Bivariate Regression Exercise #1

							                             Bivariate Regression Exercise #1
Does a respondent’s education level affect his or her attendance at religious services?
Conduct a bivariate regression, using education (educ) as your independent variable and
religious attendance (attend) as your dependent variable. Don’t forget to recode any
variables, if necessary. Answer the following questions as you go along:
        Create a scatterplot for the above analysis. Since religious attendance is an
        ordinal variable with only a few categories, use the Point Bins option to make
        sense of the scatterplot.

       What are the null and alternate hypotheses?


       List the four assumptions for bivariate regression, how you tested them, and
       whether or not they were met.
               1.

               2.

               3.

               4.

       Report the correlation coefficient, coefficient of determination, and
       unstandardized coefficients. Note which statistics were significant at the p =.05
       level.


       Write the unstandardized regression equation and discuss your findings.



After explaining the above regression analysis to your partner, conduct a second bivariate
regression to see if religious attendance has any effect on how often the respondent does
voluntary work for a charity (volchrty). Follow the steps above. Note: volcharty should
be reversed so that higher values represent more frequent voluntary work. Your partner
will be doing a similar regression analysis suing education as the independent variable.
After doing so, answer the following question with your partner.

       Recalling what we know about standardized coefficients, which has a bigger
       effect on the frequency of voluntary charitable work, education or religious
       attendance?

For fun, together conduct a multivariate regression analysis using both education and
religious attendance as your independent variables. Use charitable voluntary work as
your dependent variable. What did you find?
                             Bivariate Regression Exercise #2
Does a respondent’s education level explain the education level of his or her spouse?
Conduct a bivariate regression, using education (educ) as your independent variable and
spouse’s education (speduc) as your dependent variable. Don’t forget to recode any
variables, if necessary. Answer the following questions as you go along:
        Create a scatterplot for the above analysis. For practice, create an overlay
        scatterplot with educ as your dependent variable and maeduc (mother’s
        education) and paeduc (father’s education) as your independent variables.


       What are the null and alternate hypotheses?


       List the four assumptions for bivariate regression, how you tested them, and
       whether or not they were met.
               1.

               2.

               3.

               4.

       Report the correlation coefficient, coefficient of determination, and
       unstandardized coefficients. Note which statistics were significant at the p =.05
       level.


       Write the unstandardized regression equation and discuss your findings.



After explaining the above regression analysis to your partner, conduct a second bivariate
regression to see if education has any effect on how often the respondent does voluntary
work for a charity (volchrty). Follow the steps above. Note: volcharty should be
reversed so that higher values represent more frequent voluntary work. Your partner will
be doing a similar regression analysis using religious attendance as the independent
variable. After doing so, answer the following question with your partner.

       Recalling what we know about standardized coefficients, which has a bigger
       effect on the frequency of voluntary charitable work, education or religious
       attendance?

For fun, together conduct a multivariate regression analysis using both education and
religious attendance as your independent variables. Use charitable voluntary work as
your dependent variable. What did you find?