Assignment 3. Bivariate Relationships

Due: Wednesday, October 17 at the beginning of class

1. Use the following data to explore the relationship between genderrole attitudes and gender. Specifically, use the data below to test the null hypothesis that genderrole attitudes and gender are independent in the population (Hint: calculate chi-square). What is your conclusion? How strong is the relationship between these two variables? (Hint: calculate an appropriate measure of association).

2. In the US, do married men and single men work the same number of hours in paid employment? Also, do married women and single women work the same number of hours in paid employment? Use SPSS and data from the 2006 General Social Survey to investigate this question (i.e., perform two difference in means tests; do two-tailed hypothesis testsand set alpha=.05; use the variables SEX, MARITAL, and HRS1). Write a statement to describe your findings (be sure to state the hypotheses, etc.).

Hint: You will be performing two separate independent samples t tests. For the first, use the Data → Select Cases menu to select only male respondents (‘if condition is satisfied’; SEX=1). After obtaining the results from the independent samples t test comparing married to never married men (Analyze → Compare Means → Independent Samples t test), return to the select cases window and select all cases. Then, select only female respondents and obtain a second independent samples t test comparing married to never married women.

3. Use the One-way ANOVA procedure in SPSS and the 2006 General Social Survey to examine the relationship between television consumption (TVHOURS) and marital status (MARITAL). Specifically, test the null hypothesis that the five group means are equal. Based on post-hoc tests, which groups, if any, have significantly different means from which groups? Also, use the Means procedure in SPSS to calculate eta-squared. How strong is the relationship between tv consumption and marital status? Write a statement to describe your findings (be sure to state the hypotheses, etc.).

4. Use SPSS and the 2006 General Social Survey to study the relationship between years of father’s education (PAEDUC) and years of respondent’s education (EDUC).

a. Construct a scatterplot of these two variables in SPSS and place the best-fitting linear regression line on the scatterplot. Describe the relationship between father’s education and education.

b. Have SPSS calculate the regression equation predicting education with father’s education. What are the intercept and slope? What are the coefficient of determination and the correlation coefficient? What is the 95% confidence interval for the slope? Interpret the intercept, regression slope, confidence interval for the regression slope, and the coefficient of determination.

c. What is the predicted education for someone who has a father with 6 years of education? Show your work.

d. What is the predicted education for someone who has a father with 14 years of education? Show your work.

5. Using the following raw data, analyze the relationship between mean education and the unemployment rate (Note that the cases are occupations and not individuals; N=8). Calculate the intercept and slope, Pearson’s correlation coefficient, and the coefficient of determination. Interpret the intercept, slope, and coefficient of determination. Treat the unemployment rate as the dependent variable. Show your work.

Occupation / Mean Education / Percent Unemployed
Managers of food-serving and lodging est. / 13.1 / 3.8
Civil engineers / 16.1 / 2.9
Agricultural and food scientists / 16.5 / 2.2
Dentists / 20.2 / 0.1
Registered nurses / 15.4 / 1.5
Archivists and curators / 16.1 / 2.4
Real estate sales occupations / 14.4 / 2.2
Sales engineers / 15.5 / 4.5

- 1 -