Extra Credit Project Paper DUE: Wednesday, Dec. 6, in class
For this assignment, you will pose a research question, identify a suitable data set with which to examine this question, develop hypotheses, plan appropriate analyses, execute them, and report them in a 4-7 page paper. This paper is worth up to 10% extra credit toward your course grade.
Your research question can be related to any topic you choose. Once you have thought of a few different research questions in which you are interested, go to the SDA web site and browse the different data sets' codebooks to find variables that relate to your research questions. You can get information on the research design and data collection methods used for a given survey by looking at the abstract ("Abst") on the main SDA archive page and in the introduction of a survey's codebook. Depending on your interests, you should also spend time looking at the "Other Archives" accessible from the SDA Archive page (remember to follow the directions the subsequent web pages give you). [Note that all of the data sets in the "Other Archives" (except the Italian Election Results and Surveys) have more limited versions of SDA which do not allow regression analysis (therefore, you couldn't examine the relationship between two interval scale variables in these data sets with SDA)]. In choosing your research question and variables to examine, make sure they are not the same or slight variations of exercises already presented in this course (including the text). In other words, be creative and explore something new!
If you prefer, you may use another data set that is not in an SDA-type archive. The data set would have to be small enough and/or formatted appropriately for you to enter or paste into WebStat. You could also compile a data set from published sources of official statistics about such units as countries, states, companies, etc. (e.g., in an almanac, encyclopedia, or reputable web site, magazine, or newspaper). If you choose this option, make sure to save your data in a file on your hard disk and floppy disk, so you don't have to re-enter them each time you work on your analysis.
Your research question and hypotheses should pertain to the relationship between two or more variables. Because all of the SDA data sets and almost any other data set you are likely to find are based on observational research designs, you should investigate the nature of the association between two variables -- that is, whether they seem to be causally linked, spuriously related (confounding factors), linked by an intervening variable, or related to different degrees depending on the value of a third variable (interactive or conditional relationship). This means that you should include at least 3 variables in your analysis, either as part of an elaboration/partial tables analysis (for categorical variables) or a series of bivariate correlation/regression analyses (for interval scale variables). If you are able to find a data set from an experimental study, then including additional variables in your analysis will not be necessary for this assignment.
To complete this assignment successfully, you need to see me soon for further instruction on evaluating different interpretations of associations in observational data. I will meet with students for this additional instruction only by appointment on or before November 20. If your project involves relationships between categorical variables, you will need to read the additional material about elaboration on reserve at the library.
Your analysis should begin with a thorough description of each variable separately, including the appropriate measures of central tendency and dispersion and characterization of the distribution shape (verbally and graphically). Draw histograms, boxplots, pie/bar charts, etc. by hand or with some other tool (e.g., Excel, etc.) based on the relevant output from SDA.
Your bivariate analyses should follow, complete with appropriate crosstabulations and percentaging and measures of association. If you are analyzing interval scale variables from an SDA-type archive data set, then you should ask me to produce scatterplots for you. I will produce your scatterplots if you provide me with the name of data set, web address (URL), variable names, designation of which variables should be on the x and y axes, and the subset of cases to be included in the analysis (if you are excluding some cases [e.g., excluding men for an analysis based on women only]). The latest date that I will accept such requests is November 22. I will return scatterplots to you within 7 days of receiving your request (i.e., no later than November 29).
If you run into situations in your analyses where we have not covered the appropriate technique (e.g., measure of association between a nominal and interval scale variable), you need to see me for further instruction. In some cases, the most practical strategy will be to recode a variable into a lower scale of measurement and use techniques you already know, but you should do this only after consulting with me.
Project plan DUE: Monday, November 13, in class
Before you begin your analysis, you must turn in a proposal or plan for your study. Once I approve your plan, you may begin your analysis. Your plan should be 1-3 paragraphs long. It should:
Do not do any analysis until your plan has been approved by me. This plan is worth 10 points.
Required components for paper:
Divide your paper into four sections: introduction, methods, results, and discussion. In the introduction, give the background to your analysis. Describe your research question(s) and hypothesis(es), including the rationale.
In the methods section, describe the data set, sampling procedures, and data collection methods. Be sure to cite the URL (web address) for the data set.
In the results section, report your findings fully in verbal, numerical, and graphical terms. Explain your quantitative results--don't expect that numbers alone will make the findings clear. Be sure to interpret your results and use them to evaluate your hypothesis(es) explicitly.
In the discussion section, briefly summarize what you found and offer your conclusions about your hypothesis(es). Discuss unexpected or interesting results and speculate about their meaning and significance. Suggest further analyses (e.g., specific other variables to consider, other analytic procedures to use) that could be done to investigate the research question(s) and hypothesis(es) further. That is, describe how you could test alternative explanations of your results. Review any limitations to your analysis that stem from research design and data collection methods issues, and propose other ways (designs and data collection methods) that might be used study the research question(s). Attach all SDA/WebStat output and numerical calculations as an appendix.
The paper should be 4-7 pages in length, including any graphs or tables, but excluding the appendix. The paper should be typed, double-spaced, spell-checked, and proofread. Cut out and paste SDA tables into your paper, or recreate them in your word processor--do not simply put one or several pages of output with irrelevant information in the middle of your paper.
If you have any questions about how to conduct or report your analysis project, please see me during office hours or by appointment.
Grading components
project plan - 10 pts.
execution of analysis (logic, appropriateness, completeness) - 50 pts.
reporting of analysis in paper (required components, interpretation, clarity of writing, grammar, punctuation) - 40 pts.
total = 100
No late papers will be accepted.