Date of Award

8-1977

Degree Name

Master of Arts

First Advisor

Dr. Richard Malott

Second Advisor

Dr. Howard Farris

Third Advisor

Dr. Brian Iwata

Fourth Advisor

Dr. J. Michael Keenan

Access Setting

Masters Thesis-Open Access

Abstract

The present series of studies is a behavioral systems analysis of the problems of job performance, promptness at work, and cost control. Many studies surveyed in the business and industrial literature have methodological problems, and those employing good experimental control have not addressed the cost effectiveness of the procedures they evaluated (Pommer and Streedbeck, 1974; Hermann et. al., 1974). Those studies employing quasi-experimental designs generate inadequate data and also often fail to address the cost effectiveness of the procedures used (Brobst, 1976; Kent, 1974; Schnelle and Lee, 1974; Campbell and Stnley, 1966; Baer, Wolf and Risley, 1969). A major reason for these deficiencies in such studies is the additional time and money involved in doing a more adequate study. These same studies also lack client participation in the design, which is often desirable for ethical and practical reasons (Kent, 1974).

In the present studies, we used time-series quasi-experimental designs (Campbell and Stanley, 1966) combined with a behavioral systems approach (Malott, 1972). The critical difference between true and quasi-experimental design, the researcher controls the random assignment of the sample population to treatment and no-treatment conditions. We employed a single group in each of these studies, making numerous observations in each of at least two successive conditions. The greatest single threat to internal validity in this time0series design is the possibility of some uncontrolled simultaneous event actually causing the changes observed. Schnelle, Kirchner, McNees, and Lawler (1975) used the time-series analysis when it was impossible to gain full experimental control. We used a time-series analysis for similar reasons. A further source of confounding in these studies was that we conducted them concurrently so that at one point three independent variables were implemented simultaneously. Table 1 shows the schedule of independent variables.

Included in

Psychology Commons

Share

COinS