Some Difficulties in Longitudinal Assessment of Need Satisfaction Related to Management by Objectives

Abstract

This study reports changes in on-the-job need satisfaction as related to management training and development in Management by Objectives. Previous work (Timm, atensen, Strauss, & Babcock, 1977) indicated that the introduction of Management by Objectives may be related to positive changes in the extent to which various needs are met. One of the methodological difficulties, however, cited in the above study was the reliance upon respondents' recall as to levels of satisfaction prior to the Management by Objectives program. This srudy examines changes in job-related satisfaction before and after a 6-mo. management training program in Management by Objectives in a state correctional agency. This study describes data collected just prior to the initiation and shortly following the program The population smdied included all 50 employees of the stare correctional agency. Data were collected through the use of the Porter-Lawler Needs Questionnaire (Porter. 1961 ) . Deprivation for each of the five need levels decreased following the training program: security needs decreased from a mean of 1.52 (SD = 2.03) to 1.07 (SD = 1.72), social needs decreased slightly from .86 (SD = 1.17) to .83 (SD = 1.48), esteem from 1.60 (SD = 1.71) to 1.20 (SD = 1.54), autonomy needs from 1.61 (SD = 1.63) to 1.42 (SD = 1.58). and self-realization needs from 1.82 (SD = 1.63) to 1.26 (SD = 1.68). Although each of the need levels reflected change in the anticipated direction, none of the changes reported are at acceptable levels of significance (p = .05 for self-realization, however). Results suggest that, although the present design circumvented the previous problem of the unrcliability of respondent's recall in the measurement of change, it nevertheless presents another unanticipated problem in design. There is no reason to believe that respondents reporting a lack of need deprivation prior to the training should report changes in satisfaction; in fact, the only change which is possible is nrgatlve change. In the current design there are no means for identifying and determin~ng the influence of satisfied respondents, a task which would require the identification and matching individual responses before and after training, so that the change among dissatisfied respondenrs is probably understated, a condition which requires consideration in future research.