a) Think... Ruminate... Ponder... Speculate... Imagine... Introspect... Extrapolate.. Question...Wonder.. .Bounce your thoughts off the wall...until you have comfortably identified a research topic that you wish to investigate through secondary data analysis. Once you have identified what exactly your topic is, briefly explain why this is an important area of inquiry and what it that you hope to learn from your investigations.

b) Visit Sociological Abstracts or Ebsco.host in Murphy Library and find at least three recent studies that have at least some relevance (they need not be on exactly the same topic but should be somehow related) to your research topic. Briefly summarize each of these studies as they pertain to your research topic. There should be at least some degree of overlap between this research literature and the set of independent variables that you adopt in your research study (see part d below).

c) Next, describe your dependent variable(s). Since the dependent variable is such an important part of your study, copy the exact questions that you adopted as its operational definition. Explain precisely how you went about recoding the data for these particular questions and how you combined these items together, if indeed you did so, to create a multiple item indicator.

d) Next, identify at least ten independent variables that you believe will be significantly correlated with your dependent variable. Indicate your rationale for why you believe that each of these independent variables will be causally linked to you dependent variable. Also identify the survey question that corresponds to each of your independent variables. In addition, be sure to incorporate at least one multiple-item indicator among your set of independent variables. Explain how and why you went about combining these items into a single index.

e) Next, discuss the strengths and weaknesses of survey research methodology that are pertinent to investigating your specific research topic. In particular discuss any potential problems of bias in the validity of the data you are evaluating.

f) Next, select any one of your statistically significant correlations between one of your independent variables and your dependent variable. Now discuss why it would be appropriate to introduce a control variable to test how the strength of your previously identified bivariate correlation might change after this control variable is introduced. There are various ways to perform this statistical control function. However, I would recommendthat you do either a partial-order correlation or a multiple regression--with only two independent variables--realizing, of course, that both of these methods presume that you are working wiht interval level data for all variables. Finally, interpret the meaning of this statistical control. Specifically, does the original correlation remain significant or not? And supposing that it does remain statistically significant, does the magnitude of the bivariate correlation diminish, stay the same, or increase? Whether you choose to use a partial-order correlation coefficient or a beta coefficient (with multiple regression), in either case you will need to compare this resultant coefficient with the original bivariate correlation coefficient.

g) What type of sample is it that you are analyzing? To what population group may your study findings be generalized?

h) Discuss your specific study findings and please attach your computer printout (as an appendix) from which your findings are drawn. Which of your independent variables were significantly associated with your dependent variable and which are not? Be sure to show the actual magnitudes of association for all of your statistical associations as well. This means that you must type the actual correlation coefficient with either a positive (+) or negative (-) sign into the main body of your research report. This is the only way the reader has of knowing how strongly any of your independent variables is associated with your dependent variable.

i) Do a multiple regression analysis with all your previously identified independent variables and your selected dependent variable. How much total variance (R²) in your dependent variable are you able to explain? Which three of your independent variables explain the most unique variance in your dependent variable [Attach the one page computer printout of your multiple regression data analysis in this section of your term project report.]

j) What does it all mean? Relate your findings to some social scientific theoretical orientation. In other words, explain how you might interpret your findings at a more abstract level. In short, what broader meaning might your findings have? This will, of course, require some speculation your part.