Social Statistics 526b

Dr. Darren E. Sherkat
Department of Sociology
Southern Illinois University
3422 Faner Hall
SHERKAT@siu.edu Office Hours: Tu, W. 10:00-12:00, or by appointment
Goals
I have three goals for the course: (1) to provide graduate students with an understanding of basic inferential statistics; (2) to lay a foundation for understanding more advanced statistics; and (3) to understand the connection between statistics and theory, so that you will come to know why statistics are useful. We will use a fairly basic, non-mathematical introductory statistics book. You will not be required to derive distributions or use calculus, however, you will be required to do hand calculations and solve computer problems.

We will cover the following topics:

(1) The basics: mean, median, mode, proportion; Dispersion--variance and standard deviation; Tests of differences in means and proportions--Z- and T-tests. Dependent and independent variables.

(2) Some other distributions: Binomial, polytamous, the linearity assumption and why you need to pay attention to violations of linearity; tests of association and fit for other distributions--chi-square.

(3) Introduction to analysis of variance: group means and variances--comparisons using the F-statistic; Adjusted means; Explaining variation in a dependent variable.

(4) Bivariate regression and Correlation: Introduction to the general linear model; intercepts, slopes and standardized slopes; R2.

(5) Multiple regression: interpreting coefficients; explained variance; dummy variables; model assumptions and an introduction to some common problems in Ordinary Least Squares regression.

(6) Regression methods for categorical dependent variables: Logistic regression, ordinal and multinomial logistic regression.

Course Materials:

David Knoke and George W. Bohrnstedt. Statistics for Social Data Analysis. Third Edition.


Course Requirements:

(1) Homework assignments will be given for each class. Homework will be graded, and you are required to work independently on your assignments. There is a slight exception with the computer portions of assignments. I expect you to attempt to run the analyses independently, but I do not consider assistance with programming an honor code violation in this course. Assistance with interpreting the output is an honor code violation. Homework assignments will count for 50% of the grade in the course.

(2) There will be two in class examinations. The purpose of the examinations is to allow students to demonstrate that they know how to compute and interpret basic statistics. Each examination will count for 25% of the course grade.


Computing Issues

We will use SPSS for windows exclusively for homeworks and problems. I will assist with programming issues in class examples, but you must pay attention and take notes. I will not repeatedly answer the same questions about programming.

* Dr. Darren E. Sherkat (B.A. Tulsa, 1987; M.A. Duke, 1989; Ph.D. Duke, 1991), Associate Professor of Sociology and Director of Graduate Studies. Phone: 453-7619. My substantive work is in the sociology of religion, social movements, medical sociology, and sociological theory. Specifically, some of my recent papers have examined: regional variations in African American religious participation; the ideological foundations of moral crusades against pornography; participation in and effects of 1960s era protest movements; integrating rational choice and structuration theories, the effect of Christian fundamentalism on educational attainment, and race differences in morbidity and health service utilization and costs. My empirical work uses an array of quantitative methods (though I am also trained in comparative historical analysis and include qualitative archival and interview materials in several recent publications), including: OLS regression, ANOVA (for adjusted and unadjusted mean comparisons across groups); binary and polytomous logistic regression; log-linear and log multiplicative association models; structural equation models (with and without latent variables, and for limited endogenous indicators); and survival regression. My recent articles appear in American Sociological Review, Social Forces, Annual Review of Sociology, Social Science Research, and Journal for the Scientific Study of Religion, Journal of Aging and Health, among other places.


Homework 1

Chapter 1.

Problems: 5, 9,

Chapter 2

Problems: 1,2,6