The purpose of this course is to provide an introduction to research methods and elementary quantitative data analysis for students who do not have any experience with data analysis, statistics, or social-science computing. If you've had a statistics course and forgotten most of it (or hated it), this course may be right for you as well. If you are math phobic but understand that knowledge of quantitative data analysis constitutes the coin of the realm in many fields today, this could may also be for you. And if you want to be able to understand the argument in political science articles that appear in professional journals, this course may also be for you.

The course is organized around practical applications using the kinds of data practicing political scientists use in their research. You will also be asked to apply these applications to a research project of your own. My approach is to emphasize these practical applications in order to help students understand how to evaluate evidence. I do not focus on mathematical derivations or statistical theory for two very practical reasons. First, I am not a mathematician and do not understand these derivations and theories well enough for my own comfort, much less well enough to teach them. Second, I do not think that most political science students will learn methods for methods sake, but rather that you will learn these methods if I can show that they will be useful as you undertake interesting projects. I also believe students learn by doing, so there will be a heavy emphasis on applying the techniques to research problems. The reading for the course is not heavy, but the "laboratory" part of the class will take some time and effort.

The benefit of taking a course like this is that you acquire practical skills that can be used in a variety of contexts. These skills will certainly be relevant to research you might do in other political science or social science classes. They also are regularly used by practicing policy analysts, researchers, political consultants, journalists, and decision makers outside of academia. In fact, one book on the reading list shows how a mathematician might read the newspaper——and should give you a better understanding of why this course is useful. Because social science research is increasingly important in the world today, the skills we will study are also useful——some would say necessary——for the educated citizen to be able critically to evaluate information and conclusions offered in the media, by government, and by other interested parties to public debate. None of you should be surprised if perspective employers also want to know the extent to which you can work with these tools.

I assume no prior computer or data analysis experience. The mathematical concepts we encounter are no more complicated than simple arithmetic and percentages. But the techniques are themselves powerful, and they are fundamental to much more elaborate and complex methods that can require years to master.

This is not a course in which you can fall behind. It is not a course in which you can be a passive learner. You must come to each class having done the reading and having thought about the critical questions raised in the reading. I will expect you to discuss this material from an informed perspective.

You also must do each of the lab exercises by their due date; or they will not be graded. I do not impose this requirement to be punitive but rather because learning in a course like this is definitely cumulative. You will fall hopelessly behind if you do not keep up.

Plan way ahead for this course. Computer glitches occur. Projects are lost; print each of your outputs; goblins attack at random. None of these excuses——not even that the dog ate your papers——will be accepted.

Your major output for this course will be a research paper in which you demonstrate your skill in using quantitative methods. One way in which you will know that you understand the concepts involved here will be when you understand that the question, "How many pages should it be?" is irrelevant. Your paper must present an interesting question for analysis, describe a research design appropriate to answer the question, explain the data used in your analysis and why it is appropriate, analyze the data in order to answer the question, and discuss your results. You may "double count" this paper with one from another course with the written permission of the instructor in the other course.

Your grade will be determined according to the following schedule:

Class participation 20%

Lab exercises 30%

Research paper 50%

The following books will be read in full:

Larry Gonick and Woollcott Smith, The Cartoon Guide to Statistics

Kenneth Hoover and Todd Donovan, The Elements of Social Scientific

Thinking, 6^th edition

John Allen Paulos, A Mathematician Reads the Newspaper

Charles Prysby and Carmine Scavo, Voting Behavior: The 1996 Election

W. Phillips Shively, The Craft of Political Research, 4^th edition

Other articles are listed in the syllabus.

This is clearly a course in which the instructor must keep the pace of the course in sync with what the students are absorbing. Thus, the schedule below is tentative. At the end of each week, I will give you the next deadlines for laboratory assignments.

1. Introduction: What is political research? Why is this discipline called "political science"? (one week)

Topics: the nature of systematic research

the elements of science in political science research

the ethics of scientific research

Readings: Shively, Chapters 1, 2 ; Hoover and Donovan, Chapters 1, 2

2. What kinds of theories do political scientists test? (one week)

Topics: Federalist No. 10 as political theory

explanatory versus causal theories

precision in language in theoretical statements

Readings: Shively, Chapter 3

Gonick and Smith, Chapters 1, 2

Prysby and Scavo, Chapter 4

Lab 1: Introduction to SPSS; accessing data sets; Crosstabs

3. Political behavior: are people self-interested? An example (two weeks)

Topics: how do citizens define their self-interest

how do people decide for whom to vote

Readings: Prysby and Scavo, Chapters 1-3

Stone and Davis, "An Introduction to Quantitative

Research," on course website

Miller et al., "Systematic Assessments of Presidential

Candidates," American Political Science Review

80:521(1986)

Lab 2: Recoding; careful labeling; selecting cases

Testing the relationship between two variables

Exploring the concept of control

 

4. Research design: post-hoc studies versus experimental designs

(two to three weeks)

Topics: different types of research designs

reality testing

operationalizing variables

reliability and validity

holding a variable constant

causal thinking

Readings: Shively, Chapters 4-6; Hoover and Donovan, Chapter 3

Gonick and Smith, Chapters 4, 5, 8, and 10

Paulos, begin here and read throughout the course; not

bad night-stand reading

review Prysby and Scavo, Chapter 4

Cover and Blumberg, "Baby Books and Ballots: The

Impact of Congressional Mail on Constituent

Opinion," American Political Science Review

76:347 (1982)

Maisel and Stone, "The Politics of Government-Funded

Research: Notes from the Experience of the

Candidate Emergence Study," PS, December 1998, p. 811

Lab 3: Review of data sets that are available

Individual meetings to discuss research proposals

Transforming variables; COMPUTE and COUNT

From this point in the course on, students will be working on their own research papers. Some of you will use more sophisticated quantitative techniques than others. However, you should all be exposed to at least elementary means of data analysis. Thus, in class we will review the techniques listed below. We will concentrate on those that you are using in your own research, but you should be familiar enough with all of these to read articles in which they are used.

 

5. Elementary data analysis (six weeks)

Topics:

1. logic of causal analysis

Readings: Hoover and Donovan, Chapter 4

Review Prysby and Scavo, Chapter 4 again

Lab 4: data management

testing a causal theory

random samples

bivariate and multivariate relationships

2. inference

Readings: Shively, Chapter 9

Gonick and Smith, Chapters 3, 6, 7, and 9

3. univariate statistics

Readings: Shively, Chapters 7-8

Gonick and Smith, Chapter 12

4. regression analysis
5. bivariate and multivariate regressions