Sales Success and Job Satisfaction

IT seems to be accepted as a matter of great faith in contemporary industrialrelations theory, research, and applied activity that employee efficiency is directly related to job satisfaction, although the relationship between the two is not perfect, to be sure. There is, of course, considerable empirical evidence to merit such faith.' It is the intent of this paper to shed light on this relationship in terms of additional empirical findings. This paper describes a follow-up scale analysis performed in conjunction with an attitude survey of the sales personnel of a large candy company. The original study entailed a gross statistical analysis of the various components of job satisfaction, as indicated by the responses of the sales organization to a mailed questionnaire, particularly as the responses were related to relevant status characteristics of the respondents. In all, 1123 questionnaires, out of an original total of slightly in excess of 1600, were received by the investigators. The company furnished the investigators with sales "ratings" of its truck, or route, salesmen, all of whom worked on a directcommission basis. These ratings, based upon mean weekly sales, classified the salesmen as "superior," "excellent," "average," "below average," or "poor." The delineations were not entirely satisfactory as indices of sales efficiency, since an individual salesman might have been working at or near maximum efficiency but in a territory of limited economic potential, resulting in a rating of "average" or below. However, these were the only available efficiency ratings, and there was evidence to indicate that such exceptions were few in number. Excluding supervisors and special salesmen, a total of 805 salesmen were classified in this fashion, with 196 qualifying as "superior," 200 as "excellent," 104 as "average," 196 as "below average," and 109 as "poor." The questionnaire items considered to be reflective of job satisfaction, together with their percentage breakdowns by efficiency rating, are presented in Table 1. The responses of the 805 salesmen to questions 1, 3, 7, 9, 11, and 13 seemed to indicate that "average" salesmen exhibited slightly greater job satisfaction than did those salesmen rated above and below them. As is well known, however, crude statistical analyses of this sort are susceptible of considerable unreliability. Therefore, it was deemed advisable to use a more refined analytical technique in order to ascertain whether "average" salesmen of this concern were significantly more satisfied with their positions. The responses of a randomly selected sample of 100 such employees were arbitrarily scored and placed upon a scalogram board.2 If the criterion of 90 per cent reproducibility were used for including an individual question in the scale,3 none of the thirteen questions could be included. A visual inspection of the board indicated that a number of questions had an inconsistent pattern of responses for highand low-ranked respondents. These items were eliminated from the board, leaving seven questions. After rescoring and reranking, visual patterns of reproducibility immediately developed. However, the reproducibility of the individual questions remain-