As a precursor to viewing this study history, please read my Philosophy of the Syllabus Post [here]

  • CLASS: SOC 882
  • SEMESTER: Spring 2011
  • CLASS TITLE: Analysis of Social Data II
  • TIME & LOCATION: MW, 10:20 AM – 12:10 PM, Berkey 216
  • INSTRUCTOR: Dr. Hui Liu
  • EMAIL: liuhu@msu.edu
  • OFFICE: 517-353-3265
  • Teaching Assistant: Dilshani Sarathchandra
  • E-mail: sarathch@msu.edu

Assignments

  • 1/26/11 Assignment 1
  • 2/9/11 Assignment 2
  • 2/23/11 Assignment 3
  • 3/2/11 Assignment 4
  • 3/16/11 Mid-Term Report
  • 3/30/11 Assignment 5
  • 4/13/11 Assignment 6
  • 4/20/11 Assignment 7
  • 4/27/11 Assignment 8
  • 5/4/11 Final Report Due by noon

Course Topics (by week)

  1. Review of Bivariate Linear Regression
    • -Dependent variable and independent variable
    • -Hypothesis test for regression coefficients
    • -r and R2
    • -Predicted values and residuals
    • DO FILE
    • CLASS 1 NOTES
  2. OLS Multiple Regression
    • -Standardized and unstandardized regression coefficients
    • -F-test
    • -R, R2 and Adjust R2
    • -Categorical independent variables
    • DO FILE
    • CLASS 2 NOTES
  3. OLS Models Assumption and Violations
    • -linearity assumption.
    • -zero mean of errors assumption.
    • -homoscedasticity assumption
    • -no autocorrelation assumption.
    • -Normal distribution assumption.
    • DO FILE
    • CLASS 3 NOTES
  4. Regression Diagnostics
    • -Influence diagnostics
    • -Multicollinearity diagnostics
  5. -Modeling Non-linear Relationships
    • -Centering
  6. -Modeling Interaction Relationships
  7. Model Building and Comparison
    • -Backward elimination
    • -Forward selection
    • -Stepwise selection
  8. Review and Exercises
  9. SPRING BREAK
  10. Analyzing Associations Using OLS
    • -Progressive adjustment method
  11. Path Analysis
    • -Path diagram
    • -Direct and indirect effects
    • -Path decompositions
  12. Logistic Regression
    • -Inferences
    • -Interpretations
  13. Logistic Regression
    • -Inferences
    • -Interpretations
  14. Multinomial Logistic Regression
    • -Inferences
    • -Interpretations
  15. Ordered Logistic Regression
    • -Inferences
    • -Interpretations
  16. Review and Exercises

Required Texts

  • Hamilton, Lawrence. 1992. Regression with Graphics: A Second Course in Applied Statistics. ISBN-10: 0-534-15900-1. ISBN-13: 978-0-534-15900-9. Publisher: Brooks/Cole. (Referred to below as H1992)
  • Website with STATA/SPSS/SAS programming codes reproducing examples in H1992 can be found at: http://www.ats.ucla.edu/stat/examples/rwg/
  • Agresti, Alan & Barbara Finley. 2009. Statistical Methods for the Social Sciences (Fourth Edition). Publisher: Prentice Hall. ISBN-10: 0130272957. ISBN-13: 9780130272959 (Referred to below as A&F).
  • Website with STATA programming codes reproducing examples in A&F can be found at: http://www.ats.ucla.edu/stat/examples/smss/default.htm
  • Suggested Supplementary Text
    • Hamilton, Lawrence C. 2009. Statistics with Stata (Updated for Version 10). Publisher: Cengage. ISBN-10: 0-495-55786-2. ISBN-13: 978-0-495-55786-9.
  • Other related readings will be uploaded on the Angel course web.

Course Objectives

This is a second level graduate statistics course in social science with emphasis on multiple regressions. We emphasize the practical application of quantitative methods to answer questions in social science research. The course is composed of two parts. In the first part of the course, students are expected to get familiar with OLS models for multiple regressions. Topics in the first part include OLS model assumptions and violations, modeling interaction relationships, nonlinear relationships, model building and path analysis. The second part of the course focuses on regression models for categorical dependent variables such as logistic regression, multinomial logistic regression and ordinal logistic regression. By the end of this course, we expect you will be able to:

  • appropriately interpret relevant statistical reports in academic articles;
  • create relevant statistical reports as requested; and
  • choose appropriate statistical methods and apply them in your own research projects.

Prerequisite

Successful completion of SOC881, or its equivalent is a prerequisite for this course. A solid knowledge of algebra is also required to succeed in this course. Students are assumed to have a solid understanding of basic statistical topics such as sampling distributions, hypothesis testing, bivariate correlation and regression. Some prior exposure to multiple regression is also very helpful.

Software

STATA will be introduced as the major statistical analysis package in this course. We will teach the students how to use STATA to analyze data from social science surveys, such as General Social Survey, using the
statistical methods discussed throughout the course. You are free to finish the assignments using other statistical software if you have preference, but the major software illustrated in this course is STATA. A brief
lab section will be offered to show students to start with STATA at the beginning of the course.

Course Requirements

Reading assignments: Readings from the texts will be assigned each week. Although reading assignments do not constitute your final grade, they are extremely helpful for you to understand the lectures and finish
the homework. I strongly recommend that you read the assigned material once before class and then again afterwards.
Homework assignments: You will be expected to turn in 8 homework assignments. Each homework assignment will correspond to one or two topics coved in class. You are allowed and encouraged to collaborate with others, but the work you turn in must be your own. Homework must be received on or before the due dates. There will be a penalty of 20% of the possible points per day late for late submissions.
Research Report: All students are required to write a research report in this course. This report will be approximately 10-15 pages double spaced and will include a research question section, a data and sample
section, a measures section, a methods section and a results section. The main purpose of this report is to encourage you in the process of conducting research and enable you to apply what you have learned in the
class to an actual scientific study. We strongly encourage you to talk with the instructor or TA about this project report at least once during the semester. All students are required to turn in both a mid-term report
at the middle of the semester and a final report at the end of the semester. Detailed instructions on the final report and mid-term report will be handed out in class.

Final Grade:

  • Eight Homework Assignments: 60% Total
  • Mid-term Research Project Report: 5%
  • Final Research Project Report: 35%
  • A: 90 and 90+
  • B: 80 and 80+
  • C: 70 and 70+
  • D: 60 and 60+
  • F: <60
  • In order to get A in this class, you have to get A for both homework assignments and final research
  • report.

All assignments and reports must be turned in as hard copy before the due date or before class on the due date. In the case that you were unable to be present when the assignment/exam is due, you may send an electronic submission to show that you have completed it on time. You are still required to turn in a hard copy when you can. The hard copy you turn in should not be changed from your electronic submission. We will only grade and comment on hard copies. The last day for dropping courses with no grade reported is the middle of the semester.