answer the question from the book

Key Terms and Concepts (Johnson and Reynolds chapters 8-15)[footnoteRef:1] [1: Although the final exam emphasizes the post-midterm material, students should recognize that the material covered in the second half of the semester is built upon the material covered in the first half. Therefore, it would be a wise approach to review the midterm study guide. ]

Be prepared to define, explain the significance—and, where appropriate, provide examples—of the following:

Observation—including direct, overt, covert, participant, and non-participant

Indirection observation—including erosion and accretion measures

Interviewing/open-ended questions

Content analysis

Primary data

Secondary data

Ethical considerations of research (e.g., informed consent)

Institutional Review Board

Notes—including mental, jotted, and field

Ethnography/”going native”/field study

Non-reactive methods/reactive methods

Survey research/closed-ended questions

Reliability, as it relates to the various methods we covered (especially intercoder)

Validity, as it relates to the various methods we covered

Question types/problems—including leading, push, double-barreled, ambiguity, and slang

Written record (episodic and running)—including advantages and disadvantages of each

Frequency distribution

Data matrix

Variables (independent and dependent)

Measures of central tendency

Measures of dispersion, especially standard deviation

Normal distribution

Inferential statistics/logic of statistical inference

Null hypothesis/alternative hypothesis

Type I error/Type II error

Sampling distribution

Levels of measurement

Bivariate analysis, including types and direction of relationships (e.g., linear; positive/negative)

Measures of association

Coefficients (e.g., beta)—including interpretation

Statistical significance

Multivariate analysis/regression, including advantages of bivariate analysis (e.g., causation)

Assumptions of linear models

Dummy variable

Error, as it relates to bi- and multivariate analysis

Reactivity

Structured/unstructured observation

Central limit theorem

Essay Questions

Be sure to address every element—as completely and clearly as you can—of the writing prompts, giving examples, when appropriate, to illustrate your arguments.

1) Benjamin Disraeli, a former 19th century British Prime Minister, is credited for having stated, “There are three types of lies: lies, damned lies, and statistics.” What did he mean? How does, for instance, the film Who’s Counting? illustrate Disraeli’s assertion? How could inferential statistical, bivariate, and multivariate analysis be “misused”? How could measures of central tendency be “misused”? Consider mode, median and mean, in particular; which is much more vulnerable to skewing by large outliers? Be sure to address sampling considerations (e.g., sampling error) for inferential statistical analysis, and causation and spuriousness for bivariate and multivariate analysis. Why does it all matter?

2) The aim of this semester has been to develop a familiarity with tools and concepts that allow students of political science to bring evidence to bear on questions about politics and public policy. Furthermore, familiarity with the tools and concepts covered in this course should enable students of political science to examine social science research with a critical eye. Select three of the following methods: observation, interviewing, survey research, or content analysis—and describe the distinct disadvantages and advantages proffered by each. Draw comparisons. Include discussions of validity and reliability. Is social science research truly “value-neutral”? Should it be? Why? Give an example, using one of the methods above, of how a researcher might address the following question: To what extent were U.S. presidents in the 20th century concerned about the threat of war? How do you know? Be specific! Why are the methods you did not select less useful, in your estimation?

Note: Students should also be able to interpret SPSS output, specifically r squared, statistical significance, and unstandardized beta.