PLACE IN RETURN Box to remove this checkout from your record. TO AVOID FINE return on or before date due. DATE DUE I WE DUE DATE DUE MAGrcz L MR 011%? RUG/11993 22807 AM. 11 '3. . 1 J 1/98 Glamorous-m4 ABSTRACT THE ORAL INTERVIEW AS A PREDICTIVE DEVICE IN THE SELECTION OF MICHIGAN STATE POLICE TROOPERS The Michigan Civil Service Commission periodically announces a competitive examination to select qualified applicants for the position of trooper in the Michigan State Police. A weighted part of the examination is an oral interview conducted by a three-member interview board. This study was undertaken with a two-fold purpose. The first was to determine if there was a significant correlation between the oral interview ratings of the applicants and evaluations of their subsequent performance in the trooper training school and later as probationary troopers. The second purpose was to determine, if pos- sible, why the examination process was failing to screen out applicants who were appointed but failed to complete the training school courses. The primary data which served as the basis of the analysis were the oral interview ratings of three groups of applicants who were successful in the examination. These ratings were correlated with the grades achieved by these same men in the trooper training school class to which they were appointed, and with the probationary troop- er evaluation ratings covering the initial six months of their service as a trooper on field assignment. The primary method of analyzing the data was by use of the Pearson product-moment method of correlation. In addition, several sub-groups of the primary data were tab- ulated for comparison purposes. Parts of the findings and conclusions are based on an analysis of these tabulations. The results of the data analysis reveal a signifi- cant amount of reliability in the ratings of the oral interview board members. The reliability coefficients of correlation are of the order .6M-.73. However, the corre- lations of the interview ratings with the training school grades and the probationary trooper ratings resulted in validity coefficients indicating that the oral interview lacked validity as a device for predicting future perform- ance of applicants in the training school and as probation- ary trOOpers. Correlation coefficients of the order .14-.23 resulted from the comparison of interview ratings and train- ing school grades, and .O7-.27 from the comparison of the interview and probationary trOOper ratings. In the final chapter of the study which is a summary discussion of the data, the techniques of analysis, and the results, the writer postulates some possible flaws in the data and their effect on the results. The concluding para- graphs consist of a number of suggestions based on the find- ings, which would, in the writer's_0pinion, improve the selection of applicants for the trOOper training schools and lead to more meaningful future studies of this type. THE ORAL INTERVIEW AS A PREDICTIVE DEVICE IN THE SELECTION OF MICHIGAN STATE POLICE TROOPERS By C. C. Riggs A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF ARTS Department of Political Science College of Business and Public Service 1961 Approved by: J/C£§::JQ //- cgjéfi‘ Major/Proféssor ’ PREFACE The study reported here was undertaken with the aim of analyzing the oral interview procedure as it is used in the examination of applicants for the position of Michigan State Police Tr00per. The data analyzed was collected from records of examinations given in 1957, together with the subsequent state police training school records and probationary service reports of the officers appointed as a result of the examinations. Thus, this is not a specially designed experiment conducted for the sole purpose of this study-~rather, it is a post-audit of existing practices aimed at determining the predictive value of the interview. Since the results of Michigan Civil Service examinations and the personnel records of state employees are of a confidential nature, a special effort has been made to protect the identity of all individuals who participated in the examination process, either as an applicant or as a member of an interview board. The same objective approach has been maintained in all mention of trooper training school grades and probationary service ratings of the officers. Had not the writer been given unrestricted access to the examination records and applicant files of the 11 civil service commission and the training school records and personnel folders of the state police officers in- volved, this study could not have been made. The writer wishes to acknowledge the cooperation and assistance of Mr. C. J. Hess, Deputy Director of the Michigan Civil Service Commission, and Captain Arthur H. Long and Lieutenant Jack P. Foster, of the Michigan State Police, for making available the necessary records and reports upon which the study is based. Also, a note of appreci- ation to Dr. Stanley Gabis, major professor, for his encouragement and guidance throughout the project, Professor Henry Clay Smith for technical advice, and to my wife, Phylis, whose assistance as proof-reader and typist has added materially to the final report. The writer is a member of the examination staff of the Michigan Civil Service Commission, but readers are advised that Opinions and conclusions expressed are his own and do not necessarily reflect the views of either the Michigan Civil Service Commission or the Michigan State Police. 111 TABLE OF CONTENTS Page PREFACE O O O O O I O O O O 0 O O O O O O O O O 0 O 0 ii LIST OF TABLES. O 0 O O O O O O O O O O O O O O O O 0 V CHAPTER I. THE INTERVIEw--ITS USE AND VALIDATION. . . . . l The Interview in Employee Selection. . . . . . l The Problem of Interview Validation. . . . . . 9 Report of Interview Validation Studies . . . . 15 II. BACKGROUND OF THE STUDY. . . . . . . . . . . . 28 The Applicants O O O O O O O O O O O O O O O C 28 Pre-Interview Screening. . . . . . . . . . . . 3A The Oral Interviews. . . . . . . . . . . . . 39 The Trooper Training School. . . . . . . . . . 56 III. DATA ANALYSIS. . . . . . . . . . . . . . . . . 61 Interview Ratings, Training School Grades and Probationary Trooper Ratings . . . . . . 61 Iv. SUMMARY AND CONCLUSIONS. . . . . . . . . . . . 76 SELECTIVE BIBLIOGRAPHY. . . . . . . . . . . . . . . . 90 APPENDIX. 0 O O O O O O O O 0 O O 0 O O O O O O 0 0 O 92 iv TABLE I. II. III. IV. LIST OF TABLES Training School Performance as Related to Armed Service Experience. . . . . . . Training School Performance as Related ' to Occupational Background . . . . . . . Training School Performance as Related to Age Groups. . . . . . . . . . . . . . Comparison of Interview Ratings of 100 Applicants Who Completed the School and 38 Applicants Who Failed . . . . . . Page 30 31 33 65 CHAPTER I THE INTERVIEW--ITS USE AND VALIDATION The Interview in Employee Selection The practice of interviewing Job-seekers to assist in determining their suitability for employment has long been a practice of private employers. With the advent and growth of merit system selection of government employees, the interview technique rapidly took its place as an im- portant selection tool among the recruiting procedures of these personnel agencies. There are differences in the ways in which private employers and civil service depart- ments conduct interviews and utilize the results, but the main purpose of the interview remains the same--to assist the employer in determining the over-all suitability of the job applicant for the particular type of employment for which he is being considered. Fear discusses three types of interviews-~the direct interview, the indirect interview, and the pattern- 1 The direct interview is basically a ques- ed interview. tion and answer session. The indirect interview is one in which the interviewee is allowed almost complete freedom 1Richard A. Fear, The Evaluation Interview (New York: McGraw-Hill, 1958), p. 25. - 1- to discuss any topics that he chooses. The patterned interview is something of a combination of the other two methods. 'The discussion is directed and guided by the interviewer, but the interviewee is permitted complete freedom to discuss tOpics which the interviewer considers relevant. Of these three types of interviews, the patterned interview appears to be the more effective and appropriate method for purposes of employee selection. Control of the interview is maintained by the interviewer so that all important areas of the applicant's background can be covered, but the information is obtained in an indirect manner. The aim is to obtain spontaneous information without resorting to direct questions. This avoids giving the applicant the impression that he is being cross- examined. The patterned interview technique involves the use of a standardized rating form on which is listed the traits and characteristics considered most effective in the evaluation process. The rating form is based on factors listed in a Job specification. The interviewer knows what qualities the Job requires and guides the interview so as to obtain the desired information about the applicant. Following each interview, the applicant is rated on the basis of the interviewer's Judgment as to the ex- tent to which the applicant possesses or is lacking in the necessary qualities. Wagner states: - 3 - An interview, regardless of its length or purpose, .should be conducted according to a standardized form. This prevents aimless rambling, lengthy digressions, agd the possibility of omitting im- portant areas. Some additional aspects of the patterned interview as it relates to validity will be discussed in the section on validation of interview results. The purpose of the interview has a definite re- lationship to when and how it is used in the selection of employees. As stated earlier, its basic purpose is to assist in the screening of applicants. The importance of the interview in the final decision varies among private employers and the merit system departments depending upon the philos0phies and attitudes of those who set employment standards and policy. In general, the interview probably has a greater weight in the hiring process in private employment due to the major emphasis in merit system programs on the use of other objective testing techniques-- chiefly, the written examination. This is not to imply that private employers make little use of other selection devices. Many firms make extensive use of intelligence tests, psychological tests, and aptitude batteries; but the cornerstone supporting the whole merit system concept is the objective, competitive written examination. In civil service departments throughout government, the- 2Ralph Wagner, “The Employment Interview: A Crit- ical Summary," Personnel Psychology, Vol. II (19h9), p. 42. (This article.summarlzes vafious interview studies report- ed up to 19A9. An extensive bibliography is included.) - h - written examination constitutes the main load of the examination program. A basic view of persons working in the personnel selection field is that the interview should not be used to evaluate factors that can be better evaluated by the use of other testing procedures.3 Thus, to test an engi- neer's technical knowledge, some form of written test based on the subject matter would be used. To determine a management trainee's general intelligence level and ability to profit from the training, one of a number of standardized intelligence tests would be appropriate. To test typing or stenographic skills, a performance test would be administered. These tests would evaluate the technical skills and basic aptitudes, but in many types of employment there is another aspect of over-all compe- tence that is equally important. This is the area of personality and general personal fitness of the applicant. It is in the evaluation of these factors that the inter- view assumes an important role. According to McMurray: It must be kept in mind that all tests of intelli- gence, aptitudes and proficiencies are measures of what the man or woman can do; they are indices of the individual's equipment; the skills, apti- tudes and experience which he brings to the job. However, they provide no assurance that he will do what they show he can do, once he is hired or that he will otherwise be a desirable employee. He may have adequate intelligence, aptitudes and skills and still be highly unsuitable owing to laziness, irresponsibility, or inability to get along with others. These latter characteristics 3ibid., P. 43. - 5 - can, in the author's judgment, be best evaluated by intervifiw procedures used in conjunction with the 1:381:30 As stated earlier, there are certain aspects of the examination process in civil service agencies which re- quire that the interviews differ somewhat from those con- ducted by private employers. Bingham, Moore and Gustad make the following comments abOut the merit system exami- nation process: Applications must be received from all eligible citizens who wish to be considered; and the selective process must be conducted in a manner obviously fair to all. Safeguards against charges of bias, favoritism, or political influence are imperative._ The procedures followed in sifting applicants and arranging lists of eligibles in order of merit must be so adequate and sound that they will command public confidence and, if neces- sary, stand the scruginy of judicial review in the event of appeal. Contrary to the implication in the above quotation, every applicant considered for the examination is not necessarily accepted; he may be allowed to participate in the first stages of the examination but be eliminated prior to the scheduling of the interviews. Also, from the standpoint of efficient administration of the over-all examination program, it is not advisable to provide that the interview be a weighted part of each examination given. The criteria that governs the use of the interview ”Robert N. McMurry, "How Efficient are Your Hiring Methods?", Personnel Journal, Vol. XXVI, No. 2 (June, 1947), p. 50.. H 5Walter Van Dyke Bingham and Bruce Victor Moore, with collaboration of John W. Gustad, How to Interview hth ed. rev.; New York: Harper and Brothers, 1959), p. 121. - 6 - is the nature of the duties of the particular class of civil service positions for which the examination is being conducted. If the position involves a great deal of pub- lic contact, close teamwork with fellow employees, or distinct leadership qualities, the interview will probably be included in the examination process. Thus, the inter- view would be a part of examinations for social worker, employment and claims interviewer, conservation officer and state police trooper, but probably would not be in- cluded in examinations for highway laborer, stores clerk, tree trimmer or carpenter. 0rdway, §t_§l., have this to say concerning the interview in civil service examining: . . .When the personality of the candidate is important to performance, it is essential that someone responsible for selection interview him personally, see him in action, observe the way he conducts himself during the give-and-take of a personal conference and afford him an Oppor-6 tunity to establish his ability to do the job. 7 Fearing and Fearing refer to the civil service in- terview as occurring in a unique social-psychological con- text. They cite the required capacity of the interviewer to see a particular job as related to the interests and needs of the public as a whole, and also emphasize the 6Samual H. 0rdway, Jr., et al., Oral Tests in Pub- lic Personnel Selection, Civil ServIce Assefibly of the United States and Canada, (Chicago, 1943), p. 10. 7F. Fearing and F. M. Fearing, "Factors in the Ap- praisal Interview Considered with Particular Reference to the Selection of Public Personnel,V Journal of Psycholggy, Vol. XIV (1942). p. 138., . . _ - 7 - fact that there are certain legal requirements pertaining to the guarantee of certain procedures and rights. An important factor they mention is the requirement that the results of the interviews be interpreted to the examina- tion participants and possibly to the public at large. The objective attitude and fair-and-equal treatment of all applicants is emphasized throughout all parts of the examination preceding the interview. However, if the applicant is successful in the preliminary screening and is notified to appear for the interview portion of his examination, he may well consider that the members of the interview board have suddenly taken a very keen interest in him as an individual. Up to this point, he may have been but one in a group of several hundred who were com- peting in the examination. Now, he is by himself--the center of attention of three,four, five or even more interviewers. 0rdway, §t_§l., make the following comment on this aspect of the civil service interview: It is also recognized that, whatever the form or ostensible purpose of the interview may be, it serves one necessary end-~to humanize an otherwise bureaucratic relationship. To the applicant, par- ticipation in a civil service examination is a highly personal experience, whereas the examiner is prone to view it as a matter of impersonal, objective routine. The interview presents the one opportunity along the way to personalize the process. Moreover, it often presents an Oppor- tunity for engendering the goodwill which springs from letting the candidate know he is being con- sidered as a human entity, rather than as an grray of skills, talents, and similar abstractions. 8Ordway, et al., op. cit., p. 5. - 8 - The multiple-member interview board is standard for civil service interviewing. In some interviews the appli- cants are screened on a "pass-fail" basis; but in by far the majority of interviews, the interviewers must not only decide which of the applicants are qualified for the posi- tion but must assign ratings which in their judgment re- flect the degree of qualification. For this reason, the reliability and validity of the interview ratings are of prime importance. Within limits, the reliability of the ratings increases as the number of interview board members and the length of the interviews increase.9 This is a reflection of the well-known "two heads are better than one" approach to decision making. A board composed of several members is more likely to appraise accurately the qualifications of the applicants. This is to the advantage of those participating in the examination as well as the entire public service when those who are the better-qualified for a position are ranked at the top of the employment list and are given first consideration when an appointment is to be made. Unfortunately, an examination, including the inter- view portion, can be reliable but still lack validity when it comes to the actual performance of those selected and appointed to the positions. An examination is reliable 9Milton Mandell, "Civil Service Oral Interviews," Personnel Journal, Vol. XVIII (19h0), pp. 373-382. (An. excellent discussion of civil service interviewing, with emphasis on rating methods.) -9- when it consistently measures whatever it does measure. However, if it is measuring characteristics and qualifi— cations which have little relationship to successful per- formance on the job, the examination is not a valid selection device. The difficulty of determining interview validity is the tOpic of the following section of this chapter. The Problem of Interview Validation In any attempt to determine the validity of the interview as an employee selection device, the initial problem is one of determining the criteria by which employee performance is to be judged. What constitutes satisfactory performance on the job? What factors dis- tinguish the excellent employee from the merely adequate or the unsatisfactory? Different researchers have used different measure- ments. -Attempts have been made to validate interviews on the basis of a number of criteria. Quantity and quality of production, efficiency ratings, length of time the worker remained employed, written test scores, and super- visors' evaluations have all been used-~either singly or in combinations-~with varying degrees of success as in- dicated by the resulting correlation figures. The impor- tance of selecting the proper criteria is indicated by McMurry: - 10 - . . .This establishment of adequate criteria is at once the most important and most difficult phase of the validation of tests and igher selection instruments and procedures. The above comment is indicative of the attitude of experi- enced researchers. The results of any validation study of employee selection methods can be no more valid than the criteria against which the methods are measured. There are, according to Brogden and Taylor,11 three steps which are essential to adequate criteria con- struction. These are: (l) The determination of the elements to be measured, (2) the determination of how each element is to be measured, and (3) the determination of the relative importance of each element to over-all efficiency. Before selecting the elements to be measured, a thorough understanding of the job and the duties the em- ployees perform is necessary. Only those criteria which actually are basic to successful performance should be included. Factors which appear to be necessary for suc- cessful performance but which have little actual relation- ship will, if included in the study, result in false or misleading results. After the individual criterion has been selected, a method of rating or measuring it must be devised. Sometimes 10McMurry, Op. cit., p. 50. 11Hubert E. Brogden and Erwin K. Taylor, "The Theory and Classification of Criterion Bias," Educational and Psychological Measurement, Vol. XI (1950), p.—162. the measuring device is obvious, as in the case of piece- work production (pieces per given period) or length of employment (days, weeks, or months). In the case of ratings by supervisors, a scale listing four or five de- grees of adequacy in the trait being measured is usually used-~with a given number of points assigned each degree-- such as: Excellent, 5; Good, A; Fair, 3; Poor, 2; and Unsatisfactory, 1. When a number of different criterion are used in the rating of over-all efficiency of performance, a method must be established whereby the ratings on the various criterion are combined so as to produce a composite rating. Here again, the use of assigned weights is common; the criteria found to be most important in over-all perform- ance should receive the heavier weighting. In the case of ratings by supervisors, as described above, where a number of traits are evaluated, the weight- ing of each trait in the composite rating may be accomplish- ed at the same time the weighting is done for each specific trait scale. Thus, an "Excellent" rating in a trait of major importanCe might be alloted ten points, while an "Excellent" rating in a trait having less effect on over- all performance might receive but five points. After all traits have been rated, the composite rating is obtained by merely adding the scores on the individual trait ratings. Perhaps the most difficult aspect of attempting a study of the predictive value of any employee selection -12- device is that of obtaining an adequate record of the em- ployee's work performance. The usual employee records kept by personnel departments seldom meet this require- ment because the efficiency ratings, service ratings, or other supervisory reports available, more often than not, refer to performance in general terms. They are lacking in necessary specific information regarding positive and negative characteristics of the employee. In these cases, where supervisory ratings are a desired criterion and such records do not exist or are inadequate, special ratings must be obtained. In obtaining these ratings, the person making the study may provide for ratings based on observation of the employee during a specific test period; or he may want ratings based on employee performance under normal working conditions over an extended period of time. In either case, according to Thorndike: If ratings are to provide a relevant criterion measure of the individual, two conditions must be met. The rater must be willing to rate the 12 individual fairly, and he must be able to do so. The factors underlying the above quotation are those which are of necessity involved in a supervisoresubordinate relationship. Even when the rating supervisors are briefed on the rating process, and a composite rating is deveIOped from the individual ratings of several supervisors, the 12Robert L. Thorndike, Personnel Selection (New York; John Wiley and Sons, Incorporated, I9E9), p. 155. - 13 _ ever-present element of subjective bias may influence the hOped-for validity of criterion based on ratings. Brogden and Taylor state: The most obvious and probably the most serious source of contamination peculiar to ratings 13 arises because of the so-called halo effect. Halo effect is a peculiarity of the rating procedure whereby the rater is so strongly influenced by some major characteristic of the person being rated, either favorably or unfavorably, that his ratings of other characteristics and his final over-all rating are also influenced by his reaction to the one factor.14 This influence may result in either an increase or a decrease in the rating assigned, but in any event, such a rating will not reflect an accu- rate evaluation of the person being rated, thus impairing the validity of the rating as a criterion. McMurry15 states that minimum data on which an over- all evaluation of employee job success may be based should include volume and quality of production, length of ser- vice, and success ratings by foremen. Other researchers have utilized other criteria, depending on the particular study they were making. In by far the majority of reported studies, however, the supervisor's ratings have been a 13Brogden and Taylor, op. cit., p. 173. lb’Walter Van Dyke Bingham, “Halo, Invalid and Valid," Jogrnal of Applied Psychology, Vol. XXIII (1939), pp. 221- 22. 15McMurry, op. cit., p. 50. -1Li- criterion factor in establishing the validity of the interviewer's predictive ratings. There are many types of rating forms and rating scales, but the results of studies conducted in 195016 indicate that validity is more dependent on the rater than on the rating technique used in any given series of ratings. These studies indicated two important aspects of ratings, whether used as predictive factors or as validation criterion: (1) One of the most effective means of increasing the validity of ratings is by increasing the number of raters of comparable competence for each person being rated, and (2) when the same raters are used, the validity of their ratings using a given technique is simi- lar to the validity of their ratings when a different technique is used. Some reported "validity" studies appear to be more a measure of the reliability of the ratings. In the above- mentioned study, the authors devote a paragraph to this subject and state the case so well that it seems apprOpri- ate to quote it here so as to afford the reader a better understanding of the studies to be reported in the follow- ing section of this chapter. The authors state: Considerable doubt has existed as to the sound- ness of validating ratings against other ratings used as criteria. The fact that both predictor 16A. G. Bayroff, Helen R. Haggerty, and E. A. Rund- quist, "Validity of Ratings as Related to Rating Techniques and Conditions," Personnel Psychology, Vol. VII, No. l (1957): pp. 93-113. - 15 - and criterion ratings are obviously judgmental measures in which rater biases and capabilities must Operate raises the question as to the in- dependence of these measures. One answer has been to define validity in such cases as agree- ment among raters, especially the agreement of one rating with a consensus of ratings. Validity thus interpreted becomes similar to reliability. 7 Hundreds of interview studies have been reported. Those few summarized in the following section have been selected to give the reader a limited cross-sectional view of several of the more important studies of interviewing as well as an impression of the more typical reports of studies similar to that which is the basis of this paper. Report of Interview Validation Studies One of the earliest reported studies related to employee selection by interviewing was not a validity study as such, but was aimed at determining the reliabil- ity of the ratings of a number of interviewers. It is reported here because the results of this early study have long been cited as evidence of the unreliability of the interview as a selection device. The study was originally reported by Scott, Bingham, and Whipple in 1916 (in Volume A of the 1916 issue of Salesmanship, under the title, "Scientific Selection of Salesmen."). The writer was unable to locate this 17Bayroff, Haggerty, and Rundquist, Ibid., p. 112. -16... original source. The following account is based on a report of the study by Bingham, Moore, and Gustad.18 The experiment was concerned with the ability of sales managers to select applicants applying for sales positions. There were twenty-three interviewers; twenty sales managers and three investigators of the problems of selecting sales personnel. There were twenty-four appli- cants, all of whom were actually seeking employment. Each interviewer was assigned a room and each applicant in turn called upon each interviewer and for five minutes presented a selling talk on any line of merchandise that he chose to sell. The applicants were instructed to assume that each "merchant" was a buyer of whatever product he was attempting to sell. The interviewers were instructed regarding their "merchant" roles and were told to assume that they alone stood between the applicant and the payroll of the hypo- thetical company they were representing. They were ad- vised that they could prevent the applicant from giving his sales talk if they so desired, and use any methods they thought would best enable them to evaluate the appli- cant as a salesman. The interviewers were to rank each of the applicants as best, second-best, etc., until all twenty-four were ranked. When the results were analyzed, it was found that a wide range of rankings were assigned each applicant. 18Bingham, Moore and Gustad, op. cit., pp. 105-108. - 17 - For instance, one applicant was ranked in first place by two interviewers but he was twenty-second on the list of a third interviewer. The correlation between the rankings of the individual interviewer and the consensus of the twenty-three was computed for each. The lowest correla- tion was .55 and the highest .85. On a number of appli- cants there was fairly close agreement by the majority of the interviewers in spite of the spread of the correlation figures. The wide variation in ratings by interviewers in this and similar early studies led many people in the field of personnel work to seriously question the useful- ness of the interview as a sound tool of selection. A more recent writer on the subject of interview research expresses the following vieWpoint concerning the early studies designed to predict success in salesmen: . . .the aim (of the interview) has hardly been defined at all, and it is not surprising that the reliability of the interview used for this purpose has been consistently poor. What is surprising is that such experimental work, built ggezugfizyyigiiggri9foundation, should be so On the basis of the later deveIOped approach to the selection interview-~in which the patterned interview is stressed-~this early study had a major flaw in the technique utilized. There was no attempt made by the interviewers 19K. A. Yonge, "The Value of the Interview: An Orientation and Pilot Study," The Journal of Applied Psychology, Vol. XL, No. l (1956), p. 25. I... -18... to identify traits, qualities, or characteristics to serve as a common basis for judging the qualifications of the applicants. Each rater was free to use whatever "yard- stick" he thought best. In fact, the procedure leads the writer to question the "interview" concept itself insofar as this particular study is concerned. From the instruc- tions given the raters and applicants, one can easily infer that in some instances the rating may well have been based solely on the sales presentation made by the applicant with little actual interviewing being involved. The use of the patterned interview based on a guided discussion of pre-selected topics resulted in studies yielding results of a more reliable nature. McMurry2O presents an interesting discussion of three such studies, all of which were conducted on the same basis. A feature of these studies is that even though some of the applicants were rated as being "unsatisfactory" on the basis of the interview, all applicants were employed. This made possible an evaluation of the interview ratings as a predictor of failure on the job as well as eventual successful performance. In all three of these studies the applicants were interviewed and rated at the time of employment. In the first two studies, each employee was ranked in one of four 20Robert N. McMurry, "validating the Patterned Interview," Personnel, Vol. XXIII, No. 4 (January, 1947), pp. 263- 272 - 19 - categories: "Outstanding", "Good", "Fair or marginal", or "Unsatisfactory". In the third study, the applicants were rated on a five category scale: "Excellent", "Above aver- age", "Average", "Below average", and "Definitely unsuit- able". The first study was made at the Link-Belt Company in Chicago. The interviews were conducted by members of the regular employment office staff who had received training in the use of the patterned interview. The vali- dation study was conducted one and one-half years follow- ing employment, at which time 407 of those hired were still employed. The validation results reported here are based on a comparison of interview scores and evaluation ratings by foremen. In making their ratings, the foremen were instructed to divide the workers into two equal groupse-above average and below average. Bases for this division were produc- tivity, attitude toward supervision, and over-all desir- ability as an employee. In addition, foremen were asked to indicate the outstanding employees in the above average group and also those in the below average group who were clearly unsatisfactory. When possible, two or more inde- pendent ratings were obtained; but where disagreements were obvious, the foremen discussed the employee and reached an agreement as to the rating asSigned. Correlation of the two groups of ratings resulted in a Pearson Coefficient of correlation of .43 1.1.02. -20... Analysis of the ratings showed that the foremen and the interviewers were in complete agreement in their evalu- ations of 292 of the 407 employees. The major area of disagreement was in the "Fair or marginal" category of the interviewers when compared with the foremens' ratings-- 257 employees were rated "Fair or marginal" by the inter- viewers. Of this number, the foremen rated 175 "Below average", four "Very poor", 75 "Above average" and three "Outstanding". 0f 32 workers rated "Unsatisfactory" by the interviewers, 23 were considered "Very poor" by the foremen. 0f the other nine in this group, eight were rated "Below average" and only one "Above average" by the foremen. Of the eight workers rated "Outstanding" by the interviewers, six received the same rating from the fore- men and the other two were considered to be 'Above average". The second study was conducted for the White Motor Company in cooperation with the Aero-Mayflower Company in Indianapolis. The subject of the study was a group of 108 applicants for the position of truck driver. The inter- viewer in this study was a trained psychologist not con- nected with the employment staff of either company. He interviewed all the applicants and assigned ratings on the basis of the four categories outlined in the Link-Belt study. The progress of the applicants was carefully fol- lowed for eleven weeks while they went through a training course and were assigned out on the road as drivers. After this eleven week period, the interview ratings were - 21 - compared with the success of the drivers as indicated by their length of service. In this study, the biserial co- efficient of correlation was .61:t.ll. The results of the study revealed that had the company hired only those applicants rated as "Outstanding" by the interviewer, the employment turnover in truck- drivers would have been halved. Had they hired only those rated "Outstanding" or "Good", the turnover would have been reduced by twelve per cent. 0f the 15 applicants rated ”Unsatisfactory“, only two (13.3%) were still em- ployed at the conclusion of the study. Of the eight rated "Outstanding", six (75%) were still on the payroll. 0f 39 rated "Good", 15 (38.5%) remained, and of 46 rated "Fair or marginal", 12 (26.1%) were still employed. The third study was conducted at the York Knitting Mills, Ltd., plant in Canada. The interviewing of 84 job applicants was done by members of the regular employment staff who had received careful training in the use of the patterned interview. (Mr. J. J. Carson and Dr. H. C. Grant, of the firm of J. D. Woods, and Gordon, Ltd., of Toronto, supervised this study and trained the interview- ers.) In this instance, the ratings of the interviewers were correlated with supervisors' ratings Of the employees. The Pearson Coefficient of correlation was .6l‘i.05. The interviewers and the supervisors were in agree- ment on their rating of 54 of the 84 employees. Only one person was rated "Excellent" by the interviewers, and this - 22 - same worker was the only one to receive an "Excellent" rating from the supervisors. The greatest disagreement was in the "Definitely unsuitable" category. Interviewers gave 17 applicants this rating, but the supervisors rated just three of the workers as "Definitely unsuitable". 0f the remaining 14 workers rated “Definitely unsuitable" by the interviewers, the supervisors rated eight as "Below average", five as "Average", and one as "Above average". 0f the remaining categories, the supervisors agreed with the interviewers on 13 out of 18 rated "Below average", 31 out of 40 rated "Average", and six out of eight rated "Above average". In all three of these studies, the predictions of the interviewers resulted in a positive correlation when measured against the on-the-job criterion. McMurry‘s con- clusion is that the results indicate that a carefully conducted patterned interview has value in predicting the job success and stability of persons employed in these particular factory occupations and as truck drivers. He states that some allowance must be made for unreliability in the criteria where they consist of supervisors' ratings. This is evidently in reference to the Link-Belt study where the correlation was substantially lower than in the other two studies. He makes no conjecture as to the basis of the unreliability, but it is a logical assumption that the previously mentioned elements of subjective bias and/or the halo effect were involved in these ratings. - 23 - In the truck driver study the only criterion for validation was the length of service of the employees. This is an adequate measurement if the sole objective of the selection interview is to select those who show poten- tial for permanent employment and to eliminate the "job- hOppers" and "drifters". This criterion does not, how- ever, by itself, provide any indication as to the employee's work quality and quantity or his over-all desirability as an employee. Yonge21 takes the View that if work perform- ance is judged solely on the basis of the length of time an employee stays on the job, so many uncontrolled vari- ables are involved that its value as a criterion is very questionable. Another study concerning a civil service examination for the position of Captain of Police provides interesting data on the reliability and consistency of ratings by the various members of a multiple-member interview board. The examination was conducted by the Los Angeles City Civil Service Commission in 1940, and was reported by Fearing and Fearing in 1942.22 There were 100 applicants for the examination. Each applicant received a 40 minute interview by a four member interview board which was composed of a social worker, a chief of police from a smaller city, a captain of police 21Yonge, op. cit., p. 26. 22Fearing and Fearing, op. cit., p. 131-153. -224- who was personnel officer for a large metropolitan police department, and a professor of psychology from a large university. The interviewers rated each applicant on nine different traits and then made a summary evaluation rating which was separate and not a composite of the other nine ratings. In determining the final score assigned by each interviewer, this summary rating was assigned a greater weight than the other ratings. The authors of the study contend that an analysis of the results produced evidence that certain attitudes peculiar to each interviewer are reflected in the scores they assigned. They cite as a possible cause the occupa- tional or professional status of the board members which causes their appraisals to be conditioned by a complex of forces including attitudes which they bring to the inter- view situation. This is born out by the ratings of the two police Officers which showed preference for applicants whose primary service had been in the "uniform" rather than the detective branch of the service. These officers' ratings also showed significant positive correlations be- tween ratings on "Experience" and "Actual length of ser- vice" of the applicants whereas the ratings of the other board members showed non-significant correlations between these items. .The psychologist's ratings showed the high- est correlation between the total interview rating and the amount of education as reported by the applicant as well as between the ratings on “Education" and actual education. - 25 - The traits most clearly differentiated by the inter- viewers were "Education", "Experience", and "Summary evalu- ation". .The authors base this conclusion on the relatively larger sigma of the distribution for these traits and on the fact that they were the items on which "biases" were most markedly expressed. Although these three traits were the ones on which the raters clearly differentiated within their individual ratings, they were also the ones (plus a fourth, "Ability to present ideas") on which the greatest over~all agreement was found when ratings were correlated with each other. The r's were of the order .40qA9. The least agreement was found on traits of "Neat- ness and dress", "Tact" and "Maturity of judgment", with r’s of the order .23-.32. The correlations were relatively high between the scores of the psychology professor and the social worker, the psychology professor and the police chief, and the police chief and the police captain. All correlations for the psychology professor were relatively high, and those for the police captain rather low. Interview validation studies have resulted in corre- 1ation figures indicating that the patterned interview can be an effective tool in personnel selection. However, other studies indicate that the role of the interview is limited and other selection tools are better predictors of subsequent performance. A recently reported study bears - 26 - this out. Campbell, Prien, and Brailey23 made a study of clerical worker trainees employed by a large public utility. The subjects of the study were 95 men and women who were interviewed by staff members of the personnel department of the company. All interviewers had received some train- ing in the use of the patterned interview. The criterion consisted of supervisors' ratings made one month after the employees had completed their probationary employment per- iod. Prior to the employment interview, the employees had been given performance tests.2h An objective personality test was also given to each employee before the super- visors had rated them. As a result of the study, the authors conclude that the value of the interview as a predictive measure is limited and questionable. They acknowledge that it is a useful tool in obtaining facts and in orienting the em- ployee, but in their study the objective personality test (Gordon Personal Profile) showed substantially higher validity as a predictive measure than either the perform- ance tests or the interview. 23Joe1 T. Campbell, Erich P. Prien, and Lester G. Brailey, "Predicting Performance Evaluations," Personnel Psychology, Vol. XIII (1960), pp. 435-440. 2L1These tests were the Wonderlic Personnel Test and an arithmetic reasoning test. The writer would not con- sider either of these a "performance" test as the term is used in civil service examining. The Wonderlic is an acknowledged "intelligence" test and arithmetic reasoning is also a factor of over-all intelligence. - 27 - And so it goes. Researchers continue to publish reports--some recommending the use of the interview, others advising caution, and still others favoring the use of other predictive measures. A majority of the reports do seem to have one common element, and that is a lack of sufficient detailed information. Jones25 made a survey of over 2,100 references on employee selection in industry. She found that only 427 contained sufficient information to permit evaluation of the study. Her main criticism is that many of the reports are lacking in details concerning the statistical interpretation of the findings and also adequate information in regard to the criteria used for validation. The following chapter begins the discussion of this study of the interviews for applicants for the exami- nation for Michigan State Police Trooper. The first sec- tion consists of general background information about the over-all examination and the applicants who were successful. 25Margaret Hubbard Jones, "The Adequacy of Employee Selection Reports," Journal of Appliengsychology, Vol. XXXIV, No. 4 (August, 1950f} pp. 21942247’ CHAPTER II BACKGROUND OF THE STUDY The Applicants The examination for the position of State Police Trooper is announced at irregular intervals by the Michigan Civil Service Commission. Despite the rigid requirements, there is never a lack of applicants--a majority of whom are initially accepted and allowed to compete in the examination. In the series of examinations which resulted in the selection of the applicants who are the subjects of this study, there were a total of 2,911 applicants. Writ- ten examinations were given on September 22, 1956; Janu- ary 12, 1957; and again on February 16, 1957. Each group of applicants for one of these three written examinations continued on through the complete examination process in- dependently of the other two groups. When the September, 1956, group had completed the examination, the names of the passing applicants were placed on the employment list for eventual consideration for appointment to a trooper training school. As each group completed the examination, the names of those who were successful were added to the employment list. Position on the list was governed by the examination score of each applicant. The employment list - 28 - - 29 - consisted of 295 names at the time the first appointments were made to the training schools with which this study is concerned. The examination and training school records of the 140 young men who qualified for appointment to the training school reveals some interesting background infor- mation. Although each applicant had his individual traits and Characteristics, as a group these potential state police trOOpers had a number of similarities. The minimum educational requirement for all appli- cants was graduation from high school or satisfactory com- pletion of the equivalent General Educational Development tests. Ten of the 140 applicants qualified on the basis of these tests. 0f the group, 41 had completed from one to three years of college. Only one applicant had complet- ed college and earned a degree. Ninety of the applicants had had active duty assign- ments in the armed services. Another ten had been members of either the National Guard or a Naval Reserve unit. The remaining forty applicants had had no armed service exper- ience. Armed service experience of the applicants is shown in Table I, together with the number being graduated or failing in the training school. There had been some conjecture among both civil service staff members and state police command officers that the applicant who had previously received training in a branch of the armed forces would adapt more readily to -3o- the training school environment and would thus stand a better chance of completing the school. Table I shows that seventy-five per cent of the applicants who had been in service were graduated from the school. lacked any service experience, per cent were graduated. Of those who sixty-seven and one-half These figures would tend to bolster the conjecture, but there was actually no evidence to indicate what relationship, if any, armed service ex- perience, or lack Of it, had to success or failure in the training school. TABLE I Training School Performance as Related to Armed Service Experience Number in Number Number Percentage Branch of Service School Graduated Failing Failing Army or National Guard 54 41a 13 24.0 Navy or Coast Guard 24 14 10 41.7 Air Force 12 11b 1 8.3 Marine Corps 10 9 1 10.0 No Service Experience - 40 27 13 32.5 Totals 140 102 38 27.2 aOne Army veteran resigned after five months as a probationary trOOper. bOne additional Air Force veteran resigned two weeks after graduation. An analysis of the occupational backgrounds reveals that of the 140 applicants, 56 had been engaged primarily - 31 - in some type of factory work prior to taking the State Police Trooper examination. This was the largest group, but busineSs was a close second with 54 applicants having a baCkground of office work, selling, or other closely- allied occupation. Ten applicants were employed as local police officers when they applied for the examination. Four men were farmers. Five could claim no work experi- ence. The remaining eleven applicants were employed in miscellaneous occupations which were dissimilar and do not fit into any of the categories mentioned above. Table II reflects training school performance as related to occu- pational background of the applicants. TABLE II Training SchOOl Performance as Related to Occupational Background Number Number Number Percentage Occupation in School Graduated Failing Failing Factory Worker 56 36 20 35.7 Business 54 41 13 24.1 Police Officer vfilo Ba 2 20.0 Farming 4 4 O 0 Miscellaneous 11 8 3 27.3 No Occupation 5 5b 0 0 Totals 140 102 38 27.2 aOne former police officer resigned two weeks after graduation. bOne officer with "No Occupation" resigned after five months as a probationary trooper. - 32 - The above figures indicate that there is no "pre- ferred" occupational background insofar as success as a state police officer is concerned. There were failing applicants in either the training school or during the probationary period from every occupational category listed except that of farming. Here there were so few cases that this fact is not significant. However, the writer would recommend continuing analysis of this type on all future training school applicants to see if a trend develops. The minimum age for acceptance as a trooper appli- cant is 21 years. The maximum age is 29 years. Theoret- ically, a man could be 29 years of age when accepted for the examination but 33 years old when called for the training school. This is due to the three-year life of the employment list. Practically, however, applicants who pass the examination have seldom waited longer than one year before being considered for a school. Table III reflects training school performance as related to age. groups of the applicants. The figures shown in Table III reveal that per- centage-wise there were more failing applicants in general among the lower age groups than there were among the older applicants. The exception is in the thirty-year age group where the greatest percentage failed. There were enough failing applicants at each age, however, to indicate that age, by itself, is not a major factor in success or fail- ure in the training school situation. - 33 - TABLE III Training School Performance as Related to Age Groups Number Number Number Percentage Age in School Graduated Failing Failing 21 20 14 6 30.0 22 26 20a 6 23.1 23 17 11 6 35.3 24 15 ll 4 26.7 25 18 11 7 38.9 26 10 9 1 10.0 27 10 7 3 30.0 28 8 7b 1 12.5 29 11 9 2 18.2 30 5 3 2 40.0 Totals 140 102 38 27.2 8One additional officer, age 22, resigned two weeks after graduation. bOne additional officer, age 28, resigned after five months as a probationary trooper. We now will examine briefly in the next section the actual sequence of events in the examination process leading up to the oral interview. The interview portion of the examination will be discussed separately in a later section of the study. _ 3h - Pre-Interview Screening The initial step in the examination process is the filing of an application for examination by the applicant. Each application is reviewed by the civil service examiner to see if the applicant is qualified on the basis of the minimum qualifications stated in the public announce- ment. If information on the application reveals that the person fails to qualify, the application is rejected. A number of applicants are thus eliminated for failure to qualify on the bases of age, height, weight, education, vision, obvious physical handicaps, arrest records, and/or an admitted record of extensive traffic violation offenses. In addition to failure to meet the specific qualifications for the trOOper examination, a few applicants are always rejected for failure to meet the basic civil service re- quirements of United States citizenship and residence in the State of Michigan. In some instances-~mainly where an arrest record is involved--the applicant is conditionally accepted for the examination pending a final decision after more detailed information has been revealed by the field investigation. The written test is the first part of the examina- tion involving direct participation of the applicants. The test is given on the same date to applicants through- out the state at the various testing centers of the civil service commission. The test used is the Wonderlic - 35 - Personnel Test. Its use in the trooper examination is to screen out those applicants who lack the capability to assimilate and benefit from the concentrated and acceler- ated curriculum in the trOOper training school. The Wonderlic test has been used for a number of years as a part of the trooper examination. Analysis of over-all performance in the training school indicates that a lower- ing of the raw score "passing point" by two correct answers on this test results in a noticeable difference in the capabilities of the trainees. By far the greatest number of applicants who fail the trooper examination are eliminated by the written test. Usually, between one-half and two-thirds of the applicants are screened out at this stage of the process. Those applicants who are successful in the written test are notified to report for the agility test, height and weight check, and vision test. These unweighted parts of the examination are administered by state police officers at various locations throughout the state. Officers who conduct these tests are members of the staff at the train- ing school. The check on height and weight is given at this stage of the examination to eliminate those who obviously do not qualify. This removes them from further competi- tion and reduces the number of applicants on whom a field investigation must be conducted. Experience has shown that certain applicants will list on their application for -35.. examination a height and weight which is acceptable; but when they are actually measured and weighed, they may be as much as two inches under the minimum height and fail to meet the minimum weight by an amount sometimes as great as twenty pounds. The minimum weight was 150 pounds, and the minimum height was five feet, nine inches. The vision test is given on "sight-screener" equip- ment designed for testing vision in connection with driver- 1icensing examinations. The minimum requirement is un- corrected vision of at least 20/30 in each eye, corrected to 20/20 with glasses prior to appointment to the training school. A test for color-blindness is also given, using the American Optical Company's Pseudo-Isochromatic Test consisting of eighteen charts. The agility test is designed to screen out those applicants who are physically weak as well as those who do not possess adequate coordination and/Or balance to suc- cessfully perform the required tests. The tests consist of a six foot rope climb, six bar-chins, twelve push-ups, and a standing broad jump of six feet, six inches. A majority of the applicants have little difficulty with these tests. The rope climb and the standing broad-jump are the great- est eliminators. It is noticeable that applicants who are excessively heavy in relation to their height have diffi- culty with these two tests. The final step in the pre-interview screening of the applicants who have passed the agility and vision tests _ 37 - is the field investigation and a review of the report by the field investigation review board. Each applicant receives a questionnaire which, when completed, gives a rather complete personal history of the applicant. The applicant retains this form until an officer from the state police post nearest his home calls on him, at which time he gives it to the officer. Information in the questionnaire provides the investigator with basic background information on the applicant and serves as a starting point in his investigation of the applicant. All Officers who conduct these field investi- gations are experienced officers--either command officers of the post or senior troopers. After talking with the applicant (and his wife, if he is married), the Officer begins his investigation. He obtains information from a number of sources and by the time he is finished, is able to submit a rather comprehensive report (see the Report of Field Investigation form in Appendix). [As the field investigation reports come into the office of the training bureau commander, he reviews them and makes a list of the applicants whom he thinks should he failed in the examination on the basis of information in the reports. After his review, the reports are sent to the civil service examination section where the examiner in charge of the trOOper examination also reviews them and compiles a list of those he thinks should be failed. After all reports are reviewed, the commander of the training ~38- bureau and the civil service examiner meet as the "field investigation review board" to compare notes. All appli— cants whose names have been listed by both reviewers are eliminated from further competition in the examination. This procedure admittedly gives these two persons consid- erable discretionary authority over the outcome of the examination for any given applicant. However, their de- cisions are not based on mere whim or capricious impres- sion. They are guided first of all by a number of minimum qualification requirements spelled out in the civil service specifications for the trooper classification, and second- ly, by departmental standards and precedent cases that have develOped over the years. Any applicant who is failed by the field investigation review board has the same right of appeal to the civil service hearing board as any other failing applicant. Probably the main reason for failing applicants on the basis of the field investigation is the traffic viola- tion record the investigation brings to light. A number of these applicants are those whose applications were con~ ditionally accepted by the examiner. Usually the applicant himself does not provide enough information on which to base a decision, and the number of such cases precludes a complete check of the traffic record at the initial stage of the examination. To illustrate, the applicant may in- dicate on his application form that he has received several "tickets" for traffic violations. This in itself does not - 39 - warrant rejecting the application so it is conditionally accepted. However, the field investigation reveals that these ”tickets" were all received within a twelveémonth period and resulted in the applicant's Operator's license being suspended for a thirty-day period. Had the examiner known these facts, he would have rejected the application-- any loss of driving privilege as a result of violations automatically disqualifies the applicant. Other appli- cants fail to list any traffic violations, but the inves- tigation reveals convictions for from one or two offenses to over a dozen in some cases. In screening the field investigation reports, cases arise in which either the civil service examiner or the state police officer thinks that an applicant should he failed, but the other member does not have the applicant's name on his failing list. The report is then jointly re- viewed, and if agreement either to accept or fail cannot be reached, the applicant is scheduled to appear before the oral interview board where the final decisiOn is made. The oral interview process is the subject of the next section of the study. The Oral Interviews The State Police Trooper examination applicants whose interviews are the subject of this study were inter- viewed as three separate examination groups, in the same sequence as they were scheduled to compete in the written -40.. test portion of the examination. Applicants who took the written test on September 12, 1956, were interviewed dur- ing a four-day period in February, 1957. Those who took the January 12, 1957, written test were interviewed in April, 1957. Interviews for this group took five days. The February, 1957, written test group completed their interviews in June, 1957. Six days of interviewing were required to complete the examinations for this group. A total of 2,911 applications were received for the three examinations. Beginning with the processing of the applications, each step in the examination procedure re- duced the number of applicants still in competition. Add to these unsuccessful applicants the number who withdrew and those who just did not show up to compete in one phase or another of the examination, and those remaining to be interviewed numbered less than ten per cent of the origi- nal group. To be exact, 243 of the initial 2,911 appli- cants were still competing following the screening by the field investigation review board. The interviews were held at Michigan State Police Headquarters in East Lansing for all applicants from the lower peninsula. A11 upper peninsula applicants were in- terviewed at the Michigan State Police Eighth District Headquarters at Marquette. Because of the relatively large numbers of appli- cants to be interviewed for each of the three examinations, multiple interview boards were used. During the four days -41.. of the February interviews, two boards were in session at East Lansing for two days. On the third day, one board interviewed at East Lansing while two members of the sec- ond board drove to Marquette. 0n the fourth day, both boards were again interviewing-~one at East Lansing and one at Marquette. However, the third man on the board at Marquette was a different person than the one who served during the first two days at East Lansing. This arrange- ment may seemconfusing, but in actual practice it works out very well. A similar plan was followed when the April and June interviews were conducted. However, for the April interviews there was a larger group of applicants so two interview boards were in session during the first three days, and one board remained in session at East Lansing while members of the second went to Marquette for one day of interviews there. The June arrangements were the same as those for the February interviews. In each series of interviews, the one board in session at Marquette interviewed all upper peninsula applicants. As was previously indicated, each interview board was composed of three members. This is in line with ac- cepted civil service interviewing procedure. Board members with differing backgrounds are asked to serve so that ap- plicants' qualifications may be evaluated from differing points of view. For these interviews, each board consisted of a state police command officer, a staff or faculty mem- ber from Michigan State University, and a member of the - A2 - examination staff of the Michigan Civil Service Commission. For the Marquette interviews, the Michigan State University board member was replaced by a faculty member from Northern Michigan College. A total of thirteen interviewers were used-~six university men, four state police officers, and three civil service examiners. The university men were all experienced in inter- viewing and in psychological testing techniques. All had academic backgrounds emphasizing psychology as a major field of study. They were invited to serve as interview board members on the premise that their knowledge of psychological factors, coupled with their interviewing experience, would make possible a more valid over-all rating of each applicant. They would evaluate the appli- cants on the basis of the same general standards as the other board members and base their ratings on the same factors listed on the rating form. However, it was hOped that their different frames of reference, especially in regard to certain psychological factors such as motivation and attitudes, would result in a more penetrating analysis of the over-all suitability of the applicants. For five of these six interviewers, these examinations provided their first experience as members of a civil service inter- view board. The sixth member had assisted several times previously on trooper interview boards at Marquette. The state police members of the interview boards were all officers with between fifteen and twenty years of -u3- experience in the department. All were selected because of their demonstrated command and leadership abilities and their insight concerning problems of recruiting and training new officers. Three were directly concerned with the personnel and training functions of the department. All of these officers had played a part in establishing the qualifications for trooper applicants and in develop- ing the training school curriculum. One was a captain, two were lieutenants, and the fourth was a sergeant. All had served as interview board members for previous groups of trOOper applicants. The three board members representing the civil service commission were all senior members of the exami- nation staff. All had served on a number of previous interview boards for trooper applicants. The interview board members met several days prior to the first interviews for each of the three examinations. At these meetings the interview procedure was discussed in detail, the use of the rating forms was explained, and there was a general discussion of the traits and character- istics of applicants that experience had shown to be either desirable or undesirable on the bases of previous examina- tions and training school classes. Some of the more im; portant factors, both favorable and unfavorable, were dis- cussed in terms of how their presence or absence in an applicant's over-all qualifications might affect his rating. On the first day of interviewing, each board held a brief -im- meeting prior to interviewing their first applicant. This was to clarify their procedure, select one member to act as chairman of the interview board, and to answer any last-minute questions by any of the members. The function of the board chairman was to escort the applicant to the interview room, introduce the board members, explain to the applicant the general purpose of the interview and its part in the over-all examination process, and start the discussion by asking the first few questions. The chair- man was nominally in charge of the board of which he was a member. He largely set the pace for his board and co- ordinated its activities with that of the other interview boards. Because of their previous experience with the trooper examination, either the civil service examiner or the state police officer acted as chairman. In some cases these two members rotated the assignment, with each serv- ing in the capacity for one-half of the interviewing each day. In answering questions asked by applicants, all board members participated; however, the state police officer and the civil service examiner answered most ques- tions since the applicants inquired mainly about civil service procedure or the trOOper training school. When the interviewing schedules were made up, no thought was given as to which interview board would inter- view any given applicant. Of course, when there was but one board in session (as at Marquette), that one board in- terviewed all allocated to that day's schedule. - A5 - At East Lansing, when the applicant checked in with the receptionist, his file was placed in an "out" basket behind any files already there. When a board was ready for another applicant, the chairman would take the file for the next applicant from the front of the basket. Thus, the applicants were interviewed in the order of their arrival, and each was interviewed by the board that happened to be ready for another applicant when his file had reached the front of the basket. When the chairman had picked up the applicant's file, he returned to the interview room and briefly re- viewed its contents--both for his own information as well as the other board members. The file contained the pre- viously-mentioned questionnaire the applicant had completed as well as the report of the field investigation. As the file was reviewed, the board members could take notes if they wished to do so. A brief personal history summary was reported from the questionnaire, and key points on each category of the field investigation report were brought out. Each file contained a cover sheet prepared by a member of the training school staff in which he called attention to matters about which the board members might wish to question the applicant. Examples of these nota- tions are: "Poor grades in high school," "Disciplinary action received in service," "Check traffic violations," "Extended period of illness while in high school," "Poor 1 references,‘ and "Dismissed from employment--l956." These - A6 - cover sheets were prepared after the field investigation review board had approved the applicant to appear before the interview board. No single matter mentioned would by itself be a sufficient reason for failing the applicant. The notes were made to call the attention of the board to matters that might otherwise be overlooked in a hurried review of the file. Applicants were usually questioned about these matters that were noted. The interviewers could then consider the applicant's statements about the questionable factors or incidents as well as any official records or the investigating officer's report. In making their final rating, the board members were free to evalu- ate each situation as they saw fit. The writer, while serving as a board member, found that in some cases these matters had a negligible effect on the final rating. In other instances the discussion with the applicant brought out factors that were the basis for a failing rating. Following the review of the file, the chairman of the board brought the applicant into the interview room, introduced him, and the interview proceeded. As stated previously, the time alloted to each applicant was thirty minutes. The time spent in reviewing the applicant's file varied from five to as much as ten minutes. The actual interview took from twenty to thirty minutes, with the average time being closer to the lower figure. After the applicant left, the board members used an additional three to five minutes to make their ratings and comments. -47.. In making the ratings, each rater acted independ- ently with no consultation with other board members or group discussion preceding the rating. Occasional comments are made following the rating, and sometimes board members will discuss certain applicants at the end of the day's interviewing, but there is no discussion of the applicant prior to the rating. This is not to imply that board members are completely unaware of how other members may have rated an applicant. The writer has found that board members come to recognize that certain attributes, atti- tudes, and combinations of factors in an applicant tend to influence other board members in a manner that can be predicted with some success. The questions asked and the subjects pursued in the questioning also indicate to other interviewers the reaction of a board member after the board has worked together long enough for the members to become fairly well acquainted. This is not to imply that the board members can tell what numerical rating others will give an applicant, but it is not unusual for one board member to be fairly accurate in judging whether a fellow board member will "pass" or "fail" certain applicants. The writer sees no harm in this in itself. However, there is the possibility that an interviewer may consciously or subconsciously allow his impression of how the other board members will rate an applicant to influence his own rating. This could be cited as a possible weakness in the use of multiple-member interviewing boards. Needless to say, any - H8 _ board member who becomes aware that this factor is influ- encing his rating should take immediate positive steps to overcome the influence. If he cannot accomplish this, he should withdraw from the interviewing board. To continue to rate applicants under these conditions would constitute a double injustice-~first, to the applicants, and second- ly, to his fellow board members. Some merit system agencies use the method whereby the board members discuss the applicant's qualifications and arrive at a mutually agreed-upon rating which is then recorded on a single rating form. The Michigan Civil Serv- ice Commission uses the individual rating method with the ratings being combined and the average score computed after the interviews are completed. The rating form used in these interviews was espe- cially designed to be used in the State Police Trooper examination (see Appendix). Basically, the form consists of four parts: The instructions to the rater, the identi- fication data, seven specific qualification factors with five sub-categories in each, and the final rating scale where the actual rating is made. The instructions printed on the form are very brief and general. At the meeting of the board members prior to the interviews, these instructions were enlarged upon and discussed in detail. The name of the applicant, the interview date, and the actual time of the interview appear on the form along - Ag - with the identification number of the rater. The actual time that the applicant was in the interview room is all that was recorded (time spent reviewing the file and fill- ing in the rating forms was not included). Although the board members were introduced to the applicant, each rating form is identified only by a number that was assigned to the rater. The rules of the Michigan Civil Service Com- mission allow an applicant to review all parts of his examination with the exception of the initial written test portion. The method of identifying each rater by a number makes it impossible for the applicant to positively identify any one interview rating as having been made by any certain board member. In making his rating, the interviewer actually rates the applicant in all seven of the factors listed on the form although but one final rating score is given. The factors considered specifically are: Voice and speech, clothing and grooming, physical appearance, ability to express ideas, emotional stability, friendliness, and general attitude. Each of these is listed on a separate line and includes five categories of general descriptive phrases which may or may not be appropriate for any partic- ular applicant. In general, the undesirable descriptions of each factor appear on the left of the series and the more favorable attributes are listed on the right side of the form. Raters are instructed to underline the words or phrases which they think apply to the applicant being - 5o - interviewed. If the rater feels that none of the terms listed are applicable, he may write his own comments on a line provided for that purpose. Experience with the rating form has shown that the categories labeled "Emotional Stability" and "Friendli- ness" are most difficult to rate--at least this has been the writer's reaction. Actually, these terms are mis- nomers in that they do not identify accurately the traits evaluated. The writer submits that few psychiatrists could appraise accurately a person‘s emotional stability on the basis of an interview of less than thirty minutes. Certainly, it is a large order for a layman to undertake. The writer is of the opinion that what is being judged primarily and rated as "Emotional Stability" is the reaction of the applicant to an unfamiliar "stress" situation which in many cases results in a noticeable nervousness on the part of the applicant. Other factors do enter into the evaluation, however. For instance, the field investigation report may indicate that the applicant is "quick-tempered" or has a reputation for "coolness" in emergency situations. These comments would be weighed by the interviewer in conjunction with his own observation of the applicant's behavior. The term "Friendliness" could be labeled more apprOpriately "SOciability" or "Social Presence." Here again the field investigation report may provide clues to the gregarious nature of the applicant, or the lack of _ 51 - it, but the interview board members are interested primar- ily in the impression the man makes on the basis of a brief encounter-~such as occurs when an officer stops a motorist and issues a traffic violation summons. At the bottom of the form is a rating scale labeled "Personal fitness for the position." The range of the Scale is from 0 (zero) to 100. The so-called "passing point" on the scale is 70. Any rating of less than 70 is a failing rating and falls within the rating range labeled "Unsuited for this work-~not endorsed." The "passing" range On the scale is sub-divided into three areas, each having a SCOpe of ten points: 70 through 79 is "Endorsed", 80 through 89 is "Endorsed with confidence", and 90 through 100 is "Endorsed with enthusiasm." From 0 to 70, the rating points are in intervals of 10. From 70 to 100, there are single intervals. Whenever a rater assigned any score lower than 70, it was mandatory that he explain his failing rating in the space provided for comments below the rating scale. In actual practice the raters usually made some statement on the comment line regardless of the score they assigned. Usually, the over-all impression of the applicant that the interviewer gained in judging him on the bases of the seven categories was reflected in the rating assigned on the 100 point scale. However, the personnel fitness rating is actually independent of these seven factors in that many other factors enter into the final rating. A _ 52 - rater may indicate favorable comments near the right side of the form for nearly all of the seven factors listed but still assign a low or even a failing score because of other considerations. This would normally be as a result of mat- ters mentioned in the field investigation report, similar to those previously listed. Or, it could be because the rater considered the applicant unsuitable on the basis of a pronounced lack of favorable qualities in one of the seven factors-~such as general attitude or emotional sta- bility. In general, however, an applicant would have to rate fairly low in several of the seven factors before a failing rating would result. At the time these interviews were conducted, any two failing ratings on the interview automatically elim- inated the applicant from the examination regardless of the rating assigned by the third rater or the score the applicant achieved on the written test portion of the examination. Current practice requires that the failing ratings be unanimous by all board members before the ap- plicant is automatically eliminated. The result of this recent directive from the civil service commission has been to make more difficult the task of the interview board members since they now must rank the "failing" ap- plicants as well as those who receive "passing" scores. Previous to this change, the majority Of "failing" appli- cants usually were given a score of 60 on the rating scale on the theory that the majority decision would govern. - 53 _ If another rater also gave a score of 60, the applicant would be eliminated; but if the other two board members gave scores of 70 or higher, the one rating of 60 would not be so low that the one board member would be respon- sible for failing the applicant when the other raters considered him to be qualified. Now, since a majority de- cision does not govern, the writer would hazard a guess that some of the individual interviewer's ratings are going to be so low as to be difficult to justify if the applicant files an appeal. For this reason, as well as others which will be cited later, the writer is not in full accord with the current practice. If the interview board is "split" concerning pass- ing and failing ratings, the ratings are computed and added to the written test score. If the written test score is sufficiently high to off-set the points lost on the interview so that the total examination score is 70 or higher, the applicant's name is placed on the employ- ment list. For example, if an applicant received two ratings of 60 and one of 80 on his interview, and a score of 38.50 on the written test, he would pass the examination; while under the previous practice, he would have failed. Each part (written test and interview) counts 50% of the final examination score. For each part, the maximum score is 50 and the minimum passing score is 70% of 50, or 35. The interview score would be computed in this manner: 60 +-60 - 5h _ 4—80== 200—+ 3:: 66.667 (the average score), and 66.667 x 5q%== 33.334, which is the final interview score. This 33.334 added to the written test score of 38.50 would give a final score of 71.834, nearly two points over the re- quired passing score of 70. The applicant, who had been failed by two of the three interviewers, would gain a position on the employment list and in due time would be called to attend a training school. The procedure in filling vacancies in state service is for the civil service department to certify three names from the top of the appropriate employment list for any one position to be filled. The hiring authority in the department where the vacancy exists selects one of the three persons whose names were certified. Usually a limit- ed number of positions are filled at any one time. Often only a few appointments are made during the entire "life" of the employment list, and a number of names of qualified applicants still appear on the list when it expires. When this occurs, those who have passed the examination with median or lower scores often are not considered for em- ployment since their names have not come within "certifiable range" on the list. I The above situation has not prevailed in the recent expansion program of the Michigan State Police. Applicants have been appointed to the training school in groups as large as seventy-five at one time. The department follows the practice of appointing applicants from the list in rank - 55 - order, starting at the top of the list and working down through the scores until the desired number of men have been appointed. The employment lists rarely have been sufficiently large to cover the appointments required for more than two training schools. Sometimes an entire list will be used in making appointments to a single training school. The practice in appointing state police troopers thus differs from that used in filling a majority of state positions. Every applicant who gains a place on the em- ployment list eventually is considered for appointment. In the case of the State Police TrOOper examination, the writer, for three reasons, favors the procedure where- by the applicant is eliminated from competition if he is failed by a majority of the interviewers. The first reason has been cited-~the fact that all passing applicants are eventually certified to a training school. The second reason is that the written test consists of an estimate of intellectual ability only, and although this is an un- deniably important factor, there are many other factors necessary for competent performance in the position. These other factors lend themselves more readily to appraisal in an interview situation. Thirdly, recent discussions with training school officials indicates that applicants who are failed by a majority of the interviewers, and are later appointed to the school, usually prove to be unsatisfac- tory, or at best, "border-line" trainees. - 56 - The result of these interviews was that 150 of the 243 applicants received passing ratings from the interview board, and with their written test score added, received a final score on the examination of 70 or higher, thus gain- ing a position on the employment list. There were a number of names already on the employ- ment list as a result of an examination given earlier in 1956. However, these applicants had fairly low scores; the high scorers on the list having been appointed pre- viously to a 1956 training school. A majority of the ap- plicants who were appointed to the May 6 to June 20, 1957, training school were in the groups interviewed in February and April. Appointments had thus been made before the June interviews were held. The applicants interviewed in June were appointed to the second school considered in this study--the school in session from July 29 through September 16, 1957. Additional training schools were held later in 1957 and in 1958. The following section of this study is a general discussion of these trooper training schools. The Trooper Training School Two trOOper training schools serve as the basis of that part of this study dealing with school performance of the applicants. Each school was six weeks long. Both were conducted at the training school facilities of the Michigan State Police at their headquarters in East Lansing, Michigan. - 57 - A total of 81 applicants reported for the first school which lasted from May 6 to June 20, 1957. Sixty- two applicants were graduated from this school, and three who were graduated resigned during the six months proba- tionary period, making a total of 22 who were appointed but failed to become "confirmed" troopers. The second schoOl, from July 29 to September 16, 1957, was smaller with only 59 applicants being initially appointed. Of these applicants, 40 were graduated. One applicant re- signed two weeks after graduation, making a total of 20 applicants who were unsuccessful in this group. From the two schools, a total of 102 applicants were graduated. The interview, training school, and probationary trooper ratings of these men comprise the data which are the bases of this study. The training school is a self—contained unit. Ap- plicants live in dormitories, eat in the mess-hall, and receive all of their training right at state police head- quarters. The curriculum consists of both academic and physi- cal instruction, with the larger share being of an academic nature. A complete listing of subjects in which instruc- tion is given will be found in the Appendix. However, examples of academic subjects are: Report writing, type- writing, motor vehicle law, law of arrest, and public speaking. The two major courses of physical instruction are personal combat and water safety. Some courses involve - 58 _ both Classroom work and physical instruction. Examples of this type of subject are first aid, firearms, traffic con- trol, and motor vehicle accident investigation. In addition to the course work, the trainees are assigned to various work details throughout their stay at the school. These consist of cleaning assignments in the living quarters, the gymnasium, and the swimming pool and locker-room areas. Other trainees are assigned to the kitchen detail. Each day begins with a thirty-minute period of calisthenics at 6:00 A.M., fifteen minutes after the trainees are awakened. From then until "lights out" at 10:30 P.M., the trainees follow a set routine which allows little leisure time unless purposely scheduled. The fol? lowing schedule is illustrative of an average daily routine at the training school: Daily Activity Schedule bureau. 5:45 A.M. Reveille 6:00-6:30 A.M. Calisthenics 7:00 A.M. Breakfast 7:30 A.M. Sick Call 7:45 A.M. Inspection 8:00-11:55 A.M Morning Classes 12:00 Noon Dinner 1:00-4:55 P.M. Afternoon Classes 5:00 P.M. Supper 6:00-10:30 P.M. Study and Leisure time (6:00-8:00 P.M. Occasional Evening Classes) 10:30 P.M. Lights Out The nucleus of the training school staff is about a half-dozen officers permanently assigned to the training When a training school is in process, additional - 59 - personnel are temporarily assigned to the training bureau to aid in the instructing and to assist in supervising and directing the activities of the trainees. Officers from both the headquarters administration staff and from the field are assigned to the school. Each of the special bureaus and divisions of the department is represented, either at the initial training school or during the follow- up training period when the trainees ("confirmed" troopers by then) return for an additional four to six weeks of "refresher" and supplementary instruction. The grading system used in the training school results in a weekly grade for each trainee in each subject. Examinations are usually given every Saturday morning while the school is in session. Each trainee's grades in each subject are added and an average grade for that week's work is computed. At the completion of the training school, an average grade is computed for each trainee for the en- tire period of instruction. These grades, representing the final school average for each trainee, were used as the basis for comparison of training school performance with other ratings used as criteria in this study. Before going on through the resulting comparison of these ratings, a few final comments concerning the train- ing school might be helpful in assisting the reader to understand better the role of the training school in the over-all Operation of the department. -60... The duties of the Michigan State Police cover two major categories. The first is that of highway traffic patrol. The second, of equal importance, is the field of general law enforcement. The primary function of a trOOp- er is highway patrol and enforcement of the traffic laws. However, in performing these duties, the trooper is con- stantly coming in contact with a variety of crimes and criminals. The use of the automobile is as important a factor in the activities of the criminal element of our pOpulation as it is in the daily activities of the gen- eral public. A review of the training school curriculum will re- veal that the most time is spent on highway patrol and traffic enforcement subjects. When the trainees return for the supplementary schooling, they will have completed a six-month assignment at one of the fifty-four state po- lice posts located throughout the state. They will have had considerable experience in traffic patrol work and will have assisted in the investigations of a number of criminal complaints. This background will enable them to comprehend better the instruction concerning general law enforcement and crime investigation. It is primarily for these reasons that these latter subjects are largely re- served for the second training session. The subject of the following chapter is a discussion of the analysis of the interview ratings, training school scores, and probationary trOOper ratings. CHAPTER III DATA ANALYSIS Interview Ratings, Training School Grades and Probationary Trooper Ratings The data which serve as the basis for this study consist of scores or ratings of the state police trOOper applicants at three different stages of their progress from applicant to state police officer. The basic method- ology consists of a comparison of the oral interview ratings with the average scores achieved at the trOOper training school, followed by a comparison of these same ratings with the ratings given by the supervising senior officers dur- ing the six-month probationary training period. As relat- ed to the traditional research experimental method then, we have, in a sense, the interview ratings serving as the variable factor being measured, the training school scores as the intermediate control criteria, and the probationary service ratings in the role of the ultimate control criteria. The Pearson product-moment method of correlation was used beCause the data lend themselves to this method of analysis and also because it is the most commonly-used pro- cedure for analyzing interview ratings. -61- -52.. As was stated at the outset of this study, this was not a planned experiment in the sense that the various pro- cedures and rating techniques were devised especially for the purpose of conducting the study. The subjects had been selected, the interview techniques and rating methods had been established, the training school curriculum and grad- ing criteria were agreed upon, and the probationary rating criteria had been formulated and rating forms devised. In fact, all elements of the study had been completed before the decision was made to use the trooper interviews as the subject of the study. In this fact lies the value of the project for the purpose that the writer had in mind. The examination procedures for selecting State Po- lice TrOOpers have been develOped over the years, and are apparently doing an effective job if the caliber of the average trOOper now in the Michigan State Police is used as an evaluation criterion.26 However, the command offi- cers at the training school and the writer have noted the relatively large number of trainees who have resigned or been dismissed from recent training school groups. In some instances, the number of trainees who have failed to complete the training has been over one-third of the group 26The Michigan State Police Department is nation- ally recognized as one of the leading state organizations whose duties encompass both "general' police investiga- tion as well as the traffic patrol function. However, there is not, to the writer's knowledge, any organization that makes an official evaluation or ranking of police organizations of this type on a state~by~state basis. -63- that initially started. 0f the 138 trainees who started the schools surveyed in this study, 38 failed to finish. This was 27.5% of the total. In addition, four trainees who were graduated from the school later resigned from the department. If these are included, the number who failed increases to 30.4% of the initial group. The study was undertaken with a two-fold purpose. The firSt objective was to see if there was a significant correlation between the ratings of the applicants in the oral interview portion of the examination and their subse- quent performance in the training school and as probation- ary troopers. The second purpose was to determine from the study, if possible, evidence as to why the examination process was failing to screen out applicants who failed to complete the training school once they were appointed. Since the study lacked the pro—planning of a con- trolled experiment, the writer did not formulate hypothe- ses with the intent that the findings would prove or dis- prove them. Whether the results were positive or negative, it was hOped that their interpretation would provide at least a partial answer to the problem of the failing appli- cants. In making the data analysis, the interview ratings were the first to be considered. Because of the correla- tions planned, it was necessary to treat the interview ratings in two groups. The total number of cases was 138, with 100 in the group that successfully completed the - 64 - training course, and 38 who did not. Since there were no training school average grades or probationary ratings on these 38 men, the interview ratings could not be correla- ted with those factors of performance. In the absence of correlation figures for the group that failed to complete the training school, the analysis had to be confined to what could be determined from the few school grades avail- able as well as information concerning reasons for leaving the training school. The first analysis made was the comparison of each interview board member's ratings with those of the other members. The ratings were first tabulated to determine the range and frequencies. The mean score was ascertained and the standard deviation of the combined individual means was then computed. Table IV on the following page shows the results of this analysis as well as the results of a similar analysis of the interview ratings of the failing applicants.27 Originally two tables, they have been com- bined for purposes of ease of comparison. In rating the applicants who successfully completed the schOol, the lowest score any of the raters gave was 60, which is ten points below the minimum passing rating.28 There was a 5-point spread between the highest scores given by the individual raters, with the state police board __._——_.— 27Failing rating analysis is discussed later, beginning on page 67. 28See discussion of failing ratings on pages 52-53- - 65 - .annoa hpmnoapmnopa on» wcansp cocwamon hoped snow omozp no oxen .mwchmn mafiafimm co poncmfioo po: mm: :ofipma>op cnmpcmpmo .OOH ma 2 ocp occam powwow mucosa: on» spas Hmoapcopa ohm Ozonw wcammma pom monzwau ommpcoonomn popes ooa>nom Ha>fio I gonna apampo>HCD I .>HZD Qmo hopes ooafioa spoon 1 mass ruin m.©s so.smumm.so : om 2H a Hawk mwzHB mmaow o: mm m: m whom u--- m.cs mmaoo A an OH u Hana Ammv “many gummy .>Hz: we.» mwws mmrow 0 mm mm H whom i... c.as monoo m mm OH H Hana ammflv gamma hmwmv lama am: ~m.w awmw smuoo m mm H: a comm oonemm zen: mozam cs ownos omuam om mmaczaem soaom uo>o emmmeam cmmHozmsamma ceases on: updsoaaaac am one Hoocom cap copoflasoo 0:3 mpcmoaaam< OOH no mwcapmm 3oa>noch no Comanmasoo >H mgm¢8 - 66 - member giving the highest rating and thus having the great- est range of ratings. The most restricted range of ratings was that of the civil service board member. Six applicants who later successfully completed the training school were given failing ratings by the university board member. The civil service member failed four applicants who later completed the school. However, in the long run, his eval- uation of the applicant proved to be correct on two of them. One resigned after completing three months of the probationary period, and the second after completing five months. Each of the raters gave the majority of the appli- cants a rating in the lower one-third of the 70 through 100 point passing range. In each case, over 50% of the individual ratings were in the 70 through 80 point range although when the ratings were combined, the number dropped to 46% of the total. There was a variation of only 1.5 points in the mean scores of the raters. The standard deviation of the means differed by only .44 of one point. The ratings Of each board member were correlated with the ratings of the other members with the following results: Raters Correlation Coefficient State police rater with civil service rater .73 ji.ou7 University rater with state police rater .68 :t.054 Civil service rater with university rater .64 I:.059. -67.. These correlations are an indication of the relia- bility of the ratings. Although they are not high enough to warrant the conclusion that the ratings definitely are reliable, they are sufficiently high to indicate a substan- tial relationship. They indicate that there was a positive agreement among the raters concerning the majority of the factors that they were evaluating, at least to an extent that resulted in a substantial agreement in their final judgment of the applicants. The correlations show that the state police rater and the civil service rater were nearest in agreement on their evaluation of the applicants. The correlation was lowest in the comparison of the civil service rater's evaluations with those of the university rater. This implies that these two raters were not ar- riving at similar conclusions concerning the qualifica- tions of the applicants because one of them was being influenced by factors which did not affect the ratings of the other. The ratings of the 38 applicants who failed to com- plete the training school were tabulated by the same method used in tabulating the ratings of the passing applicants. However, since no correlations were to be made, the stand- ard deviations were not computed. The results of the tabulation are shown in Table IV on page 65. The tabulation indicates that the state police mem- ber of the interview board did the best job of predicting applicants who for one reason or another failed to complete - 68 - the training school. The civil service member of the inter- view board did the poorest in this respect since he did not assign a failing rating to any one of the 38 who eventually failed in the school. Comparing the tabulations in Table IV further validates the predictions of the state police mem- ber since it shows he predicted failure for only two ap— plicants who did not fail. Combining his ratings for both the passing and the failing applicants, he has a net fig- ure of three accurate predictions of failure. On this same basis, the civil service member has a net figure of two correct predictiOns (since two applicants eventually were Classified as failing). The university member incorrectly predicted five failures. He predicted failure for six applicants, all of whom eventually succeeded, and he failed one applicant who later did fail in the school. These pre- dictions assume a rather insignificant status when an analysis of the training school records reveals that all three board members assigned passing ratings to 35 appli- cants, all of whom eventually failed to complete the train- ing school or resigned during the probationary period. The initial plan for the analysis of the training school scores was to select certain key subjects from the curriculum and correlate the grades with the interviewers' ratings. This approach had to be abandoned when it was found that "key" subjects could not be identified on any valid basis. An alternate method was selected then where- by the six weekly average grades were added and a final - 69 - over-all school average was computed for each applicant. These average grades were tabulated and correlated with the combined oral interview ratings and the individual ratings of each board member. The range of training school grades was from 77.61 to 93.80. The mean score was 86.9, and the standard deviation, 3.14. The validity correla- tions were as follows: State police rater with school scores .19 :£.O97 University rater with school scores .14 :t.098 Civil service rater with school scores .23 21.095 Combined interview ratings with school scores .20 21.096. The above correlations indicate that the civil service rater did the better job of the three in predicting the level of performance of the applicants appointed to the training school, and the university rater did the poorest. 0f the four correlations, only the .23 and .20 figures indicate a "border-line" significance. The .1A and .19 figures fall in the area usually considered as insignifi- cant insofar as indicating a valid relationship. The probationary ratings of the trainees who were graduated and assigned to field duties presented a special problem as to their utilization in the study. The form was devised by the Michigan State Police as an evaluation tool to assist in appraising the prOgress of the probationary - 7o - troopers. Completion of the form does not result in a numerical rating. The form provides for rating the troop- er in twelve different categories, each representing an important aspect of on-the-job performance. Under each of the twelve categories appear four descriptive phrases or comments, each reflecting a degree of excellence in that trait ranging from highly desirable to unsatisfactory (see the rating form in the Appendix). The face of the form bears the probationary troop- er's name and the post to which he is assigned, the period of time covered by the rating, the signature of the super- vising officer who made the rating, and the date. 0n the reverse side appear spaces for comments by the probationary trOOper's post and district commanders. A probationary rating is completed for each new trooper at the end of each month of the six-month probationary period. There is no standard practice in the department concerning the assign- ment of senior officers to supervise the field training of probationary trOOpers. However, if the complement of sen- ior officers at the post permits, it is customary to have the probationary officer work with several different super- visors during the six-month period. Experienced officers do not use all the same techniques or have the same approach in conducting investigations, handling Citizen's complaints, or interrogating criminal suspects. By working with sever- al senior officers, the probationary officer is given the Opportunity to gain a broader knowledge of police methods - 71 - and techniques. Also, the post and district command offi- cers can evaluate better the performance of the new troop- ers after receiving progress reports from several super- visong officers. However, in some of the smaller posts having a limited complement of officers, it is sometimes impossible for each probationary trOOper to work with more than one senior officer except on brief assignments. Some of the ratings of the training school graduates who were the subjects of the study were all completed by the same senior officer. In other cases, as many as four super- visors participated in the rating procedure during the probationary period. In the analysis of these ratings, the writer's first thought was again to select certain "key" rating factors as the criteria. This idea was discarded for the same reason that it was eliminated as a technique for evaluat- ing the training school grades. However, it was obvious that some method of weighting the ratings would have to be devised if the desired numerical ratings were to be ob- tained. The most valid method would have been to assign the greatest weight to the more important factors of pro- bationary performance. Again, the question--"Which are the more important?"--and again, a satisfactory answer was not forthcoming. Proceeding on the assumption that each of the twelve factors is an essential aspect of performance, the decision was to weight them equally. A scale was de- vised whereby a check in one of the four boxes in each - 72 - category would be assigned a numerical weight in descend- ing order of desirability, as indicated by the accompany- ing descriptive phrase. The numerical values selected were: Eight (for the top box), six, four, and two (for the lower box, which indicates undesirable traits). Thus, a probationary officer who received a rating of eight in each of the twelve categories would achieve a cumulative score of 96, which is comparable to the top scores achieved in both the interview ratings and the training school grades. The numerical values were assigned to the ratings of the first dozen probationary officers and the scores were computed. It was discovered that the average ratings were essentially the same whether based on all six proba- tionary ratings or only on the first two and last two ratings. On this basis the average probationary rating score was computed for each of the 98 probationary offi- cers, using only four ratings for each. The number of subjects was reduced to 98 because of the resignation of two of the initial 100 before each had received four of the probationary ratings. The range of the probationary rating scores was from 44 to 94. The mean score was 73, and the standard deviation was 11.37. When correlated with the oral inter- view ratings, the following validity coefficients were obtained: - 73 - State police rater with probationary ratings .22 i.O97 University rater with probationary ratings .07 iglO Civil service rater with probationary ratings .27 i3094 Combined interview ratings with probationary ratings .16 i;099. These correlations resulted in the same ranking among the raters as was the case with the training school grade cor- relations. Here the figures for the state police rater and the civil service rater were the higher. The combined rating and university rater figures were in the "non- significant" category. The state police rater figure was in the "border-line" area and the .27 figure for the civil service rater approaches the point where a positive re- lationship is indicated. Before concluding this chapter, the writer would like to present some information found in the records of the training school concerning the alleged reasons for the failure of the 38 trainees who did not complete the train- ing school. The term "alleged" is used since in some in- stances the command officers at the training school ques- tion that the reasons given by the trainee in his exit interview were the true causes for his leaving the school. This contention is supported by the fact that several ap- plicants who resigned from a training school later re- applied for the trooper examination, and when interviewed - 7a - during the second examination, stated a reason for leav- ing the school that differed from the one shown on their resignation form. The length of time the failing applicants remained at the school varied from two days to 29 days. The aver- age stay before leaving was 8.15 days. There were only five applicants dismissed from the school. The other 33 who left submitted voluntary resignations although in a few of these cases dismissal would have eventually result- ed. Training school records show the following reasons for separation: Resignations (33) Lack of over-all physical ability Poor general physical condition Fear in boxing Unable to defend self in boxing Specific physical disability: Right knee injury (1) Leg ailment (1) Lack of interest in school Homesickness Miscellaneous: Fear of swimming (1) Disliked school (2) Financial problems at home (1) 4 DONUOCID -L‘—‘\lf\) Dismissals (5) Absent from school without leave 3 Cannot accept SUperviSion or discipline 1 Failed course work (also cited lack of interest) 1 Total: 38 It will be noted that of the 38 who left the school, 18 did so for reasons based on physical shortcomings. This z r , _ ‘ ' arm .11; i ‘ ‘ ' .st \ ...... ' 'r . - O \ - \ a. f \ r - - l I r .- , - 9 ,- i I----- O \ .' :12 i ii. ‘ ‘ i. - 75 - is nearly fifty per cent of the total who left. Since the interview board members do not eliminate applicants because of suspected physical shortcomings, they should perhaps be credited with a better record of predicting performance than the tabulations indicate. The interview board may fail an applicant on the basis of obvious physical de- fects, but in questionable cases the final decision is left to the physician who examines all applicants prior to appointment. The writer has had the experience of assigning a passing rating to an applicant whom he felt reasonably sure would fail the training school for physi- cal reasons, but as a layman he could not justify his Opinion. The applicant was approved for appointment by the examining physician, but the physical activities proved to be too rigorous and he resigned from the school. The following concluding chapter of this study consists of a summary of the findings and the conclusions of the writer. 1.; -':‘1._' . ‘ o I n '2 ~ -- , 1511590 " :r-rr', m o J. " - ' '- ' C: "PIT/“9’": '3 f ' ' ', 1: ‘, ”it ‘ W' , . " ':, '.' ,g '. .."T)‘:Hgt’.U8 " o I L, .. 1' _ .. , . . . ; .."'r.;.:~.;.‘.09'16: ' I ‘:‘.;‘ .‘iJJtu '.‘.'-.;,.‘.' a " . . - an” .. 121.? .l3f‘i‘ o 1 . c" p - '.f . -.' r . - ' - r. . I .. 7:4'-‘\: “'._‘ _.l v . . I ‘ ‘ ‘. .‘7‘: .53.! In! .3 to U :3 ' D 'r 5' 1 o ' I o ‘. .2300! w. xggffitg :1":';..5.'v_': _ I; ' ’3 I . 0.3" 3.19“ {I5 0.. . . ' - Q. *‘ ' 1:1 luring». _ ' ' - - - ‘ '-~";o.‘3 03" iv .- .' fit)- fr: '.2 1'. . 1’. .. . , “in 1.358,; u. .Ag (1“ firs" 4:134'3'“. . " F I- 1.. .' -. : 21108891 I 0.0 a; 7. - . ~- ; _.. .1 {so " ' I ‘c.; ,._ 1.-.. J ' ‘ . ~ ‘ . . ' g . "'Ilq¢ 21101311129110» so...‘ ..: '7' . . . «$13,103.; 1 :‘ :0'_' fi. ‘ ’. . m (.91 g CHAPTER IV SUMMARY AND CONCLUSIONS This study was undertaken with the aim of analyz- ing the current interviewing procedures for state police trooper applicants by comparing interview results with performance in the trOOper training school and as proba- tionary officers in the field. The resulting correlation figures indicate that on the bases of the criteria used, the interview procedure lacks significant validity as a means of predicting subsequent performance in either the training school or in the field as a probationary officer. The correlations of interview ratings with training school scores are so low as to indicate only a slight re- lationship between them. Any prediction that a high inter- view rating would result in a high average score in the training school courses would be subject to valid criticism on the basis of the correlations obtained. In general, a correlation coefficient of less than .20 is considered to indicate an almost negligible relationship. The combined rating correlation was just at .20, with one individual rating correlation being above this point at .23, and two being below it at .19 and .14. However, there are two factors which should be given consideration in this - 76 - _ 77 - analysis. The first of these could be considered as being a point in favor of the interview ratings, or if inter- preted differently, a point adding support to the lack of validity as apparently indicated by the correlation coef- ficients. This factor deals with the amount of pre- screening the applicants have received prior to the time they are interviewed. The examination procedure is such that the obviously unsuited applicants have already been eliminated. It is a highly selected group of men that the interviewers must evaluate. All of them have already qual- ified on the bases of a number of minimum requirements. Many major factors of acceptance or rejection in the normal civil service interview are not present in this group of applicants. Final rating decisions must hinge on the as- pects of personal fitness that are difficult to identify and even more difficult to interpret accurately on the basis of a brief interview. For these reasons, each time an interviewer makes an extreme rating, either high or low, the odds against his rating being accurate are greater than would be the case in the usual interview situation. The Opposite interpretation of this viewpoint is re- flected in statements the writer has heard several times-- "These trooper interviews should present few problems. The applicants have already been thoroughly screened, so all you have to do is pass them." This reasoning is re- futed by this study which revealed that 35 applicants were "passed" by all three interview board members-~and all 35 -78- failed to complete the training school. This indicates that there are factors involved that are not being evalu- ated prOperly at some stage of the examining process. The writer is of the Opinion that the physical condition of the applicants is one of the factors, but there are others linked with failure in the training school for reasons such as "Lack of interest" and "Homesickness." It is these lat- ter factors that the interview board members are failing to interpret correctly. Having served on oral interview boards where the only pre-screening of the applicants was on the basis of the written examination and also on boards which interviewed applicants only after rather extensive pre-screening, the writer favors the viewpoint that the interview that follows the pre-screening is the more dif- ficult in terms of appraising the applicant's over-all personal fitness. The second of the two factors mentioned which should be given some consideration in attempting to eval- uate this study is a question as to the validity of at- tempting to use the training school grades as a criteria for validating the interview ratings. The interview ratings are based to a large extent on personality factors, primarily those of the applicants, but the personalities of the interview board members are also reflected in the ratings. 0n the other hand, the training school grades are based primarily on the applicants' performances on objective tests of knowledge acquired in the various - 79 - training school courses. The subject of this study is the interview ratings so no attempt was made to correlate scores on the written portion of the trOOper examination (the Wonderlic Personnel test) with the training school grades of the applicants. The writer submits, however, that if such a comparison were made, the resulting coef- ficient Of correlation would be greater than the .20 ob- tained in the comparison of the interview ratings and training school grades. The writer attaches significance to the fact that failure in the academic work of the training courses was listed only once (and then as one of two factors) as a reason for failure to complete the training school. Fur- ther, although the majority of the 38 trainees who left the school were not there long enough to compile an ex- tensive academic record, there were sufficient grades available to give some indication of the academic perform- ance of this group. A search of the training school rec- ords resulted in a tabulation of some 34 weekly average class grades of these 38 failing applicants. Nearly all of these grades were based on the course work covered in the first two weeks of the school. The mean of these 34 grades achieved by the 38 failing trainees was 90.4. The mean of the grades achieved by all trainees graded during this two-week period (including the grades of the 38) was 90.2. This is fairly substantial proof that the 38 train- ees, as a group, were not having difficulty with the subject - 80 _ matter of the training school during the period prior to their leaving. This leads to the further conclusion that criteria for determining success in the training school involves factors other than the ability to meet academic standards as reflected in school grades. This reasoning brings us back to the original statement that the use of training school grades as a validation criterion is Open to question. The reasoning may be summarized and the question answered in the four following statements: 1. The purpose of interview ratings, based largely on subjective evaluation of per- sonality factors, is to predict trainee success or failure in the training school and as a probationary trooper. 2. Training school grades are based largely on objective tests, and training school records reveal that grades obtained are not a determining factor in the resig- nation or dismissal of trainees. 3. The ability to meet academic standards is not a determining factor in the successful com- pletion of the training school (and the pre- diction of failure or successful completion is a basic goal of the interview process). 4. Training school grades are not a valid cri- terion for use in evaluating interview ratings. - 81 - In general, the validity coefficients obtained in the correlation of the interview ratings with the proba- tionary ratings were, like the coefficients resulting from the training school grade comparisons, so low that they fail to indicate any significant relationship. The co- efficient for the combined interview rating comparison was .16, even lower than for the like comparison with training school grades, which was .20. In these compari- sons there was but one validity coefficient based on the comparison of individual ratings that indicated a degree of validity that approaches significance. That was the coefficient for the ratings of the civil service member of the interview board, which was .27. The coefficient for the state police member was .22. For the university mem- ber, it was .07. A comparison of these validity coefficients with those obtained in the correlations with training school scores (page 69) shows a higher correlation with the pro- bationary ratings for both the state police and civil service members of the interview board. The correlations are lower for the university member of the board and also for the combined ratings. The greatest difference in correlations was in those of the university member. The figure decreased from .14 on the training school correla- tion to .07 in the probationary rating correlation. This was due to a larger percentage of the interview ratings of this rater falling in the lower ranges of the 70-80 : \ ‘. 0 ”p a. r J O J — u '4 :‘A . l... .. :Q’Iol - o .L n}: 1 ".11". 9d? O‘II‘ J I": -B£ca§oo 51f” Ono j... small!“ .I "'zuo I“ . U A! ‘l', O "4 4': {10:3 ‘. . r ‘ 3"} I I)“, I . ' f-303 ~47 .I‘ Of):- ' Lot-1d 5’35 A 0'! i'txoa - 82 - point range (see Table IV, page 65), thus causing a great- er number of the ratings (46.8%) to fall below the mean score. 0n the other hand, the mean of the probationary ratings was fairly low (73), and a large percentage of the ratings (87.7%) was above the mean. The end result was a large number of negative deviations for the individual interview ratings and a large number of positive deviations in the individual probationary ratings. The results could have been mitigated somewhat (with a probable increase in all of the correlation coefficients) had the writer dis- carded several unusually low probationary ratings on the basis that they were "extreme" scores and not representa- tive of the over-all ratings. However, to have increased the mean in this manner obviously would have caused spur- ious correlation coefficients to result. This would have compromised the basic objective of the study, which was to analyze the ratings "as is" with no attempts to skew the results. A comparison of the factOrs rated by the interview- ers with those rated by the officers supervising the pro- bationary troopers indicates some differences that should not be overlooked. The rating forms are so constructed that the interviewers are rating mainly in terms of apti- tudes and potential, whereas the officers in the field are rating key factors of ability on the basis of actual ob- servation of performance. There is no flaw in this, since the objective of the interviewers is to predict successful - 83 - performance. A major source of error may enter into these ratings, however, and seriously affect their reliability. The source of this error is the matter of interpretation of standards of adequate aptitude and potential by the interviewers and the standards of adequate performance by the officers making the probationary ratings. The writer is of the opinion that the ratings of the interviewers are apt to be more reliable and valid as criteria data than are the probationary ratings. This statement is based on two considerations. First, there are fewer interviewers than supervising officers. Be- cause of this, the interview ratings possess a greater degree of internal consistency in that each interviewer rates a larger percentage of the total group of applicants. All applicants rated by any one interviewer will be judged by the same standards and interpretations as to what con- stitutes adequate qualifications. Although the interview- ers were briefed concerning the requirements, and examples of adequate and inadequate qualifications were cited, each interviewer still had to formulate his individual inter- pretation of what constituted adequacy. An examination of the interview results indicates that as the number of ap- plicants rated by any one interviewer increases, the re- liability and validity of the ratings also increases. The civil service commission assigned three men to serve as interview board members for these applicants. The state police assigned four men, and there were six from the - 84 _ university. In each of the three series of correlations, the civil service ratings resulted in the higher correla- tion figures, with the state police ratings being second highest, and the ratings of the interviewers from the university being the lowest. It is reasonable to assume that the same factors which hampered the application of uniform interpretations of qualifications in the interview situation also entered into the probationary ratings, and probably to an even greater extent. Conditions favoring rater bias in the in- terpretation of what constitutes adequate performance are greater in the case of the supervising officers than for the interviewers. Where there was a total of 13 inter- viewers, the supervising officers who made the probation- ary ratings numbered nearer 150. Each individual officer's attitudes and law enforcement experience were bound to have influenced his evaluation of the performance of the proba- tionary trOOper he was supervising. The second factor tending to make the probationary ratings less reliable and valid was the lack of any pre- briefing of the supervising officers as to uniform inter- pretation of the terminology of the comments on the ap— praisal form. It is true that there are departmental in- structions covering the use of these forms, and the depart- ment has high standards of trOOper performance and conduct with which all senior officers are familiar. However, in the final analysis, it is the judgment of the individual _ 85 - senior officer as to what constitutes adequate performance on the part of the probationary officer. The writer sub- mits that with some 150 officers involved, there is bound to be some rather wide variation in the evaluations. In a final brief summary of this study, the writer would like to make the following observations based on his interpretation of the findings: 1. The results of the study indicate that the interview was not a valid device for predicting the future performance of trooper applicants in the training school and as a probationary officer. These results are not unexpected in view of the validation criteria used-~the training school grades which are based on factors other than those considered in making the interview evaluations, and the probationary trOOper ratings which are subject to two flaws of possible major consequence: (l) A large amount of rater bias affecting the ratings on an unequal basis, and (2) a strong pos- sibility of criterion contamination in the form of factor weighting which was done on a purely arbitrary basis. The writer makes the following suggestions which would, in his Opinion, tend to improve the selection of - 86 - trooper applicants and make possible more meaningful future studies of the type reported here: 1. Continuity of membership on the interview boards should be stressed. If at all prac- tical, based on the number of applicants, one three-member interview board should in- terview all applicants. If several boards must be used, the membership of each should remain the same throughout the interviews. Interviewers should be briefed more thorough- ly, and added emphasis should be placed on evaluating applicants on the basis of the requirements for successful performance as a state police officer on field assignment rather than on the requirements for success- ful completion of the state police trOOper training school. The writer feels that since the qualification of applicants for appoint- ment to the training school is the immediate goal of the interview, some board members over-emphasize this and in so doing fail to evaluate in terms of long-range performance potential. For instance, the training school situation requires the ability to withstand strict supervision and regimentation, whereas in a field assignment as a plain-clothes of- ficer, a man works alone much of the time. ?. V H": I l ’ I".' I . ’ II ' I III 'I g A I- I " I I1 A I t . II' '| [I I' f I" ' I .I I‘ll ‘s'l I" I. I" r' ' v" I ‘) I ‘ 7". yo ' ”hi (' ' I: 'I J , . . I‘ll H1 ' k l'. ‘0‘", I I I' I I9“). I . [1. I I I‘ “'l . ) ‘l. VI 'I‘ ‘ I " E! I ‘I 'I '| 3". I 0 I. "’1". " . ‘. ' _ _ I. I' II“ A. I- . . . I I O . - )1: x P'- - 1“ ' '_I~ ‘ . . Q I v a ’_ —‘ A f’, !I‘1V .- I ‘- ' j! - ‘ , .‘ IV - u {x . ‘ . 1 - I. I I l r I fi I—- " h 3-7' I ‘ - _,. . .—..,, .7 - 4. K). II; J > I A . 'I r I ' "-. u.‘_. f I 0' _ o -' I ' I .1 ‘0‘. _ ,I‘ ‘-‘O. - r..- r'. as I . a' F25 .Ce . ("‘I 8L- ,_ -. '- - 87 - He determines his own work schedule and is usually subject to limited supervision only. An applicant might fit very well in one of these situations but have serious shortcom- ings concerning the other assignment. Wherever possible, the officer conducting the field investigation on an applicant should obtain and include in his report ap- praisals by the applicant's previous employ- ers in the Specific areas of: Dependability, judgment, work quality, attitude, and initi- ative. This would give the interview board members additional information concerning important aspects of the applicant's past performance. The writer feels that board members could safely place some reliance on such comments. Experience has shown that, in general, Michigan citizens are proud of their state police and are rather candid in their appraisal of applicants for the de- partment. The length of time for conducting the inter- views should be increased to a minimum of 45 minutes per applicant. This would allow more time for all parts of the interview-~especi- ally for reviewing the applicant's file. In some instances, in fairness to the applicant, 7. _ 88 - the actual interview could well be longer so as to give board members additional time to resolve certain points upon which they were undecided. Both the interview rating form and the pro- bationary rating form should be revised so as to provide more similar comparison in terms of factors rated and standards of adequacy. The latter form should also in- clude a summary rating scale comparable to that found in the interview rating form. (The latest revision of the interview rating scale provides for only seven "passing" scores, at intervals of five points from 70 through 100). The agility test should be revised, or a new screening device adOpted, to test more thor- oughly the over-all physical fitness of the applicants. This factor should be emphasized also at the final medical-physical examination given each applicant when he reports to the training school. Further analysis should be made as to the factors of adequate performance as a proba- tionary officer, especially concerning the interpretation of the standards of adequacy. This information Should be made available _ 89 _ to all senior officers who supervise and submit ratings on probationary officers. If feasible, this information should be given in the form of instruction at an in-service training session. SELECTIVE BIBLIOGRAPHY Articles Bayroff, A. G., Haggerty, Helen R., and Rundquist, E. A. "Validity of Ratings as Related to Rating Tech- niques and Conditions," Personnel Psychology, VII, No. 1 (1957). pp. 934113. Bingham, Walter Van Dyke. "Halo, Invalid and Valid," Journal of Applied Psychology, XXIII (1939), pp. 221-228. Brogden, Hubert E. and Taylor, Erwin K. "The Theory and Classification of Criterion Bias," Educational and Psychological Measurement, X (1950), pp. 159-186. Campbell, Joel T., Prien, Erich P., and Brailey, Lester G. ' "Predicting Performance Evaluations," Personnel Psychology, XIII (1960), pp. 435-440. Fearing, F., and Fearing, F. M. "Factors in the Appraisal Interview Considered with Particular Reference to the Selection of Public Personnel," Journal of Psychology, XIV (1942), pp. 131-153. Jones, Margaret Hubbard. "The Adequacy of Employee Selection Reports}' Journal of Applied Psychology, XXXIV, No. 4 (August, 1950), pp. 219-224. Mandell, Milton. "Civil Service Oral Interviews," Personnel Journal, XVIII (1940), pp. 373-382. McMurry, Robert N. "How Efficient are Your Hirin Methods?" Personnel Journal, XXVI, No. 2 (June, 1947 , pp. 45-53. McMurry, Robert N. "Validating the Patterned Interview," Personnel’ XXIII, N00 4 (January, 19”?): pp. 263~272. UthPOCk: Richard 3. "AnaIYSis of Employment Interviews," Personnel Journal, XII, No. 2 (1934), pp. 98-101. Uhrbrock, Richard S. "The Personnel Interview," Personnel Psychology, I, No. 3 (Autumn, 1948), pp. 273-302. - 90 - Wagner, Ralph. "The Employment Interview: A Critical Summary," Personnel Psychology, II (Spring, 1949), pp 0 17")4’60 Yonge, K. A. "The Value of the Interview: An Orientation and Pilot Study," The Journal of Applied Psychology, XL, No. 1 (1956), pp. 25-34. Books and Reports Bingham, Walter Van Dyke, Moore, Bruce Victor, and Gustad, John W. (collaborator). How to Interview. 4th ed. revised. New York: Harper and Brothers, 1959. Fear, Richard A. Employee Evaluation Manual for Interview- ers. New York: The Psychological Corp., 1943. Fear, Richard A. The Evaluation Interview. New York: McGraw-Hill, 1958. Kahn, Robert Louis, and Cannell, Charles F. The Dynamics of Interviewing. New York: Wiley and Sons, Inc., 1957. Kelly, E., and Fiske, D. The Prediction of Performance in Clinical Psychology. Ann Arbor: University of Michigan Press, 1951. Maier, Norman R. F. The Appraisal Interview. New York: Wiley and Sons, Inc., 1958. Meehl, P. Clinical vs. Statistical Prediction. Minneapolis: University of Minnesota Press, 1954. 0rdway, Samual H., Jr., et a1. Oral Tests in Public -—-————— Personnel Selection. Chicago: Civil Service Assembly of the United States and Canada,l943. Thorndike, Robert L. Personnel Selection. New York: John Wiley and Sons, Inc., 1949. - 91 - 1!'|I|l| APPENDIX -92.. State of “£1118“ THIS REPORT TO BE C5-120 Rem” STATE POLICE TROOPER CIVIL SERVICE comssmN "DE 1" DUPLICATE FIELD INVESTIGATION REPORT ON APPLICANT Name Address NOTE: UNDERLINE ANY WORD OR WORDS WHICH BEST DESCRIBE THE APPLICANT. IF NONE IS APPLICABLE. INSERT APPROPRIATE DESCRIPTIVE TERIS. IN ADDITION, BELOW EACH CATEGORY '- PERSONAL APPEARANCE. ORAL INTERVIEW AND GENERAL IIPRESSION --. A SPACE IS PROVIDED FOR A GENERAL RATING. THIS RATING SHOULD BE GIVEN AS EXCELLDNT. GOOD. FAIR OR POOR. PERSONAL APPEARANCE: DR-S: Conservative. ordinary. collegiate. flashy. rural 8) ta) FEATURES: Refined. ordinary. coarse. dissipated c) NEATNESS: Nell-groomed. neat. untidy. dirty d) BUILD: Athletic. Iediul. stocky. slender. frail. rat Te) SKIN CONDITION: Healthy. nor-a1. blemished (specify) f) STATURE: Erect. stooped. round shouldered. other (specify) 3) CLEANLINESS: Hands. fingernails. skin. teeth (underline if satisfactory) RATING: ORAL INTERVIEW: a) APPROACH: Friendly. quiet, hesitant. uninpressive l3) HANDSHAKE: Extreme, firs. average. weak 0) POISE: Well-poised. lacking d) VOICE: lell-nodulated, clear. low. too low. loud. harsh. nasal, high-pitched £2) ASSURANCE: Self-confident. average. cocky. tinid .f) NERVOUSNESS: None. slight. very nervous g) ACCENT: None. foreign. regional. slight. very noticeable h) TACT: Tactful. average. blunt. lacking. crude 1,) ENTHUSIASI: Enthusiastic. average. undemonstrative. indifferent J ) FORCE: Porceful. sufficient. self-conceited. lacks initiative 1;) A-iABILITY: Cooperative. self-centered. stubborn. resentful :1) ALERTNESS: Alert. responsive. lackadaisical. dull In) IATURITY: Nature. responsible. immature. irresponsible 11) ANSWERS QUESTIONS: Definitely. inaccurately, vaguely. evasively. slowly. quickly RATING: C5-120 state of Michigan. Rev.756 PAGE 2 STATE POLICE TROOPER CIVIL SERVICE COMMISSION Name Address The following questions are to be answered completely on succeeding pages to be inserted. using as many as necessary. and in answering refer to each question by its number and title. (In Duplicate) L 2. I. 1. ARREST RECORD CREIIT RECORD EDUCATION RECORD EMPLOYMENT RECORD HEALTH RECORD IARITAL STATUS IILITARY RECORD RECREATION HOIE CONDITIONS Any arrest record. including traffic offenses. Also any arrest record of any immediate member of the family exclusive of traffic offenses. To include present financial status and current financial obligations. Also past reputation for incurring and settling indebtedness. To include high school. college or similar educational records. exclusive of elementary school. in regard o attendance, conduct. average grades and general characteristics as a student. Also any special school activities. such as athletics. debating. class leadership etc. To include findings from past and present employers regarding work habits and reasons for leaving any former positions. To include any findings of illnesses or injuries which might interfere with effec- tive performance of police duties. Also any medical history of immediate family which might indicate hereditary tendencies. Likewise any contagious. infectious chronic or other disease or major illness with which any member of family may now be afflicted and the extent to which the applicant has been exposed ot it. To include number of children. their ages. if a child is expected. and other _persons dependent on applicant for support. If a veteran. to include type of discharge. any service disability. and if subjected to any disciplinary action. Also any present draft or reserve status. To include any particular hobbies or pastimes. To include neighborhood. dwellings. applicant's position in dwelling. condition of home. number of occupants. etc. AN! ADDITIONAL INFORMATION - To include interviews with neighbors and business or personal associates. INVESTIGATING OFFICERS COIIENTS NVESTIGATING OFFICER DATE (Signature) AUVK POST DISTRICT COHIANDER'S COMMENTS TBICT COINANDER DATE (Signature) K DISTRI CT _— umhhhttgo I. am<_m==»zm muzmo_azou oumeoozm to: _ :»_n oumceeza _ =h_. oumaoezu _ cameoozu _ egos m.z» «on eme_=mz= _ . zo_e_moa Lea eo Wfi....q.d.-.J Tfi.3q14—1fi4.....- 1 n q _ _ fi _ mmmzhl 4532;“ woe woo wee ace eon ace mom mom uOH mo flOOH umhhhzzcu .zuaOa Ina c_ o.um..ooa vco .mco_c_ao .acunaa Co oc__oow .mmoc__vcowawca «cocoa .ou«m_coauu «mococ.o_naooaoo aao> o.naao>am .o>_uoaoaoou .aaocwo .oocoaawemuc_ an ..:mucouoa «ugaoeom ace ._:*«comoa .co._:m mo:e_pp« 4_ncowam ao>o I_wwwu .v_ou .wcmumwo mmm2_4azm.ma .o uhhhhzzdo .mEmmaoc .omoo «a >_owo.a600 .s_ao anomoaeoo __o: .omoo «a >_ucoaoqa< .omcov .ommo um __. acme mao>aoc _nao>om >h_4.m._eama:: Iauo_o anon. mucomoaa acomumoac massac< o.uu». »Lo> maoeuca.o> _mowmo__w .eomachu mmumaxm ow >»_L_m< .2 Amazhtzso .uoco_u .co_xo_aeoo coca .oocoaeoa .omoao IncaOLS __oa ace—coon .mc_xoo. ensues .oaau one manage .oaaamoa .mc_aeoaaa xoos no .oaauaeew »ao> no «on». .co_xo_aeoo Inca wooao .ac.aooaan coco .moaauao» _m_oaw no .oaaumoa azoao_m no .oaaanoa Loon so .con aeo_o I oc_ucuumu:o any—co: I o>_mmoaae_ ooaao>u : acne—vac oa:«n££...o>_mmoaaenc= .*o_aaoo uan..aooa aao> uuz<¢:a .m "whhhztob .mc_£ooao oao.:o~£5_ .uoSOOLm .c~o_o can «no: .oc_ .uommoau »_oua_aa0aaao .oocmaaoaaa «use: .oounoan ..os xao> ..os.wozwo.o neon—aam uo.o mason» so .asmau 0:” .uoeooLo >.aooa 1:: .oc-zvo_o aua_o oz_xo°¢o oz< wz_:h040 .N uhhfihticb .oo_o> oc_xao m .nuoow .uoowou cocoon u._£ .aouuaua coco .oouu_:uos ..o3 .ucamao.a vce auo_o Ian 0: .o_aovcuumsouc= no .ucoooa cOEEooc: no .am"_ .cmaa; ago) :ouwam oz< mu_o> .H mmpmmezH game uh