Newsom: Data Sets

Note: Data sets may be random samples or otherwise randomly altered for additional anonymity. Findings will generally be consistent with those from the actual study data but should not be published or otherwise reported as results from the study. I will only post current quarter data sets. Depending on your computer and browser, you may need to right click and "save target as" or "save file as" to download these files.

Multilevel Regression

Highschool and Beyond data set, hsbmerged.sav, hsb.mdm

Early Childhood Longitudinal Study, ecls2011.sav, ecls2011.mdm.

HW1

Oregon head start data, headmh.sav, headmh.mdm

HW2

Work, family, and health data, wfh.sav, wfh.mdm.

abamerged.sav. These data come from a study of a community survey of 378 respondents conducted in 42 neighborhoods. The survey was conducted by the American Bar Association (ABA) and examines reports of crime, neighborhood drug use, impressions of the neighborhood, and attitudes toward police (this data is publicly available from ICPSR). I suggest as the dependent variable a question about how the resident feels overall about his or her neighborhood (ranging from 1=“really bad” to 5 “great place”; NHRATING). Level-1 predictors include reported fear of being attacked (FEAR) and a rating of neighborhood problems (PROBS1) derived from an 8-item rating scale of the seriousness of the problems in the neighborhood, and SEX (0=male;1=female). The level-2 predictors are aggregated variables (PROBS2), constructed as a level-2 variable by averaging all the responses within a neighborhood. So, the scale ranges from 1 no problem to 4 serious problems and is a neighborhood-level indicator of neighborhood erosion issues. The neighborhood identification variable is NH.