Choose a topic that interests you that you can explore either with the 2006 GSS. You should have one primary "dependent" variable that you are interested in (Y). This variable should be at the interval-ratio level of measurement. Choose at least 3 independent variables that you think may be important for your research topic. Two of these should be interval-ratio (X1 and X2), and the third should be categorical (X3). You may create alternative versions of any of these by recoding the variables originally on the file as long as you specify what you are doing. Next, carry out a mini-research project using this set of variables. Carry out statistical analyses and then write up your results in a few paragraphs according to the instructions below.
Step 1: Instructions for data analyses
1. On your SPSS output, label EACH statistical analysis listed below according to the question number so that I can tell what I'm supposed to be looking at. Highlight the labels with a highlighter or circle with a colored-ink pen before turning in. Do not turn in any irrelevant output (i.e., anything that you wouldn't want to include in your write-up).There will be points off for not clearly labeling your output.
2. Run appropriate descriptive statistics for each variable (i.e., means, medians, and ranges for interval-ratio variables, frequency tables for categorical variables, any others you think would be interesting or helpful). For frequencies be sure to calculate percents as well as cell counts.
3. Examine the relationship between your categorical variable and your dependent variable.
a. Conduct a t-test for differences in means. If your categorical variable has more than two categories, choose a single "reference" category. Then compare the mean of Y in the reference category to the mean of Y in each of the other categories in turn. (Thus if you have three categories you will do two t-tests, if you have four categories you will do three t-tests, etc.)
b. Code the categorical independent variable into a dummy variable or series of dummy variables, as appropriate. Then run a regression model with Y as the dependent variable and the appropriate number of dummy variables as your only independent variables. When you interpret this, compare both your coefficients and the results of your significance tests with part a).
4. Run a correlation matrix between the three (2 independent, 1 dependent) interval-ratio variables.
5. Run four more regression models:
c. Y as a function of X1. Interpret.
d. Y as a function of X2. Interpret.
e. Y as a function of both X1 and X2. Discuss the changes in coefficients, if any, and what they tell you.
f. Y as a function of X1, X2, and the dummy variable(s) representing X3. Interpret.
Step 2: Additional instructions for Write-up:
1. Your write-up should be in a separate, typed document, not in your SPSS output.
2. Introduce the research topic in a sentence or two. Briefly discuss your rationale for choosing the particular independent variables you used (this could be because you think there is a theoretical relationship or because you simply think they could be important variables to "control" for).
3. Explain how your outcome variable is measured. Explain each of the independent variables you use in turn, including a discussion of whether you transformed any of them or created any new variables.
4. Write up your interpretation ofALL the results you generated above, paying attention to any specific instructions for interpreting each part. Also, be sure that you are always clearly identifying what type of test or model you're talking about and what it tells you about the relationships between the variables, using language appropriate for that particular model.THERE SHOULD BE NO STATISTICAL ANALYSES THAT ARE NOT WRITTEN ABOUT IN THE WRITE-UP SECTION. No credit will be given for something that appears in your SPSS output but not in your write-up.
5. Write a brief conclusion about what you have learned about your topic from this analysis.