Search

The Position Analysis Questionnaire (PAQ) is a structured job analysis procedure that provides for the analysis of jobs in terms of each of 187 job elements, these job elements being grouped into six divisions: information input, mental processes, work output, relationships with other persons, job context, and other job characteristics. Two cluster analysis procedures were used in the clustering of jobs on the basis of data from the PAQ. The BC-TRY program was carried out with the 14 overall or general job dimensions as applied to a reasonable varied sample of 3,700 jobs and resulted in the identification of 33 job clusters. The CODAP (Coordinated Occupational Data Analysis Program) was based on the scores on 21 of the divisional job dimensions for a sample of 800 jobs, and resulted in the identification of 45 clusters. The differences in the results may be due to the differences in the nature of the job dimensions used in the two instances, rather than being associated with the clustering procedures as such. The 95-page appendix contains the tables of data (15 pages) and line graphs for the job clusters resulting from the two programs. (Author/AG)

The study deals with the relationships between job satisfaction and the following characteristics of the job and/or the incumbent: (1) Job characteristics as reflected by job dimension scores and work quality dimensions, both being based on the Position Analysis Questionnaire (PAQ); (2) job-related interests expressed by job incumbents, as measured by the Activity Preference Questionnaire (JAPQ); (3) the 'match' between an incumbent's job characteristics and his expressed job interests; (4) general affect, measured by a response disposition scale. Data were obtained from 407 workers in 29 jobs in two telephone companies and for the 9 management and 20 non-management jobs. The above-named predictors were used in stepwise regression analyses to predict 'work' and total satisfaction as measured by the Job Description Index (JDI). In addition, the effect of the response disposition was partialled out from the work satisfaction scores of the combined sample.

The report deals with the background and the development of the Position Analysis Questionnaire (PAQ), which was used as the basic job analysis instrument in the research program. The PAQ (Form A) used in the study includes 189 job elements of an essentially 'worker-oriented' nature, these elements generally characterizing work activities of a behavioral nature (or that have strong implications in behavioral terms), and elements that characterize certain aspects of the context within which human work is performed. The job elements of the PAQ have been used as the basis for deriving various sets of job dimensions, and for studies of an exploratory nature that deal with the potential use of the PAQ as the basis for developing synthetically-derived job attribute requirements, and for job evaluation purposes. The report describes the development of the PAQ, Form A, from earlier job analysis instruments, and the more recent development of a modified version of the PAQ, Form B.

The primary purpose of the study was to obtain estimates of the human attribute requirements of the job elements of Form B of the Position Analysis Questionnaire (PAQ). A secondary purpose was to explore the reliability of job-related ratings as a function of the number of raters. A taxonomy of 76 human attributes was used, and ratings of the relevance of these attributes to each of the PAQ job elements were obtained. A minimum of 8 raters and a maximum of 11 raters rated each of the attributes. The mean and median ratings for each attribute as it related to each job element were computed, these serving as the basis for the attribute profiles of the job elements. Measures of the reliability of the ratings were also obtained. (Author)

The investigation is a pilot study relating to the job evaluation of enlisted jobs within the United States Navy. Its specific purposes were those of identifying the job factors which contribute to differences in job values, and of determining the relative importance of each such factor. (Author)

It was hypothesized that, across the spectrum of jobs, there is some underlying 'structure' of human work in terms of the human behaviors involved. The project was directed toward the identification of behaviorial job elements and of their organization into job dimensions, and the exploration of certain possible practical applications of job data based on such job elements or dimensions. The project involved the development of a job analysis instrument called the Position Analysis Questionnaires (PAQ) that included 189 job elements of a worker-oriented nature. Principal components analysis procedures were used in the analyses of two types of data based on the PAQ. One set consisted of job data for 536 jobs that were analyzed with the PAQ. The other consisted of 'attribute profiles' of the job elements; (These consisted of median ratings of the relevance of 67 human attributes to each job element). (Author)

A sample of 179 positions were used. These represented 90 'jobs' for which test score data were available from the U.S. Employment Service on most of the nine tests of the General Aptitude Test Battery (GATB). For each such job, two criterion values were used as indications of the relevance of a given test to the job; one of these was the mean test score of people on the job, and the other was the coefficient of validity of the test. Two methods of deriving job requirements were used. One of these consisted of a regression analysis using scores to predict the two criteria; this method resulted in rather substantial multiple correlations with both criteria for most of the GATB tests. The other method consisted of deriving job attribute indexes for each of 67 attributes (these having been based on ratings), and subsequently correlating these across the 179 positions, with the two criterion values for each test. This method resulted in lower, and more variable correlations with both criteria. The results of the first analysis suggest quite strongly the possibility of being able to develop valid sets of job requirements on the basis of essentially job data such as that obtained with the Position Analysis Questionnaire (PAQ). (Author)

The study was designed to investigate the hypothesis that there is some structure underlying the domain of human work, and that this structure can be defined in terms of one or more sets of job dimensions. The basic approach to the derivation of these dimensions involved the characterization of job activities and work situations in behavioral or 'worker-oriented' terms using a job analysis instrument known as the Position Analysis Questionnaire (PAQ). Two major data sources were developed and structured in terms of the behavioral job elements comprising the PAQ. One set of data consisted of 536 jobs analyzed with the PAQ, while the second set of data consisted of the ratings of the relevance of 67 different human attributes to each of the job activities and work situations described in the PAQ. Three different multivariate procedures were used to construct several sets of job dimensions. There were noticeable similarities between all of the dimensions, and it was concluded that there is a certain structure to the world of work that can be identified. Implications for the use of such dimensions, particularly in the synthetic validity context, are noted. (Author)

The basic hypothesis was advanced that job-related behavior is importantly and predictably related to going rates of monetary compensation of jobs. To test this hypothesis, job data were obtained with the Position Analysis Questionnaire (PAQ), a structured job analysis format, for a sample of 340 jobs from 45 organizations. The job sample was randomly divided into two sub-samples (A and B) and three types of possible predictors were derived from the job analysis data obtained with the PAQ; these data were then subjected to a stepwise regression analysis to identify optimum sets of predictors of the wage or salary rates for these jobs. All three types of predictors yielded substantial multiple and cross-validation correlation coefficients with wage and salary rates for the jobs in the samples. In addition, it was generally found that predicted compensation rates more closely corresponded to observed rates at the lower end of the compensation scale, than at the upper end. It was suggested that with additional research, this general method might possibly be used as the basis for the development of an operational job evaluation system applicable on an across-the-board basis to jobs of many types. (Author)

The report describes one phase of a research program which is, in part, directed toward synthetically establishing job requirements by the use of a structured job analysis format and the rated attribute requirements of each of the job elements comprising the format. The phase described in this report involved: (1) the selection of human attributes relevant to job performance; (2) the obtaining of ratings of the relevance of those attributes to the job elements of the position analysis questionnaire; (3) the analysis of the reliability of such ratings; and (4) the derivation of attribute requirements profiles for the individual job elements on the basis of several sets of ratings. (Author)

This study was carried out to explore the extent to which current pay rates of jobs can be predicted on the basis of worker-oriented job variables. For this purpose the Worker Activity Profile (Gordon and McCormick, 1963) was used. It consists of 162 job activities of a check-list or rating scale nature that can be used in analyzing any job. The check-list items have numerical values of 0 and 1, and the rating-scale items values of 1 to 7. A sample of 253 jobs was used in this study; these jobs were selected because national wage or salary data had been published for them by the Bureau of Labor Statistics. The national (weighted) average rates of pay for these jobs were used as a criterion. The total sample was divided into an experimental sample (154 jobs) and a hold-out sample for cross-validation (99 jobs). Analyses with the experimental group were carried out to identify the items of the Worker Activity Profile which were related to rates of pay. Three combinations and weightings of such items were then used with the hold-out sample. While the three cross-validation procedures resulted in significant correlations, they were of only moderate size. For example, seven items, with statistical weights, gave a correlation with pay rates of .36. Various factors which might serve to attenuate the magnitude of the relationship were discusses. (Author)

Semantic problems in job description were explored. An attempt was made to classify verbs used in job descriptions along a continuum of orientation, where the end points were defined as: (1) Job-Oriented--those words which characterize what is accomplished in a work activity in terms of job objectives, but do not specifically characterize what the worker is doing. (2) Worker-Oriented--those words which characterize what the worker is doing, that is, the human behaviors that are involved in the job. A sample of approximately 1000 verbs was extracted from over 4000 job definitions of the U. S. Employment Service. These words were subjected to two sorting procedures to identify those which had some degree of worker orientation associated with them. A total of 300 words were identified in this manner and these were judged along a 7-point scale of orientation. This scale was constructed in such a manner that subjects could indicate the inapplicability of a work to this type of classification. The distinction between 'job orientation' and 'worker orientation', as applied to verbs relating to human work activities, was one which was differentiated, to a statistically significant degree, by both naive and sophisticated raters. The pooled judgments of several raters resulted in categorizations which had very substantial reliability.

This study was carried out in order to explore the possibility of the establishment of job requirements. The procedures followed in this study involved the development, for each job, of attribute scores for various human attributes. These were derived as follows: (1) Identification of attributes to be used (42 were selected for this purpose). (2) Rating of each activity of the Worker Activity Profile in terms of the degree to which each such attribute would be required. (3) Derivation of an attribute weight for each attribute on each item of the Worker Activity Profile (a single value for each attribute was derived from the several corresponding ratings). (4) Analysis of each job (in a sample of jobs) using the Worker Activity Profile. (5) Computation, for each job, of an attribute score for each attribute (this score was based on the attribute weights for the various items of the Worker Activity Profile as multiplied by the item value for the job, with these cross-product values summed across all items that were analyzed as being applicable to the job). (Author)

This report summarizes a series of exploratory studies relating to worker-oriented job variables. These variables include worker behaviors in jobs and certain job context variables (situational and environmental) that have relevance to human work, as contrasted with job-oriented variables that deal more with the technological aspects of jobs. It is postulated that human work can be characterized in terms of either type of variable. However, for purposes of establishing common denominators across jobs of varied technological types, one must describe jobs more in terms of human behaviors and related variables (what are referred to here as worker-oriented variables).

The present report concerns a further investigation of the dimensions of worker-oriented job variables through use of the Worker Activity Profile (WAP). In this study, factor analyses similar to Gordon and McCormick's were conducted, using a job sample based on percentages of people within various occupational areas. The resulting factors were then compared with Gordon and McCormick's by use of the coefficient of congruence. The results of six separate factor analyses of WAP item groups were judged to be generally meaningful. Of the 32 factors obtained in this study, 16 were classified as comparable to factors from Gordon and McCormick's study, two others appeared from inspection to be similar to two of their factors, and eight factors showed a noticeable relationship with factors obtained by Gordon and McCormick. It was concluded that there are fairly stable and meaningful dimensions of jobs that characterize work in worker-oriented terms. These dimensions are sufficiently structures so that they emerge through factor analysis despite differences in the choice of bases for job samples used. (Author)

Three different procedures were used for deriving 'attribute indices' for job factors, these attribute indices being derived for 14 different aptitudes, motor abilities, and physical capacities. These attribute indices, in turn, were used experimentally in two phases of the study. In one phase they were correlated with worker trait requirements of 401 jobs as published by the U. S. Employment Service. These published job requirements were given for the same attributes. The correlations were all significant, and of essentially the same magnitude, as those obtained in the previous study (in which the 162 items were treated individually). In the second phase, the indices of four attributes were derived for 43 Naval ratings (jobs). These, in turn, were correlated with test validity coefficients from 125 test validation studies with these 43 Naval ratings. This analysis resulted in significant relationships for most of the tests and subtests. The results of this analysis were generally more positive than a corresponding type of analysis in the previous study of 140 varied industrial jobs for which published validity data were available. In general, the results of this study lend encouragement to the possibility of being able to establish job requirements through what are essentially job analysis procedures. (Author)