TO: Students in CJ 605
FROM: R. B. Taylor
DATE: 3/21/00
RE: Paper parts 3 and 4.

PAPER PART 3 DUE 4/4: REPORT ON LEVEL 2 IMPACTS

The first part reports on Level 2 impacts. That is, you make a decision of what level 2 predictors to include, and you see what they do. The paper has the following parts:

Theory.
Explain what variables you want to include at level 2. Describe the process whereby the predictor connects ECOLOGICALLY to the outcome.

Presume that your L2 predictor is entered AFTER you have entered the 3 city dummies.

Further, given the limited number of L2 neighborhoods, you can probably only enter one or two predictors after the 3 dummies.

Table
Include a Table that shows for both Level 1 and Level 2 variables the coefficient, t-ratio, and percent of variance explained at L1 and L2.

Describe
Describe the results for your Level 2 variables

PAPER PART 4 DUE 4/11

In this analysis you will report on a residual analysis. More specifically, I would like to see mean, min, max for CHIPCT, MDIST, and MDRSVAR (see table at bottom of P. 35 in the manual. I also would like to see a probability plot (P-P plot in SPSS) of the residual dispersions (MDRSVAR). Please answer the questions: are there any extreme outliers suggested by these residual analyses (of course, if there are, be sure to identify the specific locations), and is the distribution of the residuals pretty normal or not?

This should probably run 1 - 3 pages typed double spaced, plus one table, plus one figure.