6
Psychological Reports, 1976, 38, 1175-1180. @ Psychological Reports 1976 SELLING AND THE SALESMAN: PREDICTION OF SUCCESS AND PERSONALITY CHANGE1 ALLEN A. TURNBULL, JR.' Lynchburg College and Randolph-Macon IVoma~r's College Summary.-By administering a personality inventory and atcitude ques- tionnaire to 201 college-student salesmen prior to and after a selling experience, an attempt was made to (1) measure the predictive value of several personalicy and atcitudinal variables in regard to sales success and (2) assess any personality changes after a short selling period. None of rhe main predictor variables (Ex- traversion, Self-esteem, and Machiavellianism) correlated significantly with the nvo criteria of sales success. Combining the variables via a discriminant function analysis did not result in significant discrimination of sales success among salesmen. However, at the end of the selling period, all salesmen showed a significant increase on the exrraversion (p < ,005) and self-esteem scales (p < ,005). Several possible explanations for these results were discussed. McMurray (1961) stared that ". . . selling as an art and science has made litrle progress since the early days of the industrial revolution" (p. 113). There have been almost as many traits of sales success identified as there have been predictive studies, and the artempt to relate personality scales to job-perform- ance criteria has come under some criticism (see Guion & Goctier, 1965). Nevertheless, an assessment of the predictive validity of three personality and attitude variables for sales success was undertaken in a sample population thought most likely to demonstrate a relationship between sales success and personality variables. The three factors studied were introversion-extraversion, self-esteem, and Machiavellianism. The first two factors have been frequently cited in the literature as related to sales success (e.g., Anderson, 1929; Korn- hauser, 1933; Harrell, 1960; Miner, 1962). The Machiavellianism scale designed by Christie and Geis ( 1970) taps a person's general strategy for dealing with people and the degree to which he feels other people are manipulative in interpersonal situations. Individuals scoring high on this trait (high Machs) should show a tendency to exploit situations and others for self-gain. The Mach scale has not, to the author's knowledge, been tested for its predictive value for sales success, although Christie and Geis (1970) stated that it is theoretically consistent for high Machs to be better salesmen. 'The research reported here was part of a larger study submitted in partial fulfillment of requirements for the M.A. degree to the College of William and Mary faculty. The author wishes to express his gratitude and indebtedness to Virgil McKenna. Kelly Shaver, and Stanley Williams for their help and guidance throughout the studv. A brief version of this was presented at the ~anaaian ~s~cholo~~cal ~ssociatio~ Annual ~ e e t i " ~ , Windsor. Ontario. Tune. 1974. '~equest; for rep& ihduld be sent to Allen A. Turnbull, Department of Psychology, Lynchburg College, Lynchburg, Virginia 24501.

SELLING AND THE SALESMAN: PREDICTION OF SUCCESS AND PERSONALITY CHANGE

  • Upload
    allen-a

  • View
    215

  • Download
    3

Embed Size (px)

Citation preview

Psychological Reports, 1976, 38, 1175-1 180. @ Psychological Reports 1976

SELLING AND THE SALESMAN: PREDICTION OF SUCCESS AND PERSONALITY CHANGE1

ALLEN A. TURNBULL, JR.'

Lynchburg College and Randolph-Macon IVoma~r's College

Summary.-By administering a personality inventory and atcitude ques- tionnaire to 201 college-student salesmen prior to and after a selling experience, a n attempt was made to (1) measure the predictive value of several personalicy and atcitudinal variables in regard to sales success and ( 2 ) assess any personality changes after a short selling period. None of rhe main predictor variables (Ex- traversion, Self-esteem, and Machiavellianism) correlated significantly with the nvo criteria of sales success. Combining the variables via a discriminant function analysis did not result in significant discrimination of sales success among salesmen. However, at the end of the selling period, all salesmen showed a significant increase o n the exrraversion ( p < ,005) and self-esteem scales ( p < ,005) . Several possible explanations for these results were discussed.

McMurray (1961) stared that ". . . selling as an art and science has made litrle progress since the early days of the industrial revolution" (p. 113). There have been almost as many traits of sales success identified as there have been predictive studies, and the artempt to relate personality scales to job-perform- ance criteria has come under some criticism (see Guion & Goctier, 1965). Nevertheless, an assessment of the predictive validity of three personality and attitude variables for sales success was undertaken in a sample population thought most likely to demonstrate a relationship between sales success and personality variables. The three factors studied were introversion-extraversion, self-esteem, and Machiavellianism. The first two factors have been frequently cited in the literature as related to sales success (e.g., Anderson, 1929; Korn- hauser, 1933; Harrell, 1960; Miner, 1962).

The Machiavellianism scale designed by Christie and Geis ( 1970) taps a person's general strategy for dealing with people and the degree to which he feels other people are manipulative in interpersonal situations. Individuals scoring high on this trait (high Machs) should show a tendency to exploit situations and others for self-gain. The Mach scale has not, to the author's knowledge, been tested for its predictive value for sales success, although Christie and Geis (1970) stated that it is theoretically consistent for high Machs to be better salesmen.

'The research reported here was part of a larger study submitted in partial fulfillment of requirements for the M.A. degree to the College of William and Mary faculty. The author wishes to express his gratitude and indebtedness to Virgil McKenna. Kelly Shaver, and Stanley Williams for their help and guidance throughout the studv. A brief version of this was presented at the ~ a n a a i a n ~ s ~ c h o l o ~ ~ c a l ~ s s o c i a t i o ~ Annual ~ e e t i " ~ , Windsor. Ontario. Tune. 1974. '~equest ; for rep& ihduld be sent to Allen A. Turnbull, Department of Psychology, Lynchburg College, Lynchburg, Virginia 24501.

1176 A. A. TURNBULL, JR.

It was predicted that extraversion, self-esteem, and Machiavellianism would all be positively correlated with sales success. It was additionally pre- dicted that high self-esteem, high Mach extraverts would be the best salesmen and that low self-esteem, low Mach introverts would be the worst. If indeed these personality variables are important, then someone possessing all of the seemingly desirable characteristics relating to sales success should be better at selling than someone just possessing one or two of the proposed traits relating to sales success.

A second aim of the present study was to measure any personality change after a prolonged period of door-to-door selling on two variables, extraversion and self-esteem. Daily approaching strangers in a door-ro-door selling situa- tion would demand a large degree of social extraversion. It was predicted that, after engaging in characteristic extraverted behavior for a period of time, sales- men should generally become more extraverted. I t was further predicted that only successful salesmen should show an increase in self-esteem. The unsuc- cessful salesmen should retain their original score or possibly have lower self- esteem scores.

From colleges/universities across the U.S.A. 201 male student salesmen completed the initial inventory during the firsr morning of a week-long train- ing session. The age range was 16 to 28 yr.; the median age was approximately 19. Participation was voluntary, and 90% of the students completed the con- fidential inventory. All subsequent testing was restricted to this initial pool of salesmen.

This particular company sold only during the summer months through college students. The company published a variety of books ranging in price from $9.95 to $29.95. All salesmen were paid solely on a commission basis; the company paid no expenses even during training and setting up in one's territory. After the training session, salesmen were assigned to territories within one of approximately 28 states from the Midwest to the Atlantic Seaboard. Since territories were essentially randomly distributed to all salesmen, territory differences should not affect the over-all correlations between the predictor variables and sales success. Finally, all selling was door-to-door over wide cross-section of socioeconomic neighborhoods.

Materials

All of the experimental measures were obtained either from the initial multiple inventory (pretest) or from [he multiple inventory (posttest) sent to salesmen still selling at the end of the summer. The pretest consisted of: (1) the Eysenck Personality Inventory-Form A (Eysenck & Eysenck, 1968), ( 2 ) the Self-esteem Scale, (Rosenberg, 1965), and ( 3 ) the Mach V Scale (Christie

SELLING, SUCCESS, AND PERSON ALI'IY 1177

& Geis, 1970). The posttest consisted of: ( 1 ) the Eysenck Personality Inven- tory-Form B, ( 2 ) the Self-esteem Scale, and (3) various questions concerning salesmen's summer of selling and their attitudes toward the company.

Procedure

Salesmen completing the pretest had previously signed a contract to sell with the company, so they were assured a job. The posttest was mailed to each salesman in the original sample who was still selling at the end of the summer ( a 12-wk. period). Salesmen returned the posttest to the author in an enclosed, self-addressed, stamped envelope. I t was reiterated on the cover sheet that all information would remain confidential and would not be released to the com-

pany. There were two criteria used for measuring sales performance: ( 1 ) the

total wholesale business during the 3-mo. period, June through August (whole- sale business) and ( 2 ) the wholesale business divided by the total number of calls over the summer (sales index). It was hoped that the latter measure would indicate a salesman's ability relatively uncontaminated by amount of effort expended. The company supplied the measures of criteria and other data needed for analysis of sales success.

RESULTS AND DISCUSSION Prediction of Saler Succesr

Of the 201 salesmen completing the initial inventory, 180 salesmen's in- ventories were storable. Of these 180 salesmen, 111 were still selling at the end of the 12-wk. selling period, and the prediction of sales success was based on these 111 salesmen.

Visual inspection of the two criteria of sales performance and the three personality ~redictor variables showed that all closely approached normal dis- tributions, therefore, no data transformations were necessary. Of the three pre- dictor variables, none showed a significant correlation with sales success as measured by the wholesale business criterion. The Pearsonian correlations for all salesmen completing the summer are given in Table 1. Since over 20 correlations were obtained, only coefficients significant at the .O1 level were accepted as meaningful in order to guard against chance relationships.

Two variables that, not surprisingly, were correlated with wholesale busi- ness were hours worked ( r = .47, p < .01) and number of calls ( r = .32, p < .01). The.old adage "there is no substitute for hard work" seems to have been substantiated.

Since certain personality traits may have been differentially important for nmv as opposed to veteran salesmen, the sample was divided into first-, second-, and third-year men. As can be seen in Table 1, the first-year men had essentially the same correlational pattern as all salesmen combined. N o consistent sig-

A. A. TURNBULL, JR. .

TABLE 1 WHOLESALE BUSINESS AND SALES INDEX CORRELATIONS

Variable All Salesmen First Year Second Year Third Year (N = 111) (n = 85) (n = 15) ( n = 11)

Extraversion -.Ol (.03) .OO (.07) -.2G (-,411 .30 (.38) Self-esteem .02 (-.12) -.I4 (-.05) .09 (-.I71 -.50 (-.55) Machiavellianism .OG (.02) .oG (.07) -.48 (-,461 .36 (.34) Total Calls .32* .50t .52 .12 Total Hours .48* . 5 7 t .53 -.41

Note.-All probabiliry levels are for two-tailed tests based on the Pearson 7 critical levels. Coefficients in parentheses represent sales index correlations. *p < .01. tp < .OOl.

nificant correlations were found across years for any one personality trait or attitude measure.

A second measure of sales performance, sales index, was obtained by dividing the wholesale business by total number of calls made. This index was not another independent criterion since calls were significantly correlated with wholesale business. Without extracting this factor, wholesale business was cor- related with the sales index .63, but it might be a possible indicator of sales- men's sales "ability" as opposed to sales "results." However, the sales index yielded no significant correlations with either the personality or attitudinal measures ( see Table 1 ) .

So far, analyses of the data have been restricted to linear models. TO check for any curvilinear relationships, the three personality variables were plotted against both measures of sales performance. The scatter plots ap- proximated a random distribution for all variables.

In order to check on the summary prediction that salesmen possessing all of the "desired" personality characteristics would sell more or possess more sales ability, several discriminant function analyses were performed on the data. Accurate placement into groups scaled according to success (Low, Medium, High), using both criteria, did not occur in any of the discriminant function analyses.

A review of the predictive sales results suggests chat the personality variables measured did not account for much of the variance associated with sales volume or sales ability. Most of the predictive analyses used only the data for first- year salesmen, but there seemed to be no difference between the personality scores of dropouts (salesmen not selling after 8 wk.), repeaters, and first-year men (see Table 2). There was, however, an obvious difference in sales be- tween first-, second-, and third-year salesmen. The correlation between years with the company and wholesale business was .62 ( p < ,001). In view of the present results this increase would be attributed simply to increased experi- ence and continued motivation rather than to any personality characteristics the returning salesmen possessed.

SELLING, SUCCESS, AND PERSONALITY

TABLE 2 MEANS (AND STANDARD DEVIATIONS) FOR DROPOUTS, FIRST., SECOND-,

AND THIRD-YEAR SALESMEN

Dropouts Firsr Year Second Year Third Year ( n = 6 9 ) ( n = 8 5 ) ( n = 15) ( n = 11)

Wholesale Business (Dollars) 2078 4484 5792 (1368) (2406) (2203)

Extraversion (EPI score) 12.5 13.2 13.9 l2.G (3.4) (3.5) (3.4) (3.4)

Machiavellianism (Mach V score) 87.7 88.6 88.3 94.1 (15.1) (13.3) (15.6) (23.7)

Self-esteem (Rosenberg Scale score) 15.1 14.8 16.9 15.9 (3.8) (3.4 1 (2.9) (2.0)

Calls (Entire 1 2 wk.) 1821.4 1587.5 1949.8 (835.7) (533.9) (247.0)

Hours (Entire 12 wk.) 659.5 729.9 722.7 (211.4) (265.3) (77.7)

Pe~sonulity Change T w o personality variables were remeasured at the end of the summer o n

the basis of 51 returned posttests. The prediction that all first-year men would show an increase in extraversion as a result of forcing themselves to knock on 4 0 to 6 0 strange doors a day was supported. All preselling measures were from Form A, and all posttest measures were from Form B of the Eysenck scale. Since a score difference of 2.1 becween Form A and Form B is reflected in the norms of American college students (Eysenck & Eysenck, 1768) , 2.1 was sub- tracted from every posttest Extraversion score. All analyses were performed on these transformed data. A repeated-measures analysis of variance (Kirk, 1969) gave a significant pre-post difference (F1,33 = 10.5, p < .005). N ~ n ~ u c c e ~ ~ f u l salesmen showed a n increase from 13.0 to 13.9 on the Extraversion scale, and successful salesmen showed a n increase from 11.7 to 14.0. T h e interaction was not significant. Success was operationally defined as $2000 or more whole- sale business.

I t was also predicted that only successful salesmen would show an increase in self-esteem. The successful salesmen increased on the average from 14.9 to 17.7. T h e unsuccessful salesmen increased an average of 1.2 points (15.4 to 16.0). Whi le the pre-post main effect was significant (F1.33 = 11.1, p < .005) , the interaction was not (Fl,aa = 1.7, n.s.), yielding only partial con- firmation of the prediction, although the data were in the predicted direction.

In retrospect, of the two predictions of personality change, one was con- firmed and one was partially confirmed. All salesmen became more extraverted and had a better opinion of their capabilities as measured by their self-esteem scores. I t seems normal to expect an increase in social extraversion as measured on the Eysenck scale, since the salesmen had been engaging in characteristically extraverted behavior over the past 12 wk. However, the situational specificity

1180 A. A. TURNBULL, JR.

of the setting should be kept in mind; the more extraverted salesmen may not be more extraverted once they return to rheir college settings and different norms are operating. It is also probably not unusual to find a slight increase in self-esteem scores for the unsuccessful salesmen, since simply completing the summer away from home selling door-to-door every day would be cause for satisfaction.

These data indicate overwhelmingly that in the present instance the internal personaliry and attitudinal factors measured were of minor importance in salesmanship; external factors such as hours worked and number of prospects contacted were of major importance. A generalization (keeping in mind the sample) would be that anyone who is willing to learn correct sales techniques and is motivated to work diligently stands a good chance of being successful. This should be heeded by companies spending a disproportionate amount of rheir resources on personnel selection. It may be more prudent and profitable to develop better training programs and work incentive plans rather than engage in intensive personnel selection procedures.

REFERENCES ANDERSON, V. Psychiatry in industry. New York: Harper. 1929. CHRISTIE, R., & GEIs, F. Studies in Mdchiauellianism. New York: Academic Press,

1970. EYSENCK, H. J., & EYSBNCK. S. B. G. Eysenck Personality Inventory, mnnual. San

Diego, Calif. : Educational and Industrial Testing Service. 1968. GUION, R., & GO'ITIER, R. Validity of personality measures and personnel selection.

Personnel Psychology, 1965, 18, 135-164. HARRELL, T. W. The relation of test scores to sales criterion. Personnel Psychology,

1960, 13, 65-69. KIRK, R. F. Experimental design: procedrbres for the behavioral sciences. Belmont.

Calif. : Brooks/Cole, 1969. KORNHAUSER, A. W. Personality measures of good and poor salesmen. Psychological

Bulletin, 1933. 30, 730-731. MCMURRY, R. W. The mystique of super-salesmanship. Harvard Busines~ Revinu,

1961, 39(2), 113-122. MINER, J. B. Personality and ability factors in sales performance. Journal of Applied

Psychology, 1962, 46, 6-13. ROSENBERG, M. Society and the adolescent self-image. Princecon, N . J . : Princeton

Univer. Press, 1965.

Accepted March 31, 1976.