Fall, 2001 (Roos, Soc. 311)

Final Paper

Due: Friday, December 14th (Roos office, A342 Lucy Stone Hall)

Your final paper should be on the order of 7-8 pages in length. It must not exceed 10 pages (not counting tables). For the final project you should choose the data analysis technique(s) and presentation mode(s) that best allow you to test your hypothesis. You should not do frequencies, crosstabs, means, and standard deviations. You should not make tables, draw histograms, line graphs, and circle graphs. That would be overkill and detract from the overall argument you are making. Instead, choose those techniques that best investigate the hypothesis you are testing. The point of this paper is to get you to make an argument in a concise and cogent manner, to test your theory using actual data (from the 1998 General Social Survey), and to present your results in an effective manner.

What to do: With the three variables you have been using, develop your hypothesis, and test it using the elaboration paradigm. This means you will be using the PROC FREQ procedure (see the SAS manual). You will probably need to recode at least some of your variables. (Make sure that you read the SAS manual, so that you know what you are doing). If your dependent variable can be an interval variable, you can do PROC MEANS--talk with me if you want to do this.

Important note: You will need to produce at least two tables: 1) the original relationship (the bivariate table showing the dependent variable by the independent variable), and 2) the elaborated relationship (the trivariate table showing the dependent variable by the intervening variable by the most antecedent variable). Make sure you put your variables in the correct order on the PROC FREQ statement (read the manual). Take the information from your output and refashion it to make two tables, one showing the original relationship and one showing the elaborated relationship (do not just paste the results from the computer output onto a sheet or paper, rewrite them into tabular form). Make correctly labeled and titled tables.

Supplement this basic analysis with other analyses or data presentation techniques to expand on the argument you are making. Use your imagination. The best grades will go to those who take the analysis beyond a mere discussion of the elaboration of their variables. One way to expand upon your analysis is to do a brief literature review--to do so go to the library and read several articles related to your dependent variable. As a first step, you can find some links to relevant articles from the "Links" option associated each of the GSS variables (for example, from the GSS website, click on "mnemonic" and then "abany" and then "Links" to get a bibliography for this abortion variable.

In writing your paper, follow the general format for research reports (see Appendix C in Babbie for some ideas). First, discuss the purpose of your study in an Introduction (what exactly are you studying and what hypothesis are you testing?). Researchers generally provide a brief review of the literature. Describe in detail why you expect to get the results you do. Make sure your arguments make sociological sense. This section should be combined with your introduction. If you do a literature review, put it in this section.

Second, in your Methods section, briefly describe your study design--what data are you using and how are you going to test your hypothesis? Remember, this is to let other researchers (who may want to replicate your study) know exactly how you went about testing your hypothesis. This is the section in which you describe the General Social Survey, how you coded all your variables (the conceptualization and operationalization), any recoding you did, any missing values you declared, and so forth.

Third, in an Analysis section present the results of your analysis and interpret your data. This section should be the longest since it is the guts of the report. You should provide your tables and graphs and describe your results. Number the tables and/or figures sequentially throughout the report. Relate your results back to your initial hypothesis? Is your hypothesis supported? rejected? If you reject your hypothesis, try to think through why it is wrong. Suggest what additional variables might be incorporated into the analysis in the future.

Finally, summarize the results of your study in a Conclusions section. Whether or not your hypothesis is supported, suggest where future researchers might go from here--what additional questions might be asked and answered to more fully understand the theoretical issues you have been investigating. If you cite any literature, include a set of references in a References section.

If you would like some examples of research reports, you can skim through sociology journals in the periodicals room (e.g., Social Forces, American Journal of Sociology, American Sociological Review, Social Problems).

DO NOT TURN IN YOUR OUTPUT--USE THE INFORMATION FROM YOUR OUTPUT TO MAKE TABLES AND INCLUDE THEM IN THE FINAL REPORT.