CALIFORNIA STATE UNIVERSITY LONG BEACH PPA 696--RESEARCH METHODS BINGHAM AND FELBINGER CH. 17

1. BIBLIOGRAPHIC INFORMATION

1. Author: R. H. Frank & J. D. Kaul

1. Title: The Hawthorne Experiments: First Statistical Interpretation

1. Source: American Sociological Review, 43, 1978

 

1. SUMMARY OF THE RESEARCH
1. PROBLEM STATEMENT:

Do physical conditions or interpersonal relations have more influence on worker productivity?

1. BACKGROUND:

The original Hawthorne experiments were investigating the effects of illumination on productivity, but concluded that interpersonal relations were more important

1. HYPOTHESIS:

Better interpersonal relations lead to worker satisfaction which increases productivity

1. MEASUREMENT OF VARIABLES

1. Dependent variable: Worker output quantity and quality;

1. Independent variable(s): Hours worked per day worked; days worked per week; net hours per week; weeks per period; number of scheduled rest stops per day; minutes of scheduled rest stops per day; minutes of unscheduled rest stops per day;

1. Control variable(s): managerial discipline; economic depression; defective raw materials; temporary worker replacement; small group incentive plan;

1. RESEARCH DESIGN:

Quasi-experimental design using time series regression; measures worker quantity and quality over 23 work periods; also measures all independent and control variables for the same periods.

T1	T2	T3	T4	T5	T6	T7	T8
O1	O2	O3	O4	O5	O6	O7	O8

 

1. SAMPLING:

The original data on 5 workers for 23 work periods is re-analyzed using time series regression

1. INSTRUMENTATION:

Original documentation of all variables is used

1. DATA COLLECTION/ETHICS:

Data were collected from original plant records

1. DATA ANALYSIS:

Time series regression used to estimate the effects of each of the independent and control variables on the dependent variables;

1. CONCLUSIONS:

Output quantity can be explained mainly by tighter managerial discipline, the presence of an economic depression, more frequently scheduled rest periods, and a monetary small group incentive. Output quality was explained mostly by the amount of defective raw materials, more frequent scheduled rest stops, the economic depression, and fewer weeks per work period. Only the more frequent rest stops can be considered better interpersonal relations between workers and management; the depression and incentive plans are based on economic motivation. Cannot assume that changes in output were caused by other, unmeasured variables

1. CRITIQUE

1. Possible Threats to Internal Validity

1. History:

Controlled by including a variable for the depression

1. Maturation:

Controlled by including variables for rest periods

1. Testing:

No other group was measured at the same time to control for the effects of testing; is there a "Hawthorne effect"?

1. Instrumentation:

Original records were kept consistently over time

1. Regression Artifact:

No information on whether these were high- or low-performing workers to begin with;

1. Selection bias:

Workers were not selected at random but volunteers; replacement of two poorly-performing workers was controlled by a variable in the regression equation

1. Experimental Mortality:

Two of the original workers were replaced

1. Design contamination:

Possible that workers changed their behavior in relation to something outside the workplace that was not controlled for in the regression equation

1. Possible Threats to External Validity

1. unique program features:

Very small number of workers (n=5); however, units of analysis are production periods

1. experimental arrangements:

"Managerial discipline" could be broken down into smaller components to explore how various facets of this variable affected worker performance

1. other threats: