1. Give an example of a hypothesis from one of the subfields of political science. In a couple sentences, explain the hypothesis, especially the relevant causal relationships.
2. Starting with the hypothesis just provided, select one concept contained in it and explain the concept in more detail. Provide a formal definition.
3. Continuing with this same concept, provide an operational definition. Identify two or three of the concept's empirical manifestations. Are these empirical manifestations best treated as alternative measures of the concept, or should they be combined into a single composite measure? Are there quantitative indicators that you suspect are available for purposes of measurement? Comment on the most likely sources of measurement error in your operationalization of the concept, addressing issues of validity and reliability.
4. Using data in one of the class datasets, construct and compute a some meaningful composite index or scale. This need not be the measure you describe in parts 2 an 3. As a "paper-and-pencil" illustration, show how the measure is computed using actual numbers for one observation. Does the result make sense for that observation? Save (but don't print) the datafile with your computed index.
1. You may work in pairs. Create a survey consisting of 10-15 questions designed to measure opinions on a political or social issue of interest to you. Some possibilities include:
2. Administer the survey to 30-40 individuals. Create an SPSS datafile consisting of the responses. Each individual will be represented as a case (or line of data), with the variables (or columns) corresponding to the responses to each question in the survey. Create labels for the variables, and for each variable create labels for the values that the variable can take.
3. Select one opinion measured by your survey and create a chart showing how that opinion is distributed over some demographic characteristic represented in your sample (e.g., income, age, education). Adhere to the principles of effective visual display of quantitative information.
4. Along with the questionnaire, data printout, and chart, submit a brief discussion of one or more hypotheses that the survey results might be used to test. What is the most likely source of selection bias in the results you obtained?
For this assignment, use either the dataset you created in exercise 2 or apo94.sav. Select two categorical variables that you think may be related and do the following:
1. Comment on the hypothesized relationship between them? Is this a causal relationship? Identify the dependent variable and the independent variable.
2. Create a dummy variable for a particular response (or responses), and compute the mean of that dummy variable for two values of the independent variable. (You may select two values, or recode multiple values into just two values.) Is the difference in means statistically signficant? Does there appear to a be relationship between the two variables based on this analysis?
3. Construct a contingency table. Here you are not restricted to a single response (or combined responses), nor are you limited to two values of the independent variable. Use your wordprocessor to create the table; do not copy and paste unedited SPSS output.
4. Interpret the contingency table. How do the observed frequencies compare to the expected frequencies? Do the results of this analysis suggest a relationship between the variables, and is this relationship statistically significant? State the overall conclusion that you draw from this analysis. Is there support for your hypothesis?