Be as complete as possible. Think very carefully about your model specification. You will be graded not only on how successfully you run and interpret the regression and the diagnostics but on how well you specify a model as well.

1. Your DV is going to be feelings toward Bush. This is a thermometer rating running from 0 (very "cold" feelings) to 100 (very "warm" feelings), with 50 being the neutral point. Use the variable bushT .
   Your independent variables will be level of attitudes on defense spending, Bush’s tax cuts, abortion, and gun control. See the codebook for measures.
   

1. Write down a specific Research Question relevant to this design.
2. In a few sentences explain the importance of this research.
3. Explain each of the independent variables was chosen. What are your expectations about their effect on the DV? State your clear hypotheses.
4. Run the regression. Interpret the results substantively and thoroughly. What can you conclude from the regression you run? Look at the coefficients (constant, slopes) and the Beta weights. Remember that this interpretation involves testing the significance of coefficients. Which variables are significant, which are not significant? Are there any surprises? Any coefficients with an unexpected sign, magnitude, significance? What is the model fit (and its corresponding significance)? Based on the regression, what are your conclusions regarding your hypotheses?

1. Do you detect (a) multicollinearity? (b) heteroskedasticity? (c) non normal error distribution in your model? Provide empirical evidence for whether or not it was violated in your regression model, based on graphs and other diagnostics tests you run. How can you correct for each? Correct for the violations you detect, report on whatever you chose to do. Report robust (heteroskedasticity corrected) significance tests and interpret them.
   
   b) In the second model, you will add gender and an interaction between gender and attitudes on abortion to the previous model.

1. A researcher thought that the effect of abortion attitudes on feelings for Bush is probably stronger for women. Theoretically explain this hypothesis.
2. Empirically test the researcher’s expectation: run the same model you used so far, but add a) gender, and b) the interaction between gender and abortion attitudes.
3. Thoroughly interpret the results of the model in (7). What do you conclude regarding the researcher’s argument?

Price: $18.94

Solution: The downloadable solution consists of 8 pages, 1094 words and 5 charts.
Deliverable: Word Document