SOC 206 **Quantitative Methods II **

Akos Rona-Tas

Office: SSB 488

Phone: 534-4699

Office Hours:

MW 11:30-12:20

or by appointment

Email: aronatas@ucsd.edu

In this course we will cover multivariate linear regression, and some of its
extensions. I assume that you have taken SOC 205 or an equivalent course,
and you are familiar with simple regression analysis. The course
will focus on causal models and conceptual issues, and it will de-emphasize
mathematical technicalities. We will be using STATA for data analysis.

My goal is not to turn you into statisticians. I will consider my course a success if at the end of the course you will be competent users of these techniques, and fearless readers of articles and books using these statistical tools. I also hope that you will be able to discuss critically and intelligently quantitative works from the discipline.

Apart from the textbook. Agresti
and Finlay, *Statistical Methods for the Social Sciences,* there is one
other book you will need. Tim Futing Liao, *Interpreting
Probability Models: Logit, Probit and Other
Generalized Linear Models. *The books are available in the mailroom of the
Sociology Dept.

There are a few other
articles for this course to read. Most of those are available through JSTOR or
other electronic databases you can access through the library.

We have three thematic parts (linear regression models, structural equation models and probability models) broken up into nine two day segments. The items in the Reading section must be completed before the first date of the segment and the items in the Application section before the second date.

You will have four assignments all downloadable from the course website.

http://pages.ucsd.edu/~aronatas/soc20602.html

All of you will have to make two presentations. One will be one of your assignments that you have to present at the seminar before it is due. The second is a presentation of one of the articles from the Application section that has an *. There is a final research paper where you test some theory using multivariate analysis. Each assignment will count as 15% of your grade and the final paper will count for 30%, class participation including the presentations will make up the rest.

January 5, 7

**LINEAR MODELS**

**Causation and the
Logic of Multivariate Analysis**

**Review of OLS
Regression **

Reading: A&F, Ch. 9, 10, 11

Dawes, Faust and Meehl,
Clinical versus Actuarial Judgment. *Science*
Volume 243, Issue 4899 March 31, 1989, 1668-1674.

Lave, Charles A. and James G. March. An Introduction to Models in the Social Sciences. Chapter 1,2 (copies available in the mailroom)

**Application:**

__The Great Chocolate Debate__

Maurage, Pierre,
Alexandre Heeren, and Mauro Pesenti. Does Chocolate Consumption Really
Boost Nobel Award Chances? The Peril of Over-Interpreting
Correlations in Health Studies. *The Journal of Nutrition*, April 24, 2013

__The Rossi - Zeisel Debate__

Berk, Richard A., Kenneth J. Lenihan, and Peter H. Rossi. Crime and Poverty: Some
Experimental Evidence from Ex-Offenders. *ASR*. Vol.45. No.5. pp.766-786.

Zeisel, Hans. Disagreement
over the Evaluation of a Controlled Experiment. *AJS*.
Vol.88. No.2. pp.378-389.

Rossi, Peter
H., Richard A. Berk and Kenneth J. Lenihan. Saying It Wrong with
Figures: A Comment on Zeisel. *AJS *Vol.88.
No.2. pp.390-393.

Zeisel,
Hans. Hans Zeisel Concludes the Debate. *AJS*. Vol.88.
No.2. pp.394-396.

Further

Dawes, Robyn. 1979. The Robust Beauty of Improper Linear
Models in Decision Making*. American
Psychologist*, July, pp.571-582

H.M. Blalock, *Causal
Inferences in Nonexperimental Research*__.__

H. Smith, Specification Problems
in Experimental and Nonexperimental Social Research,
In *Sociological Methodology* 1990, 20:59-91.

https://www.youtube.com/watch?v=CCiIfjm8qjw

January 12,14

**Model**** Building**

Reading : A&F, Ch. 9, 10, 11

Berk, Richard A. An Introduction
to Sample Selection Bias in Sociological Data, *ASR* vol. 48, 1983

**Application:**

Further Reading:

Winship, Christopher and Robert Mare. Models for
Sample Selection Bias*.** Annual Review of Sociology*, Vol. 18
(1992), pp. 327-350

Ross M. Stolzenberg and Daniel A. Relles. Tools for Intuition about Sample Selection Bias and Its Correction. ASR, Vol. 62, No. 3 (1997), pp. 494-507

Antonakis, John and Samuel Bendahan, Philippe Jacquart, Rafael Lalive. On making causal claims: A review and recommendations. The Leadership Quarterly 21 (2010) 1086–1120 (A short and simplified version of this article is presented here with some cheesy animation by Antonakis https://www.youtube.com/watch?v=CCiIfjm8qjw .)

Gerber,
Theodore. Membership Benefits or Selection Effects?
Why Former Communist Party Members Do Better in Post-Soviet Russia. *Social Science Research*, 2000, pp. 25–50

Rona-Tas, Akos and Alya Guseva.
The Privileges of Past Communist Party Membership in Russia
and Endogenous Switching Regression. *Social
Science Research*, 2001, pp. 641-652

January 19 Martin Luther King Day

January 21, 26

**Generalized
Linear Models**

**Non-linearity,
non-additivity, heteroscedasticity, count data, **

**Application:**

January 28, February TBD

**STRUCTURAL
EQUATION MODELS**

**Path-Analysis **

**Latent Variable
Models**

*1st Assignment
due** January 28*

A&F pp.527-541

**Application:**

Otis Dudley Duncan, Path Analysis: Sociological Examples. *AJS*,
vol. 72, 1966

*Prudence A. Widlak and Carolyn C. Perrucci, Family Configuration, Family Interaction, and Intellectual Attainment, Journal of Marriage and Family, Vol. 50, No. 1 (Feb., 1988), pp. 33-44

Further

O.D. Duncan, *Introduction to Structural Equation Models*.

P.M. Blau and O.D. Duncan, *The
American Occupational Structure*.

February TBD, February 9

**Latent Variable
Models (continued)**

A&F pp.527-541

**Application:**

*Leahey,
Erin. 2007. Not by Productivity Alone: How Visibility and
Specialization Contribute to Academic Earnings, ASR vol 72, (August) pp.533-561

*David, John Frank, John W. Meyer and David Miyahara. 1995. The Individualist
Polity and the Prevalence of Professionalized Psychology: A Cross-National
Study. *ASR* vol.60 (June) pp.360-377

Further

John C. Loehlin*, Latent
Variable Models: An Introduction to Factor, Path, and Structural Analysis. *

Bollen, Kenneth A.. *Structural equations with latent variables.*

Structural Equation Models with Latent Variables slides

** **

February 16 Presidents’ Day

February 11, 18

PROBABILITY MODELS

**Dichotomous
Dependent Variable**

**Logistic
Regression (Logit) and Probit**

*2 ^{nd}
Assignment due February11*

A&F, Chapter 15

Liao, Chapters 1, 2, 3

**Application:**

*Rona-Tas, Akos.
1994. The First Shall Be Last? *American Journal of Sociology *100/1:40-69

Further

Long, J. Scott. 1997. *Regression Models for Categorical
and Limited Dependent Variables.*

Output 1: Introducing Logistic Regression

Output 2: Explaining fit using log-likelihood

February 23, 25

**Polytomous****
Dependent Variable**

**Multinomial Logit**

Liao, Chapter 6

**Application:**

*Tak Wing Chan and John H. Goldthorpe, Social Status and Newspaper Readership' The American Journal of Sociology, Vol. 112, No. 4 (Jan., 2007), pp. 1095-1134

Output: Multinomial and ordered logit

March 2, 4

**Ordered Logit**

*3 ^{rd}
Assignment due March 2*

Liao, Chapter 4,5,

**Application:**

*Cech, Erin ,
Brian Rubineau, Susan Silbey
and Caroll Seron. 2011. Professional
Role Confidence and Gendered Persistence in Engineering. *American Sociological Review*, 76/5 pp. 641-666

Further

C. Winship
and R.D. Mare, Regression Models with Ordinal Variables, *ASR*, 1984
August, 49:512-525.

Allison, Discrete-time Methods
for the Analysis of Event Histories, in *Sociological Methodology*,
1982,

Yamaguchi, Kazuo, *Event History Analysis*,

March 9, 11

**Review**

*4th Assignment
due March 11*

John P. A. Ioannidis. 2005. Why Most Published Research Findings Are False. PLoS Med 2(8): e124. doi:10.1371/journal.pmed.0020124

Charles Tilly. 2008. *Why? What Happens When People Give
Reasons... and Why*. Princeton UP

Adrian E. Raftery. 2001.
Statistics in Sociology, 1950-2000: A Selective Review, *Sociological Methodology*,
Vol. 31. (2001), pp. 1-45.

Andrew Abbott, Transcending General Linear Reality. *Sociological
Theory*, vol 6:2 1988

**Assignments:**

Three variable regression. Plot, residuals.

Structural Equations

Probability Models.

Data Analysis for Final Paper.

**Final Paper:**

Construct multivariate models to test some hypothesis. The paper should follow the ASR format, roughly, this outline:

Abstract

Introduction that states the problem

Literature review reflecting some familiarity with the literature outlining alternative theories

Your own theory and the derived hypotheses

Data and Method

Discussion of Findings

Conclusion

References (ASR format)

The paper should be 15- 25 pages in length without the tables and the references. Please use a 12 point font, double-space and number each page and submit the paper as an e-mail attachment.

*Final Paper Due
03/20/2015 12:00 am*