For this assignment, you will propose a research question, formulate a theory, derive a testable hypothesis, collect the data (you can find a dataset on your own or download Twitter data), and test your hypothesis (e.g., using a difference-of-means test between two groups). Here is an example for you to use as a template (You must think of your own research question, theory, hypothesis, etc. This example is solely to give you a template.) After reading the example, complete questions 1–8 below.
Suppose the research question is “What factors determine the policy agenda of world leaders?* You could then propose an “urgency” model of policy issue prioritization that goes something like this:
Every world leader has a vast agenda they want to implement. Implementing reform, however, is a very time-consuming process: leaders have to collect the necessary information, formulate a policy, and work with their party (and the opposition) to enact the policy into law. As a result, they may only be able to accomplish a handful of specific reforms throughout their term. In addition to pursuing their own policy goals, leaders also have to respond to major events outside of their control, e.g., wars, terrorism, economic crises. These exogenous events crowd out any pre-existing agenda because they often demand an immediate response. This means that world leaders who experience a major political crisis early in their term will have to shift their policy priorities from the issue they campaigned on to responding to the event.
Based on this theoretical model, we can derive the following testable research hypothesis:
World leader’s policy priorities will be defined by major political events early in their term.
We can test our hypothesis using the following research design. Our unit of analysis will be tweets by President Obama. Our independent variable—the defining political event early in the leader’s term—will be the economic recession during Obama’s first term (Obama’s second term did not include a similar defining event, so we will treat his second terms as a control group). We will analyze substantive words in the President’s tweets, during each term, as well as prior to his election (to measure his policy priorities). If our hypothesis is true, we will see no meaningful difference in tweet content prior to 2008 and the second term — as he could focus on his campaign agenda. We will, however, see a difference in topics between the campaign period and the first terms, as the first term’s agenda was crowded out by the financial crises.
1. Research Question [10 words]: State a general research question that you can answer by using one of the tools learned in class (network analysis or Twitter data). (Note how, in the above example, the research question and the theory are general, as they do not name specific leaders or policy areas. Also note that a research question must be a “why” question, i.e., do not ask yes/no questions).
2. Theory [150 words]: Propose a logically consistent theoretical model that answers your question. Make sure to explicitly state all your assumptions. (Note that it is more important that your theory is logically consistent than correct. If you cannot fit within the word limit, simplify your theory.)
3. Make a causal chart that represents your theory. Use squares to represent the independent variables and ovals to represent the dependent variables. See the below example.
4. Hypothesis [15 words]: Derive a prediction from your theory. Make sure that your hypothesis includes the independent variable, the dependent variable, and specifies the relationship between them. In the above example, the independent variable is the presence of a critical event, the dependent variable is the content of the policy agenda, and the relationship is “when a critical event is present, it dominates the policy agenda; when no critical event occurs, the leader can shape her own agenda.”
5. Data [100 words]: You must use at least one of the tools learned in class to test your hypothesis. Explicitly state which tool(s) you are using. Describe the data will you use to answer your question. Where will you obtain the data (e.g., website)? What is the unit of analysis? How will you measure your independent and dependent variable?
6. Download/save the data. Clean up the data and perform your analysis.
7. Results [100 words plus any graphs, figures]: Note: all figures and tables must be formatted, do not include R code or output. All tables and figures must be described in text. Present the results of your analysis (e.g., tables, graphs). Describe the results.
8. Did you get the answer you expected? Why or why not? [100 words]
Note: The word limits indicate strict maxima (any extra words will not be treated as a part of your answer). You must submit (1) the data you used (a zip file if large) (3) a file that contains detailed answers to each of the above questions (MS Word, pdf, etc.). Please put the word counts next to each answer.