When random assignment fails: Some lessons from the Minneapolis Spouse Abuse Experiment

Show other versions (1)

In this paper, we consider what may be done when researchers anticipate that in the implementation of field experiments, random assignment to experimental and control groups is likely to be flawed. We then reanalyze data from the Minneapolis Spouse Abuse Experiment in a manner that explicitly models...

Full description

Saved in:  
Bibliographic Details
Main Author: Berk, Richard (Author)
Contributors: Smyth, Gordon K. ; Sherman, Lawrence W. 1949-
Format: Electronic Article
Language:English
Published: 1988
In: Journal of quantitative criminology
Year: 1988, Volume: 4, Issue: 3, Pages: 209-223
Online Access: Volltext (lizenzpflichtig)
Volltext (lizenzpflichtig)
Journals Online & Print:
Drawer...
Check availability: HBZ Gateway
Keywords:
Selection Bias
Domestic Violence
random assignment
field experiments
Description
Summary:In this paper, we consider what may be done when researchers anticipate that in the implementation of field experiments, random assignment to experimental and control groups is likely to be flawed. We then reanalyze data from the Minneapolis Spouse Abuse Experiment in a manner that explicitly models violations of random assignment. As anticipated, we find far larger treatment effects than previously reported. The techniques developed should be useful in a wide variety of settings when random assignment is implemented imperfectly.
ISSN:1573-7799
DOI:10.1007/BF01072450