collected between May 2013 and June
2015.

Materials and Design

The study room was divided with a wooden
fence resulting in two roughly equally sized areas
(see Figure 1). The main apparatus consisted of a
tilted transparent tube that produced a sound if
children threw wooden balls down the tube. To
make the game additionally attractive, a rope light
was wrapped around the tube (see Figure 2A).
Additional materials included a wooden table with
a slanted top and a transparent box that contained
the wooden balls for the child during the test
phase. On the higher end of the table a little woo-
den stick prevented the balls from rolling down
(see Figure 2 B and 2C).

Dyads participated in a between-subject design
and were randomly assigned to one of three
experimental conditions (n = 16 in each): mutualis-
tic-helping, altruistic-helping, and a no-need con-
trol condition. In addition, each child was
randomly assigned to either the player or the
observer role (see below). In the mutualistic-help-
ing and the no-need conditions, the game was
equally engaging for both children. In the altruis-
tic-helping condition, the game was only engaging

for the player because a curtain covered the appa-
ratus. The adult experimenter lifted the curtain at
the beginning of the study such that the observer

Figure 1. Experimental room setup during test trials of the mutualistic-helping condition (A), the no-need control condition (B), and the
altruistic-helping condition (C). The fence in the center of the room had an approximate length of 390 cm and a height of 53 cm. The
apparatus was placed in the center of the room for the no-need control (B) and mutualistic-helping (A) conditions but behind a curtain
in the altruistic-helping condition (C). The wooden balls could be thrown into the tube on the player’s side and the apparatus lit up
and produced a sound on the observer’s side. In the altruistic-helping condition, the sound of the apparatus was muf”ed and only the
player could see the effect of placing a wooden ball into the tube. The two wooden balls were placed on the table such that they were
only accessible to the observer. In the mutualistic-helping (A) and altruistic-helping (C) conditions, the box on the player’s side was
empty, whereas several additional wooden balls were placed in the box in the no-need control condition (B). [Color !gure can be
viewed at wileyonlinelibrary.com].

Figure 2. (A) The game apparatus used in the study
(length = 145 cm, diameter 7 cm). (B and C) The table
(length = 60 cm, width = 29 cm, height = 52 cm) onto which the
experimenter put the balls in the experimental conditions. [Color
!gure can be viewed at wileyonlinelibrary.com].

1644 Hepach, Kante, and Tomasello

could brie”y see the apparatus that was addition-
ally muted to make the game less attractive for
the observer. Each dyad participated in a warm-
up phase, a demonstration phase, and a test
phase. The test phase consisted of two identical
test trials. After the second test trial, children
switched sides along with their parents such that
the observer child also got a chance to play the
game. Parents, who sat in a chair reading a maga-
zine, were asked to not engage with their child or
to comment on the situation.

Procedure

An adult female and male experimenter carried
out the study. One experimenter (E1) was responsi-
ble for interacting with the observer child, while
the other experimenter (E2) interacted with the
player child. During a warm-up phase in which
both families were introduced to one another, the
fence was not set up and the main apparatus was
not visible. Once both children were comfortable,
E1 set up both the fence and the main apparatus. In
the altruistic-helping condition, the subsequent
demonstration phase proceeded as follows: the
experimenters introduced the game (involving the
main apparatus) to the children. E1 placed two
wooden balls on a chair (see Figure 1). The player
child could access the balls and throw them into
the tube. Next, two additional wooden balls were
made available on the wooden table for the player.
For the !nal run of the demonstration phase, E1
moved back to the chair and placed on it two addi-
tional balls available for the player child. In the
mutualistic-helping and the no-need conditions, the
demonstration phase was engaging for both chil-
dren who saw the effect of the ball rolling down
the shoot. In the altruistic-helping condition, the
chair was on the player’s side and only the player
was engaged in the game. The experimenter did
address the observer if they began to lose interest
to ensure that they still paid attention to the game.

For the subsequent test phase, E1 started the !rst
test trial by drawing children’s attention to a new
set of wooden balls. E1 moved to the wooden table
and grabbed the transparent box. In the mutualis-
tic-helping and altruistic-helping conditions the box
contained two wooden balls. E1 took the balls out
of the box and placed the empty box onto the
player’s side of the fence. Then, E1 attempted to
place the balls on the table but because the experi-
menter began to turn away from the table, the balls
“accidentally” rolled to the observer’s side and out
of the player’s reach. In the no-need control

condition, the box was !lled with several wooden
balls, from which E1 took two out. E1 placed the
full box on the player’s side of the fence and placed
the two balls on the observer’s side of the table
repeating exactly what E1 said in the other two
conditions. In all conditions, E1 waited for the
observer to face them before placing the balls on
the table. The two experimenters moved away from
the situation and turned their backs to the children
(see Figure 1). The test trial ended after the obser-
ver provided at least one wooden ball; the player
was given time to operate the apparatus. If the
observer did not provide any of the out-of-reach
balls, the trial ended after 60 s. In that case, the
experimenter removed the balls from the table and
repeated the demonstration phase. Correspond-
ingly, the second test trial was identical to the !rst.

Coding and Reliability

During the test trial, the observer could help the
player by either throwing a ball over the fence, put-
ting the ball directly into the player’s hand, or put-
ting the balls in their initial position. We accounted
for whether or not the player requested help from
the observer (see Table 1 for coding details). We
further divided the player’s request into two cate-
gories: (a) reaching for the wooden balls without
any other form of communicative request (reach-
only) and (b) referring to the apparatus, box, or
wooden balls by means of pointing or verbal
expressions (communicative), for example, saying
“There” in conjunction with a reach for the wooden
balls. All sessions were videotaped. For the 18-
month-olds, the main experimenter coded all the

Table 1
Coding Scheme for Children’s Behavior During the Test Trial

Behavior
category Examples

Helping The observer Throws the ball over the
fence

Puts the ball into the player’s
hand

Places the ball on top of the
table where the player can
reach it

Requesting The player Reaches for the wooden balls
Points to the balls or to the
apparatus

Verbalizes need, for example,
“Ball!” or “Here!”

Toddler Peer Helping 1645

data and a second coder (blind to the study’s
hypotheses) coded a random sample of 25% of all
dyads. The dependent variables of interest were
helping (j = .85), the type of help (j = 1), request-
ing (j = .79), and the type of request (j = .86). For
the 30-month-old children, the main experimenter
coded a random sample of 25% of all dyads and an
additional second coder (blind to the study’s
hypotheses) coded all the data. Similar to the 18-
month-old children, the dependent variables of
interest were helping (j = 1), the type of help
(j = .96), requesting (j = .87), and the type of
request (j = .77).

Data Analysis

In the following section, we outline the separate
analyses and specify the details of each statistical
model. We carried out six separate analyses to
investigate both the observer’s helping behavior as
well as the player’s requesting during each test
trial. We analyzed data with generalized linear
mixed models (GLMMs, binomial error distribu-
tion) in R (Bates, Maechler, & Bolker, 2011; R Core
Team, 2015). The statistical signi!cance of each
independent factor was tested with a likelihood
ratio test comparing a full model to a reduced
model without the respective factor(s). In cases
where there were no trial effects, the dependent
measure for the post hoc comparison was whether
or not the child helped on at least one test trial.
Post hoc comparisons were carried out using Fish-
er’s exact tests. To determine an effect size estimate
comparable to the R2 of linear regressions, we cal-
culated effect sizes for each analysis by dividing the
variance of the !xed factors by the sum of the vari-
ance of the !xed factors, the variance of the random
factors, and the residual variance (adapted from
Nakagawa & Schielzeth, 2013).

Helping

The main analysis focused on whether or not
the observer provided the player with at least
one of the wooden balls during a given test trial.
The dependent measure was a binary coded vari-
able (yes or no). The independent factors were
condition, age, gender, and trial. In addition, the
model included random intercepts for dyad and
the identity (male or female) of the experimenter
on players’ side (Schielzeth & Forstmeier, 2009).
Preliminary analyses revealed no interaction of
age and condition on the observer’s helping
behavior.

Requesting

We investigated the frequency of the player’s
requests on each test trial for the mutualistic and
altruistic conditions given that the player never
requested in the no-need control condition. The
GLMM included condition, age, gender, and trial as
main effects, as well as random intercepts for both
dyad and experimenter identity. The dependent
measure was a binary coded variable for requesting
(yes or no). Preliminary analyses revealed no inter-
action of age and condition on the player’s request-
ing behavior.

Requesting (Type)

For trials on which the player made a request,
we analyzed the type of the request on each test
trial. The dependent measure was a binary coded
variable (communicative request or not) and the
independent factors as well as the random inter-
cepts were the same as for the analysis of the fre-
quency of requests.

Requesting and Helping

We investigated whether the player’s request
in”uenced the observer’s helping through carrying
out a GLMM analysis including only the mutualis-
tic and altruistic conditions given that the player
never requested in the no-need control condition.
Note that this analysis took into account individual
participant differences (for both requesting and
helping) across the two test trials. The model
included condition, gender, age, request (yes or no),
and trial as main effects, as well as random inter-
cepts for both dyad and experimenter identity. The
dependent measure was whether the observer
helped on a given trial. Preliminary analyses
revealed no three-way or two-way interactions of
age, condition, and request on the observer’s help-
ing behavior.

Type of Request and Helping

We investigated the effect of the type of the
peer’s request on the observer’s helping behavior
using a GLMM, including only the mutualistic and
altruistic conditions given that the player never
requested in the no-need control condition. The
model included condition, gender, age, type of
request (communicative or reaching only), and trial
as main effects, as well as random intercepts for
both dyad and experimenter identity. The

1646 Hepach, Kante, and Tomasello

dependent measure was whether the observer
helped on a given trial. Preliminary analyses
revealed no three-way or two-way interactions of
age, condition, and type of request on the obser-
ver’s helping behavior.

In!uence of Siblings and Kindergarten Attendance on
Helping

In a !nal analysis we included as main effects
the factors kindergarten or caretaker attendance
(yes or no) and presence of siblings (yes or no) into
the initial model (see Helping) regarding the obser-
ver’s helping behavior. This analysis is relevant in
order to investigate the degree to which children’s
helping is in”uenced by their experience with sib-
lings and with peers in kindergarten (see also Dunn
& Munn, 1986; Sara!no, 1985). Preliminary analyses
did not reveal an interaction of kindergarten atten-
dance and presence of siblings, no interaction of
kindergarten attendance and condition, and no
interaction of presence of siblings and condition.

Results

Helping

Children’s helping behavior varied between the
experimental conditions, GLMM: v2(2) = 24.38,
p < .001, R2(overall model) = .27 (see Figure 3). There was neither an effect of trial, v2(1) = 0.04, p = .85, nor of gender, v2(1) = 0.005, p = .95. In

addition, the rate of children’s helping did not differ
between the two age groups, v2(1) = 2.35, p = .13.
More children helped their peer on at least one test
trial in the mutualistic-helping condition (Trial 1: 17
of 32, 53%; Trial 2: 17 of 31, 55%) compared to the
no-need control condition (Trial 1: 6 of 32, 19%; Trial
2: 3 of 31, 10%), Trial 1 and Trial 2: p < .01. Like- wise, more children helped at least once in the altru- istic-helping condition (Trial 1: 17 of 32, 53%; Trial 2: 17 of 31, 55%) compared to the control condition, Trial 1 and Trial 2: p < .01. There was no difference between the mutualistic- and altruistic-helping con- ditions in the number of children who helped their peer at least once, Trial 1 and Trial 2: p = 1. With regard to the type of helping, on the !rst test trial, 27 (67.5%) of 40 children placed the ball(s) into the player’s hand, 2 (5%) of 40 placed the wooded ball (s) on top of the table or into the player’s box, and 11 (27.5%) of 40 children reached over the fence and placed the ball(s) onto the player’s side. On the sec- ond test trial, 21 (57%) of 37 children placed the ball (s) into the player’s hand, 3 (8%) of 37 placed the wooded ball(s) on top of the table or into the player’s box, and 13 (35%) of 37 children placed the ball(s) onto the player’s side.

Requesting

The frequency of children’s requesting behavior
did not vary between the mutualistic and altruistic
conditions, v2(1) = 0.02, p = .89, R2(overall
model) = .003. There was neither an effect of trial,

Trial 1 Trial 2

0

4

8

12

16

0
4
8
12
16

18-m
onth-olds

30-m
onth-olds

Mutualistic
-Helping

No-need
Control

Altruistic
-Helping

Mutualistic
-Helping
No-need
Control
Altruistic
-Helping

Condition

N
um

be
r

of
C

hi
ld

re
n

Type of Help
Hand

Position

Throw

Figure 3. The number of children who helped their peer plotted separately for each age group and test trial. For further illustration and
in addition to plotting the number of children who helped, we provide information regarding the type of help provided by the obser-
ver. The type of help was not a factor in the statistical analysis given that we did not have prior hypotheses regarding the way in
which toddlers would help their peer.

Toddler Peer Helping 1647

v2(1) = 0.01, p = .92, nor of gender, v2(1) = 0.1,
p = .75, or age group, v2(1) = 0.03, p = .87.

Requesting (Type)

For those trials in which the player made a
request, the type of request (communicative vs.
reaching-only) varied between the mutualistic-help-
ing and altruistic-helping conditions, v2(1) = 3.88,
p = .049, R2(overall model) = .09. Speci!cally, chil-
dren showed more communicative requests in the
mutualistic-helping (30 of 42; 71%) compared to the
altruistic-helping (21 of 41; 51%) condition (see Fig-
ure 4).

Requesting and Helping

Analyses investigating the effect of requesting on
helping behavior did not yield statistically signi!-
cant results. There was no effect of condition,
v2(1) = 0.2, p = .9, age, v2(1) = 0.28, p = .92, request-
ing, v2(1) = 0.03, p = .87, gender, v2(1) = 0.21,
p = .65, or trial, v2(1) = 0.4, p = .53, on the obser-
ver’s helping behavior, R2(overall model) = .04.
More speci!cally, on the !rst test trial, 11 (50%) of
22 children helped if the player did not make a
request, and 23 (55%) of 42 children helped after
the player did make a request. Similarly, on the sec-
ond test trial, 11 (52%) of 21 children helped if the
player did not make a request and 24 (59%) of 41
children helped after the player did make a request.

Type of Request and Helping

Children’s helping during the test trial was in”u-
enced by the type of request (communicative or
not) shown by the peer, v2(1) = 5.05, p = .02,
R2(overall model) = .09. There were no effects of
condition, v2(1) = 0.0001, p = .99, age, v2(1) = 0.37,
p = .54, gender, v2(1) = 0.01, p = .94, or trial,
v2(1) = 0.01, p = 0.91, on the observer’s helping
behavior. For the subsequent analyses, and given
that there was no effect of trial, we averaged the
dependent variable across both trials. When the
peer showed a noncommunicative request but only
reached for the wooded balls, children helped more
often (Mnumber of trials = .74, SD = .44) compared to
when the request was communicative (Mnumber of tri-
als = .46, SD = .45).

In!uence of Siblings and Kindergarten Attendance on
Helping

Children who grew up with at least one sibling
at home helped their peer more often (Trial 1: 23 of
43, 53%; Trial 2: 19 of 41, 46%) than children who
did not grow up with a sibling at home (Trial 1: 17
of 53, 32%; Trial 2: 18 of 52, 35%), v2(1) = 7.31,
p = .007. On the other hand, children who attended
kindergarten helped their peer less often (Trial 1: 29
of 77, 38%; Trial 2: 27 of 76, 36%) than children
who did not attend kindergarten (Trial 1: 11 of 19,
58%; Trial 2: 10 of 17, 59%), v2(1) = 7.98, p = .005.
Similar to the initial analysis on the observer’s help-
ing behavior, there was a statistically signi!cant
effect of condition, v2(2) = 24.53, p < .001, and no effect of age (p = .19), gender (p = .44), or trial (p = .9).

Discussion

This study investigated whether 18- and 30-month-
old children help a same-aged peer to complete an
instrumental goal. When participants witnessed
another child as unable to continue a task, they
readily helped them by providing the necessary
objects, which were only accessible to the helper.
Children acted prosocially in a variety of ways,
either by depositing the object where the player
could reach it or by putting the object directly into
the player’s hand. Crucially, children helped less
when no help was needed because the potential
bene!ciary had access to additional objects being
able to continue the task on their own and was
therefore not dependent on the peer’s help. One

sdlo-htnom-03sdlo-htnom-81

Age Group

N
um

be
r o

f C
hi

ld
re

n
W

ho
S

ho
w

ed
A

C
om

m
un

ic
at

iv
e

R
eq

ue
st

0
5

10
15

20

Altruistic-Helping
Mutualistic-Helping

Figure 4. The effect of condition on the player’s (i.e., recipient’s)
type of request. For additional illustration, the data are plotted
separately for both age groups.

1648 Hepach, Kante, and Tomasello

conceivable explanation for this helping behavior
could be that the task was engaging and attractive
for both children, such that the helper also bene!t-
ted from providing the out-of-reach objects to see
the task continued. However, when the situation
was changed, such that the game was only visible
and engaging for the player, the peer still continued
to help. Together, these !ndings suggest that chil-
dren are not only motivated to act prosocially
toward their peers but that they also do so in a
coordinated way when help is actually needed.

These results extend previous research !ndings
by demonstrating that young children are not only
motivated to help adults (e.g., Dun!eld, Kuhlmeier,
O’Connell, & Kelley, 2011; Warneken & Tomasello,
2009) but that their prosociality also extends to
same-age peers. When an adult needs help, children
readily help by sharing resources and information
(e.g., Brownell et al., 2009; Liszkowski et al., 2006),
comforting (Zahn-Waxler et al., 1992), and instru-
mentally helping (Rheingold, 1982; Warneken &
Tomasello, 2006; Warneken & Tomasello, 2007). In
such situations the adult typically expressed their
need through structured cues until the child
responds (e.g., Svetlova et al., 2010; see also Warne-
ken, 2013 for examples of proactive helping in 2-
year-olds). In addition, in the majority of those
cases (a) children interacted with a competent and
motivated adult, and (b) help was needed in situa-
tions that are structured to be familiar to children,
such as everyday household tasks (e.g., Rheingold,
1982; though see Warneken & Tomasello, 2006 for
the use of novel tasks). The current study suggests
that children’s helping is neither limited to adult
partners nor does helping only occur in a struc-
tured or familiar context. In fact, in the present
study, many participants proactively helped their
peer and correctly identi!ed the actual need.

This is particularly interesting given that at the
age of 18 months, children do not appear to suc-
cessfully collaborate with one another (Brownell &
Carriger, 1990; Brownell et al., 2006; Eckerman
et al., 1989; Hay, 1979). One crucial difference in
comparison to previous work is that children in the
present study did not have to spatially and tempo-
rally coordinate their complementary roles. For
example, in the study by Brownell and Carriger
(1990), one child had to wait for the peer to be in
the correct position and to carry out a complemen-
tary action at the same time. This required children
to monitor both their own actions as well as the
behavior of their partner. On the other hand, chil-
dren in the present study had to merely monitor
the player’s need for assistance with reduced

cognitive demands of coordination. The need could
be ful!lled so long as children made one of the
wooden balls available. In fact the type of help
included both more coordinated (placing the ball in
the peer’s hand) and less coordinated behaviors
(placing the ball on the peer’s side of the fence).

It is important to note that although the rates of
helping were similar between the mutualistic and
altruistic conditions, the type of the player’s
requesting behavior was not. In the altruistic condi-
tion, the player more often reached for the objects,
whereas in the mutualistic condition, the request
was more communicative including pointing and
verbalizations. Despite similar degrees of helping,
children may have perceived the situation in the
mutualistic condition differently viewing the mutu-
alistic condition as a more collaborative activity in
which requesting involved communicating the joint
need. The fact that children in the present study
showed more communicative requests in the mutu-
alistic-helping condition may indicate that they took
into account what the observer knew about the sit-
uation, thus using pointing and verbal expressions
to refer to a common ground. On the other hand,
in the altruistic-helping condition the apparatus
was hidden from the observer’s view such that
there was no joint goal for the player to communi-
cate about.

Furthermore and despite the form of the player’s
request varying between conditions, we did not
!nd a contingency between requesting and the
observer’s helping behavior. Children helped their
peer independent of whether a request was made
or not. This is comparable to toddlers’ dif!culties to
collaborate as reported in previous work (Brownell
& Carriger, 1990; Brownell et al., 2006; Eckerman
et al., 1989; Hay, 1979). Although children in the
present study were sensitive to the peer’s need, the
interaction between the two children was not one
in which both parties coordinated their complemen-
tary actions. In such a case one would have
expected the player’s requesting to be more in”uen-
tial on the observer’s helping. Although adults use
communicative cues to scaffold their activities with
children and assume complementary roles, commu-
nication in peer interactions may be more ambigu-
ous. It is conceivable that the main function of the
player’s request was to draw attention to his or her
need. Overall, in comparison to previous work, our
results suggest that 1- and 2-year-old children’s dif-
!culty to collaborate is not due to a lack in motiva-
tion to engage with others or from being insensitive
to others’ needs. Rather, the discrepancy between
toddlers helping others instrumentally and failing

Toddler Peer Helping 1649

to collaborate may be rooted in the more demand-
ing nature of collaborative tasks that require “exible
coordination. The crucial addition from the present
study is the altruistic-helping condition, which
shows that children help a peer even when they
themselves do not immediately bene!t from help-
ing.

Our !ndings are also consistent with previous
theories on the development of prosocial behavior
in toddlers. The fact that children help each other
instrumentally does not contradict !ndings that
children are competitive and sometimes aggressive
toward their peers. It is often in object-centered
activities where a valued toy has to be shared such
that con”icts between toddlers and their peers may
arise. However, this is not because toddlers are
indifferent toward peers but rather the con”icts
stem from an inability to communicate and negoti-
ate in taking turns (Eckerman & Peterman, 2001;
Hay & Ross, 1982). One explanation for why chil-
dren in the present study helped their peer could
be that helping was not tied to an immediate cost.
The observing child could not herself or himself
play with the wooden balls. The motivation to pro-
vide instrumental help may be different from shar-
ing a more valued resource, including access to a
toy (e.g., Bar-Tal et al., 1982; Eckerman et al., 1975;
Hay, 2006; Hay & Ross, 1982). It would therefore
be an interesting avenue for future research to
observe and compare children’s instrumental help-
ing to their sharing behavior in free-play conditions.
A similar direction for future studies is to systemat-
ically vary the cost associated with helping using a
similar experimental design as the one used in the
present study. Indeed, costly and noncostly helping
among peers may follow different developmental
trajectories (see also Dun!eld et al., 2011).

We did not !nd support for previous !ndings
reporting a decline in toddlers’ helping behavior
during the 2nd year of life although the empirical
evidence is mixed (Eckerman et al., 1975; Hay,
Caplan, Castle, & Stimson, 1991). In the present
studies, across all experimental conditions both 30-
month-old and 18-month-old children helped their
partner at equal rates. Previous work had found
that although children’s motivation to engage with
peers increases over the 2nd year of life (Eckerman
et al., 1975), toddlers’ motivation to help after a
request but not their spontaneous helping, declines
during a similar age period (Hay et al., 1991). How-
ever, in the present study, we found children’s
helping behavior to systematically vary neither
between situations where the peer made a request
(elicited helping) and did not make a request

(spontaneous helping) nor to be in”uenced by age
(or an interaction of age and request). At the same
time, the present study was not designed to speci!-
cally test a contingency between requesting and
helping given that the request was not experimen-
tally manipulated. It is an interesting and relevant
question for future research to investigate whether
the underlying motivation to help both between
conditions (mutualistic and altruistic) and across
ages is different. Although the main aim of the pre-
sent study was to investigate whether toddlers help
to speci!cally ful!ll their peer’s need, a relevant
task for future research is to more systematically
investigate the individual differences, for example,
the in”uence of executive functions, on children’s
coordinated prosocial behavior toward peers (Miller
& Marcovitch, 2015).

A further relevant and related issue concerning
individual differences is the in”uence of peer expe-
rience on toddlers’ prosocial behavior. In the pre-
sent study, we found an effect of condition when
controlling for peer experience and whether a child
grew up with a sibling. We did not !nd a cumula-
tive effect of both growing up with a sibling and
attending kindergarten. In fact the in”uence of
experience with peers was more nuanced. Children
who grew up with a sibling showed more prosocial
behavior across conditions. This parallels previous
work that found a positive association of having a
sibling and helping (Dunn & Munn, 1986). Growing
up with siblings may facilitate children’s own
prosocial behavior because siblings, especially older
ones, provide help on a regular basis (see also
Lamb, 1978). On the other hand, children who
attended kindergarten showed overall less helping
behavior than those who did not. Children at the
age of 1 and 2 years will have only recently been
introduced to kindergarten, and interactions among
stranger peers are not exclusively prosocial (Sara-
!no, 1985). However, more work is needed to sys-
tematically investigate the in”uence of signi!cant
social context variables (see also Howes, 1987;
Howes & Rubenstein, 1981). For example, is chil-
dren’s behavior in free-play conditions related to
their helping in experimentally controlled studies?
In addition, including both need and no-need con-
ditions may be fruitful for procedures aimed at fos-
tering and promoting prosocial behavior among
peer toddlers.

In summary, when 18- and 30-month-old tod-
dlers are confronted with a situation in which a
peer cannot continue a task, they help either with
or without a request from the peer in ways that
appropriately ful!ll the need, even when the task

1650 Hepach, Kante, and Tomasello

itself is not engaging for the helper. This form of
prosocial peer behavior emerges before children
are able to successfully coordinate collaborative
activities (e.g., Brownell & Carriger, 1990). The
prosocial motive to help same-age conspeci!cs
may be foundational to the later emerging and
more complex forms of human prosociality and
cooperation.

References

Bar-Tal, D., Raviv, A., & Goldberg, M. (1982). Helping
behavior among preschool children: An observational
study. Child Development, 53, 396–402. doi:10.2307/
1128982

Bates, D., Maechler, M., & Bolker, B. (2011). lme4: Linear
mixed-effects models using S4 classes. R package version
0.999375-42.

Bischof-K!ohler, D. (1991). The development of empathy
in infants. In M. E. Lamb & H. Keller (Eds.), Infant
development: Perspectives from German speaking countries
(pp. 245–273). Hillsdale, NJ: Erlbaum.

Brownell, C., & Brown, E. (1992). Peers and play in
infants and toddlers. In V. B. Van Hasselt (Eds.), Hand-
book of social development: A lifespan perspective (pp. 183–
200). New York: Plenum Press.

Brownell, C. A., & Carriger, M. S. (1990). Changes in
cooperation and self-other differentiation during the
second year. Child Development, 61, 1164–1174. http://
www.jstor.org/stable/1130884

Brownell, C. A., Nichols, S., & Svetlova, M. (2005). Early
development of shared intentionality with peers. Behav-
ioral and Brain Sciences, 28, 693–694. doi:10.1017/S0140
525X05240128

Brownell, C., Ramani, G., & Zerwas, S. (2006). Becoming
a social partner with peers: Cooperation and social
under- standing in one- and two-year-olds. Child Devel-
opment, 77, 803–821. doi:10.1111/j.1467-8624.2006.00
904.x

Brownell, C., Svetlova, M., & Nichols, S. (2009). To share
or not to share: When do toddlers respond to another’s
needs? Infancy, 14, 117–130. doi: 10.1080/15250000802
569868

Dahl, A. (2015). The developing social context of infant
helping in two US samples. Child Development, 86,
1080–1093. doi:10.1111/cdev.12361

Deynoot-Schaub, M. G., & Riksen-Walraven, J. M. (2006).
Peer interaction in child care centres at 15 and
23 months: Stability and links with children’s socio-
emotional adjustment. Infant Behavior and Development,
29, 276–288. doi:10.1016/j.infbeh.2005.12.005

Dun!eld, K., Kuhlmeier, V. A., O’Connell, L., & Kelley,
E. (2011). Examining the diversity of prosocial behav-
ior: Helping, sharing, and comforting in infancy.
Infancy, 16, 227–247. doi:10.1111/j.1532-7078.2010.00
041.x

Dunn, J., & Munn, P. (1986). Siblings and the develop-
ment of prosocial behaviour. International Journal of
Behavioral Development, 9, 265–284. doi: 10.1177/01650
2548600900301

Eckerman, C. O., Davis, C. C., & Didow, S. M. (1989).
Toddlers’ emerging ways of achieving social coordina-
tions with a peer. Child Development, 60, 440–453.
doi:10.2307/1130988

Eckerman, C. O., & Peterman, K. (2001). Peers and infant
social/communicative development. In G. Bremner &
A. Fogel (Eds.), Blackwell handbook of infant development.
Handbooks of developmental psychology (pp. 326–350).
Malden, MA: Blackwell.

Eckerman, C. O., Whatley, J. L., & Kutz, S. L. (1975).
Growth of social play with peers during the second
year of life. Developmental Psychology, 11(1), 42.
doi:10.1037/h0076131

Hay, D. F. (1979). Cooperative interactions and sharing
between very young children and their parents. Devel-
opmental Psychology, 15, 647. doi:10.1037/0012-1649.15.
6.647

Hay, D. F. (2006). Yours and mine: Toddlers’ talk about
possessions with familiar peers. British Journal of Devel-
opmental Psychology, 24(1), 39–52. doi:10.1348/0261510
05X68880

Hay, D. F., Caplan, M., Castle, J., & Stimson, C. A. (1991).
Does sharing become increasingly” rational” in the sec-
ond year of life? Developmental Psychology, 27, 987.
doi:10.1037/0012-1649.27.6.987

Hay, D. F., Castle, J., Davies, L., Demetriou, H., &
Stimson, C. A. (1999). Prosocial action in very early
childhood. Journal of Child Psychology and Psychiatry, 40,
905–916. doi:10.1111/1469-7610.00508

Hay, D. F., & Ross, H. S. (1982). The social nature of early
con”ict. Child Development, 53, 105–113. http://www.js-
tor.org/stable/1129642

Hepach, R., Vaish, A., & Tomasello, M. (2012). Young
children are intrinsically motivated to see others
helped. Psychological Science, 23, 967–972. doi: 10.1177/
0956797612440571

Hepach, R., Vaish, A., & Tomasello, M. (2013). A new
look at children’s prosocial motivation. Infancy, 18(1),
67–90. doi:10.1111/j.1532-7078.2012.00130.x

Howes, C. (1987). Social compentency with peers: Contri-
butions from child care. Early Childhood Research Quar-
terly, 2, 155–167.

Howes, C., & Rubenstein, J. L. (1981). Toddler peer
behavior in two types of day care. Infant Behavior and
Development, 4, 387–393. doi:10.1016/S0163-6383(81)
80039-2

Lamb, M. E. (1978). Interactions between eighteen-month-
olds and their preschool-aged siblings. Child Develop-
ment, 49, 51–59. http://www.jstor.org/stable/1128592

Liszkowski, U., Carpenter, M., Striano, T., & Tomasello,
M. (2006). Twelve- and 18-month-olds point to provide
information for others. Journal of Cognition and Develop-
ment, 7, 173–187. doi:10.1207/s15327647jcd0702_2

Toddler Peer Helping 1651

http://dx.doi.org/10.2307/1128982

http://dx.doi.org/10.2307/1128982

http://www.jstor.org/stable/1130884

http://www.jstor.org/stable/1130884

http://dx.doi.org/10.1017/S0140525X05240128

http://dx.doi.org/10.1017/S0140525X05240128

http://dx.doi.org/10.1111/j.1467-8624.2006.00904.x

http://dx.doi.org/10.1111/j.1467-8624.2006.00904.x

http://dx.doi.org/10.1080/15250000802569868

http://dx.doi.org/10.1080/15250000802569868

http://dx.doi.org/10.1111/cdev.12361

http://dx.doi.org/10.1016/j.infbeh.2005.12.005

http://dx.doi.org/10.1111/j.1532-7078.2010.00041.x

http://dx.doi.org/10.1111/j.1532-7078.2010.00041.x

http://dx.doi.org/10.1177/016502548600900301

http://dx.doi.org/10.1177/016502548600900301

http://dx.doi.org/10.2307/1130988

http://dx.doi.org/10.1037/h0076131

http://dx.doi.org/10.1037/0012-1649.15.6.647

http://dx.doi.org/10.1037/0012-1649.15.6.647

http://dx.doi.org/10.1348/026151005X68880

http://dx.doi.org/10.1348/026151005X68880

http://dx.doi.org/10.1037/0012-1649.27.6.987

http://dx.doi.org/10.1111/1469-7610.00508

http://www.jstor.org/stable/1129642

http://www.jstor.org/stable/1129642

http://dx.doi.org/10.1177/0956797612440571

http://dx.doi.org/10.1177/0956797612440571

http://dx.doi.org/10.1111/j.1532-7078.2012.00130.x

http://dx.doi.org/10.1016/S0163-6383(81)80039-2

http://dx.doi.org/10.1016/S0163-6383(81)80039-2

http://www.jstor.org/stable/1128592

http://dx.doi.org/10.1207/s15327647jcd0702_2

Miller, S. E., & Marcovitch, S. (2015). Examining executive
function in the second year of life: Coherence, stability,
and relations to joint attention and language. Develop-
mental Psychology, 51, 101. doi:10.1037/a0038359

Nakagawa, S., & Schielzeth, H. (2013). A general and
simple method for obtaining R2 from generalized linear
mixed-effects models. Methods in Ecology and Evolution,
4, 133–142. doi:10.1111/j.2041-210x.2012.00261.x

R Core Team. (2015). R: A language and environment for
statistical computing. Vienna, Austria: R Foundation for
Statistical Computing. Retrieved from http://www.R-
project.org/

Rheingold, H. (1982). Little children’s participation in the
work of adults, a nascent prosocial behavior. Child
Development, 53, 114–125. doi:10.2307/1129643

Ross, H., Tesla, C., Kenyon, B., & Lollis, S. (1990). Mater-
nal intervention in toddler peer con”ict: The socializa-
tion of principles of justice. Developmental Psychology,
26, 994.

Sara!no, E. P. (1985). Peer-peer interaction among infants
and toddlers with extensive daycare experience. Journal
of Applied Developmental Psychology, 6(1), 17–29. doi:10.
1016/0193-3973(85)90013-9

Schielzeth, H., & Forstmeier, W. (2009). Conclusions
beyond support: Overcon!dent estimates in mixed
models. Behavioral Ecology, 20, 416–420. doi:10.1093/be-
heco/arn145

Svetlova, M., Nichols, S., & Brownell, C. (2010). Toddlers’
prosocial behavior: From instrumental to empathic to
altruistic helping. Child Development, 81, 1814–1827.
doi:10.1111/j.1467-8624.2010.01512.x

Warneken, F. (2013). Young children proactively remedy
unnoticed accidents. Cognition, 126, 101–108. doi:10.
1016/j.cognition.2012.09.011

Warneken, F., & Tomasello, M. (2006). Altruistic helping
in human infants and young chimpanzees. Science, 311
(5765), 1301–1303.

Warneken, F., Chen, F., & Tomasello, M. (2006). Coopera-
tive activities in young children and chimpanzees. Child
Development, 77, 640–663. doi:10.1111/j.1532-7078.2007.
tb00227.x

Warneken, F., & Tomasello, M. (2007). Helping and coop-
eration at 14 months of age. Infancy, 11, 271–294.
doi:10.1111/j.1532-7078.2007.tb00227.x

Warneken, F., & Tomasello, M. (2008). Extrinsic rewards
undermine altruistic tendencies in 20-month-olds.
Developmental Psychology, 44, 1785. doi:10.1037/a001
3860

Warneken, F., & Tomasello, M. (2009). The roots of human
altruism. British Journal of Psychology, 100, 455–471.
doi:10.1348/000712608X379061

Waugh, W., Brownell, C., & Pollock, B. (2015). Early
socialization of prosocial behavior: Patterns in parents’
encouragement of toddlers’ helping in an everyday
household task. Infant Behavior and Development, 39, 1–
10. doi:10.1016/j.infbeh.2014.12.010

Williams, S. T., Mastergeorge, A. M., & Ontai, L. L.
(2010). Caregiver involvement in infant peer interac-
tions: Scaffolding in a social context. Early Childhood
Research Quarterly, 25, 251–266. doi:10.1016/j.ecresq.
2009.11.004

Zahn-Waxler, C., Radke-Yarrow, M., Wagner, E., & Chap-
man, M. (1992). Development of concern for others.
Developmental Psychology, 28, 126. doi:10.1037/0012-
1649.28.1.126

1652 Hepach, Kante, and Tomasello

http://dx.doi.org/10.1037/a0038359

http://dx.doi.org/10.1111/j.2041-210x.2012.00261.x

http://www.R-project.org/

http://www.R-project.org/

http://dx.doi.org/10.2307/1129643

http://dx.doi.org/10.1016/0193-3973(85)90013-9

http://dx.doi.org/10.1016/0193-3973(85)90013-9

http://dx.doi.org/10.1093/beheco/arn145

http://dx.doi.org/10.1093/beheco/arn145

http://dx.doi.org/10.1111/j.1467-8624.2010.01512.x

http://dx.doi.org/10.1016/j.cognition.2012.09.011

http://dx.doi.org/10.1016/j.cognition.2012.09.011

http://dx.doi.org/10.1111/j.1532-7078.2007.tb00227.x

http://dx.doi.org/10.1111/j.1532-7078.2007.tb00227.x

http://dx.doi.org/10.1111/j.1532-7078.2007.tb00227.x

http://dx.doi.org/10.1037/a0013860

http://dx.doi.org/10.1037/a0013860

http://dx.doi.org/10.1348/000712608X379061

http://dx.doi.org/10.1016/j.infbeh.2014.12.010

http://dx.doi.org/10.1016/j.ecresq.2009.11.004

http://dx.doi.org/10.1016/j.ecresq.2009.11.004

http://dx.doi.org/10.1037/0012-1649.28.1.126

http://dx.doi.org/10.1037/0012-1649.28.1.126