Assignment Goal: This assignment asks you to develop a testable research question and hypothesis and apply the appropriate statistical techniques to test your hypothesis. You will use the scientific research process to answer a question that is interesting to you and determine whether two concepts are statistically related to one another. You will incorporate class concepts throughout the paper and interpret the statistics and models that you produce. Assignment Specifics: The set-up of the paper should be as follows: Introduction (1 page, broken into two paragraphs): The introduction should begin with a statement of your topic, why it is important, what your research questions are (conceptual and concrete), and your theory about how your variables are related. In other words, whats your causal explanation? What is your reasoning behind thinking that the independent variable is going to cause the dependent variable? In the second paragraph, you should describe your data source, variables (identify the independent and dependent variables and their levels of measurement), and state a testable hypothesis (based on how the variables are measured in the dataset). In other words, in this section, I want you to go from conceptual theory with your causal explanation to operationalization to stating your testable causal hypothesis. Proposed mediator (one or two paragraphs): Propose a plausible mediator of the relationship. This doesnt have to be a variable thats in the dataset (you wont be doing a statistical analysis to test whether the variable is, in fact, a mediator). Describe your reasoning behind proposing that the variable might mediate your hypothesized relationship between your independent variable and dependent variable. Note that your causal explanation (in the introduction section) may very well take you in the direction of a possible mediator. Proposed moderator (one or two paragraphs): Next, propose a plausible moderator of the relationship. This doesnt have to be a variable thats in the dataset (you wont be doing a statistical analysis to test whether the variable is, in fact, a moderator). Describe your reasoning behind proposing that the variable might moderate your hypothesized relationship between your independent variable and dependent variable. of data (1 page, plus a graph): In this section, provide descriptive statistics about your dependent and independent variables. Provide information on measures of central tendency and the dispersion (i.e., spread or variation of the values around the central tendency) of your independent and dependent variables. Also, provide a visual representation of your dependent variable and assess whether the distribution is symmetric or skewed. Bivariate hypothesis test (2 pages, plus tables and graphs): You should next move into a discussion of how your main independent variable of interest relates to your dependent variable. You should provide two bivariate hypothesis tests. First, turn both your independent variable and your dependent variable into 2-level categorical variables, and compute a X 2 test (chi-square test, also known as a tabular analysis). Second, using the original coding of the independent variable and dependent variable, conduct either at t-test (also known as a difference of means test) or a correlation, depending on how the measurement metric of your independent variable (hint: if its categorical, do a t-test; if its continuous do a correlation). Next, present a graphical representation of the relationship (from the t-test or correlation analysis; a bar graph, line graph, or scatterplot). You should describe the relationship between the variables, including direction, strength, linearity, etc. (if applicable with your levels of measurement) in the text of your paper. Bivariate regression analyses (1-2 pages, plus a table): Next, test your hypothesis using a bivariate regression. Present a table with the results of your regression analysis. Interpret the results, and assess whether your hypothesis is confirmed or not. Address the relationships between the independent variable and the dependent variable. In this section, also report and interpret the value of your R2. Are there omitted variables that you think may potentially serve as confounding variables in your established relationship? Briefly describe what you think might be a confound and how this may affect your findings. The four hurdles to establishing a causal relationship (1.5 pages): Now that youve conducted your analyses, evaluate your theory and evidence with regard to the four hurdles discussed in Kellstedt and Whitten (pg. 54-62). Describe what each hurdle is, and discuss whether you think your theory and findings pass each hurdle along the way In other words, given your theory, hypothesis, and results, how confident are you that theres a causal relationship between your independent variable and dependent variable? Conclusion (1 page): Your conclusion should consist of two parts. First, provide a one paragraph summary of your results. Include a discussion of whether or not a relationship exists between your independent and your dependent variables at the bivariate level (referring to both your bivariate tests and bivariate regression) and whether you suspect that the presence or absence of a relationship would remain after you controlled for possible confounds. Your second paragraph should assess how well your independent variables explain the variance in your dependent variable. Is the relationship between the independent and dependent variables strong or weak? If the relationship is weak, provide some reasoning for why it may be weak and some other factors that future researchers should consider. Ultimately, the first paragraph should assess whether your hypothesis is supported while the second paragraph should assess the strength of your evidence. Formatting: The paper should be in 12 point Times New Roman Font with 1 inch margins. The paper should be approximately 8-10 double-spaced pages, plus the graphs and tables, and the pages MUST be stapled. The paper cannot be longer than 13 pages of text (not including graphs and tables). A hard copy of the assignment needs to be handed in with the .do and .log files stapled to the back. The .do and .log files do not count in your page count. Electronic copies will not be accepted.
