AN iNVESTlGATiON ENTG THE PERFORMANCE ARE FEEDBACK " MONITORtNG ABEUTEES 0F ‘ ‘SALESMEH USlfiG SELECTE: INTERACTION VARIABLES Thesis for the Degree of Ph. D. MiCHIGAN STATE UNIVERSITY DGNALD WILLIAM JACKSON. JR. 1973 LIBRARY Michigan Stall: Univem'ty This is to certify that the thesis entitled AN INVESTIGATION INTO THE PERFORMANCE AND FEEDBACK MONITORING ABILITIES 0F SALESMEN USING SELECTED INTERACTION VARIABLES presented by Donald William Jackson, Jr. has been accepted towards fulfillment of the requirements for Ph.D (1991.99 in Business — Marketing Major professor 0-7539 7 1 2: amomo av ~ HMS & SIJNS' liEA’l‘Wfi‘LJEE- ' ‘ llfl’mflfl- Em: ml ABSTRACT AN INVESTIGATION INTO THE PERFORMANCE AND FEEDBACK MONITORING ABILITIES OF SALESMEN USING SELECTED INTERACTION VARIABLES BY Donald William Jackson, Jr. Marketing scholars have been concerned with finding traits of the successful salesman for nearly six decades. Traditional studies concentrated on the salesman in iso- lation. They searched for quantifiable differences in background, personality, or motivation which would effec- tively differentiate between high and low perfonming salesmen. A more modern viewPoint examines selling from an interaction or communications standpoint. Therefore, the salesman is viewed as only one part of a dyad. To inter- act, he must send and receive messages from the prospect. The problem, then, is to isolate those traits possessed by high performing salesman and salesmen who are able to effectively monitor feedback. Interaction or communica- tions-linked variables were chosen to accomplish this task. These variables are "empathy," other-directedness, self-acceptance and Machiavellianism. Donald William Jackson, Jr. The data were obtained through a series of three questionnaires. First, several accounting machine sales- men from throughout the United States were subjected to an audio visual selling simulation deve10ped by Gary Grikscheit.l Responses to this simulation were later content analyzed to determine the number of verbal and nonverbal cues which the salesman reported. Next, the supervisor of each salesman involved was contacted and asked to complete a questionnaire. Finally, each salesman was asked to complete a similar question- naire. Several significant findings from the research may be noted. The coding method used in the Grikscheit study to content analyze the open-ended responses proved reli- able.2 HOwever, the replication of the significant Grikscheit hypotheses, while almost uniformly in the predicted direction, did not possess the same statistical significance attached to the original research.3 The present study was able to differentiate between effective cue monitoring salesmen and less effective monitors in several ways. Bivariate analysis indicated that the number of verbal cues reported was related positively to "empathy,” while the number of nonverbal cues reported was positively correlated with self- acceptance. The study also showed that high and low performers Donald William Jackson, Jr. could be differentiated in several ways. Performance ranking correlated positively with "empathy” while dollar sales were positively related to other-directedness. Although many of the other relationships were in the pre- dicted direction, none was of a statistically significant magnitude. Factor analyses of the data indicated that "empathy," Machiavellianism and performance ranking were positively related to one factor. Another factor had high negative loadings on performance and self-acceptance and a high positive loading on other-directedness. These findings indicate that "empathy" is important for successful selling, but must be tempered by a task orientation or the salesman may overidentify with the prospect and lose sight of his goals. They also point out the fact that the salesman who is looking to others for his orientation and is unsure of himself, will not be a high performer, since he will not be able to con- centrate on the prospect and satisfy his needs. Canonical analysis was used to determine the maximum correlation between the cue monitoring variables and predictor variables and between the performance variables and predictor sets. It was shown that the canonical r was not significant in either case. Because of the many correlations used in this study, care should be taken when interpreting the results. Donald William Jackson, Jr. Spurious correlations may exist which could be indigenous only to these data. Furthermore, this study was conducted on a sub-sample of one company's salesmen at a particular point in time. Therefore, generalizing to other popula- tions should be done only with great care. Replication of this study is essential to determine its reliability. FOOTNOTES 1. Gary M. Grikscheit. ”An Investigation into the Ability of Salesman to Monitor Feedback." (Unpub- lished Ph.D. dissertation, Michigan State University, East Lansing, 1971). 2. Ibido ' pp. 203-2050 3. Ibid., pp. 18-22. AN INVESTIGATION INTO THE PERFORMANCE AND FEEDBACK MONITORING ABILITIES OF SALESMEN USING SELECTED INTERACTION VARIABLES BY Donald William Jackson, Jr. A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Marketing and Transportation Administration 1973 Capyright by DONALD WILLIAM JACKSON, JR. 1973 ACKNOWLEDGMENTS Several individuals and groups deserve recognition for their contributions to this thesis. This research was conducted with the assistance of the National Cash Register Company, Dayton, Ohio. NCR provided both access to their sales force and many valuable hours of managerial and clerical assistance in data collection. Without this assistance the present research would not have been possible. A very special note of thanks is apprOpriate for Mr. J. M. Boyle. Vice President-~Domestic Sales and Mr. T. H. Biggs, Assistant Vice Presidente-Marketing, Education and Publications at NCR. Their dedication to increased knowledge and administrative support are greatly appreciated. Max Bozarth and Larry HaWkins of the Field Sales Training organization at NCR also provided invaluable assistance in data collection. Dr. William J. E. Crissy, Professor of Marketing and Transportation Administration, Michigan State University iii served as research committee chairman and academic advisor and has guided me since the outset of my doctoral program. This direction has been far more than academic, however. He has served as a close friend and deep inspirational force. This pervasive influence shall continue throughout my life. For this I am indebted to him. Dr. Kenward L. Atkin, Chairman of the Department of Advertising, Michigan State University, as a member of the Research Committee, provided an invaluable source of assistance and encouragement during the course of this research. His "Open door" and willingness to give of his time and energy are deeply appreciated. Dr. Richard J. Lewis, Professor of Marketing and Transportation at Michigan State University, also served on the Research Committee. His methodological assistance has saved many hours of valuable time and his comments and suggestions led to many improvements in this research. Gary Grikscheit. Assistant Professor of Marketing at the university of Utah, also deserves a special note of thanks. He provided access to his selling simulation and many hours of explanation which made this study possible. His willingness to share his findings and methodology are particularly appreciated. I am indebted to Ramon J. Aldag of Michigan State iv University and Carol Good of Arizona State University for their patience and assistance in the computer programming aspects of this research. I would also like to thank my colleagues at Arizona State University, and especially Dr. Robert F. Gwinner, Chairman of the Marketing Department, for their unselfish help in providing me with the time to finish this research. The young ladies of the clerical staff of the Marketing Department who typed the manuscripts under the direction of Mrs. Irene Baker also deserve credit. Also, Mrs. Jo McKenzie of Michigan State university who typed the final draft of this thesis should be commended for her efforts. Her patience, despite my many deadlines, should not be unrecognized. Finally, I wish to thank my wife, Sylvia, for her encouragement and support throughout the duration of this research. TABLE OF CONTENTS LIST OF TABLES. . . . . . . . . . . . . . LIST OF Chapter I. II. III. FIGWES O C C I O O O O O O O O 0 INTRODUCTION . . . . . . . . . . Statement of the Problem . . . Purposes of the Study. . . . . Basic Premises of the Research Null and Alternative Hypotheses. . . . Methodology. . . . . . . . . . Limitations of the Study . . . Potential Contributions of the to Marketing Theory . . . . . Potential Contributions of the to Marketing Practice . . . . LITERATURE REVIEW. . . . . . . . "Traditional" Studies. . . . . Communications and Interaction Independent Variables. . . . . Empathy. . . . . . . . . . Other-Directedness . . . . Self-Acceptance. . . . . . Machiavellianism . . . . . Performance. . . . . . . . . . RESEARCH DESIGN. . . . . . . . . Independent Variables. . . . . I "Empathy" . . . . . . . II Other-Directedness. . . III Self-Acceptance . . . . IV Machiavellianism. . . . vi Study Study Studies Page viii u)h*a)\lO\U| H F'H 17 21 21 34 50 SO 63 66 69 74 82 82 83 84 85 86 Chapter Dependent Variables. . I Number of Verbal Cues. II Number of Nonverbal Cues . III Total Number of Cues IV Sales Manager's Evaluation V Dollar Sales Volume. VI Earnings . . . . Sample Selection . . . Data collection. . . . Analysis of the Data . IV. PRESENTATION OF FINDINGS The Reliability of the Analysis Method . . . Content Testing of the Significant Findings of the Grikscheit Research. Hypothesis l-l Findings. Hypothesis 1-3 Findings. Hypothesis 1-4 Findings. Hypothesis 1—7 Findings. Hypothesis 1-10 Findings Hypothesis 2-1 Findings. Hypothesis 2-3 Findings. Hypothesis 2-4 Findings. Hypothesis 2-6 Findings. Summary of Findings in the Replication . . . . . Nonparametric Analyses Findings Concerning "Empathy". . Summary of Findings on "Empathy" Findings Concerning Other- Directedness. . . Summary of Findings on Other- Directedness. . . Findings Concerning Self- Acceptance. . . . Summary of Findings on Self- Acceptance. . . . Findings Concerning Machiavel- 1 ianism O O O C 0 Summary of Findings on Machiavellianism. Multivariate Relationships . Summary of Multivariate Findings vii Page 87 87 88 88 88 89 89 90 90 92 105 105 106 107 107 108 109 109 110 111 113 114 115 116 116 118 119 121 122 124 124 126 127 133 Chapter V. SUMMARY AND CONCLUSIONS. . . . . . . Truncated Sample . . . . . . . . . Review of the Findings . . . . . . A Caution Regarding the General- izations to be Drawn from These Findings. . . . . . . . . . . . . Implications of the Study to Market- ing Practice. . . . . . . . . . . Implications of the Study to Market- ing Theory. . . . . . . . . . . . Suggestions for Further Research . BIBLIOGRAPHY. . . . . . . . . . . . . . . . APPENDIX A. Questionnaire used on Salesmen in Conjunction with the Audio Visual Selling Simulation, "A Sales Call on the Holt Company" . . . . . . . . Cover Letters and Follow up Question- naire for Selected Salesmen. . . . . Cover Letter and Questionnaire for Sales Managers . . . . . . . . . . . Rules for Coding Responses to the Audio Visual Selling Simulation "A Sales Call on the Holt Company" . Summary Tables for Data Analysis . . viii Page 136 136 138 140 140 143 144 148 156 160 172 184 188 Table LIST OF TABLES The Reliability of the Grikscheit Coding Method. . . . . . . . . . . . . . Comparison of Mean Total "Idea Scores" . Comparison of Mean Nonverbal "Idea Scores" Comparison of Encounter Variances. . . . Comparison of Mean Number of Tactical Changes . . . . . . . . . . . . Correlation Between Salesmen's Responses to Question No. 2 and No. 5. . . . . . . Decline in Mean "Idea Scores". . . . . . Mean Nonverbal "Idea Scores" Over Time . Comparison of Encounter Rating Variances Across All Testpoints in Time I and Time II . . . . . . . . . . . Comparison of Mean Number of Tactical Changes in Time I and Time II. . . . . . Summary of Findings on Grikscheit Replication. . . . . . . . . . . . . . . Hypotheses Concerning Empathy. . . . . . Hypotheses Concerning Other-Directedness Hypotheses Concerning Self-Acceptance. . Hypotheses Concerning Machiavellianism . ix Page 106 107 108 108 109 110 111 112 113 114 115 116 120 123 125 2E 3E 4E Highest Loadings of Ten Variables on Four Orthogonal Factors. . . . . . . . . . Factors for Independent Variables. . . . . Factors for Dependent Variables. . . . . . Correlation Matrix for Performance criteria I O O O O I O O O O O O O O O O O Canonical Correlations . . . . . . . . . . Intercorrelations of Independent Variables, Age and Length of Employment . . . . . . . Kolmogorov-Smirnov TWO SAMPLE TESTS Comparing Salesmen and Sales Managers. . . Summary of Nonparametric Findings. . . . . Factors for Ten Variables on Four Orthogonal Factors . . . . . . . . . . . . Page 128 130 131 132 133 188 189 190 192 LIST OF FIGURES Figure Page 2-1 A List of Desirable Qualities in Salesmen . . . . . . . . . . . . . . . . . . 22 2-2 The Process Model of Communication . . . . . 42 2-3 Need-Satisfaction Theory . . . . . . . . . . 45 2-4 A Possible Relationship Between "Empathy" and Performance. . . . . . . . . . 61 2-5 Criteria for Measuring Performance of Salesmen. . . . . . . . . . . . . . . . . 76 xi CHAPTER I INTRODUCT ION Social scientists, marketing scholars and practi— tioners have long been concerned with isolating those characteristics which would allow them to effectively discriminate between high and low performing salesmen. Armed with this information, they could do a much better job of selecting and training salesmen, thereby increasing sales efficiency and lowering turnover. However, solu- tions to this problem have not been easy. In 1948, one author said that personnel research specialists found the prediction of success in selling to be especially diffi- cult. He concluded that, “the difficulty encountered in devising improved selection techniques for saleswork is probably not equaled in any other group of occupations."1 It seems as though this conclusion could safely be ex- tended to the present with little or no fear of criticism. lEarl A. Cleveland, "Sales Personnel Research 1935- 1945, A Review," Personnel Psychology, Volume 1 (Summer 1948), p. 247. 2 While this is a difficult problem, its solution is one with large payoffs since, "a small increase in selec- tivity of salesmen can lead to a large decrease in control 2 Most marketers know that the costs of ineffi- costs." cient hiring processes are high. One study showed that it may cost as much as $8200 direct out-of-pocket costs to place a trained salesman in the field.3 Historically, many studies have searched for those traits possessed by the successful salesman. The majority of these studies have concentrated on the salesman in iso- lation. They have searched for quantifiable differences in background, personality, or interests which would effectively differentiate between high and low performing salesmen. Webster and Davis say, for example, that, "until recently, personal selling was almost exclusively analyzed from the point of view of the salesman. This 'salesmanship' approach assumed, at least implicity, that the outcome of the sales interview was a function of the the specific actions and abilities of the salesman. . . This traditional emphasis on the salesman's influence 2Amitai Etzioni, Modern Organization (Englewood Cliffs, New Jersey: Prentice Hall, Inc., 1964), p. 64. 3Wall Street Journal, November 19, 1964, p. l. 3 also led to a major concern with isolating the 'traits' of a successful salesman."4 However, these studies have been largely inconclusive and difficult to generalize from. Reasons for this include variations in sales jobs and variations in the personalities of successful sales- men.5 As one writer states, "Sales work is indeterminant. Who knows the right way to sell? Some men sell by in- flating the product; some by kidding it: some sell by being humorous: some by being earnest; some by being aggressive; some by being yes-men. Whatever works is the right way."6 A more modern viewpoint of selling views the selling situation from either a social interaction perspective7 or from a communications standpoint.8 Recent studies 4Kenneth R. Davis and Frederick E. Webster, Jr., Sales Force Management (New YOrk: The Ronald Press Com- pany, 1968), p. 152. 5Raymond W. Mack, "Who Is the Salesman?", 223 Marketing Revolution (Proceedings of the 37th National Conference of the American Marketing Association, December 27-29, 1955), pp. 8-14. 61bid.. p. 12. 7Franklin B. Evans, "Selling as a Dyadic Relation- ship: A New Approach," American Behaviora1_§cientist, Vol. 6, No. 9 (May 1963), pp. 76-79. 8Theodore Levitt, "Communications and Industrial Selling," Journal of Marketing, Vol. 31 (April 1967), pp. 15-210 4 have recognized that the buying situation is, in fact, a social situation and must be looked at as such. This modern viewpoint puts emphasis on prospect-salesman interaction. One writer characterizes interpersonal interaction as behavior influenced by, "how one person thinks and feels about another person, how he perceives him and what he expects him to do or think, how he reacts to the actions of the other."9 Therefore, viewing the salesman in isolation is often not enough. The salesman must be viewed as only 10 Because of this one part of an interacting dyad. thrust toward viewing selling as an interpersonal inter- action, much emphasis has also been placed upon com- munications theory. Communication includes not only the effective sending of messages which are deemed credible and memorable,11 but also the receiving of feedback in an effective manner so that strategies may be altered to fit the needs of the prospect. In 1970, Gary Grikscheit developed a video taped selling encounter which depicted 9Fritz Heider, The Psychology ofglnterpersonal Rela- tions (New York: John Wiley and Sons, Inc., 1958), p. 1. 10Evans, op. cit., p. 76. 11Harold C. Cash and‘W. J. E. Crissy, The ngchology of_§elling, Vol. 12 (1965), p. 62. 5 interaction between salesman and prospect in the account- 12 . . In an experimental env1ronment, ing machine industry. Dr. Grikscheit exposed selected salesmen to this simula- tion and investigated their ability to monitor feedback along with their sales performance characteristics. Dr. Grikscheit and the sponsoring company have made this audio visual selling simulation available for this research. Thus, this study will attempt to eXpand upon the initial Grikscheit research by investigating selected human interaction and communications-linked variables and their relationship to both sales performance and the ability to monitor feedback. Statement of the Problem The prdblem, then, is twofold; first, trying to find human interaction or communications-linked variables which are possessed by the high performing salesman and, second, trying to find variables possessed by the sales- man who is able to effectively monitor feedback. The findings of this research may have a dual impact. First, they may serve as a link between the old, trait-searching 12Gary Michael Grikscheit, "An Investigation of the Ability of Salesmen to Monitor Feedback," (unpublished Ph.D Dissertation, Michigan State University, East Lansing, 1971), pp. 39-46. 6 approach which focused solely on the salesman and the more modern interactive approach by isolating interper- sonal traits which correlate with sales success. Second, they may enrich the modern literature by adding new dimen- sions to the interaction dyad and establishing some ex— ploratory parameters for further investigation. Specifically three major questions are asked: I. Do high effect salesmen differ from low effect salesmen in their: a. Empathic ability? b. Other-Directedness? c. Self-acceptance? d. Machiavellianism? II. Do those salesmen who are able to more effec- tively monitor feedback differ from the less effective monitors with respect to their: a. Empathic ability? b. Other-Directedness? c. Self-acceptance? d. Machiavellianism? III. Do sales managers differ from salesmen with respect to their: . Empathic ability? . Other—Directedness? . Self-acceptance? . Machiavellianism? 006'!” Purposes of the Study The purposes of this research are as follows: I. To attempt to validate the coding method used to content analyze the responses to the audio visual selling simulation used in the Grikscheit study.13 13Ibid.. pp. 83—86. 7 II. To replicate the original Grikscheit research and thus act as a reliability check on the re- sults obtained. III. To refine and expand upon the performance criterion used to differentiate between "high effect" and "low effect" salesmen. IV. To attempt to isolate certain interaction or communications—linked attributes possessed by the "high effect" salesman. V. To attempt to isolate certain interaction or communications-linked attributes possessed by the salesman who is able to do a superior job of monitoring feedback. VI. To attempt to isolate certain interaction or communications-linked attributes possessed by the sales managers of the selected salesmen. Basic Premises of the Research Several premises were used to develOp the following hypotheses: First, the effectiveness of a salesman's communica- tions in the face to face situation is dependent upon his ability to monitor feedback. Second, the "high effect" salesman possesses more of certain social or communications-linked traits than the "low effect" salesman. Third, the salesman who is able to effectively monitor feedback possesses more of certain social or com- munications-linked traits than the salesman who is less able to monitor feedback. 8 Null and Alternative Hypotheses The following are the null and alternative hypotheses being tested in this study. They are broken down into four subgroups: I. Empathy II. Other-Directedness III. Self-Acceptance IV. Machiavellianism I. Empathy IAO There is no relationship between the salesman's empathy and the number of verbal cues he will record in the communications simulation. 1A1 Salesmen who identify more verbal cues in the communications simulation will possess more empathy. IBO There is no relationship between the salesman's empathy and the number of nonverbal cues he will record in the communications simulation. IBl Salesmen who identify more nonverbal cues in the communications simulation will possess more empathy. ICO There is no relationship between the salesman's empathy and the total number of cues he will record in the communications simulation. ICl Salesmen who identify more total cues in the communications simulation will possess more empathy. IDO There is no relationship between the salesman's empathy and his performance ranking. IDl High performing salesmen will possess more empathy than low performing salesmen. IE0 There will be no difference between the amount of empathy shown by the supervisors and the salesmen. IE1 9 Supervisors will show more empathy than salesmen. II. Other-Directedness IIAl IIBO IIBl IICO IICl IIDO IIDl IIE IIEl There is no relationship between the salesman's other-directedness and the number of verbal cues he will record in the communications simulation. Salesmen who identify more verbal cues in the communications simulation will be more other- directed. There is no relationship between the salesman's other-directedness and the number of nonverbal cues he will record in the communications simulation. Salesmen who identify more nonverbal cues in the communications simulation will be more other- directed. There is no relationship between the salesmen's other—directedness and the total number of cues he will record in the communications simulation. Salesmen who identify more total cues in the communications simulation will be more other- directed. There is no relationship between the salesman's other-directedness and his performance ranking. High performing salesmen will be more other- directed than low performing salesmen. There is no difference between the amount of other-directedness shown by supervisors and salesmen. Supervisors will show more other-directedness than salesmen. III. Selqucceptance IIIAO There is no relationship between the salesman's self-acceptance and the number of verbal cues IIIAl IIIB IIIB IIICO IIICl IIID IIIDl IIIEO IIIE IVAO IVAl 10 he will record in the communications simulation. Salesmen who identify more verbal cues in the communications simulation will possess more self- acceptance. There is no relationship between the salesman's self-acceptance and the number of nonverbal cues he will record in the communications simulation. Salesmen who identify more nonverbal cues in the communications simulation will possess more self- acceptance. There is no relationship between the salesman's self-acceptance and the total number of cues he will record in the communications simulation. Salesmen who identify more total cues in the communications simulation will possess more self- acceptance. There is no relationship between the salesman's self-acceptance and his performance rankipg. High performing salesmen will possess more self- acceptance than low performing salesmen. There is no difference between the amount of self-acceptance shown by supervisors and salesmen. Supervisors will show more self-acceptance than salesmen. IV. Machiavellianism There is no relationship between the salesman's Machiavellianism and the number of verbal cues he will record in the communications simulation. Salesmen who identify more verbal cues in the communications simulation will be less Machiavellian. There is no relationship between the salesman's Machiavellianism and the number of nonverbal 11 cues he will record in the communications simulation. IVB Salesmen who identify more nonverbal cues in the communications simulation will be less Machiavellian. IVCO There is no relationship between the salesman's Machiavellianism and the total number of cues he will record in the communications simulation. IVCl Salesmen who identify more total cues in the communications simulation will be less Machiavellian. IVDO There is no relationship between the salesman's Machiavellianism and his performance ranking. 1 High performing salesmen will possess more Machiavellianism than low performing salesmen. IVE0 There is no difference between the amount of Machiavellianism shown by supervisors and salesmen. IVE Supervisors will show more Machiavellianism than salesmen. Methodology The data were obtained through a series of three questionnaires. First, several different groups of sales- men from throughout the United States were subjected to the audio-visual stimulus, "A Sales Call on the Holt 14 Company". This filmed recording was a simulated sales encounter and was developed by Dr. Gary Grikscheit of the 14£2$Q.. pp. 66-71 (may be consulted for a descrip- tion of this). 12 University of Utah. Each salesman viewed a series of sales encounters and after each segment recorded his thoughts and perceptions in a test booklet (see Appendix A). Next, the supervisor of each salesman involved in this research was contacted and asked to complete a questionnaire in which he was to evaluate his salesman, size up his personality, and also record some of his own feelings toward several questions (see Appendix C). Finally, each salesman was asked to complete a similar questionnaire. He was to size up his sales manager's personality and also to record his feelings toward the same set of questions (see Appendix B). The final two questionnaires were divided into four sections. Each section was designed to measure a specific attribute. The first section was designed to measure the empathy between salesman and supervisor. The second section was to find the amount of inner or other- directedness each possessed. The third section dealt with the self-acceptance of the subjects while the final section examined their Machiavellianism. The data Obtained from the questionnaire were analyzed using several techniques. First, the hypotheses were tested utilizing several nonparametric tests. Second, the data were factor analyzed to determine what 13 common factors, if any, might underly the manifest data. Finally, canonical analysis was used to relate the many dependent variables to a combination of the independent variables. Limitations of the Study The limitations of this study, as with any experi- mental research, center mainly around the sampling and data collecting procedures. The sample chosen for this research was not randomly selected. To test the hypotheses under consideration, it was necessary to subject each of the respondents to an experimental environment. Salesmen were subjected to the audio-visual stimulus in a controlled environment during their stay at a company-sponsored sales school. Each sales school class consisted of salesmen from throughout the United States and Canada. Cost con- siderations and the need for a controlled environment prohibited this type of research from being done on a more decentralized basis in the field. Since the test could not be given in the field, all company salesmen could not be considered. Therefore, because those included were, by and large, rather new salesmen to the company sponsoring this research, it is possible that certain more experienced and proficient salesmen were excluded 14 from the sample. This is the primary reason for including sales managers as a separate subgrouping of high perform- ing salesmen. Without the cost and environmental con- straints it might have been preferable to sample entire sales districts. This would have provided two benefits. The performance criterion could have been improved since one supervisor could have evaluated an entire subsample of salesmen, thus eliminating inter-rater variability in evaluations. Also, the sample might have included a greater range of sales ability by including both the "seasoned pro" and the "raw recruit". However, in spite of the non-randomness of selection, a very heterogeneous group of salesmen was included in the research. Salesmen selected for inclusion came from forty-two states and four provinces of Canada, ranging from Hawaii and Alaska to Maine. Another problem with the sample drawn is that it represents only one industry and only one company's sales- men. Therefore, care should be taken when generalizing the research findings to other types of selling. One other type of limitation is inherent in this type of data collection. Because there is a high rate of turnover in this particular type of sales position, and because mail questionnaires were utilized, there were 15 quite a few non-respondents to the questionnaire follow- ing the selling simulation. Even though each of the departed salesmen was sent a questionnaire, the response rate from this segment of the sample was rather low. This segment could have included both very high and very low performing salesmen. Finally, because of the time element involved between the audio visual simulation and the return of the followup questionnaire, there is a possibility that a learning process occurred. That is, a salesman could have learned empathy or self-acceptance, etc., between the time he was subjected to the simulation and the time he returned the followup questionnaire. Potential Contributions of the Study to Marketing Theory There are several potential contributions of this research to marketing theory. This study represents a link between the more traditional studies of personal selling, which sought traits possessed by highly effective salesmen, and the more modern studies, which view selling from a social interaction or communications perspective.15 15Evans, op. cit., pp. 76-79. Levitt, 0p. cit., pp. 15-21. 16 More specifically, this study has several possible con- tributions: I. I1. Since interaction and communications-linked variables are related to effectiveness of per- sonal selling, this research may provide added dimensions to the study of social inter- action and communications as they affect per- sonal selling. Part of this study also will serve as a repli- cation of the Grikscheit research.16 Therefore, it acts as a measure of the reliability of that study. The reliability of the content analysis method is especially important: Content analysis is used as the basic method of retrieving data concerning the ability of sales- men to monitor feedback from the audio visual selling simulation. If this method is not reliable, in other words, if it is not able to yield the same results when repeated under the same conditions, then this potentially valuable method for investigating personal selling is fruitless. Therefore, this part of the research 16 Grikscheit, op. cit. 17 is of special importance. The entire Grikscheit study will also be checked for reliability, using another sample from the same population. III. Another potentially valuable contribution of the study will be to relate a predictive measure of empathy to the ability of salesmen to monitor feedback. Many marketers agree that empathy is an important quality for the salesman to possess,17 yet the measurement of empathy has 18 Therefore, this study, by been evasive. relating empathy to Objective measures of feed- back and performance, will hOpefully provide marketing theory with an added dimension into the relationship between empathy and sales success. Potential Contributions of the Study to Marketing Practice There are also several potential contributions of this research to marketing practice. Marketing 17For example see David Mayer and Herbert M. Green- berg, ”What Makes a Good Salesman," Harvard Business Review, Vol. 42, No. 4 (July-August 1964), pp. 119-125. 18For example see Albert H. Hastorf and Irving E. Bender, "A Caution Respecting the Measurement of Empathic Ability," JOurnal of Abnormal and Social Psycholquo XEVII, No. 2 (April, 1952), pp. 574-576. l8 practitioners may be provided with several more dimensions to narrow the subjectivity inherent in the selection and training of salesmen. The following represent several possible contributions. I. Sales training may benefit from this research. One of the major purposes of this study is to link certain variables to selling performance and feedback monitoring ability. Since several of the interpersonal variables are potentially learnable, e.g., empathy19 and self-acceptancezo, it may be useful for the sales trainer to have a measure of these variables to use as a bench- mark of learning. If these variables do, in fact, correlate with performance and cue monitoring ability, then the benefits of role playing, trait analysis and other behavioral training methods can be measured by using pre- and post-tests. II. Selection techniques may also be beneficiaries 19For example see Robert L. Katz, Empathy: Its Nature and Uses (New YOrk: The Free Press of Glencoe, 1963), p. 61. 20For example see James T. Mangan, How to Win Self Confidence for:§ellipg (Englewood Cliffs, New Jersey: Prentice Hall, Inc., 1957). 19 of this research. If correlates are found for the high performing salesmen, the payoffs are large in improved performance and reduced turn- over rates. Due to the high turnover in the particular industry under consideration, effec- tive screening of candidates is especially im- portant. Since the testing procedures used in this research are easily administered and scored, field utilization of these instruments is a distinct possibility. One problem which would immediately present itself would be the inter- active nature of the traits under study. There- fore, multi-variate procedures might be necessary for discrimination between potentially high and low performing salesmen. This problem could be overcome through the use of a central scoring office or a "canned" computer program at the local level accompanied by decision criteria. III. Another potential benefit from this research is further study into the nature of performance evaluation. Although this area has been re- searched quite heavily21, the conclusions are 21For example see Richard S. Barrett, Performance Ratings (Chicago: Science Research Associates, Inc., 1966). 20 rather tentative. Further study is therefore valuable. Since a number of performance criteria are available, the interrelationships between the various measures may be determined. In practice, performance evaluation is often overlooked as a source of error in sales management. However, it is very difficult to improve performance without a thorough knowledge of the nature of performance evaluation.22 Further knowledge into the various inter-relationships and trade- offs involved in the different performance mea- sures should help the marketing practitioner more effectively evaluate and select salesmen. Since many of the criteria are not correlated, almost any hypothesis concerning sales success can be proved if the right performance criterion is chosen.23 If the marketing practitioner is to make effective decisions, he must know which criterion to use in a particular situation. This research will allow the reader to see which of the variables under consideration is related to each of the performance criteria available. 22Davis and Webster, op. cit., p. 407. 23J. weitz, "Criteria for Criteria," American Psychologist, Vol. 16, (1961), pp. 228-231. CHAPTER II LITERATURE REVIEW The primary Objective of this chapter is to establish a theoretical base for the research by reviewing some relevant literature, thus placing this study into the framework created by past studies. To accomplish this, the following structure will be used. First, the "tradi- tional" trait-oriented approach will be examined. This will be followed by the more modern interaction or com- munications-oriented viewpoint. Third, literature asso- ciated with each of the independent variables will be cited which pertains to the hypotheses being studied. Finally, literature pertaining to performance evaluation will be reviewed. "Traditional" Studies Marketing scholars have been concerned with finding traits of the successful salesman for nearly six decades. PrObably the first serious scholarly attempt to differen- tiate good from poor salesmen on the basis of psychological 21 traits was quality of complexity prise that keters are ferentiate salesman. 22 the work of Walter Dill Scott. Considering the psychological measurement available and the of the task he undertook, it comes as no sur— no significant results were achieved.1 Mar- still searching for qualities which wi.l dif- the high performing from the low performing For example, Tillman and Kirkpatrick have noted several social, mental, emotional, and human qualities needed to succeed in selling. Their list of desirable characteristics for salesmen include: FIGURE 2-1 A List of Desirable Qualities in Salesmen Aggressiveness Ambition Appearance Confidence Courage Courtesy Decisiveness Dependability Determination Empathy Energy Enthusiasm Ethics Friendliness Health Honesty Imagination Initiative Intelligence Interest Intuition Judgment Knowledge Leadership Loyalty Maturity Optimism Perserverance Personality Persuasiveness Poise Rapport Reliability Resourcefulness Responsibility Sincerity Tact Understanding 1Samuel N. Stevens, "The Application of Social Science Findings to Selling and the Salesman." Aspects of Medern Marketing, AMA Management Report #15 (New York: American Management Association, Inc., 1958), p. 86. 2Rollie Tillman and C. A. Kirkpatrick, Promotion: Persuasive Communication in Marketing, Revised Edition (Homewood, Illinois: Richard D. Irwin, Inc., 1972), p. 117. 23 Although there are numerous examples of these lists of qualities which the good salesman should possess, they are, by and large, intuitive and often lack empirical findings to support them.3 On the other hand, there are a great number of empir- ically based research studies showing the physical, psycho- logical, or personality attributes possessed by the good salesman. Webster and Davis state that, "traditional emphasis on the salesman's influence upon the sale led to a major concern with isolating the 'traits' of a successful salesman. This concern was primarily with personality traits rather than situational factors."4 Although these studies have met with limited success, their authors usually warn against generalizing the research findings beyond the salesmen studied. There are several reasons for the failure or limitations of these studies. I. Lack of a consistent or meaningful definition of performance There are nearly as many definitions of performance as there are studies undertaken. Dollar value of sales, sales managers' evaluations or percent of quota are only 3For example, see the comments of Davis and Webster, Op. cit., p. 389. 4Ibid., p. 152. 24 a few of the many criteria used to measure performance. Since many of these measures of selling performance do not have high inter-correlations, great differences in results may appear depending upon which performance criterion is chosen. For instance, a certain trait may correlate very highly with one measure of performance and only slightly with another performance measure of the same selling job. II. Differences in selling jObs. Different selling jObs may require different types 5 For instance, there may be differences be- of salesmen. tween the qualities desired in a tOpnotch retail salesman and those required of an equally well qualified industrial salesman. Different retail selling jobs themselves may also require different types of salesman. Selling auto- mObiles and selling ladies handbags may require different types of individuals. And, there may very well be dif- ferences between new automObile salesmen and used car salesmen. Because of the differences in selling jObs, a 5For instance, Dunnette and Kirchner found an ability measure predicted success among industrial sales- men but not among retail salesmen. See M. D. Dunnette and W. K. Kirchner, "Psychological Test Differences Be- tween Industrial Salesmen and Retail Salesmen," Journal Wm. Vol. 44. No. 2 (1960). pp- 121- 135. 25 factor which may prove a useful predictor of selling suc- cess in one area may be irrelevant for another type of sales position. For example, Kerr and Tabolski found that empathy was a fairly high correlate of performance for new car salesmen but not of used car salesmen's per- formance.6 III. Differences ingprospective buyers. Not only do selling jobs vary but buyers also vary in different selling situations. If the buying process is viewed as an interaction between seller and purchaser, then salesmen may have to possess very different types of skills or traits to make a sale, depending upon the buyer involved. Franklin Evans has stated that, "A sale is a social situation involving two persons. The inter— action of the two persons depends upon the economic, social, physical and personality characteristics of each of them. To understand the process it is necessary to loOk at both parties to a sale as a dyad, not individu— «7 ally. With this in mind, it is easy to see how differ- ences in buyers would necessitate different qualities in 6Francais P. TObolski and Willard A. Kerr, "Predic- tive Value of The Empath Test in AutomObile Salesman- ship," Journal of Applie Psychology, Vol. 36, No. 5 (October 19, 1952), p. 311. 7Evans, Op. cit., p. 76. 26 salesmen. IV. Interaction of traits which cause univariate analysis to fail when used alone. Since many of the variables used in these studies are interrelated and, since many other variables influence the selling situation, there may be a confounding of re- sults. Unless these interactions are studied directly and an attempt is made to identify moderating variables, many meaningful relationships may go unnoticed. V. Lack of consensus on an operational definition of certain traits. There are various definitions for many of the traits used in these studies. There is also a great deal of de- bate as to which of the many instruments available really measures the trait in question.8 Since these debates con- tinue, it is difficult to compare various studies. There- fore, reliable measurement of the presence or absence of a given trait in an individual is difficult. Keeping these limitations in mind, several relevant studies may be examined. Since there are so many studies 8As an example see the controversy over the measure- ment of empathy in Norman Kagan and David R. Krathwohl, Studies in Human Interaction: Interpersonal Process Recall Stimulated by Videotape (East Lansing: Michigan State university, Educational Publication Services, 1967), pp. 460-463. 27 which attempt to isolate desirable qualities in salesmen, the studies cited were merely deemed representative. In- clusion of relevant studies was made on the basis of breadth of areas covered rather than trying to include all studies concerned with a particular area. 9 studied 522 life insurance Clarke and Merenda agents and found that Activity Vector Analyses (AVA), an instrument used to find temperament attributes of respondents, was able to effectively discriminate between successful and non-successful life insurance salesmen. They found that applicants for life insurance sales were not likely to be successful if their self-perceptions were as passive and submissive individuals rather than as socially confident persons. They also found that cer- tain combinations of personal-social data could be used as effective discriminators of success or failure as a life insurance agent. Interestingly, they found that the two measures were also statistically independent and, therefore, could be used in combination to increase predictive efficiency. 9Peter F. Merenda and Walter V. Clarke, "The Pre- dictive Efficiency of Temperament Characteristics and Personal History Variables in Determining Success of Life Insurance Agents,” JOurnal of Applied Psychology, Vol. 43, No. 6 (December, 1959), pp. 360-366. 28 In 1960, Harrelllo studied 21 petroleum salesmen to determine the relationship between certain test scores and three criteria of proficiency in selling. These criterion measures of sales success were; sales manager's evaluations, percent of quota and administrative appraisals of success. He found that production records were the most predictable of the three. Several test scores showed significant differences between high and low-producing salesmen. The tests which effectively discriminated were: 1. The Otis Test of Mental Ability 2. Four scales of the Bernreuter Personality Inventory: Stability, Dominance, Self- Confidence, Aggressiveness 3. Tact and Diplomacy of the Moss-Hunt Social Intelligence Test 4. The Sales Manager Scale of the Strong Voca- tional Interest Blank. His results also showed that the Self-Sufficiency, Objec- tive Mindedness and Sociability scales of the Bernreuter Inventory did not discriminate nor did the sizing up peOple or the judging behavior elements of the Mess-Hunt Test, nor did other elements of the Strong Vocational Interest Blank. In addition, the washburne S. A. 10T.‘W. Harrell, "The Relation of Test Scores to Sales Criterion,“ Personnel Psychology, Vol. 13 (Spring, 29 Inventory did not differentiate between either the produc- . . . 1 tion or other ratings given the salesmen. In 1962, John Miner,12 using a group of sixty-five dealer salesmen employed by a major petroleum company, found that two tests were significant predictors of sales success. The Wechsler Adult Intelligence Scale, Arith- metic Subtest yielded correlations with performance mea- sures in the low .30's. Also, the Thompkins-Horn Picture Arrangement Test13 yielded correlations in the high .SO's. In this second test several measures were found to be associated with successful performance. These were; 1. Dependence--any indication that support in the form of praise, attention, instruction or assistance is a condition for activity or positive affect. 2. Sociophilia--any indication of a wish to be with other peOple or a tendency to maintain close personal relationships with others. 3. Self-Confidence--any indications that success or approval is expected as a function of one's own efforts. llIbid., p. 69. 12John B. Miner, "Personality and Ability Factors in Sales Performance," JOurnal of Applied Psychology, Vol. 46, NO. 1 (1962), pp. 6-13. 13S. S. Thompkins and J. B. Miner, The Thompkins- Horn Picture Arrangement Test (New YOrk: Springer, 1957). 30 4. Happiness--any indications of a characteristic happy mood. On the other hand, low aggression, low sociOphilia and strong superego were found to be associated with lesser performers. Miner states that the successful salesmen all showed an ability to express emotion, espe- cially positive emotion, freely, while the lesser per- formers were inhibited.14 He further notes that these dimensions of emotional freedom should be significant in many types of sales work.15 In a separate study, Miner also shows some evidence indicating that salesmen should be in the top 30th percentile of the general p0pulation in verbal ability in order to be successful.16 James Mosel analyzed 170 women department store sales clerks. He found that twelve personal data items significantly distinguished between high and low selling cost employees. He describes the "ideal" low selling cost saleswoman, in order of discrimination as, "between 35 and 54 years of age, 13 to 16 years of formal education, 14Miner, op. cit., p. 12. 15Ibid., p. 13. 16John B. Miner and John E. Culver, "Some Aspects of the Executive Personality," Journal of Applied Psychology, vol. 39 (OctOber, 1955), pp. 348-353 at p. 352. 31 over five years previous selling experience, over 160 pounds, five years or less in next to last job, lives in boarding house, over five years on last job, minor execu- tive as principal previous experience, between 59 and 62 inches in height, one to three dependents, widowed and no lost time in the last two years."17 Clearly, generalizing from this type of study is dangerous. Yet, the author states that these results, when combined with other findings on department store sales personnel, are useful predictors of jOb success. The prOblem is that many of these variables would be impractical for managerial use in selecting applicants. For example, using weight as a criterion for hiring sales clerks does not have much face validity. In 1960, Cecil French18 conducted a study in which he attempted to find correlates of successful salesman- ship. He studied retail salesmen in a large midwestern city and found that there were no statistically significant differences between high and low producers with respect to 17James N. Mosel, "Prediction of Department Store Sales Performance from Personal Data," Journal of Applied Psychology, Vol. 36, No. 1 (February, 1952), p. 9. 18Cecil L. French, "Correlates of Success in Retail Selling," American Journal of Sociology, Vol. LXVI, No. 2 32 age, time on the job or formal education. His findings did show, however, that high producers did not intend to stay on the present job until retirement. High perform— ance was thus, correlated with upward occupational mObility and a higher reference group than lower perform- ance . Samuel N. Stevens has summarized the contributions of the behavioral sciences to selling. He noted that the following were the major conclusions which social scientists had reached in regard to the social and psychological characteristics of salesmen.19 1. There is no significant relationship between intelligence-test scores and sales success. No significant relationship has been found between independent measures of personality traits and sales success. No correlation exists between age and sales success. There is no correlation between measurable character traits and sales success. There is no significant correlation between level of education and sales success. No significant correlation exists between level of sales activity and sales success among individual salesmen. Each of the above factors has significance when studied in relation to all others in individual salesmen. 19Stevens, op. cit., pp. 86-88. 33 8. Such study as that indicated in point 7 above can provide a useful tool for selection and develOpment. 9. Salesmen are more likely to succeed when chosen with regard to the kinds of customers they will deal with rather than in terms of the types of products sold. 10. Salesmen differ from non-salesmen in four important ways: a. Salesmen are persuasive rather than critical. b. Salesmen are intuitive rather than analytical. c. Salesmen have higher average energy levels (expressed in activity). d. Salesmen are more strongly motivated by the desire for prestige, power, and material gain than by a service ideal or the need for security. 11. Salesmen's interests cluster around a dominantly persuasive common core. Although many of these studies are, by their very nature, quite restricted in sc0pe, there are results that indicate that some tests for salesmen may go across occu- pational or product boundaries. James E. Kennedy used a group of personal, personality, interest and attitudinal variables to see if a general device was as efficient for selecting automObile salesmen as a more specific instru- ment. The more specific instrument was based upon the type of product which the car salesman sold, i.e., luxury car salesman vs. economy car salesman. He found that the more elaborate procedure of develOping specific keys for specific subvarieties of car salesmen did not result in 34 any significant improvement in validity.20 That is, it did no better jOb of selection than the less elaborate procedure of developing a single key for car salesmen, in general, without regard for the various subvarieties of salesmen. It appears that for this restricted subsample and for those subvarieties chosen, product was not an important variable as a criterion of success. General- izing to other populations or to other products, however, would seem very dangerous. Mayer and Greenberg also found that their measures of'empathy and ego drive" were able to differentiate between high and low performing salesmen in a number of industries. They found these tests to be effective in the retail automObile industry, the insurance industry 21 This would seem to indi- and the mutual fund industry. cate that perhaps there are some universals of salesman- ship that go beyond product and industry. Communications and Interaction Studies Having explored some of the traditional studies of the salesman, it can be noted that "these approaches are 20James E. Kennedy, "A General Device vs. More Spe— cific Devises for Selecting Car Salesmen," Journal of Applied Psychology. Vol. 42 (June, 1958), pp. 206-209. 21Mayer and Greenberg, op. cit., pp. 119-125. 35 unidirectional; that is, they look only to the salesman, not to the two person interaction that makes the cale... As opposed to selling, most studies of interviewing in social research have dealt with the interaction problems of interviewer and respondent.“22 Borrowing from the other social sciences, the basic thrust of the more modern interaction viewpoint towards selling is that the salesman can only be viewed as one part of an interacting dyad. Because the salesman is in a position where he must interact with the prospect, it is necessary to view both the prospect and the salesman. This interpersonal interaction has been characterized by one writer as behavior influenced by "how one person thinks and feels about another person, how he perceives him and what he does to him, what he expects him to do or think and how he reacts to the actions of the other..."23 Another author says that the "interaction between the salesman and the prospect can be described as a dual learning situation. While the salesman is learning about 22Franklin B. Evans, "The Sociological Analysis of the Selling Situation: Some Preliminary Findings," Emerging Concepts in Marketing, William S. Decker (Ed.) Chicago: American Marketing Association (December, 1962), p. 476. 23Heider, op. cit.. p. l. 36 the needs of the prospect, the prospect also must become conscious of the present state of his needs. He must then evaluate his current means for meeting these needs and weigh the merits of the products being offered."24 In addition, the interactional demands of the salesman are numerous. "First, the jOb demands a person with a wide' range of behaviors and flexibility in adapting his be- haviors...In short, his role requires him to adapt to a large number of differing personalities with a wide range of possible behaviors. Second, the intensity of social contact fluctuates widely."25 This means that the effectiveness of a salesman depends upon the kinds and personalities of his company's customers.26 Therefore, studying the salesman in isola- tion may not be sufficient. What may be necessary for sales success is the ability to effectively interact with a prospect. This ability to interact may be partially dependent upon the backgrounds of the members of the dyad. 24G. David Hughes, "A New Tool For Sales Managers," Journal of Marketing Research, Vol. 1 (May, 1964), p. 32. 25James A. Belasco, "The Salesmen's Role Revisited," Journal of Marketing, Vol. 30 (April, 1966), p. 7. 26E. J. McCarthy, Basic marketing: A Managerial Approach, Third edition (Homewood, Illinois: Richard D. Irwin, Inc., 1969), p. 518. 37 Homans notes that, "because people with similar back- grounds tend to hold similar values, they are apt to be- have in ways that each will find rewarding, especially when they first meet--and what gets done then seldom gets undone later...When pe0p1e behave in ways that they find mutually rewarding, they are apt, as we know, to like one another and continue the interaction."27 This type of thinking has led Franklin Evans to hypothesize that, "The sale is a product of the particular interaction situation of a given salesman and prospect rather than the result of the specific qualities of either alone."28 He further states that some early results of 118 study, "indicate differences in the way sold and unsold prospects view the particular salesman who called upon them, how the salesman views his role, and differences in pair similarity between sold and unsold dyads. Similarity of attributes within the dyad appears to increase the likelihood of a sale."29 Successful dyads are more alike 27George C. Homans, social Behavior: Its Elementary Forms (new York: Harcourt, Brace and World, 1961), p. 218. 28Franklin B. Evans, "Selling As A Dyadic Relation- ship: A New Approach," American Behaviora14§cientist, Vol. VI, No. 9 (May, 1963), p. 76. 29Ibid., p. 79. 38 internally than unsuccessful ones in terms of "age, height, income, religious affiliations, education, politics and even smoking."30 Several other studies have been completed which look at the interaction between buyer and seller. Lombard studied twenty saleswomen in the children's clothing department of a large department store and found that salesgirls perceive customers who reject the merchandise as rejecting them and vice versa, that customers in a hurry perceive salesgirls as not being interested in them. However, the salesgirl who feels secure in beliefs about herself perceives the customer as someone who needs help.31 William.F. Whyte also pointed out the importance of the interaction between the restaurant waitress and her 32 customers. He found that the behavior of the waitress varies with the social status of the customer she serves.31 That is, the higher the social status of the 3°Ibid.. pp. 78-79. 31George F. Lombard, Behavior in a Sellinngroup (Boston: Harvard, 1955), pp. 187-217. 32William F. Whyte, Human Relations in the Restau- rant Industry (New YOrk: Mccraw Hill Bodk Company, 1948). 33Ibid., p. 92. 39 restaurant's clientele, the less friendly and personal the waitress must act. The well-adjusted waitress did not re- act to her customers' “moods." She, instead, controlled her behavior.34 Whyte also noted that "if the cook and waitress have a fight or if the waitress clashes with her supervisor, then the waitress is likely to take out her aroused feelings on the customer through poor service or discourtesy...“35 Gadel, in a study of life insurance agents, also found that salesmen tend to concentrate on particular kinds of prospects. He notes that younger life insurance agents tend to concentrate on prospects in the same age group as themselves but as the salesman matures he sells to a more heterogeneous group of prospects.36 Charles King and JOhn Summers after reviewing the literature on the interaction dyad conclude that - "Although the dimensions of analyses and the methodologies used have varied between studies, the research findings across studies are remark- ably consistent: 1. The interaction dyad appears to be relatively homogeneous across many interaction contexts. 34Ibid., p. 119. 351bid., p. 18. 36M. S. Gadel, "Concentration By Salesmen on Con- genial Prospects," JOurnal of Marketing, Vol. 28 (April, 40 Studies comparing the social status and age of participants in an interaction dyad indicate that people tend to exchange information with other age and social status peers. 2. Perceived credibility and/or expertise of the referent as an informant on a topic is an im- portant dimension in information-seeking be- havior. Seekers search for referents "more qualified" than themselves on a tOpic. In context where expertise is not perceived avail- able within the seekers' peer level, sources higher or lower in age and social status may be consulted. 3. The familyyplays an important role in inter- _personal communication in the socialization of children and in interaction within the extended family. The specific functions of family versus nonfamily interactions may be different but this area has not been explored. 4. Proximity is important in facilitating inter- action. Proximity, as a variable, is two- dimensional including physical proximity and social proximity. Obviously, physical proximity, e.g., living in the same neighborhood makes possible physical contact and the settings for interpersonal exchange. Physical proximity also suggests a minimum social proximity in terms of some overlap of social status, interests, life style, etc."37 This evidence tends to verify what might be called the characteristic similarity hypothesis. That is, the more a salesman possesses the same characteristics as his customers, the greater the probability of his success."38 37Charles W. King and John O. Summers. "Dynamics of Interpersonal Communication: The Interaction Dyad," in Donald F. Cox Ed. Risk Taking and Information Handling in Consumer Behavior (Boston: Division of Research, Graduate School of Business Administration, Harvard University, 1967), p. 261. 33David Kollat, Roger D. Blackwell and James F. Robeson, Strategic Marketing (New YOrk: Holt, Rinehart and Winston, Inc., 1972), p. 380. 41 It should be noted, however, that it is not neces- sary for the salesman to possess exactly the same feeling set as the prospect to sell effectively. Perceived similarity on the part of the prospect is what is really important. Evans notes that "the perceived similarity of religion and politics is much higher and of greater im- portance to the sale than the true similarity."39 Therefore, it is necessary for the salesman to read the prospect well in order to effectively "size him up" and then to communicate feelings which are similar to the prospects. It is also evident that, "Receivers of mes- sages vary in their ability to perceive the attitudes, intentions, feelings, needs, and wishes of others. In our interpersonal communication most people make reason- ably accurate judgments of one another: yet this slight misjudgment may cause considerable difficulties."4o This points out the second great thrust in the more modern approach to selling: that is, the communications viewpoint. Stidsen notes that "the fundamental characteristic and function of personal selling is 39Evans, op. cit., p. 79. 40Kim Griffin and Bobby R. Patton. Fundamentals of Interpersonal Communication (New Ybrk: Harper & Row Pub- lishers, 1971), p. 71. 42 intercommunication... And the essential strength to be sought for and encouraged in individual salesmen is their ability to take into account and communicate to the indi- vidualized and specific aspects of a buying situation and to evolve with that situation over time."41 Interpersonal communication may be viewed as a pro- cess with the following structure: FIGURE 2-2 The Process Model of Communication x xacx x1: x:cx.x:cx zcx xzcx x x.x1cx.x1< xacx x:cx.x 1: SOURCE 19..., x ENCODER x-)x MESSAGE x—-) x DECODER x—9 x RECEIVER x x x1:x:x1( xzcx:x1cx accumnmm x x x: