Discussion Board Chi-Square Analysis

 

Within the Discussion Board area, write 300–500 words that respond to the following questions with your thoughts, ideas, and comments. This will be the foundation for future discussions by your classmates. Be substantive and clear, and use examples to reinforce your ideas.

It is frequently presumed that as individuals get older, they become more politically conservative. A political science student wants to verify this postulate. Using secondary data, the student ran a Chi-Square analysis of the age group (18–35, 36–55, 56–80) and self-described political affiliation (liberal, moderate, or conservative). The results of his analysis are provided in the tables below, but the student is having difficultly explaining the results.

• Describe the overall findings of the Chi-Square in the output, including the cell contributions, based upon the standardized residuals.
• What conclusions can the student make concerning this postulate?

Complete Question

Primary Task Response: Within the Discussion Board area, write 

3

0

0–

5

00 words that respond to the following questions with your thoughts, ideas, and comments. This will be the foundation for future discussions by your classmates. Be substantive and clear, and use examples to reinforce your ideas.

It is frequently presumed that as individuals get older, they become more politically

conservative

. A political science student wants to verify this postulate. Using secondary data, the student ran a Chi-Square analysis of the age group (

1

8

–35, 36–55, 56–80) and self-described political affiliation (

liberal

,

moderate

, or conservative). The results of his analysis are provided in the tables below, but the student is having difficultly explaining the results.

· Describe the overall findings of the Chi-Square in the output, including the cell contributions, based upon the standardized residuals.

· What conclusions can the student make concerning this postulate?

N

Percent

N

Percent

60

100.0%

Case Processing Summary
Cases
Valid	Missing		Total
		N			Percent
Political leanings * Age category			60		10 0 .0 %	0	0.0%

6.0

6.0

Count

4

Expected Count

6.0

6.0

6.0

Standardized Residual

-.8

Count

4

Expected Count

8.0

8.0

Standardized Residual

Total

Count

20

20

Expected Count

20.0

20.0

Political leanings * Age category Crosstabulation
Age category
young adult	middle aged	older adult
Political leanings	conservative				Count					4	3	11
			Expected Count						6.0
		Standardized Residual		-.8	– 1.2	2.0
moderate	9	5
1.2	-.4
liberal	12	8
		8.0
1.4	.0	-1.4
		20
		20.0

4

4

60

Chi-Square Tests
	Value	df	Asymptotic Significance (2-sided)
Pearson Chi-Square	12.667a	.013
Likelihood Ratio	12.350	.015
Linear-by-Linear Association	8.016	1	.005
N of Valid Cases
a. 0 cells (0.0%) have expected count less than 5. The minimum expected count is 6.00.

Value

Political leanings Dependent

Age category Dependent

.106

.059

.014c

Directional Measures
Asymptotic Standard Errora	Approximate Tb	Approximate Significance
Nominal by Nominal	Lambda	Symmetric	.263	.116	2.070	.038
	Political leanings Dependent	.222	.139	1.438	.150
	Age category Dependent	.300	.111	2.372	.018
Goodman and Kruskal tau		.106		.059		.014c
a. Not assuming the null hypothesis.
b. Using the asymptotic standard error assuming the null hypothesis.
c. Based on chi-square approximation

Answer from one of the Student

 Looking at the data that was gathered from members of different ages (18-35, 36-55, and 56-80) regarding their political affiliations, there were 60 participants with 20 in each group and each one submitted a response.  These age groups were labeled “young adult”, “middle age”, and “older adult.”  Young adults had 3 times more responses for liberal and an equal amount for conservative and moderate.  The expected responses, meaning that based on the percentages of normally-distributed population members, the numbers that were expected to vote liberally, moderately, and conservatively were 6, 6, and 8 respectively.  It can be seen that they voted low for conservative and moderate while considerably over-voting for liberal.  Middle-aged adults had less of a conservative representation than the young adult and more moderate voters.  Older adults had the most conservative members.  

            The Pearson Chi-Square test was used because frequencies are being compared instead of percentages.  The values that are being evaluated are political affiliations: conservative, moderate, and liberal.  Our null hypothesis that we are determining to reject or not is that people become more conservative as they get older.  Looking at the expected values, only 1 out of the 9 age and political combinations have values that came out to what was expected.  The effect size comes out to 0.106 (12.667 / (60*2)).  This points to a difference in the hypothesized values and p = .013 tells us that the difference is significant.

The question the political science student was trying to answer was whether individuals became more conservative as they got older.  Conservative numbers would be expected to increase from young adult” to “middle-aged” and again to “older adult.”  The data does not support that.  While older adults as a whole are more conservative than moderate or liberal, the trend did not increase as they age.  Young adults to middle-aged actually become less conservative.