SOCY 5021: DATA
ANALYSIS
SPRING 2006
M 3-5:50; KTCH 33
Instructor: Professor Jason Boardman Email:
boardman@colorado.edu
Office: 206A Ketchum Hall Phone:
303-735-0172
Office Hours: Friday 1-3 Fax:
303-492-8878
Texts
1) Allison,
Paul D. 1999. Multiple Regression: A Primer. Thousand Oaks, CA: Pine Forge Press.
2) Hoffmann,
John P. 2004. Generalized Linear Models: An Applied Approach. Boston: Pearson.
3) Kohler,
Ulrich, and Frauke Kreuter. 2005. Data Analysis Using Stata. College Station, TX: Stata Press.
Objectives
This course will introduce students to multivariate
statistical analysis and it will stress the application of these methods. I
will assume that all students have had an elementary statistics course at the
undergraduate level. Thus, I will review basic descriptive statistics and
inferential statistics but we will transition into regression techniques by the
third week of the course. The goal of the course is to teach students the
skills and concepts that are necessary to perform and interpret elementary,
intermediate, and some advanced statistical analyses in the Social Sciences. Accordingly,
the primary project of the class is to write an original research paper using
any of the techniques that we discuss in the class. We will work on these papers
throughout the course.
The course is divided into three sections: 1) introduction
to STATA; 2) regression analysis for continuous dependent variables; and 3)
regression analysis for categorical dependent variables. The bulk of our time
will be spent on the second section of the course in which we will thoroughly
examine all aspects of the OLS regression model (estimation, assumptions,
interpretation, and application). The last section of the class will extend
these models to binary, multinomial, and count dependent variables.
Assessment
Your final grade in the class will be based on the
following:
1) Participation
(10%): this includes class attendance and participation.
2) Assignments
(40%): assignments are typically made for each week of class involving the
analyses of data for your project.
3) Final
paper (50%): your paper must include-the following sections a) literature
review; b) research questions/hypotheses; c) methods (data, measures,
statistical model); d) findings; e) conclusions, f) references, and g)
tables/figures.
|
Date
|
Topic
|
|
1/16
|
NO CLASS-Martin Luther King Day
|
|
1/23
|
Data Analysis Using STATA
Topics:
·
What does a data set look like?
How to read data into STATA? Managing files in STATA (.do, .log, .dat).
·
Descriptive statistics;
Inferential statistics; Bivariate statistics
·
Sampling weights
Reading (to be read by class 1/23):
·
Kohler & Kreuter Ch. 1, Ch 3 (pp. 45-59), Ch. 4, Ch. 5 (pp. 75-80).
Assignment (Due next Monday, 1/30):
·
See handout and provide a brief
description of your research project (no more than 1-2 paragraphs).
|
|
1/30
|
Correlation and Bivariate Least Squares
Topics:
·
Model estimation/assumptions,
parameter estimates, standard error, Type I error, one-tail vs. two-tail test
Reading:
·
Kohler & Kreuter 177-192
·
Allison 97-108
Assignment
(Due 2/13 in class):
·
See handout.
|
|
2/6
|
Visually exploring data: figures, graphs, and
predicted values
Topics:
·
Predicted values, Scatter plots,
Histograms, Q-Q plots, Box-whisker plots
Reading (to be read by
class 1/30):
·
Kohler & Kreuter Ch 6-7
Assignment (Due 2/6 in
class)
·
See handout.
|
|
2/13
|
Measuring and interpreting the independent variable
Topics:
·
Scale (NOIR), dummy variables, non-linearity
and model fit, creating scales, Chronbach’s alpha
Reading:
·
Allison 156-159
·
Bryman and Cramer “Constructing
variables” * (CH2 Hardy and Bryman)
Assignment
(Due 2/20 in class):
·
See handout.
|
|
2/20
|
Multiple Regression
Topics
·
The logic and estimation of the
multiple regression model
Reading:
·
Allison Ch 1, 2, 4
·
Kohler & Kreuter 192-195
·
Hoffman Ch 1,2.
Assignment (due 2/27 in
class):
·
See handout
|
|
2/27
|
Assumptions and Problems
Topics:
·
Regression diagnostics
Reading:
·
Allison Ch. 3, 6, 7
·
Hoffmann pp. 10-17
·
Kohler & Kreuter 199-216
Assignment (due 3/6 in
class):
·
Draft of the literature
review for the paper. Roughly 5 pages
(double spaced). This should have a minimum of 5 papers (although it is not
uncommon to cite 10-20 papers in a literature review).
|
|
3/6
|
Model Building and Interaction terms
Topics:
·
Stepwise regression, standardized
coefficients, path analysis, using the NESTREG command in STATA, seemingly
unrelated estimation techniques (SUEST), substantive examples of
interactions/moderators, interactions for continuous and categorical
variables, separate models, fitted values
Reading:
·
Allison: 166-171
·
Hoffman pp. 17-19
·
Kohler & Kreuter 220-223
·
McLeod & Kaiser. 2004.
Childhood emotional and behavioral problems and educational attainment. American
Sociological Review, 69: 636-658.
·
Ellison, Christopher G., Jason
D. Boardman, David R. Williams, and James Jackson. 2001. “Religious
Participation and the Life-Stress Paradigm: Findings from the 1995 Detroit Area Study.” Social Forces 80:215-49.
Assignment (due 3/13 in
class):
·
see handout.
|
|
3/13
|
Generalized Linear Models
Topics:
·
The logic of the model, estimating
the model
Readings:
·
Hoffman: Ch 2
·
Allison Ch. 9
Assignment (due 3/20 in
class):
·
Handout and draft of research
questions and hypotheses.
|
|
3/20
|
Linear Probability, Logistic, Probit Regression
Topics:
·
Model specification, interpreting
the results
Reading:
·
Kohler & Kreuter
246-249
·
Hoffman Ch. 3
·
Boardman, Jason D. 2004. “Health
pessimism among black and white adults: The role of interpersonal and
institutional maltreatment.” Social Science & Medicine 59:
2523-2533.
Assignment (due in class
4/3):
·
see handout.
|
|
3/27
|
NO CLASS-Spring Break
|
|
4/3
|
Ordered Logistic and Ordered Probit Regression
Topics:
·
Model specification,
interpreting the results
Reading:
·
Hoffman Ch. 4
·
Keene & Reynolds. 2005. The
Job Costs of Family Demands: Gender Differences in Negative Family-to-Work
Spillover. Journal of Family Issues, 26 (3):275-299
Assignment (due in class
4/10):
·
Draft of the Methods section
of your paper. This should include:
a) Data; b) Measures; and c) Statistical analyses/plan.
|
|
4/10
|
Multinomial Logistic Regression Model
Topics:
·
Model specification,
interpreting the results
Reading:
·
Hoffman Ch. 5
·
Juby et al. 2005. Sharing roles,
sharing custody? Couples' characteristics and children's living arrangements
at separation. Journal of Marriage and Family 2005 67:1 157
Assignment (due in class
4/17):
·
see handout.
|
|
4/17
|
Poisson and Negative Binomial Regression Models
Topics:
·
Model specification
·
Interpreting the results
Reading:
·
Hoffman Ch. 6
·
Skarupski et al. 2005.
Black–White Differences in Depressive Symptoms Among Older Adults Over Time. The
Journals of Gerontology Series B: Psychological Sciences and Social Sciences
60:P136-P142
Assignment (due in class
4/24):
·
Provide a draft of the
findings section of your paper.
This should include at least two tables: a) Descriptive statistics: provide
the mean/s.d. (or %/N) for all variables used in your analyses; and b)
Regression model estimates.
|
|
4/24
|
Additional considerations and subsequent models
Topics:
·
Spatial analysis, multilevel
analysis, growth models, event history models
Reading:
·
Hoffman Ch. 8
Assignment:
·
No assignment this week
|
|
5/1
|
Individual Meetings
|
|
5/10
|
Final papers are due by 12 pm (noon ) on Wednesday, May 10th.
|