A study was undertaken to investigate the relationships between objective and subjective measures of job characteristics using a modified version of the Job Characteristics Model (JCM. Hackman and Oldham, 1975). Role clarity and challenge were proposed as new psychological states. It was proposed that the personal work outcomes influenced the outcome variables. The dimensionality of the job charac­teristics was tested, as was the moderating effect of growth need strength (GNS). Personal and contextual variables and knowledge and skill were tested as predictors of the psychological states and outcome variables.

The psychology course work was rated by 328 stage one psychology students using the Job Diagnostic Survey (JDS) and by the course supervisors using the Job Rating Form (JRF; N= 12) and the Position Analysis Questionnaire (PAQ; N= 5). Perfor­ mance was measured using the formal course requirements over four assessments.

Objective measures were effective for high GNS subjects In the prediction of outcome variables, particularly for attendance. No relationship was found between the PAQ dimensions with performance and general satisfaction. General satisfaction predicted both performance and attendance. No moderating effect by GNS was discovered for the personal work outcomes or for performance. GNS was found to moderate the relationship between the psychological states and attendance. The multi-dimensionality of job characteristics was confirmed for this sample. Personal and contextual variables and knowledge and skill were useful predictors of the psycholog­ ical states and outcome variables. A new Job Characteristics Model is proposed.