CALIFORNIA STATE UNIVERSITY LONG BEACH
PPA 696--RESEARCH METHODS:
BINGHAM & FELBINGER CH. 19
  1. BIBLIOGRAPHIC INFORMATION
    1. Author: G. E. Moran
    1. Title: Regulatory Strategies for Workplace Injury Reduction: A Program Evaluation
    1. Source: Evaluation Review, 9(1), 1985:21-33.

      2.  
  1. SUMMARY OF THE RESEARCH
    1. PROBLEM STATEMENT:
Does the OSHA program of mine inspections reduce injuries?
    1. BACKGROUND:
There is a need for effective and inexpensive methods of reducing workplace injuries
    1. HYPOTHESIS:
The PAR inspection program reduces workplace injuries in mines
    1. MEASUREMENT OF VARIABLES
      1. Dependent variable: Injury rates; death rates (per million hours worked)
      1. Independent variable(s): PAR inspection program
      1. Control variable(s): Number of miners; number of hours worked per miner
    1. RESEARCH DESIGN:
A quasi-experimental, time series comparison group design; compares mines inspected by the PAR program and mines not inspected but also with high injury rates.
Group T1 T2 T3 PAR Program T4 T5 T6
G-1 O1 O2 O3 X O4 O5 O6
G-2 O1 O2 O3 O4 O5 O6
 
    1. SAMPLING:
Non-random selection of mines in the PAR program and mines not in the PAR program but with injury rates +2 standard deviations above the average
    1. INSTRUMENTATION:
Government data on mines and injuries remains stable over time
    1. DATA COLLECTION/ETHICS:
Collects government data on mines and injuries
    1. DATA ANALYSIS:
Uses regression analysis to determine how much of the reduction in injuries in PAR-inspected mines is due to the PAR program, and how much is due to the tendency of regression to the mean.
    1. CONCLUSIONS:
The PAR program produces a greater reduction in injuries in inspected mines than could be expected from non-inspected mines
  1. CRITIQUE
    1. Possible Threats to Internal Validity
      1. History:
History controlled by comparison group of mines
      1. Maturation:
Controlled by comparison group of mines
      1. Testing:
No testing administered
      1. Instrumentation:
Instrumentation and data collection stable over time
      1. Regression Artifact:
Controlled by performing regression and including data on comparison mines with high injury rates which could be expected to decrease over time naturally
      1. Selection bias:
Non-PAR mines were smaller, but size was controlled by entering it into the regression equation
      1. Experimental Mortality:
Data was available on all mines for the period studied
      1. Design contamination:
Non-inspected mines may have learned about safety techniques from the inspected mines
    1. Possible Threats to External Validity
      1. unique program features:
      1. experimental arrangements:
Mines were chosen for extreme scores (high injury rates); PAR program is applied after injury rates have risen to a higher than normal level, but is not directed at mines with low to average rates of injuries, to prevent them from getting too high.
      1. other threats: