THE GENERALITY OF RATING LEVEL Them ‘09 the Degree 0‘ M. A. MICHIGAN STATE UNIVERSITY Elliot R. Siegel 1966 unusuéan onus, University I : ABSTRACT THE GENERALITY OF RATING LEVEL by Elliot R. Siegel The purpose of the present study was to determine the generality of rating level across different rating orientations of a judge, and within orientations across heterogeneous others. A perceiver's level was defined as his tendency to make favorable or unfavorable ratings. The study was divided into two phases. The first phase consisted of a test questionnaire, in which a subject rated an inter- viewee on the basis of six rating orientations. The hypothesis tested was that when rating an other on dimensions of desirability—undesirability, a judge's level tends to be general across different orientations. The findings partially supported the hypothesis in that clear-cut generality of level was demonstrated between the following intrajudge rating orientations: 1. Self and First Person Orientations 2. Personal and Self Orientations 3. Personal and First Person Orientations 4. Personal and Third Person Orientations Elliot R . Sie gel No significant relationship was found between ten other possible orien- tation pairs. The second phase consisted of a test questionnaire, in which a subject rated three interviewees on the basis of two orientations. The hypothesis tested was that the level of a judge's ratings across heterogeneous others, tends to be general within orientations. The findings partially supported the hypothesis in that generality of level was demonstrated within Third Person Orientation ratings. A self-image projection model was developed in an attempt to explain these findings. The model posits that a judge's self-image forms the basis for the generality of level across certain intrajudge rating orientations, and within certain orientations across hetero- geneous others. The projected self-image can be thought of as the "nucleus" which "binds" a judge's ratings, thereby resulting in level generality. Projection is said to occur in response to the limited information presented to the judge. It was concluded that generality of level across intrajudge rating orientations should occur only in those rating orientations pairs having the projected self-image as a common element. Similarly, generality of rating level across heterogeneous others should occur only withinthose rating orientations in which a judge's self-image is projected. Approved:\ LIL xxx“ (KLK ,gvjik\ Committ e Chair an Date:i\;uu H \Ol 4‘31““) ( THE GENERALITY OF RATING LEVEL BY Elliot R. Siegel A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF ARTS Department of Psychology 1966 ACKNOWLEDGMENTS The author wishes to express his sincere appreciation to Dr. Henry Clay Smith, whose ideas and guidance have been instru- mental in the conception and completion of this thesis. I would also like to thank Drs. Donald M Johnson, and John E. Hunter for their valuable suggestions and assistance. ii TABLE OF CONTENTS Page INTRODUCTION. . . . . . . . . . . . . . , . . . . . . . 1 HISTORY 2 Generality Across Different Orientations . . . . . . . . 2 Projection of Traits . . . . . . . . . . . . . . . . . l4 PROBLEM......................... 16 METHOD l9 Studyl.......................19 Subjects . . . . . . . . . . . . . . . . p 19 The Case of Harry George . . . . . . . . . . . . . . 19 General Procedures . . . . . . . . . . . . . . . . . 20 MethodofDataAnalysis 22 StudyZ....................... 25 Subjects . . . . . . . . . . . . . . . . . . . . . 25 The GAWL Test . . . . . . . . . . . . . . . . . . . 25 General Procedures . . . . . . . . . . . . . . . . . 26 MethodofDataAnalysis 26 RESULTS........................ 28 Study 1 . . . . . . . . . . . . . . . . . . . 28 Generality of Level Across Orientations . . . . . . . . 28 Study 2 . . . . . . . . . . . . . . 34 Generality of Level Across Orientations . . . . . . 35 Generality of Level Within Orientations Across Others . . 36 DISCUSSION . . . . . . . . . . . . . . . . . . . . . . . 39 SUMMARY AND CONCLUSIONS . . . . . . . . . . . . . . 47 REFERENCES . . . . . . . . . . . . . . . . . . . . . . 50 APPENDICES...................... 53 iii Table LIST OF TAB LES Reliability of rating orientations . . . . . . . . . . Reliability of rating orientations . . . . . . Correlations of rating level between orientations Significant correlations with self-image as a common element 0 O O O O O O O O O O O O O p O O 0 Correlations of rating level between second person and third person orientation pairs . . . . . Correlations of second person orientation rating levelsacrossothers . . . . . . . . . . . . . . Correlations of third person orientation rating levels acrossothers................ Means and standard deviations of study 1 rating scales Means and standard deviations of study 2 ratins scales iv Page 25 27 29 3O 35 37 38 69 71 INTRODUCTION The current literature concerned with the problem of inter- personal sensitivity (the ability to understand individuals and to make accurate predictions about their behavior and experience) , has shown that we tend to perceive a person as a unified whole. The wholes we perceive, however, seem to be heavily influenced by differences in our levels. A perceiver's "level" is his tendency to make favorable or unfavorable ratings. The generality of this levelling tendency is a key element in determining its importance. The aim of the present research was to determine the generality of this levelling tendency across different rating orientations of a judge and within orientations acorss heterogeneous others. HISTORY Casual observation reveals many examples of the operation of level. One instructor sees a student as mediocre, while another sees him as outstanding. Further, the first instructor may find few students that are commendable, while another finds many. One sales manager searches intensely for qualified salesmen but finds few, while another manager finds many; and one conscientious professor sees few instructors that are qualified for promotion, while another, equally conscientious, finds many. The influence of level on accuracy is dependent upon its generality. If a perceiver's level on one trait is unrelated to his level on another, its overall influence on his own predictions is small; if closely related, its influence is great. Therefore, by generality, we do not mean that most raters tend to have a high level or tend to rate others low. Rather, we are here referring to the consistency of a perceiver'slevel. The higher the generality, the greater the consis- tency of rating levels. Generality Across Different Orientations Typically, most of our practical rating situations are based upon a Personal Orientation, i. e. , How do I think he rates ? This is the set assumed by a supervisor evaluating an employee, or 2 by a student evaluating a new roommate. However, most researchers attempting to uncover the dynamics of interpersonal sensitivity fre- quently utilize several orientations in rating the same trait. Six pos- sible orientations can be distinguished: 1. Self Orientation: How do I rate myself? 2. First Person Orientation: How do Ithink he rates me? 3. Second Person Orientation: How do Ithink he rates himself? 4. Third Person Orientation: How do I think others rate him? 5. Personal Orientation: How do I think he rates ? 6. Behavioral Orientation: How do I think he behaves ? The generality of level can be explored within each of these orientations, as well as between them. As some investigators frequently utilize one or more orienta- tions in rating several ratees, it is difficult to classify studies as dealing either with generality across intrajudge orientations, or generality within orientations across heterogeneous others. Therefore, we will group these studies under the appropriate orientation headings, and specify when they are concerned with generality within orientations as well. Self Orientation Widely varying efforts to outline the basic dimensions of per— I sonality, regularly reveal a self-evaluation factor: "self-confidence, ' "self-acceptance, ” "self-image, " etc. Differences in the level of self evaluations appear to be one of the most central and stable as- pects of personality (Smith, 1966) . Kelly (1955) found that scores in self-confidence after a twenty year period had a correlation of . 61 higher than for any other trait. Per sonal Orientation Gross ( 1961) had 62 judges rate half-minute films of 30 different men and women on the basis of a Personal Orientation. The scales were based on seven traits: general activity, rhathymia, ascendance, sociability, emotional stability, masculinity-feminity, and friendliness. The author found that a judge's rating level tended to be general across traits, and within these traits across the ratees. The generality of personal levels can seriously affect employee attitude studies based upon rating scales. Gottleib and Kerr (1959) had workers at a large Mid—Western company rate the favorableness of their attitude toward the company and the favorableness of their attitude toward the union. The results indicated that workers who were most favorable toward the union tended to be mo st favorable toward management. Similarly, workers most satisfied with their union steward also tended to be most satisfied with their supervisor. Total scores on management and union scales correlated . 74. The authors con- cluded that the high emotional acceptance and cooperative attitudes between the management and union resulted in a positive integration of employee attitudes. However, a second interpretation is possible. The authors failed to consider that their results may have been in- fluenced by the generality of level. That is, the more favorable a worker's attitude is to his company, the more favorable his attitude toward everything else is likely to be -- union, wife, children, church, car, etc. (Smith, 1966) . Stagner ( 1954) correlated average employee attitudes to management and average employee attitudes to the union, be- tween eight different companies; thereby, effectively eliminating the influence of individual differences in rating level. Unlike the results reported by Gottleib and Kerr, Stagner found only a slight relationship between these attitudes. Second Per son Orientation Gage, Leavitt, and Stone (1956) were interested in the relationship between the sensitivity of teachers and their teaching effectiveness. Filmed interviews of eight fifth graders were shown to judges who then predicted each child's yes-no responses to one hundred questions on a personality inventory. The authors found that the generality of predictive accuracy was very low. However, accuracy was found to be determined, not by interpersonal sensitivity, but by the chance interaction of a judge's rating level with the rating level of a particular child. For example, a high levelling teacher would accurately predict the responses of a high levelling child, thus suggesting under- standing of that child. In a later study by Gage (1958), he attempted to measure Second Person sensitivity, this time based upon the attitudes of a teacher's own pupils toward the teacher, curri- culum, school plant, classroom activities, etc. Once again, however, sensitivity was found to be a chance interaction be- tween the teacher's rating level, and the average level of the teacher's class. Similarly, Crow ( 1954) has demonstrated that the generality of a judge's level over others can account for the generality over others of predictive accuracy. That is, in Crow's data, accuracy seemed to be general over others to the degree that those others were homogeneous, and the judges had levels that were general over the others. In a later study, Crow and Hammond ( 1957) found more support for the generality of Second Person levels. They showed 72 medical students a total of thirty-six six-minute filmed interviews between a patient and a doctor. The films were shown at three six month intervals, ten films at each test ad- ministration. The students saw different doctors and different patients on each occasion. The judges were required to pre- dict the self—ratings of the patients on seven personality scales. The scores on each administration were intercorrelated with the scores on the other two administrations. The results in- dicated significant but low relationships among predictive accuracy scores over time. However, generality of Second Person level scores over this extended time period was very high. The highest generality among the accuracy scores was less than the lowest generality among the level scores. In view of these findings, the authors concluded by emphasizing the im- portance of a perceiver's level: . . the reliabilities reported for accuracy of inter- personal perception should be seriously questioned because the consistency of the scores may be due to consistency in response set rather than to consistency in ability. Since investigators have rarely eliminated response set components from their accuracy scores, it is maintained that differential accuracy of inter- personal perception as a reliable characteristic of per- formance is yet to be established. (Crow and Hammond, 1957) Self and Personal Orientations Of the fifteen possible combinations of rating orien- tations, the relation between Self Orientation ("self-acceptance") and Personal Orientation ("other-acceptance") has been most fully studied (Smith, 1966). Phillips ( 1951) developed an attitude scale of fifty items, with twenty-five referring to self-attitudes and twenty- five referring to attitudes toward others. Scores for accep- tance of self and acceptance of others were obtained. Correla- tions between the two scales of . 61 and . 74 were found, when the questionnaire was administered to high school and college students. The author's results support Roger's postulate of a positive relationship between attitudes toward self and others. This study was replicated by McIntyre (1952) who found a correlation of .46 between the acceptance of self and acceptance of others scale. However, he failed to demon- strate that better attitudes toward self and others are positively related to acceptance by others. Self, Personal, and First Person Orientations Similarly, Fey (1955) measured the level of self ratings and acceptance of others for 58 third year medical students, and found a correlation of . 50. Individuals with high self—acceptance levels tended also to accept others, to feel accepted by others (e. g. , First Person Orientation), but actually to be neither more nor less accepted by others than those with low self-acceptance levels. While, individuals with high other -acceptance levels tended to feel accepted by others, and were being accepted by them. Therefore, it appears that self—levels were related to levels in rating others, which in turn were both related to feelings of acceptance by others. However, actual acceptance by others tended to be related only to acceptance of others when acceptance of others was significantly different from acceptance of self. Self, Personal, First Person, and Second Person Orientations Dymond (1949, 1950) had, as the basis of her re- search, two primary objectives. The first was to develop a technique to be used in the assessment of interpersonal sensi- tivity. Her second objective was to then study the personality differences between the generally sensitive and the insensitive. To do so, she divided 80 subjects into twenty groups of four members each, with each subject a member of three different 10 groups. The subjects were then required to rate themselves and the others ona 5-point scale composed of six traits: superior-inferior, friendly-unfriendly, leader-follower, shy—self—assured, sympathetic- unsyrnpathetic, and secure-insecure. These ratings were made on the basis of four different orientations: 1. a rating of himself (Self Orientation), 2. a rating of each of the other three (Personal Orientation), 3. a rating of how he thought the other three would rate him (First Person Orientation), 4. a rating of how he thought each of the other three would rate themselves (Second Person Orientation) . The author reports a correlation of . 63 between Self and First Person Orientations. The six individuals achieving the highest sensitivity scores, and the seven individuals scoring the lowest) were then given a battery of four personality tests (WAIS, Murry ¢ Morgan TAT, Rorschach, and the California Ethnocentrism Test). We will re- turn to her personality findings in a later section. However, the results of later investigations suggest that the personality differences she found were largely due to differences, not in general sensitivity, but in level (Smith, 1966) . Lindgren and Robinson ( 1953) simplified Dymond's procedure, and tested the hypothesis that an individual's predictive accuracy was not a project of his sensitivity, but a product of his 11 level. The social norm for each trait was determined, and the re- sponses of 100 subjects were then keyed in terms of their deviation form these norms. When these scores were compared with the scores obtained by keying the responses according to Dymond's method, the resulting correlation was . 56. The authors concluded that Dymond's test of interpersonal sensitivity was, in fact, largely a measure of the tendency of an individual to respond on the basis of social norms. Furthermore, Smith (1966) states that it seems almost certain that Dymond's measure of sensitivity was a measure of level accuracy. Since the average student gave himself a high rating on these desirable traits, the measure of level accuracy was, therefore, primarily a measure of the tendency of a student to make high ratings of himself and others on desirable traits. Further support for the hypothesis that subjects tend to give what are considered to be socially desirable responses was found by Edwards (1953) . He had 152. subjects judge the desirability of 140 personality trait items on a 4-point scale. Another group of 140 sub- jects responded to these items in the form of a personality inventory. The results clearly indicated that the probability of endorsement of an item increases with the judged desirability of the item (r = . 87). Personality Differences BetWeen Low and High Levellers In view of these findings, it seems quite reason- able to interpret the stated results of Dymond's p e r s onalit y 12 test battery, as indicative of the personality differences between low and high levellers. According to Dymond, the low levellers (insensitive) were: rather rigid, introverted people . . . who prefer to get along without strong ties to other people. Thier own early emotional relationships within the family seem to have been so disturbed and unsatisfactory that they feel they cannot afford to invest their love in others as they need it all for themselves. They seem to mistrust others . . . (Dymond, 1950) The high levellers (sensitive), on the other hand, were: outgoing, optimistic, warm, emotional people whose emotional relations have been sufficiently satisfying so that they find in- vesting emotionally in others rewarding. (Dymond, 1950) Her findings are consistent with those of Grossman (1963) . Similarly, Guilford ( 1959) found low levellers to be generally less confident and more conservative than high levellers. We have thus far shown how an individual's level, when unaccounted for, can seriously influence the validity of an investigation which has been based upon the use of rating scales. A perceiver's level is a central and stable quality of his personality and has a per- vasive influence upon his judgments of others (Smith, 1966) . This influence, as we have seen, is not limited to any single rating orien- tation, but tends to generalize both within and acorss one or more orientations. As Smith points out, a comprehensive study of all areas within and between which rating levels might generalize is needed. 13 Projection of Traits There is some evidence which seems to indicate that the mechanism of projection may operate in the generalization of an indi- vidual's rating level. Landis ( 1936) found that tall raters tended to overrate the height of others; fat raters tended to overrate the weight of others; and emotionally unstable raters tended to overrate the in- stability in others. It has also been found that individuals high on the F Scale, tend to rate others high in authoritarianism as well (Crockett and Meidinger, 1956; Scodel and Freedman, 1953) . Similarly, Weingarten ( 1949) demonstrated that judges sometimes project their own insecurities into their clinical evaluations of others. The positive relationship between autobiographical ratings and test scores is taken as evidence of projection. The author defines projection as a process by which an individual selects and organizes his perceptions in accordance with his own life experiences, and thereby imposes upon the behavior of others an interpretation which reflects the feelings and emotions which he experiences himself. A Judge's Self-imag Tannenbaum, Weschler and Massarik ( 1961) state that one's self-image provides a kind of psychological "base of operations" that inevitably affects relations with family, friends, business associ- ates, and strangers. Furthermore, one's self-image is by no means 14 clear-cut. Some aspects are at the surface of personality, these are the publically held attitudes--the things we do not mind telling other people about ourselves and our views of the world. There are some feelings about the self of which we are aware, but which we do not want to share with others--these are the privately held attitudes to the self. Burried still deeper are the subconscious and unconscious aspects-~feelings about "who" we are and "what" we are that somehow we cannot fact up to, even to ourselves. The degree of a rater's ego control has a strong effect upon his ratings. Couch and Keniston (1960) administered a large battery of objective personality tests to 61 subjects in a series of four sessions. The subjects were required to rate themselves on a 7-point scale for a total of 360 items. The scales were constructed so as to control for the effect of content variables. An agreeing- disagreeing tendency was found, which was also highly related to "true-saying" and "yes-saying. " The measures showed considerable internal reliability, consistency over time, and generality over tests. A cluster of scales characterized the positive end of the agreeing tendency: impulsivity, dependency, anxiety, mania, anal preoccupation, and anal resentment. The opposite end, the disagreeing tendency, was defined by such scales as ego strength, stability, responsibility, tolerance and impulse control. From the original group of subjects, the 10 with the highest agreeing scores (yeasayers) and the 11 with 15 the lowest scores (naysayers) were then given a clinical interview which was structured around a 55 item sentence completion test. According to the authors, yeasayers were: . individuals with weak ego controls, who accept impulses without reservation, and who "agree" and easily respond to stimuli exerted on them. The passive and receptive egos of yeasayers seem to be an outgrowth of their lack of internali- zation of parental control. While maintaining the need for external control, yeasayers can freely indulge in impulse gratification. (Couch and Keniston, 1960) . While the naysayer: . inhibits and suppresses his impulses, rejecting all emotional stimuli impinging on him. The egos of naysayers take over the controlling functions of the parents, and the suppression of impulses is subsequently self-maintained and self-rewarded. (Couch and Keniston, 1960) PROBLEM The aim of the present research was to determine the generality of this levelling tendency across different rating orientations of a judge and within orientations across heterogeneous others. The hypotheses tested were: 1. When rating an other on dimensions of desirability- undesirability, a judge's level tends to be general across differ ent orientations . Z. The level of a judge's ratings across heterogeneous others tends to be general within orientations. A great deal of consideration was also given to a projec- tive process which may be operative in the generalization of an indi- vidual's rating level. In this conceptualization, a judge's projected self-image is thought of as the "nucleus" which "binds" his ratings across orientations; thereby, resulting in the generality of rating level. For example, generality of level has been shown across Self, First Person, and Personal Orientations. This relationship can be represented schematically (see Figure 1). In response to the . limited information presented to a judge, a judge's self-image is said to be projected, that is, to form a basis for ratings from each of these orientations. A judge's self-image, as a common element in each of these rating orientations, is, therefore, said to account 16 17 Key: Self Orientation ¢ First Person Orientation Personal Orientation ¢ Self Orientation , Personal Orientation ¢ First Person Orientation Self-image //////// Figure 1. A judge's self-image as the basis for generality of level. 18 \ for the generality of rating level which is found across these three rating orientations. The greater the degree of self-image projection, the greater the generality of rating level. METHOD The investigation was divided into two phases. .5391} The first phase was concerned with the problem of generality of rating level across six rating orientations. Subjects Fifty-one students (N = 51) enrolled in an Introductory Psychology class at Michigan State University served as subjects. The Case of Harry George The test of rating level used in this study is made up largely of materials developed by Cline and Rivhards ( 1960) in their study of interpersonal perception. The materials were modified somewhat, however, to fit our purposes. (See Appendix A.) The test directions, in part, are as follows: This is a test of your ability to make accurate predictions about people. It consists of six parts, and each part asks you to make predictions from a different point of view. Your task is to make accurate predictions about Mr. George [the interviewee]. Mr. George is a middle-aged married man with one child. As part of a research project on understanding people, he was given a brief interview. A typescript is given below:" 19 2.0 A typescript of an actual five minute filmed interview was presented to the subjects. The interviewee is asked questions con- cerning the following areas: personal values, personality strengths and weaknesses, hobbies and activities, self-conception, and temper. General Procedures The six rating orientations that the subjects were asked to assume in making their predictions were presented in the following order: Part I Second Person Orientation: How do you think he rates himself ? Part 11, Third Person Orientation: How do you think others rate him? Part III Behavioral Orientation: How do you think he behaves ? Part IV Personal Orientation: How do you think he rates ? Part V First Person Orientation: How do you think he would rate you? Part VI Self Orientation: How do you rate yourself ? The subjects were not given the actual names of each orientation, only the set they were to assume in making predictions on each part. Their predictions were not scored for accuracy, but keyed to reflect the level of ratings in each part. We will later dis- cuss the actual scoring procedure in detail. The items making-up each of the six parts, a total of 65 items in all, are essentially of two types: (a) open-end statements, Whose three alternatives range from (1) undesirable to (3) desirable; 21 and (b) adjective traits, six desirable and six undesirable: cooperative, confident, friendly, affectionate, careful, ambitious, stubborn, ego- tistical, shy, rebellious, impractical, and unrealistic. The adjectives are rated on a 5-point scale ranging from ( 1) "Rather UNlike" or "Very UNlike. . ." to (5) "Very like . . ." For example, the directions for Part I (Second Person Orientation) are as follows: Mr. George has checked one alternative on each of the statements below to describe himself. You are to check the alternative you think he checked. Item: When they offered me help I . (l) was somewhat embarassed (2) thanked them but refused (3) accepted The directions for Part 11 (Third Person Orientation) are as follows: Mr. George was rated on each of the traits below by persons who knew him well. They used the following rating system. (1) "Rather UNlike" or "Very UNlike him" (2) "A little UNlike him" (3) "A little like him" (4) "Rather like him" (5) "Very like him" Rate Mr. George as you think he was ratediby others. Item: FRIENDLY Parts I and III consist of a combined total of seventeen open-end statements; while Parts 11, IV, V, and VI consist of the list of twelve adjectives (i. e. , the same adjectives are used in each part). 22 The degree of desirability of both the statement alternatives and adjectives in largely socially determined. To determine desir- ability, all items were first presented to three judges (graduate students), whose consensus on desirability was the criterion used for classifying each item. However, despite the fact that a "panel" of three judges was used to rank the relative desirability of the alterna- tives to the open-end items, it is believed that these items do not adequately represent to the subjects dimensions of desirability- undesirability. The alternatives are, in some cases, ambiguous in terms of social desirability; and therefore, the interpretations they elicit are probably not consistent. It seems unlikely that Second Person Orientation and Behavioral Orientation rating levels were really being measured. Method of Data Analysis The rating level for each subject -judge was determined by constructing separate keys for each scale position of each part. For example, Part I, in which there were three alternatives possible for each item, was scored by using three different keys. Part I-key I scored the response frequency of the first alternative for each item in Part I; while, Part I-key 2 scored the response frequency of the Second alternative for each item in Part I, and so on. 23 In those parts consisting of adjectives, the first six items were desirable traits, and the last six items were classified as un- desirable. Since a judge was required to rate the interviewee on the basis of a 5-point scale, there were, therefore, five different keys used for each part. For example, Part II-key l, scored the frequency in which ( 1) was chosen for the first six adjectives in this part. Similarly, Part II-key 2, scored the the frequency in which (2) was chosen, and so on. Since a response of ( 1) for a desirable trait is indicative of a low rating level, while a response of ( 1) for an un- desirable trait is indicative of a high rating level; the key position of the last six adjectives on four of the five keys was reversed. Thus, Part lI-key l scored the frequency of response ( 1) for the first six items, and scored the frequency of response (5) for the last six items. Therefore, a judge who is an extreme low leveller would rate the interviewee on the trait "COOPERATIVE, " as (1) "Very UNlike him"; and would rate him on "UNREALISTIC" as (5) "Very like him. " Then, Part II-key 1 gives a frequency measure of the extreme low level responses for this part. The same technique was used in three other keys of Part II. It was not necessary to reverse positions in Part 11- key 3, as the middle-most point on the scale is essentially neutral. Thus, each judge's answer sheet was keyed a total of twenty-six times. There are no "right" answers on this test, only the position (level) of response is of interest. 24 Having thus obtained the frequency of response for each alternative position on each part for each subject, the next step was to calculate a judge's level on each part. This was accomplished by finding the average response on each part, for each judge. For a given part, the frequency of response of each alternative position was multiplied by that position number, and then divided by the number of items in that part. By this method, a judge's weighted average for a given part is his rating level for that part. We then have a level measure for each judge on each part. A measure of the generality of level acorss orientations is then obtained by correlating the levels of all judge's on all six parts. Reliability The reliability of each of the six parts was determined by a method very similar to that used in calculating a judge's level on each part. Having obtained the frequency of response for each alterna- tive position on each part for each subject, the frequency of response of each alternative position was multiplied by that position number, and a sum for each part half was calculated. Each sum was then divided by half the number of items in that part. By correlating the two sums of each part for each subject, a split -half reliability coefficient was derived for each of the six parts (see Table l) . (See Appendix C for means and standard deviations.) 25 Table 1. --Reliability of rating orientations. N = 51 Orientation r (corrected by Spearman-Brown) Personal . 51 Third Person . 40 First Person . 37 Second Person . 32 Self . 31 Behavioral . 28 Study 2 The second phase of the investigation was concerned with the problem of generality of rating level within orientations, across heterogeneous other 5 . Subjects The subjects were 177 female students (N = 177) enrolled in a Personality class at Michigan State University. The GAWL Test This test is a modification of the test used in Study 1, with additional materials originally developed by Cline and Richards ( 1960) . (See Appendix B. ) Unlike the first rating test, this test consists of only two rating orientations: 26 Part (A) Second Person Orientation: How do you think he rates himself ? Part (B) Third Person Orientation: How do you think others rate him ? General Procedures The subjects were required to rate not one, but three heterogeneous interviewees on the basis of the two rating orientations. There are, therefore, three sections of the test; each section consisting of the two rating orientations and a typescript of an interview with Mr. George (Section 1), Mr. Walter (Section 2), and Mr. Allen (Section 3) . i The three interviews presented are actual, and the struc- tural content of each is the same as that used in Study 1. The inter- view of Mr. George in Section 1 is the same as that presented in Study 1. The directions are otherwise the same as those used in The Case of Harry George (Study 1) . Method of Data Analysis A judge's level for each of the two parts in each of the three sections was obtained as was done in Study 1. A measure of the generality of level across orientations was derived by correlating the levels of all judges on both parts of each section. 27 Of most importance, however, a measure of the generality of level within orientations across heterogeneous others was derived in this study, by correlating the levels of each judge within each of the two parts, across the three sections. Reliability A split --half reliability measure of each part in each section was calculated by the same method used for Study 1 (See Table 2). (See Appendix D for means and standard deviations. ) Table 2. —-Reliability of rating orientations. N = 177 Orientation r (corrected by Spearman-Brown) Section 1: Mr. George Second Person . 21 Third Person .43 Section 2: Mr. Walter Second Person . 17 Third Person . 68 Section 3: Mr. Allen Second Person . 26 Third Person . 50 RESULTS .5251} Fifty-one judges rated a single other on the basis of six different rating orientations. Level measures were then obtained for each judge on each of the six parts. The purpose was to measure the degree of level generality displayed by these judges across the six orientations . Generality of Level Across Orientations The hypothesis tested was that when rating an other on dimensions of desirability-undesirability, a judge's level tends to be general across different rating orientations. The orientations measured were: Part I Second Person Orientation: How do you think he rates himself? Part 11 Third Person Orientation: How do you think others rate him? Part III Behavioral Orientation: How do you think he behaves ? Part IV Personal Orientation: How do you think he rates ? Part V First Person Orientation: Haw do you think he would rate you? Part VI Self Orientation: How do you rate yourself? 28 29 Table 3. --Correlations of rating level between orientations. Third First Second Orientation Personal Person Person Person Self Behavioral Personal - - . 60*‘1‘ . 42* . 03 . 38* . 22 Third Person -- .13 . 38* . 00 . 07 First Person -- -. 12 . 74’1”? -. 08 Second Person -- -. 12 -. 03 Self -- . 00 Behavioral -— Note: All significance levels are for a two -tailed test. >== r significant at .01 level. ** r significant at .002 level. Coefficients of correlation were calculated between each of the fifteen possible orientation pairs. These correlations are shown in Table 3. The highest correlation (r = . 74) found was between the way a judge rates himself (Self Orientation), and how he thinks an other would rate him (First Person Orientation). The next highest correlation (r = . 60) indicated that there is a very strong relationship between the way in which a judge rates an other (Personal Orientation), and how he thinks a third party would rate this other (Third Person Orientation). The third highest correlation (r = .42) was the signifi- cant relationship between how a judge rates an other (Personal Orienta- tion) , and how he thinks this other would rate him (First Person 30 Orientation). The fourth highest correlation (r = . 38) indicated that a judge shows a definite tendency to rate an other (Personal Orienta- tion) , as he would rate himself (Self Orientation). These four significant relationships appear to have one common element between them-~the judge, more specifically, the judge's self-image. We will examine this more closely, by consider- ing each of the four significant relationships separately (see Table 4); and by outlining the proposed developmental sequence of self-image projection. The actual sequence is dependent upon the order in which the rating orientations are presented to a judge. The following sequence is assumed to represent the implicit relationships between orientations that are brought into the rating situation by a judge. The degree of generality of rating level across orientations is dependent upon the degree of self-image projection that occurs in each rating orientation. Table 4. --Significant correlations with self-image as a common element. Orientation Pairs r Self and First Person . 74 Personal and Third Person . 60 Personal and First Person . 42 Personal and Self . 38 31 Self Orientation and First Person Orientation A judge's self-rating is based upon his self-image; and, as would be expected, the judge tends to feel an other would rate him as he (the judge) would rate himself. The judge projects his self- image onto the other, from which the other is expected to rate him. Self Orientation and Personal Orientation The judge has an expectancy of the other rating him as he would rate himself; the judge then, in turn, rates the other as he would rate himself. ("He sees me as I see myself; therefore, I see him as I see myself. ") Personal Orientation and First Person Orientation This relationship, in effect, ties together the first two. A judge rates an other as he would rate himself, because the other is expected to rate the judge on the same basis as the judge would rate himself. In brief, what has been said thus far is that a judge's self- image forms the basis for the generalization of level across Self Orientation, First Person Orientation, and Personal Orientation ratings. The judge's self-image can be thought of as the "nucleus" which "binds" his ratings across orientations, thereby, resulting in generalization of level across these three orientations . 32 Personal Orientation and Third Person Orientation This relationship, as well, fits in nicely with the model we have proposed. A judge rates an other as he would rate himself. Now, a judge has no real basis, from the limited information given by the typescript on which to predict how a third party will rate the other. However, we have proposed that a judge rates an other as he would rate himself, because the other is expected to rate the judge on the same basis as the judge would rate himself. Therefore, it seems likely that a judge projects his self- image onto the third party, so as to impose a basis for the third party's rating of the other. ("Others see him as I see him. ") Since the basis for rating the other on the part of the judge and on the part of the third party is the same (i. e. , the judge's self-image); then these ratings shouldbe, and are related. The fifth significant correlation poses some difficulty for the model. There appears to be generality of level between Second Person and Third Person Orientations. However, a judge's self- image does not seem to be a common factor in both orientations, since only the third party's rating can be taken as a projection of the judge's self-image. Second Person ratings do not involve the other's evaluation of the judge, nor do they involve the evaluation of the other from the judge's point of view. Rather, the judge is asked to evaluate an other as he thinks the other would evaluate himself. The judge 33 has no "stake" in these ratings. The necessity for consistency is absent, along with the necessity for self-image projection. Therefore, despite the limited information given, the judge need not seek a basis, or frame of reference, from which to structure his ratings. ("How he sees me, how I see him, or how a third party sees him, can differ from the way he sees himself. ") A second inconsistency of the model is that we would ex— pect some generality of rating level between Third Person and Self Orientations, in view of the strong relationship between Third Person and Personal Orientations. If a judge projects his self-image onto a third party so as to impose a basis for the third party's rating of an other; then rating levels between Third Person and Self Orientations should be Similar. All of the remaining correlations between orientations do not show generality of rating level. This is consistent with the pro- posed model, as a common element of "judge self—image" is not present in any of these orientation pairs . The lack of relationship between Behavior Orientation ratings and ratings for any other orientation seems most probably due to the lack of reliability of this measure (see Method). Consider- ing the operational similarity between this orientation and Personal Orientation (for which a great deal of generality was found), and the nature of the Behavioral Orientation items, this explanation seems most reasonable . 34 Study 2 Judges (N = 17?) rated three heterogeneous others on the basis of two different rating orientations. Level generality measures were obtained within orientations across these three heterogeneous others. A second measure of the generality of level across two orien- tations was also derived. The primary concern was to measure the generality of level within orientations so as to test the second hypothesis that the level of a judge's ratings across heterogeneous others tends to be general within orientations. The two orientations measured were: Part (A) Second Person Orientation: How do you think he rates himself? Part (B) Third Person Orientation: How do you think others rate him? We chose these two orientations as it appeared desirable to explore further the relationship between them, in view of the results obtained in Study 1. Before considering the results obtained for level generality within orientations across others, we will fir st examine the relation- ship between orientation pairs in each of the three sections. 35 Generality of Level Across Orientations Coefficients of correlation were calculated between the two orientation paris in each of the three sections. These correla- tions are shown in Table 5. Table 5. --Correlations of rating level between second person and third person orientation pairs. Orientation Pairs r Section 1: Mr. George Second and Third Person . 08 Section 2: Mr. Walter Second and Third Person , 36* Section 3: Mr. Allen Second and Third Person , 30* Note: All significance levels are for a two -tailed test. =3 r significant at .002 level. The relationship between orientation pairs is significant in only two of the three sections. In both cases the correlations are very similar to the relationship found between these orientations in Study 1 (r = . 38) . Surprisingly, however, the typescript used in Section 1, in which no relationship was found between rating pairs, is identical to that used in Study 1. In both studies Mr. George is the other that is rated. 36 These conflicting results are probably due to the methodo- logical weaknesses of the Second Person Orientation scale itself. This scale, as well as the Behavioral Orientation scale, is composed en- tirely of open-end items. As previously stated (see Method), we believe that these items may not adequately represent to our subjects a dimension of desirability-undesirability. The alternatives are, in some cases, ambiguous in terms of social desirability; and therefore, the interpretations they elicit are probably not consistent. It seems un- likely that we were really measuring Second Person rating level; and thus, our results cannot either support nor refute the expectancy of our model for this rating orientation. Generality of Level Within Orientations Across Others The primary concern in Study 2 was to measure the generality of level within rating orientations across heterogeneous others . Second Per son Orientation Measures of Second Person rating levels were calculated for Mr. George (Section 1), Mr. Walter (Section 2), and Mr Allen (Section 3). These rating levels were then correlated, and the re- sults are shown in Table 6. 37 Table 6. --Correlations of second person orientation rating levels across others. Section (1) George (2) Walter ('3) Allen (1) George -- -.10 .11 (2) Walter -.10 -- -,12 (3) Allen. .11 -,12 -- The results do not indicate level generality of Second Person Orientation ratings across heterogeneous others. Of the three possible relationships, one is positive, and two are negative. However, none of these relationships is statistically significant. In terms of the self-image projection model we would not expect generality of rating level within this orientation. Again, however, the nature of the scale items does not permit confirmation of this expectancy. Third Per son Orientation Measures of Third Person rating levels were calculated for Mr. George (Section 1) , Mr Walter (Section 2) , and Mr. Allen (Section 3). These rating levels were then correlated, as was done for the Second Person ratings. The results are shown in Table 7. These results do indicate level generality of Third Person Orientation ratings across heterogeneous others. Of the three pos- sible relationships, all are positive, as well as statistically significant. 38 Thus, it appears that how a judge thinks a third person would rate an other is related as to how he thinks a second and third other would be rated by third parties. Table 7. --Correlations of third person orientation rating levels across others. Section ( 1) George (2) Walter (3) Allen (1) George -- ,14>:< . 2220:“: (2) Walter . 14* —- . 16:10:: (3) Allen . 22=======k , 1620:: .. Note: All significance levels are for a two -tailed test. I; =- r significant at .1 level. ** r significant at .05 level. *** r significant at .01 level. This finding can be explained in terms of the self-image projection model. A judge's self-image is said to be projected onto the third party, so as to impose a basis for the third party's rating of the other. Therefore, the consistency of a judge's projected self- image accounts for the generality of Third Person ratings across heterogeneous other 5 . DISCUSSION The research consisted of two studies designed to measure the generality of level. The first study tested the hypothesis that when rating an other on dimensions of desirability-undesirability, a judge's level tends to be general across different rating orientations. The results offer support for this hypothesis in that they show clear-cut generality of level across: 1. Self Orientation and First Person Orientation 2. Personal Orientation and Self Orientation 3. Personal Orientation and First Person Orientation 4. Personal Orientation and Third Person Orientation Study 2 measured the generality of level within two rating orientations across three heterogeneous others in order to test the hypothesis that the level of a judge's ratings across heterogeneous others tends to be general within orientations. The results offer partial support for this hypothesis. Generality of level within Second Person ratings was not found. However, the questionable validity of the Second Person scale does not permit acceptance of this finding. Generality of level within Third Person Orientation ratings was found. When compared to the 39 40 significant correlations obtained in Study 1, these relationships are rather small. The logic of significance testing holds that these rela- tionships are statistically significant-—but are they worthwhile ? This is a subjective judgment that can be made only by the investigator, and the readers of his paper. This judgment, in the final analysis, is based upon the nature of the variables in question, and the personal biases of the author and his readers. As for these variables, and this author's bias, it is felt that the determination and measurement of any amount of relatively permanent variance inherent in the rating situation is worthwhile. For some rating orientations, the generality of level may be large; for others, quite small; and for still others, nonexistent. Nevertheless, the influence of generality of rating level must be taken into consider- ation by anyone who attempts to effectively utilize the rating method. The results concerning generality of rating level are con- sistent with the literature. The relationship found between Self and Personal Orientations is in keeping with the findings reported by Phillips ( 1951) , and McIntyre ( 1952) . Similarly, the strong relation- ship demonstrated between Self and First Person ratings is consistent with that reported by Dymond ( 1950) . Fey ( 1955) found that Self, Personal, and First Person ratings were related, as was found to be the case in this investigation. 41 We have attempted to explain our findings oflevel generality and lack of generality by means of a projection model. This model asserts that in response to the limited information presented, a judge projects his self-image, so as to impose a logical basis for his ratings. Thus, a judge's projected self-image forms the basis for the generality of level across certain intrajudge rating orientations, and within cer- tain orientations across heterogeneous others. The projected self- image can be thought of as the "nucleus" which "binds" a judge's ratings, thereby resulting in level generality. Generality of level across intrajudge rating orientations is said to occur in those rating orientations having the projected self- image as a common element. For those rating orientations in which the self-image is not projected, we would not expect to find generality of level. Generality would be predicted between Self and First Person Orientations, Personal and Self Orientations, Personal and First Person Orientations, and Personal and Third Person Orientations. These are, in fact, the orientations between which we did find clear-cut evidence of level generality. We would not expect to find generality between any of the other orientation pairs (with the possible exception of Third Person and Self Orientations). Therefore, according the the results of Study 1, and the explanation offered, generality of rating level across intrajudge rating orientations would not occur for all rating orientations; but rather, 42 there is felt to be a precise pattern of orientations in which level generality would occur to a varying degree. Generality of level within rating orientations was studied only for Second Person and Third Person Orientations. Generality of rating level across heterogeneous others, was found within the Third Person Orientation. This is in keeping with our self-image projection model. Regarding the remaining orientations not investigated, the self-image projection model would predict generality of level within First Person, Personal, and possibly Behavior Orientation ratings. The notion that projection may operate in the generality of an individual's rating level is by no means new. Landis ( 1936) , Crockett and Meidinger (1956) , and Scrodel and Freedman ( 1953) report evidence to this effect. Similarly, Weignarten ( 1949) demonstrated that judges sometimes project their own insecurities into their clinical evaluations of others. He states that this process is an unconscious one, since the individual is unaware that his judgments are influenced by his personal frame of reference. In explaining the operation of our self- image projection model, we have deliberately avoided, until now, the question of whether this process is conscious or unconscious. 43 Tannenbaum e_t__a_l. have proposed that the make-up of one's self-image is composed of both conscious and unconscious elements. Therefore, it is very likely that the process of self-image projection is both conscious and unconscious. That is, the judge may consciously project those elements of his self-image of which he is aware; and similarly, unknowingly project those elements of his self- image of which he has no conscious awareness. Admittedly, the entire projection process that we have proposed is hypothetical, and deserves careful research in and of itself. Our purpose in presenting this projection model, is merely an attempt to explain our current research findings. The type of projection that we have been discussing has variously been called naive projection (Tannenbaum §t_al. 1961) , or similarity projection (Sears, 1936; Rokeach, 1945; Guilford, 1959; Hornberger, 1960; Feshbach and Feshbach, 1963, and Campbell, Miller, Lubetsky, and O'Connell, 1964) . This projection process is seen as the attributing by the perceiver of his own characteristics to the perceived. If few cues are available to the perceiver (as is the case in our study), or if he is unable to utilize those that are avail- able, or if his feelings toward the other person are, in fact, similar to those he has about himself, then projection may become his signifi- cant modus operandi. As Tannenbaum et_al_. bring out, the vacuum that might be filled by meaningful cues is taken up by assumptions implying that the other resembles the judge. 44 If self-image projection is said to occur in rating situations where the amount of given information is limited, the question then arises as to the nature of the relation between the amount of information and the degree of projection. It is our feeling that this relation is likely to be curvilinear. However, this assumption leads to two con- flicting hypotheses: ( 1) the degree of projection is highest with maximum and minimum information; or (2) the degree of projection is lowest with maximum and minimum information. It remains the task of future research to test these hypotheses. This, in essence, is the projection process that we posit as accounting for the generality of level across and within certain rating orientations. The projection of a judge's self-image can be both a conscious and unconscious process. Therefore, if the judge is much like the other, with respect to the projected traits and characteristics, then prediction accuracy will be high. If the traits and characteristics of the judge and other are markedly different, then interpersonal sensitivity ratings are hindered. Thus, it appears that a judge should be given insight into the nature of this process, and be made, as much as possible, con- sciously aware of the extent to which his self-image influences his ratings. Similarly, those who utilize the rating method in general should either develop appropriate designs to minimize the influence of level generality, such as the forced-choice technique; or eliminate 45 these effects statistically, by taking into account the percentage of variance due to the generality of rating level. Implications for Future Research As noted, future research would do well to investigate the efficacy of the self-image projection model that we have here proposed. Of equal import, however, is the methodology that should be employed in measuring the generality of rating level. 'The short- comings of the epen-end scale items used in this study have seriously limited the interpretations that could be drawn from our results. Furthermore, the basic operations used in deriving an index of a rater's level have proven to be quite cumbersome and time consuming. These two methodological difficulties may be eliminated by the use of a modified Thurstone scale. A large number of items, drawn from various inventories (e.g., MMPI, EPPS, Strong, etc. ) , and which are believed to represent all degrees of social desirability, are presented to a large number of judges. These judges are then to rank-sort the items into several category positions, where the end- points range from very undesirable to very desirable. Those items with which there is much disagreement are discarded; and the remaining items are assigned a scale value which is the median of the scale positions assigned by the various judges. 46 The items are arranged in random order, and a rater is then instructed to rate the other (on the basis of the six orientations). by checking off those items which he feels are applicable. A'rater's level is then simply the average of the scale values of all the items he has checked. Generality of rating level can then be determined simply by correlating the average scale values obtained on each rating orienta- tion, by each judge. SUMMARY AND CONC LUSIONS The purpose of the present study was to determine the generality of rating level across different rating orientations of a . judge, and within orientations across heterogeneous others. Con- sideration was also given to a projective process which may be operative in the generalization of an individual's rating level. The study was divided into two phases. The first phase consisted of a test questionnaire in which a subject rated an inter- viewee on the basis of six rating orientations. The hypothesis tested was that when rating an other on dimensions of desirability- undesirability, a judge's level tends to be general across different orientations. The findings partially supported the hypothesis in that clear-cut generality of level was demonstrated between the following intrajudge rating orientations: 1. Self and First Person Orientations 2. Personal and Self Orientations 3. Personal and First Person Orientations 4. Personal and Third Person Orientations. No significant relationship was found between ten other possible orientation pairs . 47 48 The second phase consisted of a test questionnaire, in which a subject rated three interviewees on the basis of two orienta- tions. The hypothesis tested was that the level of a judge's ratings across heterogeneous others, tends to be general within orientations. The findings partially supported the hypothesis in that generality of level was demonstrated for one of the two orientations tested. A judge's rating level tends to generalize within Third Person Orienta- tion ratings across heterogeneous others. A self-image projection model was developed, in an attempt to explain these findings. The model posits that a judge's self—image forms the basis for the generality of level across certain intrajudge rating orientations, and within certain orientations across heterogeneous others. The projected self-image can be thought of as the "nucleus" which "binds" a judge's ratings, thereby resulting in level generality. Projection is said to occur in response to the limited information presented to the judge. It was concluded that generality of level across intrajudge rating orientations should occur only in those rating orientation pairs having the projected self-image as a common element. Similarly, generality of rating level across heterogeneous others, should occur only within those rating orientations in which a judge's self-image is projected. 49 For some rating orientations, the generality of level may be large; for others, quite small; and for still others, non-existent. Nevertheless, the influence of generality of rating level must be taken into consideration by anyone who attempts to effectively utilize the rating method. REFERENCES Campbell, D. T., Miller, N. Lubetsky, J. and O'Connell, E. J. Varieties of projection in trait attribution. Psychol. Monographs: General and Applied. , 1964, 7_8_, 1-33. Cline, V. B. and Richards, J. M. Accuracy of interpersonal per- ception--a general trait? J. abnorm soc. Psychol., 1960, El, 1-7. Crockett, W. H. and Meidinger, T. Authoritarianism and inter- ' personal perception. J. abnorm. soc. Psychol., 1956, _5_3_, 378-380. Crow, W. A methodological study of social perceptiveness. Un- published Doctor's dissertation, University of Colorado, 1954. , and Hammond, K. The generality of accuracy and response sets in interpersonal perception. J. abnorm. soc. Psychol., 1957, E, 384-390. Dymond, R. A scale for the measurement of empathic ability. 1. consult. Psychol. , 1949, _1_3_, 127-133. Personality and empathy. J. consult. Psychol., 1950, _l_4_, 343-350. Edwards, A. L. The relationship between the judged desirability of a trait and the probability that the trait will be en- dorsed. J. appl. Psychol. , 1953, _3_7, 90-93. Feshbach, S. and Feshbach, N. Influence of the stimulus object on complementary projection of fear. J. abnorm. soc. Psychol., 1963, 66, 498-503. Fey, W. Acceptance by others and its relation to acceptance of self and others: a reevaluation. J. abnorm. soc. Psychol., 1955, 2, 274-276. 50 51 Gage, N. L. , Leavitt, and Stone. The intermediary key in the analysis of interpersonal perception. Psychol. Bull., 1956, 53, 258-266. Explorations in teacher's perceptions of pupils. J. Teacher Education, 1958, 2, 97-101. Gottleib, B. and Kerr, W. An experiment in industrial harmony. Personnel Psychol., 1950, 3, 445-453. Gross, C. Intrajudge consistency in ratings of heterogeneous persons. J. abnorm. soc. Psychol., 1961, 6_2_, 606-610. Grossman, B. The measurement and determinants of interpersonal sensitivity. Unpublished Master's thesis, Michigan State University, 1963. Guilford, J. P. Personality. New York: McGraw-Hill, 1959. Hornberger, R. H. The projective effects of fear and sexual arousal on the ratings of pictures. J. clin. Psychol. , 1960, 16, 328-331. Kelly, E. L. Consistency of the adult personality. Am. Psychol., 1955, l_0, 659-681. Landis, C. Questionnaires and the study of personality. J. nerv. ment. Dis., 1936, 83, 125-134. Lindgren, H. and Robinson, J. An evaluation of Dymond's test of insight and empathy. J. consult. Psychol., 1953, 11, 172-176. ' McIntyre, C. Acceptance of others and its relation to acceptance of self and others. J. consult. Psychol., 1952, fl, 624-626. Phillips, E. L. Attitudes toward self and others: a brief question- naire report. J. consult. Psychol., 1951, 1_5_, 79-81. Rokeach, M. Studies in beauty: II. Some determiners of the per- ception of beauty in women. J. soc. Psychol., 1945, 22, 155-169. Scodel, A. and Freedman, M. L. Social percpetions of authoritarians and nonauthoritarians. J. abnorm. soc. Psychol., 1953, 4_8_, 181-184. 52 Sears, R. R. Experimental studies in projection: I. Attribution of traits. J. soc. Psychol., 1936, _7_, 151-163. Smith, H. C. Sensitivity to People: A Component Approach. New York: McGraw-Hill, 1966. Stagner, R. Dual allegiance as a problem in modern society. Personnel Psychol., 1954, _7_, 41-47. Tannenbaum, R., Weschler, I. R., and Massarik, F. Leaderslfi and Organization: A Behavioral Science Approach. New York: McGraw-Hill, 1961. Weingarten, E. A study of selective perception in clinical judgment. J. Personality., 1949, _l__7_, 369-406. APPENDIX A sslns Hay, 1965 THE CASE OF HARRY GEORGE Directions: This is a test of your ability to make accurate predictions about peOple. It consists of six parts, and each part asks you to make predictions from a different point of view. Read the directions for each part carefully. Each part of the test will use a different portion of your answer sheet. Be sure to note which spaces on your answer sheet.are to be used. . Your task is to make accurate predictions about Mr. George. Mr. George is a middle-aged married man with one child. As part of a research project on under- standing peOple, he was given a brief interview. A typescript is given below: Psycholggist: "What sort of person are you? How would you characterise yourself as a person?" Mr. Geor e: "Just an average person. I like the normal things most peeple do. I like to dance and play around that way. Of course I don't run around, I‘d say I was getting into a stable class. I'm over the younger fling." Psychologist: "What would you consider to be your greatest personality difficulty or handicap?" Mr. George: "Well, maybe too reserved." Psychologist: "You find that you're too reserved, in what way?" Mr. Georg : Well, especially in business. I think I take too much what the boss says, and do it. And maybe I can do it better, I do it the way he says to avoid trouble. In other words, I try to get along with people, which is good. But maybe sometimes I should say more about it to maybe help me or the others." Psychologist: "Assert yourself a little more?" Dir. George: "Yes." Psychologist: "Do you ever lose your temper?" Mr. George: "Well, very seldom with the person. I may become upset. I try my best not to let them know it." Psychologist: "What would you do if someone told a lie about you?" Mr. George: "Well, what kind of lie - that I did something I didn't?" Psychologist: "Yes. A lie that perhaps would be damaging to your character." Mr. George: "Well, I don't know, but I imagine I'd try and find out why the person said it. Maybe, as far as he knew, he was telling the trut ." Psychologist: ”Do you think that you would go to him and talk to him about it?" Mr. Geor e: "If it was of importance, otherwise I would forget it." Psychologist: "What sort of hobbies do you particularly enjoy?" Mr. George: "Well, I like to make things. Oh, woodwork and hunting I guess are the main things." a» e '1: * PART I Directions: Mr. George has checked one alternative on each of the statements below to describe himself. You are to check the alternative you think be checked. (Use spaces 1-10) 1. At the party, I was... (I) a little shy and reserved (2) the life of the party (3) quite smooth and polished 2' I could hate a person who... (1) is a hypocrite and two faced (2) is cruel and ridicules others (3)...I don't hate anybody (over) 3. When they told me what to do... (1) I did just the opposite (2) I listened politely but did nothing (3) I did it 4. When they offered me help I... (l) was somewhatembarrassed- (2) thanked them-but refused (3) accepted 5_. I felt my lack of success was due to. ... (l) laziness and lack of initiative (2) not getting along too well with peOple (3) adverse conditions and bad luck 6. When I'm criticized, I.. . (1) defend myself (2) take it, - (3) appreciate it 7. When I make a mistake I... (1) don't "give a damn" (2) am embarrassed (3) laugh it off - 8. When my conscience begins to bother me... ' (1) I'm' ashamed . (2) I analyze myself (3) I try to do the right thing 9. I am most anxious about. .. (1) my personality (2) world peace (3) my career- 10. I boiled up when... (l) I was criticized unjustly (2) I was cheated (3) I saw peOple hurting others , a * a * PAR!“ II Directions: lira George was rated on each of the traits below by persons who knew him well. They used the following rating system: . (1) "Rather UNlike" or "Very'UNlike him" (4) "Rather like him" (2) "A little UNlike him" .(5) "Very like him" (3) "A little like him" Rate‘Mr. Geo:g§ as you think he was rated by others. (Use spaces 29-40) 29. COOPERATIVE - : 35. STUBBORN 30. CONFIDENT ' ' 36. EGO'I'ISTICAL 31. FRIENDLY 37. SKY 32. AFFECTIONATB 38. REBELLIWS 33. CAREFUL 39.51MPRACTICAL 34. AMBITIOUS ' 40. UNREALISTIC it *‘ '1: 15 M Directions: Rate Mr. George as ou think he behaves in real life. For each of the following statements, check the one right alternative. Zfise spaces 57-63) 57. In his social life he: (1) is rather inactive and only rarely goes to social activities. (2) enjoys "dress-up" parties. - - (3) entertains informally frequently at home. 58. PeOple who know him describe him as being: ~ (I) somewhat dominating. (2) somewhat irresponsible. (3) easy to get along with. Mr . George (contd . ) -3- 59. He tends to be: (1) somewhat indifferent and lasy. (2) about average in ambition. (3) quite ambitious. 60. When he gets in an argument with his wife he: (1) may occasionally strike her. (2) shouts loudly and obscenely at her. (3) raises his voice a little but maintains good control. 61. In handling his 12-year-old son he is: (1) not too interested. (2) rather strict. (3) fairly easy going. 62. When he goes to a party he: (1) tends to flirt with women other than his wife. (2) sometimes gets drunk and then becomes the life of the party. (3) enjoys himself but is not much noticed. 63. When it comes to money: (1) he buys a great deal on credit, and is often late with his payments. (2) he buys a great deal on credit, but keeps his payments up. (3) he buys mainly with cash. a a a * PART IV Directions: Use the following rating system on each of the traits below: (1) "Rather UNlike" or "Very UNlike him" (4) "Rather like him" (2) "A little UNlike him" (5) "Very like him" (3) "A little like him" Rate'Mr. George as you think he is really like. (Use spaces 81-92) 81. COOPERATIVE 87. STUBBORN 82. CONFIDENT 88. EGOTISIICAL 83. FRIENDLY 89. SHY 84. AFFECTIONAIE 90. REBELLIGJS 85. CAREFUL 91. IMPRACTICAL 86. AMBITIOUS 92. UNREALISTIC * a a e PART V Directions: Use the following rating system on each of the traits below: (1) "Rather UNlike" or "Very UNlike you" (4) "Rather like you" (2) "A little UNlike you" (5) "Very like you" (3) "A little like you" Rate yourself as you think.Mk. George WOULD RAID YOU. (Use spaces 109-120) 109. COOPERATIVE 115. STUBBORN 110. CONFIDENT 116. EGOTISTICAL 111. FRIENDLY 117. SE! 112. AFFECTIONAIE 118. REBELLIOUS 113. CAREFUL 119. IMPRACTICAL 114. AMBITIOUS 120. UNREALIS‘I‘IC * * ~* * (over) mu: v1 Directions: Use the following rating system: (1) "Rather UNlike" or "Very 0111114 me" ’ (4) "Rather like me" (2) "A little UNlike me" . 1 (5) "Very like me" (3) "A little like me" " ‘ " . RATE YOURSELF on each of the traits below. (Use spaces 137-148) 137. coopmrrvr 143. srunnonn 138. comment 144. sconsncai. 139. 1:313er f 145. say 140. monomrs . , .j 145-. RBBELLIOUS 141. cmm , ' ' 147. IMPRACTICAI. 142. mmws ' ‘ ‘ ' 14s. umxsrtc, ' mc5/21/65 APPENDIX B The Case of Mr. George Your task is to make accurate predictions about Mr. George. Mr. George is a middle-aged, married man with one child. As part of a research project on under- standing peOple, he was given a brief interview. A typescript is given below. Psychologist: "What sort of person are you? How would you characterize yourself as a person?" IMr. George: "Just an average person. I like the normal things most people do. I like sports, I like to dance and play around that way. Of course, I don‘t run around, I'd say I was getting into a stable class. I'm over the younger fling." Psychologist: "What would you consider to be your greatest personality difficulty or handicap?" 1h» George: "Well, maybe too reserved." Psychologist: "You find you're too reserved, in what way?" Mr. George: "well, especially in business. I think I take too much of what the boss says, and do it. And, though maybe I can do it better, I do it the way he says to avoid trouble. In other words, I try to get along with people, which is good. But maybe sometimes I should say more about it to maybe help me and the others." Psychologist: "Assert yourself a little more?" Mr. George: "Yes." Psychologist: "Do you ever lose your temper?" Mr. George: "Well, very seldom with the person. I may become upset. I try my best not to let them know it." Psychologist: "What would you do if someone told a lie about you?" Nun George: "Well, what kind of a lie - that I did something I didn't?" Psychologist: "Yes. A lie that perhaps would be damaging to your character." ZMr. George: "well, I don't know, but I imagine I'd try and find out why the person said it. Maybe, as far as he knew, he was telling the truth." Psychologist: "Do you think that you would go to him and talk to him about it?" Mr. George: "If it was of importance, otherwise I would forget it." Psycholsgist: "What sort of hobbies do you particularly enjoy?" Nku Georgs: "Well, I like to make things. 0h, woodwork and hunting I guess are the main things." Psychologist: "How important do you feel religion is to people in these times?" Hr. George: "I don't go in for religion too much. I believe that it is necessary for everybody to have a basic belief. As far as the religious part goes, in my own living I don't place that as a major issue." Psychologist: "Then religion is not too important to you personally?" Pun George: "No." Psychologist: "But you do feel that people should have some sort of basic faith?" lir. George: "Yes, they have to have a code to live by, and that's the best one I San think of." The Case of Mr. George Part A lzirections: Mr. George has checked one alternative on each of the statements below in? describe himself. You are to check the alternative you think he checked. Use space 3 1-10 . l. 3. 9. 10. At the party, I was ... (l) a little shy and reserved 0 4.2,.“ '1... v ‘...a as 0... +119 r1142? —W. (3) quite smooth azxd polish ed I could hate a person who ... (l) is a hypocrite and two faced (2) is crvcl and ridicules others (3) ... I don't hate anybody When they told me what to do ... (l) I did just the opposite (2) I listened politely but did nothing (3) I did it When they offered me help I ... (l) was somewhat embarassed (2) thanked them.but refused (3) accepted I felt my lack of success was due to ... (l) laziness and lack of initiative (2) not getting along too well with people (3) adverse conditions and bad luck ‘When I'm criticized, I ... (1) defend myself (2) take it (3) appreciate it When I make a mistake I ... (I) don't "give a damn" (2) am embarassed (3) laugh it off When my conscience begins to bother me ... (I) I'm ashamed (2) I analyze myself ' (3) I try to do the right thing I am most anxious about ... (1) my personality (2) world peace (3) my career I boiled up when ... (l) I was criticized unjustly (2) I was cheated (3) I saw people hurting others Part B ~ Directions: Hr. George was rated on each of the traits below by persons who knew him well. They used the following rating system: "Rather UNlike" or ”Very UNlike him” "A little UNlike him" "A little like him" "Rather like him" "Very like him" 'Rate Mr. George as you think he was rated by others. Use spaces ll-22. 11. COOPERATIVE 12. CONFIDENT 13. FRIENDLY 1h. AFFECTIONATE 15. CAREFUL 16. AMBITIOUS 17. STUBBORN 18. EGOTISTICAL 19. SHY 20. REBELLIOUS 21. IHPRACTICAL 22. UNREALISTIC HIHI The Case of Mr. walter Your task is to make accurate predictions about Mr. Halter. Mr. walter is a young married man with two children. As part of a research project on understanding people, he was given a brief interview. A typescript is given below. Psychologist: "Just what sort of a person are you?" Mr. Walter: "That's hard to determine. I'm one type of person to myself and another type of person to society. I'd have to give two definitions to answer that correctly - how I am to myself, and how I am to people who know me." Psychologist: "Wbuld you like to discuss that a little further? What sort of person are you to yourself?" Mr. Walter: "Well, I think I'm a person of probably over-average intelligence, with ambitions to be able to better myself and my society, to be able to help society." Psychologist: "What sort of person do you feel you are to other people?" Mr. walter: "Well, I hope I'm pretty nearly the same kind of person to other people as I am to myself. I get along well with most people, I don't have a great nmny friends; I have a few intimate friends, and with these people I'm quite close. I get along well with these people. And, I can be pretty compatible with most people." Psychologist: "What do you feel is your greatest personality handicap?" Hr. welter: "The fact that I try too hard to do things, I believe. This hinders me from being able to do things - by being under certain tensions." Psycholggist: "Do you ever lose your temper?" Mr. Walter: "Rarely." Psychologist: "What sort of thing would cause you to lose your temper?" Mr. walter: "Whll, never having lost my temper completely - I've always been able to hold my emotions pretty well in check - it would have to be a fairly devastating thing, I think, to make me lose it, or to become completely out of control of myself." Psychologist: "What sort of hobbies do you particularly enjoy?" Mr. walter: "Golf, music, Spectator sports - I am not too athletic - tennis, things such as this." Psychologist: "Well, how important do you feel religion is to people in these times?" Mr. welter: "That's a pretty deep subject. Not being a deeply religious man my- self, it isn't too important to me. The moral teachings of religion help man be able to live better with himself, and with other people in society. I think today it's quite important for most people — not for the supernatural aspects of it, but for the moral teachings of religion." E§ycholsgist: "You don't feel that it's necessary for you?" Mr. walter: "Not necessary, no." The Case of Mr. walter Part A Directions: Mr. Walter has checked one alternative on each of the statements below to describe himself. You are to check the alternative you think he checked. Use spaces 23-32 23- Religion seems to me ... (l) unnecessary (2) a problem (3) necessary and important guilty about .0. (1) nothing (2) 13"?" (3) a lot of things (1) too often considered as filthy. (2) great (3) reserved for marriage When they didn't invite me I ... (I) didn't care and said, the hell with it (2) was unhappy and felt bad (3) knew they had a reason My friends consider me ... (1) fair (2) a hard worker (3) a nice person I enjoy .... (1) great music (2) being with people (3) sports My philosophy of life is ... (l) uWhatever you do, do well," (2) "Enjoy today, think of tomorrow." (3) "Do unto others, as you would have them do unto you." I would go mad if ... 2D. I feel 25. ;: 26. --—_— 27. 28. 29. -————- 30. -——-—-— Directions: him well. (1) somebody nagged me all the time (2) I had nothing to do (3) I thought there were no purpose in life meet people, I generally feel ... (l) indifferent (2) uneasy and self-conscious (3) at ease and genial party, I was ... (l) a little shy and reserved (2) the life of the party (3) quite smooth and polished Part B Mr. walter was rated on each of the traits below by persons who kneW' They used the following rating system: "Rather UNlike" or "Very UNlike him" "A little UNlike him" "A little like him" "Rather like him" ( ( E ( "Very like him" mum to i—J vvvvv Rate Mr. Walter as you think he was rated by others. Use spaces 33-hh. 33. COOPERATIVE 39. STUBBORN 3h. CONFIDENT no. EGOTISTICAL 3S. FRIENDLY t1. SHY 36. A"FIBC'I‘.?OE?AT3 t2. REBLLLIOUS 37. CAREFUL L13. IIIPRACTICAL 38. AJLBITIOUS uh. UIULJALISTIC HHH The Case of Mr. Allen - Your task is to make accurate predictions about Mr. Allen. Mr. Allen is a young, single man. As part of a research project on understanding people, he was given a brief interview. A typescript is given below. Psychologist: "Just what sort of person are you?" Mr. Allen: "Wbll, I guess an easy-going one. I'm easy to get along with." Psychologist: "Well, what else can you tell me about yourself as to what sort of person you are?" Mr. Allen: "Well, I guess that's about all. I have some temper - not much." Psychologist: "well, what would you consider to be your greatest personality handicap?" Mr. Allen: "well, I guess just paying attention when there are people talking to me. Just paying attention to them." Psychologist: "Do you have difficulty paying attention to people when they talk to you?" . Mr. Allen: "No, No, I don't have no difficulty, it's just that whenever I walk into a place, I just don't speak, I'm quiet." Psychologist: "Oh, I see, you find yourself quiet. Do you have difficulty making friends?" Mr. Allen: "No, No, I don't find no difficulty making friends." Psychologist: "After you once get to know them, then. But to begin with, you feel a little reserved about it, is that it?" Mr. Allen: "Yuh." Psychologist: "Well, do you ever lose your temper? What sort of thing would cause you to lose your temper?" Phu Allen: "Once in a great while. It has to be something pretty mean, I guess, or something pretty big. One I guess is just - I don't know - couldn't tell you that until I lost my temper. Well, for instance, my little brother taking off “with my car." Psychologist: "That would make you unhappy?" Fhu Allen: "Yuh." Psychologist: ”What would you do if someone told a lie about you?" Ifir. Allen: "I guess that would make me a little sore too, if it wasn't true.” Psychologist: "What would you do about it? Wbuld you go to the person and talk to him about it?" Mr. Allen: "I wouldn't do nothing. Just sort of keep it to myself." Psychologist: "Well, how would you feel, and what would you do if someone gave you a million dollars?" IMr. Allen: "I'd be pretty happy, I guess. I guess I've never thought about what I’d do with it. I'd spend it I guess." Psychologist: "What sort of things do you do in your spare time, when you're not working?" Mr. Allen: "Oh, usually drive around; I like to drive around quite a bit." Psychologist: "Do you participate actively in any sports, or are you a spectator?" Mr. Allen: "No, I participate in it. Basketball, for instance." Psychologist: "Well, tell me, how important do you feel religion is to people in these times?-How is it important to you?" Mr. Allen: "Yes, I really do think that religion is important. I don't know, I guess just being good, people go out, and that ain't so bad, just going out and partying, but after that, the way they gather--." Psychologist: "And you think that religion would affect that sort of thing?" Fir. Allen: "I think so, because of conscience - people have a conscience, and that would be on it." Psychologist: "In what way is religion important to you?" Mr. Allen: "I don't know, well, sometimes when you go out partying, doing something else, and yet you don't." E :3chologi:€: " n3n':: 3. ye“: r--3vion, is that it?" Mr. Allen: "Uh-huh." The Case of Mr. Allen Part A you feel like Directions: Mr. Allen has checked one alternative on each of the statements below to describe himself. You are to check the alternative you think he checked. Use spaces hS—Sh. hS. If I can't get what I want, I ... (l)wflt (2) get it in another way (3) usually get along without it I I no. I feel "down in the dumps" when ... (1) ... I don't (2) I say the wrong thing (3) I don't succeed h7. When I make a mistake I ... (1) don't give a damn (2) am embarassed (3) laugh it off h8. The thing that gets me into hot water most is ... (l) trusting people too much (2) opening my "big mouth" too much (3) indecision ... can't make up my mind h9. I enjoy ... (1) great music (2) being with peOple (3) sports 50. When they told.me what to do ... (l) I did just the opposite (2) I listened politely but did nothing (3) I did it 51. When they didn't invite me I ... (l) didn't care and said, the hell with it (2) was unhappy and felt bad (3) knew they had a reason 52. At the party, I was ... (l) a little shy and reserved (2) the life of the party (3) quite smooth and polished 53. Religion seems to me ... (l) unnecessary (2) a problem (3) necessary and important _ St. I boiled up when ... (1) I was criticized unjustly (2) I was cheated (3) I saw people hurting others Part B Directions: Mr. Allen was rated on each of the traits below by persons who knew him well. They used the following rating system: (1) "Rather UNlike" or "Very UNlike him" (2) "A little UNlike him" (3) "A little like him" (h) "Rather like him" (5) "Very like him" Rate Mr. Allen as you think he was rated by others. Use spaces 55-66. SS. COOPERATIVE 61. STUBBORN 56. CONFIDENT 62. EGOTISTICAL S7. FRIENDLY 63. SHY 58. AFFECTIONATE 6h. REBELLIOUS S9. CARBF 65. IMPRACTICAL l l 60. AMBITIOUS 66. UNREALISTIC APPENDIX C 69 Table 8. --Means and standard deviations of Study 1 rating scales. Orientation Mean SD Personal 44. 71 4.61 Third Person 45. 20 3. 94 First Person 42.18 5.06 Second Person 23. 37 2. 37 Self 42. 20 5. 12 Behavioral 17. 47 1. 65 Note: The mean and standard deviation of each scale was computed prior to division by scale constant. APPENDIX D 71 Table 9. --Means and standard deviations of Study 2 rating scales. Orientation Mean SD Section 1: Mr. George Second Person 23. 26 2. 23 Third Person 45. 45 4. 23 Section 2: Mr. Walter Second Person 19. 80 3. 03 Third Person 43. 20 6. 78 Section 3: Mr. Allen Second Person 21. 7O 2. 21 Third Person 34. 31 5. 59 Note: The mean and standard deviation of each scale was computed prior to devision by scale constant. TATE UNIVERSITY LIBRARIES 111 MICHIG Hlll I III II! HM 74 7896 3 u; u 293