An improved job dimension scale to measure job satisfaction in sales reps

Journal of the Academy of Business and Economics, March, 2004 by James H. Turner, Gene Brown

Similarly, the extrinsic scale was expanded to 12 items by including five additional items (see Table IV, below). These additional items more completely describe the constructs as defined by Herzberg (1966) and Deci (1985).

3. ANALYSIS SAMPLE

Survey instruments were returned by 277 employees representing 31 firms operating at 41 locations. The resulting response rate was 61.8 percent. Several returned instruments were incomplete or were deemed unusable for other reasons. It addition, one manufacturing company was eliminated from analysis after it was determined that their employees did not receive any contingent or performance-based compensation and had no customer contact. Moreover, this company was the only participant from that industry. Sampling frame information is summarized in Table V.

After eliminating unusable instruments and excluding the manufacturing company, the resulting sample included 28 companies with 255 employees, operating at 36 locations. This represents 56.9 percent of the sampling frame. The analytical sample information is presented in table VI.

The analysis sample was comprised of approximately 60 percent sales representatives, 8 percent sales managers, 17 percent bank employees, 2 percent finance & insurance representatives (at auto dealerships), 10 percent technicians (auto mechanics), and 2 percent service managers. Slightly more than 1 percent could not be categorized.

In this sample, approximately 28 percent (71) were salaried or hourly personnel (zero compensation contingency), 34 percent (87) were paid entirely by commission (100 % compensation contingency), with the remaining 38 percent (97) distributed between 1 percent and 85 percent compensation contingency. The mean compensation contingency was 47.6 percent.

Seventy-seven percent of the sample were Caucasian, 14 percent were African-American, 5 percent were Hispanic, 1 percent were Asian/Pacific Islanders, and 5 percent reported 'other' (mostly native American). The sample was 63.5 percent male and 36.5 percent female. Other demographic information is shown in table VII.

The sample was approximately 61 percent married, 22 percent single, 16 percent divorced, and 2 percent widowed (rounded).

Reliability of measurement scales is a prerequisite for the validity of measures. A comparative summary of scale reliabilities contrasting prior measures with those developed in this study is presented in Table VIII. All scale reliabilities, as measured by Cronbach's alpha, were greater than recommended by Cronbach (1951)i.e. > ~.70.

5. JOB SATISFACTION--DISTRIBUTION

Job satisfaction was not normally distributed. The mean job satisfaction score was 5.55 with a standard deviation of 1.01. The mode was 6.59. The minimum job satisfaction score was 2.38; the maximum score was 7.00. Intrinsic job satisfaction had a mean of 5.88 with a standard deviation of 1.01. Extrinsic job satisfaction had a mean of 5.28 and a standard deviation of 1.14. Lucas, et al (1987) used a slightly shorter version of these scales in a study of sales force turnover. In two samples of insurance salespeople (1,412 and 1,045 sales representatives) they developed means of 5.32 and 5.52 for intrinsic job satisfaction and means of 4.81 and 5.08 for extrinsic job satisfaction (after converting to seven point scales). The standard deviations developed in their study were also similar to those in the present study: .87 and .75 for intrinsic job satisfaction and 1.88 and 1.77 for extrinsic job satisfaction. The weighted means for their combined scores were 5.07 and 4.82. The combined-scale mean of 5.55 in the present study would indicate a high level of job satisfaction.