TO: Students in CJ 605, SP 00
FROM: R. B. Taylor
DATE: 2/28/00
RE: Assignment for paper 2

PURPOSE
The purpose of this assignment is for you to begin to pull together a draft of a good section of your final paper. This gets you up through the Level 1 model. You want to be sure you have talked to me before you actually start work on this.

DUE: 3/23/00

PARAMETERS
Be sure the paper is typed, double spaced. You can submit this electronically if you wish. If you do, send it to BOTH ralph@blue.temple.edu AND rbt@rbtaylor.net

SECTION 1. This section, which should run a couple of pages to four pages, lays out a theoretical rationale for your key predictor. You want to be sure to develop a theoretical rationale that is both individual level, i.e., describes the differences between neighbors, AND a theoretical rationale that is aggregate level. These two pieces are essential. You probably want to look at some work in the area. The whole "motivator" for the multilevel models, in addition to being appropriate for the clustered data, is that you can figure out what is happening with multilevel impacts without having to worry about mis-specification at level 1.

In addition, if you want you can develop a rationale for why the slope of the key predictor should vary across contexts, and what ecological factors should affect that slope.

In short, there are at least two parts: rationale for individual level predictor; rationale for aggregate level impact on the intercepts (group means). In addition, you may develop the following two parts: why you expect the slope of the Level 1 predictor to vary across contexts; and what ecological factors might influence that slope.

In developing your rationales, I include some references below that you may find of interest. In addition, I will get one copy of Davis' final report copied and put in the grad office, and will let you know when that is available.

SECTION 2. This is your description of the original data collection procedure, the sample characteristics, your independent variables, and your dependent variables. You will want one cleanly formatted table showing min, max, mean, median, and sd for all of your variables.

I strongly suggest you include a table showing the wording and response categories for all your predictor variables, as well as the dependent variable. Of course somewhere in the text or the table you also will include the Cronbach's alpha for the outcome and any indexes you choose to build.

SECTION 3. Describe the variance decomposition results for your outcome, and your key predictors. These are the results from two ANOVA runs. You want the table set up as follows:

	Dependent variable	Key predictor
Between
Within
Total
Percent between
Chi squared for between variance

SECTION 4. Describe the results of your level 1 model. This may or may not have a varying slope. I strongly suggest you FIRST estimate this model with no slopes varying at all; fix all the slopes first, THEN look at a result where the slope varies and see how deviance has changed. (IF and ONLY IF you want to have a varying slope.) I strongly suggest that you think carefully about the centering issues. You may or may not want to group mean center your key predictor, depending on how your theory is set up. I strongly suggest you include the following control variables:

• female (would not center this variable at all)
  

• education (grand mean or group mean centered)
  
• either age or years of living in the neighborhood (either grand mean centered or group mean centered, depending on your theory)
  
• At least one race variable. I would suggest against using both Africam and Hispanic. When we put in the ELPASO dummy at level 2, it will be too collinear with the Hispanic dummy. If you wanted to combine (African+Hispanic) that would be fine, or if you wanted to make a white dummy, that would be fine too. Your rationale here is going to be critical in terms of how you set up the race variable.

This section will have a table laid out as follows (one row for each predictor):

Predictor	MODEL A Model w all slopes fixed	MODEL B as before, but now add additional covariates	MODEL C Model w one varying slope	MODEL D as before, but now add additional covariates
a...	b (se) t		b (se) t
...p
Variance results * Between Explained * Significance of remaining between * Within Explained * Total Explained

MODEL A has your key predictor, with a fixed slope, and other Level 1 demographic predictors. MODEL B then adds in additional covariates such as DSOCIAL and DCOMPROB. MODEL C is just like MODEL A but allows the slope of your key predictor to vary. MODEL D is just like model C but adds in additional covariates, just like MODEL B did.
EXTREMELY IMPORTANT POINT: IN ALL OF THESE MODELS INCLUDE THE THREE CITY DUMMIES (ELPASO, CHICAGO, NEWARK) so that we are sure we are controlling for between city differences.

REFERENCES YOU MAY WANT TO SEEK OUT - ones of particular interest are italicized. Articles in Crime and Delinquency are available through PROQUEST DIRECT.

Belenko, S., J. Fagan, and K. Chin 1991 "Criminal justice responses to crack." Journal of Research in Crime and Delinquency 28:55-74.

Bennett, T. 1989 "Factors related to participation in neighborhood watch schemes." British Journal of Criminology 29:207-218.

Davis, R. C., A. J. Lurigio, and D. P. Rosenbaum 1993 "Introduction." In Drugs and the community: Involving community residents in combatting the sale of illegal drugs, eds. R. C. Davis, A. J. Lurigio, and D. P. Rosenbaum, xi-xviii. Springfield, IL: Charles C. Thomas.

Davis, R. C., B. E. Smith, and S. W. Hillenbrand 1990 Reporting of drug-related crimes. Final Report to the National Institute of Justice. Grant No. 88-IJ-CX-0032

DuBow, F. and A. Podolefsky 1982 "Citizen participation in community crime prevention." Human Organization 41:307-314.

Fagan, J. 1993 "The Political economy of drug dealing among urban gangs." In Drugs and the community: Involving community residents in combatting the sale of illegal drugs, eds. R. C. Davis, A. J. Lurigio, and D. P. Rosenbaum, 19-54. Springfield, IL: Charles C. Thomas.

Grinc, R. M. 1994 "'Angels in marble': Problems in stimulating community involvement in community policing." Crime & Delinquency 40:437-468.

Hagedorn, J. M. 1994 "Neighborhoods, markets, and drug gang organization." Journal of Research in Crime and Delinquency 31:264-294.

Hillenbrand, S. W. and R. C. Davis 1993 "Residents' perception of drug activity, crime and neighborhood satisfaction." In Drugs and the community: Involving community residents in combatting the sale of illegal drugs, eds. R. C. Davis, A. J. Lurigio, and D. P. Rosenbaum, 5-18. Springfield, IL: Charles C. Thomas.

Hope, T. 1995 "Community Crime Prevention." In Building a safer society, eds. M. Tonry and D. Farrington, 21-90. Chicago: University of Chicago Press.

Lab, S. P. 1990 "Citizen crime prevention: Domains and participation." Justice Quarterly 7:467-492.

Lurigio, A. J. and R. C. Davis 1992 "Taking the war on drugs to the streets: Perceptual impact of four neighborhood drug programs." Crime & Delinquency 38:522-538.

Lurigio, A. J. and L. Klein 1989 "Controlling crime in the community: Citizen-based efforts and initiatives." Crime & Delinquency 35:.

Perkins, D., B. Brown, and R. B. Taylor 1996 "The Ecology of empowerment: Predicting participation in community organizations." Journal of Social Issues 52:85-111.

Powers, S. 1993 "Community responses to drugs: Manhattan and Brooklyn case studies." In Drugs and the community: Involving community residents in combatting the sale of illegal drugs, eds. R. C. Davis, A. J. Lurigio, and D. P. Rosenbaum, 106-122. Springfield, IL: Charles C. Thomas.

Rosenbaum, D. P. 1988 "Community crime prevention: A review and synthesis of the literature." Justice Quarterly 5:323-395.

Simon, D. and E. Burns 1997 The Corner: A Year in the life of an inner-city neighborhood. New York: Broadway Books.

Skogan, W. and Bennett, J. (1994) Community policing: Chicago style. New York: Oxford.

Skogan, W. and A. J. Lurigio 1992 "The Correlates of community anti-drug activism." Crime & Delinquency 38:510-521.

Smith, B. E. and R. C. Davis 1993 "Successful community anticrime programs: What makes them work?" In Drugs and the community, eds. R. C. Davis, A. Lurigio, and D. P. Rosenbaum, 123-137. Springfield, IL: Charles Thomas.

Taylor, R. B. 1996 "Neighborhood responses to disorder and local attachments: The systemic model of attachment, and neighborhood use value." Sociological Forum 11:41-74.

Visher, C. and D. Weisburd 1998 "Identifying what works: Recent trends in crime prevention strategies." Crime, Law and Social Change 28:223-242.

Weingart, S. N. 1993 "A Typology of community responses to drugs." In Drugs and the community: Involving community residents in combatting the sale of illegal drugs, eds. R. C. Davis, A. J. Lurigio, and D. P. Rosenbaum, 85-105. Springfield, IL: Charles C. Thomas.

Weingart, S. N., F. X. Hartmann, and D. Osborne 1994 Case studies in community anti-drug efforts (National Institute of Justice Research in Brief). Washington: US GPO.