#### POLS 542: Intermediate Analysis of Political Data, Spring 2002

Instructor:  Jim Schubert

Office:  ZU 306

Email:  t70jns1@wpo.cso.niu.edu

Office Hours:  T, Th 2-3

Phone (w/voice mail):  815-753-9675

##### Course Description

This seminar presents an intermediate level course in bivariate and multivariate correlation and regression.  For students who have not taken a statistics course in some time, we begin with a review of the fundamental concepts of introductory statistics and build new concepts and methods upon this foundation.  Our approach to the material is not mathematical, but applied and data analytic.  The emphasis in the course is explicitly “hands on.”  I firmly believe in the value of learning data analysis by doing data analysis.  Therefore, applications provide a central focus of activity in the class.   Students are expected to leave the course capable of solving problems of data analysis using appropriate computer software.  The class will use SPSS for Windows.

Norusis, SPSS Guide to Data Analysis

Cohen & Cohen, Appled Multiple Regression/Correlation Analysis for the Behavioral Sciences

Beck, Regression Analysis

##### Requirements

Requirements for this seminar include:  attendance and participation, completion of assign exercises, a regression based research paper, and a take home final exam. Attendance is not optional, but required, unless you are hospitalized, taken hostage or jailed.

POLS 542:  Topical Outline

# A.    Data:4 levels of measurement

B.    Distributions:  frequencies, graphic displays, modality

C.    Moments of a distribution:  variance, skewness and kurtosis

D.    Points in a distribution:  z scores, mean v. median, quartiles, deciles, %s

E.     Measures of variation/dispersion: standard, interquartile, etc.

F.     Normalacy

## II.               Bivariate Analysis

A.    Correlation

1.     Pearson’s r, Spearman’s rho, Kendall’s tau, eta

2.     Partial correlations: first, second and nth order

### B.    Regression

1.     Least squares line, residuals

2.     “b” v. beta

3.     Residual variance and R Square as a PRE technique

4.     Assumptions

5.     Residuals, outliers & leverage

### III.Multivariate Analysis

1. Multiple Regression

1.     Two Independent Variables:

a.       Effect size

b.     Assumptions

c.      Residual Diagnostics

2.     Three+ Independent Variables:

a.       Effect size, partial plots

b.     Dummy variables

c.      Two & multi-way interactions

3.     Significance and Goodness-of-Fit

4.     Stepwise techniques

1. Corrections for Heteroscedasticity and Multicollinearity

1.     Principal components (factor) analysis

2.     Two-Stage Least Squares

1. Causal Analysis/path analysis

1. Analysis of Variance

1.     Cronbach’s Alpha

2   Repeated Measures

Week   Date     Cohen&Cohen             SPSS               Beck

1          1/18                                         1,2

2          1/25                 1                      4,6

3          2/1                   2                      5,7,8

4          2/8                                           18,19,20

5          2/15                 6                      21

6          2/22                 3                      22

7          3/1                                           23

8          3/8                   5

break

9          3/22                 8

10        3/29

11        4/5                   9

12        4/12

13        4/19                 11

14        4/26

15        5/3

16           5/10 final exam week