Course Outline -- Fall, 2007
Instructor:
Hiroshi Fukurai
337 College Eight
x 9-2971 (office), hfukurai@ucsc.edu (email)
Office Hours - Thursday 2:00-5:00pm or by appointment
Class:
Monday 2:00 - 5:00 p.m., College Eight 301
Textbooks:
Course Objectives:
1. Understanding of, and practical facility with, the following four areas of methodological and sociological inquiries: (1) contingency table analysis (categorical data analysis), )(2) basic general linear models (GLM) such as multiple regressions and regression with dummy variables (ANOVA & ANCOVA), (3) logistic (logit) regression and log-linear modeling, (4) reliability-validity assessments with multiple indicators and factor analytic techniques (exploratory & confirmatory factor analysis), and (5) causal analyses using single or multiple indicators (i.e., path analysis and structural equation modeling with EQS).
2. Development of an appreciation for the matrix-based mathematical model of reliability--validity assessment, causal modelings, and confirmatory analyses.
3. Familiarity with two computer programs that perform statistical analyses commonly used in the social and behavioral sciences -- SPSS and EQS. Depending upon the students' abilities and desires, other programs may be also used in class presentations and for assignments (such as SAS, LISREL).
Weekly Assignment:
You must turn in a weekly assignment by 5 p.m. Friday (please put it in my mailbox by noon and a late assignment will not be accepted). You do not have to wait until the last minute to turn in your assignment. You may work on it over weekend and complete it by Monday. No assignment will be accepted after the deadline. Each assignment should include the brief explanation of the following: (1) your hypothesis, (2) variables/statistical method to be used for your analysis, (3) analysis/result of the findings, and (4) conclusion -- whether your hypothesis is supported or rejected (min. of 3 pages – double spaced -- for each assignment, excluding the printout). Make sure to submit your printouts as well.
Readings:
You must do each week's readings before coming to the class. The reading for this course will not be easy for most of you. Write down as many questions as you have, and bring them to class or to my office hours. However, it is essential that you do the reading. If this course only teaches you a set of data-analytic techniques, it will have failed in that you will find your methodological knowledge badly outdated 15 to 20 years from now. The major goal of the course is to teach you to be able to learn new concepts and empirical understandings -- primarily by reading about them, and running programs on your own. All readings are required. You may skip the extensive numerical examples: that's why we use computers.
Computer Analyses:
Three SPSS data sets are available: (1) 2006 General Social Survey (GSS2006) data (n<1,500) , (2) UC-wide Affirmative Action survey (RCJ) data (n=977), and (3) Whiteness survey (1999-2000). The user should consult the Reader for details about each of the variables and their attributes. The reader that contains codebooks, questionnaires, etc. will be available at the UCSC Copy center.
COURSE OUTLINE
Week 1:
Course Overview: Introduction and Data Analysis
Knoke, chapters 1 thru 3 (those are review materials of undergraduate statistics course(s) – Make sure to have solid understandings of chapters 2 materials (i.e., variables and their kinds) and 3 (probability and inference, such as what is a statistical significance?).
Week 2:
Statistics for Cross-Classification (Contingency Table Analysis)
Knoke, chapter 5 (cross-tabs for the analysis of two variables) and chapter 7 (introduction of third "control" variable and elaboration)
Assignment: Please refer to the Reader (i.e., SPSS Examples for both frequency and cross-tabulation tables and their interpretations -- (4) thru (7)).
Week 3:
General Linear Model (GLM I): Bivariate and Multiple Regressions
Knoke, chapter 6 (entire chapter) & chapter 8 (8.1 through 8.5).
Week 4:
GLM II: Analysis of Variance (ANOVA) and Analysis of Covariance (ANCOVA)
Knoke, chapter 4 (skip through numerical examples) and chapter 8 (8.6 & 8.7).
Week 5:
GLM III: Special Case – Non-linear & Logistic (logit) Regression (or Extended Discussions on Multiple Regression)
Knoke, chapter 9 (read this first) and chapter 10.
Week 6:
Factor Analysis I: Introduction to Reliability and Validity -- Multiple Indicators
Carmines and Zeller, chapters 1--4
Knoke, chapter 12.6 & 12.6.1 Item reliability (471-475) (also review Box.8.1 for Cronback's alpha)
Week 7:
Factor Analysis II: Factor Analysis
Kim and Muller, pp. 1-70 (Factor Analysis)
Week 8:
Factor Analysis III: Confirmatory Factor Analysis -- Introduction to Causal Modelings (EQS)
Knoke, chapter 12 (12.7 -- read this first)
Bentler, pp. 26-33 (read this before Long book).
Long, pp.5-88 (Confirmatory Factor Analysis)
Week 9:
Causal Analysis I: Path Analysis and Covariance Structure (EQS)
Knoke, chapter 11.
Week 10:
Causal Analysis II and Other Innovative Statistical Methods for Sociologists
Knoke, chapter 12 (entire chapter except 12.7).