U.S. flag

An official website of the United States government, Department of Justice.

NCJRS Virtual Library

The Virtual Library houses over 235,000 criminal justice resources, including all known OJP works.
Click here to search the NCJRS Virtual Library

USE OF MULTIPLE REGRESSION ANALYSIS TO TEST FOR A DETERRENT EFFECT OF CAPITAL PUNISHMENT - PROSPECTS AND PROBLEMS (FROM CRIMINOLOGY REVIEW YEARBOOK, V 1, 1979, BY SHELDON L MESSINGER AND EGON BITTNER - SEE NCJ-60767)

NCJ Number
60769
Author(s)
L S FRIEDMAN
Date Published
1979
Length
27 pages
Annotation
THE POSSIBILITIES AND LIMITATIONS OF THE USE OF MULTIPLE REGRESSION ANALYSIS TO TEST THE DETERRENT EFFECT OF CAPITAL PUNISHMENT ARE DISCUSSED.
Abstract
MULTIPLE REGRESSION ANALYSIS (MRA), WHICH ATTEMPTS TO USE ALL THE CAUSAL FACTORS (INDEPENDENT VARIABLES) THAT RELATE TO THE PRODUCTION OF AN EFFECT (DEPENDENT VARIABLE) IN THE CONSTRUCTION OF A STATISTICAL RELATIONSHIP, ALLOWS FOR THE DETERMINATION OF THE EFFECT OF A SINGLE CAUSAL FACTOR UPON THE DEPENDENT VARIABLE AS THE OTHER CAUSAL FACTORS ARE CONTROLLED. THE USE OF MRA IN MEASURING THE DETERRENT EFFECT ON MURDER OF CAPITAL PUNISHMENT AS A SANCTION HAS OBVIOUS POSSIBILITIES. WITH THE MURDER RATE AS THE DEPENDENT VARIABLE, ALL OF THE CAUSAL FACTORS THEORIZED TO RELATE TO THE MURDER RATE, INCLUDING CAPITAL PUNISHMENT, CAN BE USED IN THE EQUATION. CONTROLLING FOR THE OTHER INDEPENDENT VARIABLES, CAPITAL PUNISHMENT'S DETERRENT EFFECT ON THE MURDER RATE CAN THEN BE DETERMINED. SINCE ANY GIVEN USE OF MRA IS NOT SELF-EVIDENTLY VALID, IT DEPENDS UPON CONSENSUS IN THE SCIENTIFIC COMMUNITY REGARDING THE PARTICULAR STATISTICAL EQUATION USED AND CONSENSUS ABOUT THE INCLUSION OF ALL THE INDEPENDENT VARIABLES THAT MIGHT CONCEIVABLY BE IMPACTING UPON THE DEPENDENT VARIABLE. BARRIERS TO THE SUCCESSFUL USE OF MRA TO TEST FOR A DETERRENT EFFECT OF CAPITAL PUNISHMENT INCLUDE: (1) PROPER USE OF THE IMPLICATIONS OF THE HYPOTHESIZED RELATIONS, (2) PROPER LINKING OF THE PSYCHOLOGICAL INFLUENCES ON INDIVIDUAL BEHAVIOR TO GEOGRAPHIC DATA, (3) AVOIDING AGGREGATION BIAS IN THE CHOICE OF THE UNIT OF GEOGRAPHIC ANALYSIS, AND (4) AVOIDING PROBLEMS WHICH MIGHT BE EXPECTED FROM THE USE OF ANY TIME-SERIES OR CROSS-SECTIONAL DATA. VALID STATISTICAL INFERENCES BASED ON MRA DEPEND ON REPLICABILITY, PROPER CHOICE OF FUNCTIONAL FORM, PROCEDURES USED FOR HANDLING MISSING DATA, THE APPROPRIATENESS OF APPROXIMATE STATISTICAL TESTS, AND THE INCLUSION OF ALL RELEVANT VARIABLES. THERE IS A CONSENSUS AMONG VIRTUALLY ALL OF THE STUDIES OF ALL METHODOLOGIES THAT THERE IS NO EVIDENCE OF ANY MEASURABLE DETERRENT EFFECT IN THE USE OF CAPITAL PUNISHMENT. THE ONLY MAJOR EXCEPTION TO THIS CONSENSUS IS THE ISAAC EHRLICH STUDY, WHICH HAS BEEN DEMONSTRATED TO DEPEND ON A MYRIAD OF ASSUMPTIONS. POLICY DECISIONS WITH REGARD TO THE ABOLISHMENT OF CAPITAL PUNISHMENT CAN THUS BEST BE MADE BY FOCUSING ON OTHER CRITERIA THAN DETERRENCE. NOTES ARE PROVIDED. (RCB)

Downloads

No download available

Availability