Empirical Research Methods II

Spring 2000

3- 4:20 T, H

Adamson Wing & Wean Hall 5419

 

George Loewenstein

Richard Scheines

 

Syllabus

 

http://www.andrew.cmu.edu/course/88-241/

 

1        Instructors

 

George Loewenstein

PH 319-D, x8-8787, GL20@andrew.cmu.edu

Office Hours: by appointment

 

Richard Scheines

BH 135, x8-8571, scheines@andrew.cmu.edu

Office Hours: by appointment

 

2        TA

 

Uri Simonsohn

Porter Hall 321

x8-2869

uws@andrew.cmu.edu

Office Hours: by appointment

 

 

3        Overview

 

This is a class on empirical research design and analysis, emphasizing causal inference.  Empirical research involves a number of broad steps, including formulating the problem, developing specific hypotheses, collecting data relevant to the hypotheses, analyzing the data using statistical tools, considering alternative explanations for the empirical findings, and collecting additional information with which to distinguish among these alternatives. 

A rough guide to the topics we will cover is as follows.

 

1.     Introduction and Motivation

2.     Collecting Data

- Qualitative Research: Focus Groups & Interviews

-  Surveys

3.     Representing and Analyzing Multivariate Relationships

4.     Specification Error

5.     Experimental Design and Methods

 

For each  topic, we will be studying the statistical tools germane to the topic as well as empirical studies illustrating the use of these tools.  In order to cover the statistical topics, we will be assigning portions of the main statistical text: Regression with Graphics, by Lawrence C. Hamilton, as well as The Cartoon Guide to Statistics, by Gonick and Smith. We will do several computer-labs with Minitab on both simulated and real data sets. 

A packet of reading materials supplementing the main texts will be available for purchase from Rosa Stipanovic, PH 319-F. The readings will also be available on reserve at Hunt Library.  In most weeks, there will be an assignment on the readings and questions from the lab. The reading assignments and their associated questions are due the day the readings will be covered, which will typically be on Tuesdays in the Adamson Wing of Baker Hall.  Turn these in as hard-copy, but also bring a copy to class, as your answers will be discussed there. They will be graded on a 0-100 point scale.  These assignments are meant to insure that you have thoughtfully engaged the material before the class.   Assignments will be posted on the web, and announced in class. 

Labs will take place on Thursdays during class time.  Some labs will be held in the Adamson Wing, and some in the computer cluster in Wean Hall 5419.  See the schedule of topics (on the web-site) for where class is being held.

There will be two in-class tests and a final exam. The exams will cover material similar to both the weekly questions on the readings and the weekly lab questions. The exams will be administered in the computer lab. 

There will also be five projects.  The first three will deal with interviews, focus groups, and surveys, the fourth will deal with statistical analysis of data, and the fifth will involve designing and conducting an experiment. 

 

 

4        Research Requirement. 

 

To learn first-hand about experimental methods, all students must complete a research participation requirement.  There are two alternatives for completing the requirement.  The first alternative is participation in 2 credits of SDS-sponsored experiments.  The second alternative is reading a pre-selected research article and writing short responses to questions about the research methods employed.  Within the first 2 weeks of the semester, you will receive detailed, written information on these options and have the opportunity to select one. Although your participation will not be formally graded, failure to satisfactorily complete this requirement by the last day of classes will result in a course grade of incomplete.

 

5        Required and Recommended Books

 

Required:

·       Regression with Graphics, Lawrence Hamilton

·       The Cartoon Guide to Statistics, Larry Gonick and Woollcott Smith

·       Individual Readings (Available for purchase from Rosa Stipanovic in Social and Decision Sciences, PH319-F, or on reserve in Hunt Library)

 

6        Grading

Grades will be based on assignments, two class tests, one final exam, five projects, and class participation, with weights as follows:

 

 

Weight for Final Grade

Reading & Lab Assignments

20%

Projects

20%

2 In Class Exams

15% each

Final Exam

20%

Class Participation

10%

 

Class participation will be factored into your grade as follows.  Students will often be called on randomly to answer questions in class.  If you are absent, you will get a 0. If you are unprepared, you will get a 1. If you are prepared, you will get a 2. 

Questions on the readings and labs will be graded on a numerical scale of 1-100.  A few questions will be randomly selected for grading, and the TA will assign a numerical score based on the answers to these questions.