Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

The validity of personality perceptions in discussion groups Bruce, Melodie Nadine 1991

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
831-UBC_1991_A8 B77.pdf [ 2.1MB ]
Metadata
JSON: 831-1.0098680.json
JSON-LD: 831-1.0098680-ld.json
RDF/XML (Pretty): 831-1.0098680-rdf.xml
RDF/JSON: 831-1.0098680-rdf.json
Turtle: 831-1.0098680-turtle.txt
N-Triples: 831-1.0098680-rdf-ntriples.txt
Original Record: 831-1.0098680-source.json
Full Text
831-1.0098680-fulltext.txt
Citation
831-1.0098680.ris

Full Text

THE VALIDITY OF PERSONALITY PERCEPTIONS IN DISCUSSION GROUPS by MELODIE NADINE BRUCE B.A., The University of Prince Edward Island, 1989 A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF ARTS in THE FACULTY OF GRADUATE STUDIES Department of Psychology We accept t h i s thesis as conforming to the required standard THE UNIVERSITY OF BRITISH COLUMBIA September 1991 © M e l o d i e Nadine Bruce, 1991 In presenting this thesis in partial fulfilment of the requirements for an advanced degree at the University of British Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the head of my department or by his or her representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission. Department of Psychology The University of British Columbia Vancouver, Canada Date September 4, 1991 DE-6 (2/88) Abstract The a b i l i t y of group members to judge personality i n long-term discussion groups was investigated. Participants were randomly assigned to groups of 5 to 7 members who met once a week for seven weeks. None of the participants i n any group was previously acquainted. Prio r to the f i r s t meeting, they completed a battery of s e l f - r e p o r t measures including the NEO Five Factor Inventory and the revised Interpersonal Adjective Scales. After the f i r s t , middle, and f i n a l meetings, group members rated each other on the Big Five personality t r a i t s and group p a r t i c i p a t i o n . Self-reports showed moderate correlations (M=.31) with final-week s i n g l e -item peer ratings. In addition, the correlations increased s i g n i f i c a n t l y over time. The same l e v e l of p r e d i c t a b i l i t y was found when the c r i t e r i a were peer free descriptions rather than ratings. Results are discussed i n terms of the perception of personality i n the context of group discussions. i i i Table of Contents Abstract i i L i s t of Tables. . v Acknowledgments v i Introduction 1 Effects of F a m i l i a r i t y on Peer Ratings 1 The Big Five T r a i t s 3 Personality i n Discussion Groups 4 Hypotheses 5 Peer Ratings 5 Free Descriptions 6 Method 7 Subjects 7 Procedure 7 Self-Reports 9 Peer Ratings 9 Free Descriptions 10 Results 11 The Prediction of Peer Ratings: Peer V a l i d i t i e s 13 Extraveriosn 15 Agreeableness 15 Conscientiousness 16 Neuroticism 16 Openness 16 Expert Ratings of Free Descriptions 18 Discussion 21 Peer Ratings as the C r i t e r i o n 21 i v Extraversion 22 Agreeableness 23 Conscientiousness 2 3 Neuroticism 24 Openness 25 Summary 25 Expert Rated Free Descriptions as the C r i t e r i o n 26 P a r t i c i p a t i o n and Observability of the Big Five 27 What's Special About Discussion Groups? 27 Limitations and Recommendations 28 References 31 V L i s t of Tables 1. Peer Rating Sheet 34 2. Descriptive S t a t i s t i c s for Self-Report Scales 35 3. Intercorrelations of Self Reports 36 4. Descriptive S t a t i s t i c s for Peer Ratings 37 5. Intercorrelations of F i n a l Peer Ratings 3 8 6. Intercorrelations of F i n a l Peer Ratings Controlled for Time Talking 39 7. Average Inter-rater Agreement Within Groups 40 8. Prediction of Corresponding Peer Ratings from Self-Reports Over Time 41 9. Prediction of F i n a l Peer Ratings from Self-Reports 42 10. Prediction of Peer Ratings from Corresponding Self-Reports After Controlling for Time Talking 43 11. Intercorrelations of Peer Ratings Over Time 44 12. Mean Ratings of Free Descriptions by Expert Raters 45 13. Inter-Rater Agreements of the Expert Ratings 46 14. Intercorrelations of Big Five Expert Ratings 47 15. Intercorrelations of Expert Ratings Blocked on Verbal P a r t i c i p a t i o n 48 16. Predictions of Expert Ratings from Self-Reports Blocked on Verbal P a r t i c i p a t i o n 49 17. Relative V a l i d i t i e s of Peer Ratings, Ranked Peer Ratings, and Expert Ratings of Free Descriptions 50 Acknowledgments v i I would l i k e to convey my appreciation to several individuals whose support and encouragement have helped me immeasurably during the l a s t two years. F i r s t , I would l i k e to give special thanks to my mother, Marie Manderson, both for the obvious and the intangible. Her high regard for continuing education, hers and my own, has taught me that nothing worth having comes e a s i l y . l i a n a Katz, Laura Loewen, and P a t r i c i a Tollestrup have made the process so much more tolerable -even enjoyable-simply by t h e i r presence i n my l i f e ; Thank you, friends. My thanks are also due to Elaine McKay, Steve Moon, Georgia S t a v r i d i s , and Paul Trapnell, who assisted i n the coding and analysis of data. I would also l i k e to acknowledge my committee members, Jerry Wiggins and Dan Perlman, for t h e i r helpful comments and c r i t i c i s m s . Thanks also to Rebecca C o l l i n s . F i n a l l y , I thank my advisor, Del Paulhus, for a l l that he has taught me. 1 The V a l i d i t y of Personality Perceptions in Discussion Groups The group meeting plays an important role i n many domains—business, education, and recreation to name a few. Individual differences i n the ways people inter a c t within these settings are c l e a r l y important, and yet, with some notable exceptions (e.g., Bales, 1970), l i t t l e work on the emergence of personality i n the group s e t t i n g has been done to date. The present study i s concerned with one facet of t h i s issue—'-the way personality i s perceived i n leaderless groups. S p e c i f i c a l l y , can personality be judged accurately given the complexity of interpersonal dynamics i n such groups? Furthermore, do group members become more accurate as they gain f a m i l i a r i t y with the target? In general, l i t t l e i s known about the longitudinal e f f e c t s of f a m i l i a r i t y on the accuracy of peer ratings. I w i l l begin with a review of that l i t e r a t u r e . E f f e c t s of F a m i l i a r i t y on Peer Ratings Some years ago, Newcomb (1961) addressed the acquaintance process i n a book by that name. The book was based on his studies of groups l i v i n g together for four months i n a large house. The participants were required to rate each other at weekly i n t e r v a l s on various attitudes, values and behaviors. 2 Newcomb found that p a r t i c i p a n t s ' i n i t i a l ratings were not predictive of l a t e r ratings even over a short span of f i v e weeks. Ratings at Week 5, however, were highly predictive of l a t e r ratings, suggesting that impressions eventually s t a b i l i z e d . Although Newcomb's peers d i d not rate personality t r a i t s per se, t h i s r e s u l t i s i n s t r u c t i v e . I t suggests that, i n several domains (attitudes, values, and behaviors), increasing f a m i l i a r i t y with a target has an important influence on judgments. Several more recent studies have looked d i r e c t l y at the effe c t s of f a m i l i a r i t y on accuracy of personality impressions. For example, Funder and Colvin (1988) administered s e l f - r e p o r t Q-sorts, and had two close acquaintances and two v i r t u a l strangers Q-sort each target. They found that acquaintances agreed more with one another and with the target than did strangers. Extending t h i s work, Colvin and Funder (1991) found that acquaintances predicted peer ratings better than strangers did, but acquaintances and strangers were equally accurate at predicting behavior. This pattern, however, seems attributable to the fac t that the behaviors were observed i n a laboratory s e t t i n g where the strangers, but not the acquaintances, had past experience i n observing the target. In another cross-sectional study, Paunonen (1989) looked at the ef f e c t s of acquaintanceship and t r a i t o b servability on accuracy. The re s u l t s showed a main e f f e c t for acquaintance, but no e f f e c t for obse r v a b i l i t y . In additions 3 he found that acquaintanceship interacted with l e v e l s of t r a i t o bservability, such that for low to moderately acquainted dyads, observability was p o s i t i v e l y related to self-peer agreement, but for highly acquainted dyads agreement was greater on less observable dimensions. Unfortunately, a l l the above studies looked at the e f f e c t s of f a m i l i a r i t y on personality perceptions using a cross-sectional (between-subjects) design. In that design many factors are necessarily correlated with acquaintanceship (e.g., involvement with target, target s c r u t a b i l i t y ) . To date, no one has attempted to follow the development of such impressions over t i m e — t h a t i s the goal of the present study. The Big Five T r a i t s To draw strong conclusions from such an investigation, the core personality t r a i t s should be the focus. Best validated are the "Big Five" personality t r a i t s : A set of broad t r a i t domains that each subsume a number of s p e c i f i c facets of personality (Norman, 1963). The labels t y p i c a l l y given to the f i v e domains are Extraversion (or Surgency), Agreeableness, Conscientiousness, Neuroticism/Emotional S t a b i l i t y , and Openness to Experience (or Culture). Evidence for the v a l i d i t y of these t r a i t s i s substantial (e.g., Costa & McCrae, 1989; McCrae & Costa, 1987). 4 In the interpersonal realm, the major dimensions are the two axes of the interpersonal circumplex, v i z . dominance and nurturance (Wiggins, 1979). Rather than c o n f l i c t i n g with the Big Five, they may be considered rotations of Extraversion and Agreeableness. (For a summary of the evolution of the Big Five i n personality research, see John [1990] or Trapnell and Wiggins [1990]). Some evidence i s available about the e f f e c t s of f a m i l i a r i t y on Big Five judgments. For example, Passini and Norman (1966) demonstrated that stranger-ratings showed lev e l s of i n t e r r a t e r agreement comparable to those of close acquaintances. Strangers-ratings, however, were much less correlated with s e l f - r a t i n g s than those of close acquaintances. Evidently, f a m i l i a r i t y had the e f f e c t of increasing accuracy. Watson (1989) reported much the same findings i n his recent r e p l i c a t i o n of the Passini and Norman (1966) study, with one very important difference: Strangers were as accurate as acquaintances i n predicting the target's s e l f -ratings on a l l dimensions except Emotional S t a b i l i t y . In short, the two relevant studies on the Big Five seem to c o n f l i c t . Combined with the studies on other t r a i t s , however, the bulk of the evidence supports the claim that f a m i l i a r i t y improves accuracy. Personality i n Discussion Groups 5 We could f i n d no research to date on the formation of personality impressions i n small-group discussions. Fortunately, we had an ide a l opportunity to study t h i s phenomenon i n an introductory personality c l a s s . Once a week for seven weeks, students met i n discussion groups designed to f a c i l i t a t e getting acquainted and to review course material. Discussion topics were selected to bring out a wide range of personality differences. After meetings 1, 4, and 7, participants rated themselves and each other on the same set of t r a i t s . In addition, af t e r week 7, students wrote one-page free descriptions of each member of t h e i r group. This procedure thus afforded us the opportunity to t e s t the a b i l i t y of group members to judge personality i n a novel s e t t i n g . We were also able to compare three sets of c r i t e r i a , namely peer ratings, ranked peer ratings, and peer free descriptions rated by expert judges. Hypotheses • Note that I w i l l use the term " v a l i d i t y " as a short form to r e f e r to the observed c o r r e l a t i o n between predictor and corresponding c r i t e r i o n . Peer Ratings. Based on the above l i t e r a t u r e review, several hypotheses were advanced. F i r s t , s e l f - r e p o r t s should predict peer ratings of the f i v e major personality dimensions. Second, the v a l i d i t y should be highest for the 6 dimension of Extraversion, since extraverted t r a i t s have consistently been found to be the easiest to judge interpersonally (Park & Judd, 1989). Third, v a l i d i t i e s for a l l t r a i t s should be higher among extraverted individuals because t h e i r t r a i t s are more observable. Fourth, the siz e of the v a l i d i t i e s should increase over time. Just as studies with between-subjects designs have shown close acquaintances to be better judges of personality than strangers, we predicted that ratings at Week 7 would be better predicted than ratings at Week 1. F i f t h , there should be increasing orthogonality among the peer-rated Big Five factors over time. Thus, as f a m i l i a r i t y increases, peers should become more adept at making f i n e -grained d i s t i n c t i o n s among behaviors, and at a t t r i b u t i n g each behavior to the correct dimension of personality. Free Descriptions. From the l i t e r a t u r e review, three hypotheses were suggested. F i r s t , as with the peer ratings, the i n t e r c o r r e l a t i o n s of the expert-rated Big Five should also approach orthogonality. In addition, expert ratings should be more orthogonal than peer ratings, because experts should be more a l e r t to possible confounds. Third, i t i s anticipated that s e l f - r e p o r t s w i l l predict expert ratings as well as or better than they predict peer ratings. This follows from the judicious choice of t r a i t s given to expert raters to evaluate the free descriptions. The choice was based on Trapnell's (1991) factor analysis of 7 t r a i t adjectives that correlate most highly with each of the f i v e factors. One would expect a higher self-report/peer-r a t i n g convergence between ratings made with s i m i l a r underlying concepts. Peer ratings, on the other hand, were made on t r a i t adjectives that were s i m i l a r to the Big Five but did not as c l o s e l y duplicate them as did the expert ratings. Method Subjects Participants were 89 students enrolled i n a third-year summer course i n personality. They included 55 females and 34 males ranging i n age from 19 to 45, with the majority between 20 and 24. Procedure During the f i r s t week of class and p r i o r to the f i r s t group meeting, participants completed a battery of s e l f -report measures. During the second week of classes, participants were randomly assigned to one of sixteen heterogeneous groups based on the age, gender, race, and academic interests of the student. Each group consisted of 5 to 7 individ u a l s who were previously unacquainted. Pr i o r to the f i r s t group meeting, participants were given the following overview of the weekly procedure: (1) There would be a discussion topic or topics assigned for each 8 week's meeting; (2) Each meeting would be approximately one half-hour i n length; (3) Students were to complete the assigned r a t i n g tasks as soon afte r each meeting as possible; (4) Completed ra t i n g forms were to be sealed i n the envelopes provided and returned to the instructor before the next meeting; (5) A l l exchanges that occurred within the meetings were to be kept i n s t r i c t confidence; and (6) Participants were to r e f r a i n as much as possible from s o c i a l i z i n g with other group members outside of the group meeting times. Groups then proceeded to a preassigned, r e l a t i v e l y private, meeting area. Discussion topics were chosen to meet three s p e c i f i c objectives: (a) to p a r a l l e l course topics; (b) to encourage group discussion and debate, and (c) to draw out behavior that would help participants evaluate each other on various personality t r a i t s , p a r t i c u l a r l y the Big Five. For example, to encourage behavior relevant to Openness to Experience, Week Seven's discussion topic required group members to t a l k about an absorbing moment they had experienced, and something creative they had accomplished within the l a s t year. The discussion topics were, i n order of assignment: 1) Early memories and Freud; 2) A l l p o r t ' s aspects of adjustment; 3) Sternberg's types of i n t e l l i g e n c e ; 4) Social issues and phobias; 5) Stress and test-taking; 6) Positive and negative q u a l i t i e s , and academic stereotypes; and 7) Creative and absorbing experiences. 9 To permit assessment of changes i n personality impressions as well as of f i n a l impressions, peers were asked to rate group members on the Big Five at three times, once immediately a f t e r the f i r s t week's discussion, again at the midpoint a f t e r Week 4, and once more af t e r the f i n a l meeting, at Week 7. F i n a l l y , participants wrote free descriptions of each group member. Self-Reports. The most widely used measures of the Big Five were developed and validated by P. Costa and R.R. McCrae (Costa & McCrae, 1988; McCrae & Costa, 1983, 1987). Both the 60-item NEO Five Factor Inventory (NEO-FFI), and the 300-item NEO Personality Inventory (NEO-PI), have been extensively validated. Due to the NEO-FFI's shorter length and i t s demonstrated r e l i a b i l i t y and v a l i d i t y (Costa & McCrae, 1989; McCrae & Costa, 1987), the NEO-FFI was chosen as the s e l f - r e p o r t inventory to measure the Big Five i n t h i s study. To assess the interpersonal sphere, the revised Interpersonal Adjective Scales were used (IAS-R; Wiggins, Trapnell & P h i l l i p s , 1988). Sixty-four adjectives are rated on f i v e - p o i n t scales ranging from "Disagree" to "Agree". This measure permits the scoring of eight octants of interpersonal t r a i t s : These include PA, which I w i l l term dominance. and LM, which I w i l l term nurturance. Peer Ratings. After weeks 1, 4, and 7, participants rated one another on s i x unipolar scales. A l l s i x were scored on ten-point scales with endpoints labeled "Low" and 10 "High". The f i r s t f i v e scales were designed to tap the Big Five dimensions. To help c l a r i f y the construct for participants, each adjective label (except Insecurity) was followed by two related adjectives. In the usual Big Five order, the labels were: Assertiveness (vocal, dominant), Prosocial Orientation (cooperative, l i k e a b l e ) , Work  Orientation (deliberate, organized), Insecurity. and I n t e l l e c t ( o r i g i n a l , c l e v e r ) . The s i x t h rating scale was labeled Time Talking. The page of rating scales that participants completed i s displayed i n Table 1. Insert Table 1 about here Free Descriptions. Following the l a s t discussion group, a l l p a rticipants were asked to prepare a project i n which they described t h e i r f i n a l impressions of each group member, and the behavioral evidence they used as the basis for these conclusions. Two teams of two raters were trained to code Big Five descriptions. Half of the projects were coded by one team, and half by the other. Each of these free descriptions was then coded by raters (one graduate and one undergraduate student). Ratings of the Big Five were made on seven-point L i k e r t scales with endpoints labeled "Low" and "High". As a r e l i a b i l i t y check before beginning to code the free descriptions, a random set of f i v e descriptions was chosen for each dimension and coded by a l l four raters, as well as 11 by a personality psychologist (D.L. Paulhus). The i n t e r r a t e r r e l i a b i l i t i e s were found to be quite high. Results The descriptive s t a t i s t i c s for the s e l f - r e p o r t scales are l i s t e d i n Table 2. Means, standard deviations, and alpha r e l i a b i l i t i e s of the NEO-FFI scales are i n keeping with a large student sample c o l l e c t e d by Trapnell (1990) Insert Table 2 about here The i n t e r c o r r e l a t i o n s among the se l f - r e p o r t s are given i n Table 3. Note that the pattern of correlations c l o s e l y resembles those reported i n the manual for the NEO-FFI (Costa & McCrae, 1989). Insert Table 3 about here The means and standard deviations of the peer ratings are provided i n Table 4. A l l the means range between 6.0 and 7.0 on 10-point scales. Insert Table 4 about here Unless otherwise s p e c i f i e d , only the f i n a l (Week 7) peer ratings w i l l be addressed. Their i n t e r c o r r e l a t i o n s are 12 presented i n Table 5. The absolute value of the int e r c o r r e l a t i o n s are a l l rather large, ranging from .43 to .78. Whereas the ra t i n g of time spent t a l k i n g i n the group should only be strongly correlated with extraversion, i t i s instead related to a l l dimensions. This suggests that p a r t i c i p a t i o n rate i s acting as a general confound i n the peer ratings. Insert Table 5 about here To explore t h i s p o s s i b i l i t y further, the int e r c o r r e l a t i o n s were recomputed with the variance due to time spent t a l k i n g p a r t i a l e d out (Table 6). Following t h i s correction, most of the in t e r c o r r e l a t i o n s dropped to the modest l e v e l of .20 and below. Insert Table 6 about here Table 7 displays the i n t e r r a t e r agreements for peer ratings. Agreement i s indexed by mean i n t e r - r a t e r c o r r e l a t i o n s . Contrary to our prediction, i n t e r r a t e r agreement across a l l participants did not increase over time. Overall, agreement was highest—.41 averaged across a l l f i v e dimensions—in Week 1. Weeks 4 and 7 produced roughly equivalent agreement means of .34 and .32, respectively. Agreement was greatest on ratings of extraversion, .56, and least for ratings of openness, .23. Insert Table 7 about here The Prediction of Peer Ratings: Peer V a l i d i t i e s Correlations of NEO-FFI scales with the appropriate peer rated c r i t e r i a ( i . e . , the v a l i d i t i e s ) are shown i n Table 8. As predicted, the v a l i d i t i e s at week 7 (mean r = .31) were s i g n i f i c a n t l y higher than the v a l i d i t i e s at week 1 (mean r .25), Z = 2.45, p < .02. This f i n a l figure (.31) i s only s l i g h t l y less than the mean c o r r e l a t i o n of .40 between sel f - r e p o r t s and peer ratings reported by Costa and McCrae (1989). Extraversion was strongly predicted, while Agreeableness, Conscientiousness and Openness were moderately predicted. Although s t i l l s i g n i f i c a n t , s e l f reports were least e f f e c t i v e at predicting peer-rated l e v e l s of Neuroticism. Insert Table 8 about here A l l the correlations of the s e l f - r e p o r t scales with peer ratings of The Big Five and amount of time t a l k i n g are displayed i n Table 9. With the exception of neuroticism, the highest value i n each row i s with the corresponding rating c r i t e r i o n . Note, however, that FFI extraversion i s s i g n i f i c a n t l y predictive of a l l peer ratings, although 14 highest with ratings of extraversion and time t a l k i n g , as hypothesized. Insert Table 9 about here Table 10 marks the r e l a t i o n of sel f - r e p o r t s to corresponding peer ratings (the v a l i d i t i e s ) a f t e r c o n t r o l l i n g for p a r t i c i p a t i o n . As expected, a l l correlations were decreased by p a r t i a l i n g out such variance. V a l i d i t i e s for Extraversion and Neuroticism became nonsignificant. V a l i d i t i e s for Agreeableness and Openness, however, remained s i g n i f i c a n t at .20 and .21, respectively. Conscientiousness was marginally s i g n i f i c a n t , r = .13, p < .10. 15 Insert Table 10 about here Extraversion (E). Correlations of se l f - r e p o r t s with peer ratings for the Big Five E and the IAS-R PA s e l f - r e p o r t scales are both strongly influenced by time t a l k i n g (E: .41 to .10; PA: .54 to .20). As well, there appears to be minimal improvement i n prediction over time, but only for the o r i g i n a l data (.35, .28, and .40 re s p e c t i v e l y ) . The correlations controlled for t a l k i n g time do not show such improvement, and ac t u a l l y decline (.17, .05, and .09). Agreeableness (A). Correlations of sel f - r e p o r t s with peer ratings for the Big Five A and the IAS-R LM s e l f - r e p o r t scales are both moderately influenced by time t a l k i n g . In contrast to the other r e l a t i o n s of sel f - r e p o r t s with peer ratings, Agreeableness ratings are actually improved by c o n t r o l l i n g for time t a l k i n g (A: .23 to .28; LM: .12 to .18). Regarding the o r i g i n a l peer rating predictions of agreeableness, A correlated with se l f - r e p o r t s at .01, .25, and .27 respectively, showing marked improvement. LM also showed some improvement, at .07, .05, and .18 over the three ratings. C o n t r o l l i n g for time t a l k i n g had a somewhat negative e f f e c t upon these correlations (A: -.07, .24, .20; LM: .01, .02, .14). 16 Conscientiousness (C) The r e l a t i o n of se l f - r e p o r t s with peer ratings of C was also somewhat affected by time t a l k i n g , decreasing from .36 to .28. Regarding prediction over time, the o r i g i n a l correlations were moderately strong at each ra t i n g time and showed no tendency toward improvement (.33, .25, .29). The controlled correlations a c t u a l l y decreased over time, from .23 to .11 to .13. Neuroticism (N) The observed v a l i d i t y of N, modest to begin with (.18), may an a r t i f a c t of the confounding of the l a t t e r with time t a l k i n g — t h e p a r t i a l c o r r e l a t i o n was, at -.01, nonexistent. Regarding prediction over time, both the o r i g i n a l and the controlled correlations showed a minimal decrease i n prediction over time (.25, .10, and .18; .18, -.09, and .04 r e s p e c t i v e l y ) . Openness (0) The re l a t i o n s h i p of se l f - r e p o r t s with peer ratings of 0 was also somewhat affected by time t a l k i n g , decreasing from .36 to .21. Regarding the prediction of 0 over time, o r i g i n a l correlations increased s l i g h t l y (.27, .35 and .31 r e s p e c t i v e l y ) , while the correlations controlled for p a r t i c i p a t i o n increased markedly, from .05 to .12 to .21 across the three r a t i n g periods. Insert Table 11 about here 17 Regarding orthogonality of the Big Five (see Table 11), there were high l e v e l s of i n t e r c o r r e l a t i o n among most of the variables. At Week 1, a l l variables except N were highly intercorrelated, ranging from .55 (A with C) to .75 (E with 0). Intercorrelations of N with the other variables ranged from .09 (with C) to .20 (with 0). The average i n t e r c o r r e l a t i o n among the Big Five was .44. At Week 4, the Big Five were s l i g h t l y less i n t e r -correlated, with the average being .38. Intercorrelations of N with the other variables ranged from -.02 (with C) to .22 (with A and O). A l l other variables were moderately intercorrelated, ranging from .37 (A with E and A with C) to .68 (E with C). At Week 7, the int e r c o r r e l a t i o n s had increased n e g l i g i b l y over the previous r a t i n g period, with the average being .41. Intercorrelations of N with the other variables ranged from .09 (with O) to .19 (with E and C). A l l other variables were moderately to strongly correlated, ranging from .38 (A with C) to .75 (C with 0). These re s u l t s suggests that i n t e r c o r r e l a t i o n s among peer-rated Big Five scores remain moderately high even a f t e r several weeks' acquaintance. Insert Table 12 about here 18 Expert Ratings of Free Descriptions There were no s i g n i f i c a n t differences among expert raters on any of the Big Five dimensions (see Table 12). In addition, the i n t e r r a t e r r e l i a b i l i t i e s were a l l adequate (see Table 13). For the f i r s t pair of raters, r e l i a b i l i t i e s ranged from .74 for Openness to .88 for Agreeableness. The second- set of raters showed even higher agreement, ranging from .83 for Neuroticism to .92 for Extraversion. Insert Table 13 about here The i n t e r c o r r e l a t i o n s among the expert-rated Big Five are presented i n Table 14. I t i s worthwhile comparing these r e l a t i o n s to those among the peer-ratings ( i . e . , Table 5). Insert Table 14 about here Overall, the mean i n t e r c o r r e l a t i o n (absolute value) i s lower among the expert ratings (M = .32) than the peer ratings (M = .55). The sole exception i s a strong c o r r e l a t i o n remaining between extraversion and neuroticism (r = -.67). Degree of P a r t i c i p a t i o n and Observability of the Big  Five. I t has been suggested by Park and Judd (1989) that the ease of making personality judgments about a target increases as the target's l e v e l of e x t r a v e r s i o n — i n t h i s 19 instance operationalized as t a l k i n g time—increases. To investigate t h i s claim, participants were blocked on peer-rated time t a l k i n g . Table 15 displays the peer in t e r c o r r e l a t i o n s separately for the two groups. Insert Table 15 about here Among participants rated as low i n time t a l k i n g , the mean in t e r c o r r e l a t i o n among the expert-rated Big Five was .32; For high time t a l k i n g , the mean co r r e l a t i o n was only .25. Although not s i g n i f i c a n t , t h i s difference i s consistent with Park and Judd's hypothesis i n that p a r t i c i p a t i o n aids group members i n separating the Big Five. In the low pa r t i c i p a t i o n group, a high c o r r e l a t i o n , -.67, remains between extraversion and neuroticism. Experts tended to rate introverts who participated minimally as generally insecure, and extraverts who participated minimally as secure. This trend was also present i n the high p a r t i c i p a t i o n group, but the co r r e l a t i o n (r = -.31) was comparable to that i n s e l f - r e p o r t s . Insert Table 16 about here The convergence of sel f - r e p o r t s with expert ratings ( i . e . , the v a l i d i t i e s ) were also affected by l e v e l of p a r t i c i p a t i o n . As evident i n Table 16, the e f f e c t depends upon the dimension under consideration. The mean v a l i d i t y 20 increased from .19 i n the low p a r t i c i p a t i o n group to .24 i n the high p a r t i c i p a t i o n group. Again, although not s i g n i f i c a n t , the finding i s consistent with Park and Judd's hypothesis. Insert Table 17 about here Table 17 permits a comparison of the v a l i d i t i e s using peer ratings, ranked peer ratings, and expert ratings. Extraversion and Neuroticism were predicted about equally well whether the c r i t e r i o n was peer ratings or expert ratings. There were moderate differences, however, between ranked peer ratings and the other c r i t e r i a on t h i s measure. Extraversion was less well-predicted by ranked ratings, whereas Neuroticism was better predicted. Ranked scores on the other dimensions did not d i f f e r overmuch from the other two c r i t e r i a . However, Conscientiousness scores were somewhat more highly correlated with normal peer ratings than with expert ratings, and, conversely, NEO-FFI Agreeableness and Openness scores were more strongly predictive of the relevant expert ratings than of peer ratings. The correlations between peer ratings and expert ratings ranged from moderately to strongly related. They were, for E, A, C, N, and 0, respectively, .81, .54, .59, .73, and .51. 21 Discussion Overall, the re s u l t s of the present study were encouraging. They suggested that the NEO-FFI i s a r e l a t i v e l y v a l i d predictor of peer-ratings of the Big Five personality t r a i t s i n a small-group-interaction s e t t i n g . Moreover, the v a l i d i t i e s increased s i g n i f i c a n t l y with acquaintanceship. This finding confirms l o n g i t u d i n a l l y the r e s u l t found previously with between-subjects indicators of acquaintanceship. Nevertheless, the pattern of v a l i d i t i e s appeared to be influenced by the context—the discussion group s e t t i n g . As detailed below, the a b i l i t y of group members to discern the Big Five t r a i t s was constrained—but constrained i n a coherent fashion. Peer Ratings as the C r i t e r i o n On average, Big Five s e l f - r e p o r t s correlated .31 with f i n a l peer ratings. While t h i s figure i s somewhat less than the mean v a l i d i t y reported by Costa and McCrae (1989) i n the manual (M = .46), i t should be noted that the t h e i r figures resulted from the use of validimax factor analysis, a procedure intended to maximize such cor r e l a t i o n s . By contrast, the factors i n the present study are unrotated. A second major difference involves the peer-rating procedure used. Costa and McCrae t y p i c a l l y require targets and raters to complete the same measure, the 60-item NEO-FFI that contains 12 items f o r each factor. We asked peers to make a single global r a t i n g for each Big Five dimension. The f a c t that a moderate association was s t i l l obtained is, highly encouraging. A s t r i k i n g though unfortunate aspect of the peer ratings i s the confounding e f f e c t that p a r t i c i p a t i o n rate had upon a l l peer-rated t r a i t s . P a r t i c i p a t i o n was correlated with a l l the Big Five ratings to varying degrees, a confounding that affected both i n t e r - r a t e r agreement and the observed v a l i d i t i e s . We consider t h i s issue i n d e t a i l below. Extraversion (E) Of the Big Five, Extraversion showed the highest i n t e r - r a t e r agreement and the highest v a l i d i t i e s . Note also that peer-rated extraversion was so strongly correlated with time t a l k i n g (.89) that they could be considered synonymous. This r e s u l t i s not surprising considering the descriptor words used to describe extraversion to the raters included the term "vocal". The highest v a l i d i t y observed i n f i n a l peer ratings was the prediction of extraversion from the IAS-R dominance scale (r = .51). Again, a look at the rating d e s c r i p t o r — Assertiveness (vocal, dominant)—suggests an ideal correspondence between predictor and c r i t e r i o n . Indeed, as noted i n the Method section, the peer r a t i n g i s better described as Surgency. Note that the su p e r i o r i t y of the IAS-R dominance scale over the FFI extraversion scale cannot be attributed to a longer scale. Indeed, the predictor we 23 used (the PA octant) comprises only eight adjectives; The FFI extraversion scale contains 12 questionnaire statements. F i n a l l y , there appears to be minimal improvement i n prediction of extraversion over time. I t seems that a f t e r a single meeting, group members were able to grasp with precision each others' l e v e l of extraversion. Recent studies suggest that the high observability of extraversion derives from i t s v i s u a l and auditory nature (Berry, 1990; Borkenau & L i e b l e r , i n press). Agreeableness (A). In t h i s case the FFI Agreeableness scale outperformed the IAS-R nurturance scale ( v a l i d i t i e s were .27 vs. .18). This advantage makes sense given the r e l a t i v e orientation of the two dimensions on the circumplex (see Wiggins & Trapnell, i n press). In contrast to the other re l a t i o n s of s e l f - r e p o r t s with peer ratings, Agreeableness ratings are actually improved by c o n t r o l l i n g for time t a l k i n g . The dimension was described to judges as prosocial orientation, being cooperative and l i k a b l e . Unlike judgments of extraversion, peers seemed able to d i s t i n g u i s h between quietness and group cooperation, and they appeared to base ratings of likableness on the actual content of the group interactions. Across time the v a l i d i t i e s were .01, .25, and .27 respectively, showing marked improvement. The IAS-R nurturance subscale also showed some improvement over the three ratings. Conscientiousness fC) The confounding of peer ratings of conscientiousness with time t a l k i n g decreased from an 24 o v e r a l l c o r r e l a t i o n of .36 to .28. The dimension was described to peers as work orientation, being deliberate and organized. The only way individuals could demonstrate t h e i r conscientiousness was to a c t i v e l y contribute to group discussion. However, since there was s t i l l a moderately strong r e l a t i o n s h i p between sel f - r e p o r t s and peer ratings a f t e r p a r t i c i p a t i o n was controlled, the ratings seem to have picked up on other aspects of conscientiousness that are s u f f i c i e n t l y unrelated to verbal p a r t i c i p a t i o n . Regarding prediction over time, the o r i g i n a l correlations were moderately strong at each r a t i n g time and showed no tendency toward improvement. The controlled correlations decreased over time. This suggests that there was actually a decrease i n a b i l i t y to predict Conscientiousness over time, or that p a r t i c i p a t i o n i s a relevant component of conscientiousness i n such a s e t t i n g . Neuroticism (N) Because i t i s not e a s i l y observed, neuroticism i s generally the most d i f f i c u l t Big Five t r a i t to rate (Watson, 1989). Nevertheless, the v a l i d i t y evidence was reasonable. The r e l a t i o n s h i p of s e l f - r e p o r t s with peer ratings of N was strongly affected by time t a l k i n g . An i n i t i a l c o r r e l a t i o n of .21 was due e n t i r e l y to the e f f e c t s of p a r t i c i p a t i o n , because the p a r t i a l c o r r e l a t i o n was, at -.01, n e g l i g i b l e . Peers seemed to have been unable to d i s t i n g u i s h between genuine t r a i t anxiety and s i t u a t i o n a l nervousness or i n s e c u r i t y . 25 Regarding prediction.over time, both the o r i g i n a l and the controlled correlations showed a minimal decrease i n prediction over time. Contrary to the hypothesis, peers' a b i l i t y to predict self-reported l e v e l s of Neuroticism over time did not improve. These p a r t i c u l a r r e s u l t s r a i s e an int e r e s t i n g question? What exactly was i t about the group s i t u a t i o n that rendered peers unable to rate t r a i t neuroticism? There are at least two possible explanations. F i r s t , i t i s already known that anxiety i s d i f f i c u l t f or peers to discern, regardless of the extent of the acquaintance (John, 1990). A l t e r n a t i v e l y , due to the negative s o c i a l connotations of anxiety and insecurity, participants may be attempting to hide t h e i r neuroticism from t h e i r peers. Openness (O). The r e l a t i o n of sel f - r e p o r t s with peer ratings of O was also somewhat affected by time t a l k i n g , decreasing from .36 to .21. Recall that openness was described to peers as i n t e l l e c t , being o r i g i n a l and clever. I t i s reasonable to suppose that peers may have perceived the more t a l k a t i v e group members as more i n t e l l i g e n t . Again i t i s d i f f i c u l t to manifest one's philosophical and absorption tendencies without p a r t i c i p a t i n g . Summary. Even a f t e r p a r t i a l i n g out the ef f e c t s of p a r t i c i p a t i o n , s e l f - r e p o r t s were s i g n i f i c a n t l y predictive of peer-rated agreeableness, conscientiousness, and openness. In addition, since time spent t a l k i n g i s highly related to 26 both self-reported and peer-rated extraversion, i t was expected that l i t t l e would remain of t h i s r e l a t i o n s h i p a f t e r time t a l k i n g was controlled, and t h i s expectation was borne out. However, even with the e f f e c t s of verbal p a r t i c i p a t i o n p a r t i a l e d out, peer-rated neuroticism continued to be unrelated to self-reported Neuroticism and, moreover, moderately negatively correlated with FFI Extraversion. This r e s u l t i s perhaps unsurprising considering the work of Watson (1989) who found neuroticism to be the least interpersonally obvious of the Big Five t r a i t s . I t i s in t e r e s t i n g , however, to note the v a l i d i t i e s based on spouse ratings reported i n Costa and McCrae (1989): Of the Big Five, Neuroticism had the highest v a l i d i t y ; I t has the lowest v a l i d i t y i n the present study. Apparently, neuroticism cannot be hidden from one's spouse. Expert Rated Free Descriptions as the C r i t e r i o n Although expert ratings of free descriptions were s t i l l based on peers' subjective impressions, we hoped that experts might be less subject to cert a i n rating biases For example, expert raters could draw t h e i r own, possibly more objective, opinions from the behaviors described by peers i n t h e i r free descriptions. For example, they could overlook inferences of t r a i t s from i r r e l e v a n t behavior. 27 Overall, we found that comparable v a l i d i t i e s for the peer and expert ratings. Nonetheless, expert ratings did show an improvement over peer ratings i n some respects. The orthogonality of the Big Five was improved. Although they s t i l l confounded extraversion and neuroticism, they were c l e a r l y able to di s t i n g u i s h between extraversion and standing on the other three dimensions. Therefore, the v a l i d i t i e s for peer ratings are more i n f l a t e d by the confounding of a l l f i v e factors than are the v a l i d i t i e s for expert ratings. Hence the comparable mean v a l i d i t y for expert ratings represents more unique variance accounted f o r . P a r t i c i p a t i o n and Observability of the Big Five. The examination of the expert-rated Big Five blocked on p a r t i c i p a t i o n proved i n t e r e s t i n g . When targets were judged as high i n time t a l k i n g within the group, expert raters were better able to di s t i n g u i s h among the Big Five t r a i t s . What's Special About Discussion Groups? Is there anything unique about the study of personality i n discussion groups? Our conclusions follow from three central questions. F i r s t , does the discussion group a l t e r the manifestation of t r a i t s ? New facets of personality may emerge from t h i s s i t u a t i o n . If so, the observed d e f i c i e n c i e s i n v a l i d i t i e s may suggest that the s e l f - r e p o r t predictors are 28 inappropriate whereas the peer-ratings represent emergent ind i v i d u a l differences. We dispute t h i s view. The v a l i d a t i o n of the s e l f - r e p o r t predictors was extensive. For example, the FFI was validated on a variety of c r i t e r i a . Spouses see the target i n a wide variety of situations (including group situations) over a longer period of time, whereas the peers i n the present study are lim i t e d to one s i t u a t i o n . Second, does the discussion group a l t e r the perception of personality? For example, impression management might a l t e r the manifestation of selected t r a i t s (Paulhus, 1986). If so, the observed data would underestimate both v a l i d i t y and orthogonality of the Big Five. Unfortunately we have no d i r e c t evidence on t h i s issue. Third, does the group induce moderator effects? Clearly, extraversion was a moderator for the other variables. This e f f e c t was demonstrated by blocking subjects on time t a l k i n g Apparently when extraversion was low, the other t r a i t s were st e r e o t y p i c a l l y i n f e r r e d . By contrast, the high t a l k time group showed high Big Five orthogonality and v a l i d i t y . Limitations and Recommendations The problems raised throughout the discussion section have suggested some recommendations for improvements i n the design of future studies. 29 F i r s t , r e c a l l that peer ratings of the Big Five were made on the basis of one target adjective along with two subdescriptor adjectives for each of the f i v e dimensions. For example, Openness was described as " I n t e l l e c t ( o r i g i n a l , c l e v e r ) " . Although the present v a l i d i t i e s were reasonable, they were c l e a r l y underestimated by the use of a single r a t i n g . Other studies that have used peer ratings (e.g., McCrae & Costa, 1987) have invariably used multiple ratings. Apart from the obvious benefits for r e l i a b i l i t y , aggregation can increase the l i k e l i h o o d of a strong correspondence between se l f - r e p o r t s and peer ratings by sampling from as wide a range of domain behaviors as possible. However, there i s also a p r a c t i c a l issue to consider: Participants are r a t i n g up to f i v e other individuals plus themselves on each dimension, as well as being asked to make extensive comments on the group interactions on a weekly basis. The possible fatigue due to increased demand for p a r t i c i p a n t s ' time and e f f o r t has to be borne i n mind. To r e i t e r a t e , i n future studies, several sets of t r a i t descriptors should be used i n the ratings. The o v e r a l l peer ra t i n g of each dimension would be the average across the three t r a i t s . For example, for the Big Five t r a i t of Agreeableness, participants should rate the members of t h e i r group on three bipolar scales; The anchors of the scales would be "kind" and "ruthless", "unmanipulative" and "manipulative", and "sympathetic" and "cold-hearted". 30 The second recommendation arises from a frequent complaint from the p a r t i c i p a n t s — u s e of the 10-point r a t i n g scale with no t i e s allowed. In groups with six or more members, t h i s necessarily r e s u l t s i n one or two persons being rated below (or above) the midpoint of any p a r t i c u l a r t r a i t . Several individuals indicated that, i n some cases, they f e l t that a l l persons were actually below (above) the midpoint, but that the scale was not s u f f i c i e n t l y long enough to convey t h i s pattern. Therefore, i n future studies participants should be allowed to rate t h e i r group members on a 15-point scale. This change should also serve to increase the opportunity for participants to make more f i n e -grained d i s t i n c t i o n s among t h e i r peers and themselves. As i n the present study, the participants should be required to rate the group members without giving t i e d scores, i n order to encourage them to think about and report on more subtle differences among peers. Perhaps most important i s the lack of orthogonality of the peer ratings. In future studies, the tasks must be designed to allow groups members to discriminate more e a s i l y the Big Five t r a i t s . For example, conscientiousness may be better revealed i f members are require to do homework, which they must bring to the next meeting. Hence they w i l l be judged more on what they produce than on how a c t i v e l y they p a r t i c i p a t e . S i m i l a r l y , appropriate tasks must be designed to bring out i n d i v i d u a l differences i n neuroticism. References Bales, R.F. (1970). Personality and interpersonal behavior. New York: Holt, Rinehart, & Winston. Berry, D.S. (1990). Taking people at face value: Evidence for the kernel of truth hypothesis. Social Cognition, 8 , 343-361. Borkenau, P., & L i e b l e r , A. (in press) T r a i t inferences: Sources of v a l i d i t y at zero acquaintance. Journal of Personality and Social Psychology. Colvin, C.R., & Funder, D.C. (1991). Predicting personality and behavior: A boundary on the acquaintanceship e f f e c t . Journal of Personality and Social Psychology, 60, 884-894. Costa, P.T., J r . , & McCrae, R.R. (1988). Personality i n adulthood: A six-year longitudinal study of s e l f - r e p o r t s arid spouse ratings on the NEO Personality Inventory. Journal of Personality and Social Psychology, 54, 853-863. Costa, P.T., J r . , & McCrae, R.R. (1989). The NEO Personality  Inventory manual. Odessa, FL: Psychological Assessment Resources. Funder, D.C, & Colvin, CR. (1988). Friends and strangers: Acquaintanceship, agreement, and the accuracy of personality judgment. Journal of Personality and Social  Psychology. 55, 149-158. John, O.P. (1990). The "Big Five" factor taxonomy: Dimensions of personality i n the natural language and i n 32 questionnaires. In L.A. Pervin (Ed.), Handbook of  personality; Theory and research (pp. 66-100). NY: G u i l f o r d Press McCrae, R.R., & Costa, P.T., J r . (1983). Joint factors i n s e l f - r e p o r t s and ratings: Neuroticism, extraversion, and openness to experience. Personality and Individual  Differences, 4, 245-255. McCrae, R.R., & Costa, P.T., J r . (1987). Validation of the f i v e - f a c t o r model of personality across instruments and observers. Journal of Personality and Social Psychology.  52. 81-90. Newcomb, T.M. (1961). The acquaintance process. New York: Holt, Rinehart, and Winston. Norman, W.T. (1963). Toward an adequate taxonomy of personality a t t r i b u t e s : Replicated factor structure i n peer nomination personality ratings. Journal of Abnormal and Social Psychology, 66. 574-583. Park, B., & Judd, CM. (1989). Agreement on i n i t i a l impressions: Differences due to perceivers, t r a i t dimensions, and target behaviors. Journal of Personality and Social Psychology, 56, 493-505. Passirii, F.T., & Norman, W.T. (1966). A universal conception of personality structure? Journal of Personality and Socia1 Psycho1ogy. 4, 44-49. Paulhus, D.L. (1986). Self-deception and impression management in t e s t responses. In A. Angleitner & J.S. 33 Wiggins (Eds), Self-report v i a questionnaire. New York: Springer-Verlag. Paunonen, S.V. (1989). Consensus i n personality judgments: Moderating e f f e c t s of target-rater acquaintanceship and behavior obs e r v a b i l i t y . Journal of Personality and Social Psychology. 56, 823-833. Trapnell, P.D. (1990). Item factor analysis of the Five Factor Inventory. Unpublished data, University of B r i t i s h Columbia. Trapnell, P.D. (1991). Factor analysis of the NEO-FFI with  s e l f - r a t i n g s . Unpublished manuscript, University of B r i t i s h Columbia. Watson, D. (1989). Strangers' ratings of the f i v e robust personality factors: Evidence of a surprising convergence with s e l f - r e p o r t . Journal of Personality and Social  Psychology. 57, 120-128. Wiggins, J.S. (1979). A psychological taxonomy of t r a i t -d escriptive terms: The interpersonal domain. Journal of  Personality and Social Psychology, 37, 395-412. Wiggins, J.S., & Trapnell, P.D. (in press). Personality structure: The return of the Big Five. In S.R. Briggs, R. Hogan, & W.H. Jones (Eds.), Handbook of personality  psychology. Orlando, FL: Academic Press. Wiggins, J.S., Trapnell, P.D., & P h i l l i p s , N. (1988). Psychometric and geometric c h a r a c t e r i s t i c s of the revised Interpersonal Adjective Scales (IAS-R). Multivariate Behavioral Research, 23, 517-530. 34 Table 1 Peer Rating Sheet Your Name Group No. Report on Discussion No.l: Early Memories F i r s t Impressions Rate your f i r s t impressions of the other members plus yourself on the dimensions below. Make a l l ratings on a scale of 1-10 with no t i e s allowed. Very Low Moderate Very High 1 2 3 4 5 6 7 8 9 10 F i r s t I n i t i a l You of Member Assertiveness (vocal, dominant) Prosocial Oriented _ (cooperative, likeable) Work Orientation (deliberate, organized) Emotionality _ (emotional, anxious) I n t e l l e c t ( o r i g i n a l , clever) Time Talking Effectiveness _ (contributed to group goals) Self-Admiration Insecurity 35 Table 2 Descriptive S t a t i s t i c s f or Self-Report Scales Variable N Mean Standard R e l i a b i l i t y Deviation (alpha) Big Five T r a i t s Extraversion 89 30.44 6.67 .80 Agreeableness 89 32.04 6.69 .79 Conscientiousness 89 30.69 6.90 .83 Neuroticism 89 24.18 9.27 .88 Openness 89 30.67 6.16 -70 Circumplex T r a i t s PA 83 40.82 8.34 .81 LM 84 40.72 7.99 .88 Note. PA = IAS-R dominant/assertive; LM = IAS-R warm/agreeable. 36 Table 3 Intercorrelations of Self Reports E A C N O PA Extraversion — Agreeable .29 — Conscientious .11 .15 — Neuroticism -.28 -.14 -.37 — Openness .18 .02 .03 -.19 — PA .33 -.30 .39 -.43 .14 — LM .18 .52 .06 .10 .17 -.30 Note. Correlations exceeding .19 are s i g n i f i c a n t at p=.05; correlations exceeding .25 are s i g n i f i c a n t at p=.01; correlations exceeding .32 are s i g n i f i c a n t at p=.001. PA = IAS-R dominant/assertive; LM = IAS-R warm/agreeable. 37 Table 4 Descriptive S t a t i s t i c s for Peer Ratings Variable N Mean Standard Deviation Extraversion Weekl 81 5.99 1.82 Week4 84 6.57 1.58 Week7 88 6.52 1.79 Combined 89 6.34 1.61 Agreeableness Weekl 81 6.44 1.29 Week4 84 6.80 1.24 Week7 88 6.75 1.33 Combined 89 6.66 1.07 Conscientiousness Weekl 81 6.17 1.32 Week4 84 6.47 1.29 Week7 88 6.58 1.34 Combined 89 6.45 1.10 Neuroticism Weekl 81 4.81 1.70 Week4 84 4.85 1.53 Week7 88 5.16 1.72 Combined 89 4.89 1.49 Openness Weekl 81 6.24 1.41 Week4 84 6.35 1.32 Week7 88 6.41 1.36 Combined 89 6.36 1.14 Time Talking Weekl 81 6.22 1.87 Week4 84 6.67 1.73 Week7 88 6.67 1.79 Combined 89 6.58 1.67 Note. Numerals 1,4,7 ref e r to weeks during which peers completed Big Five ratings. A l l ratings were on 10-point L i k e r t scales. 38 Table 5 Intercorrelations of F i n a l Peer Ratings Extraverted Agreeable Conscientious Neurotic Open Time Talking Extraverted Agreeable .54 Conscientious .57 .38 Neurotic -.67 -.44 -.42 Open .66 .55 .75 -.55 Time Talking .89 .52 .54 -.70 .58 Note. A l l correlations are s i g n i f i c a n t at p=.001. 39 Table 6 Intercorrelations of F i n a l Peer Ratings Controlled for Time Talking Extraversion Agreeable Conscientious Neurotic Openness Extraversion Agreeable .19 Conscientious .25 .14 Neuroticism -.06 -.06 -.11 Openness .39 .35 .64 .01 Note. Correlations exceeding .19 are s i g n i f i c a n t at p=.05; correlations exceeding .25 are s i g n i f i c a n t at p=.0l; correlations exceeding .32 are s i g n i f i c a n t at p=.001. 40 Table 7 Average Inter-rater Agreement Within Groups E A C N 0 Average Week 1 .68 .29 .33 .43 .30 .41 Week 4 .58 .30 .34 .28 .21 .34 Week 7 .56 .30 .19 .31 .23 .32 Mean .61 .30 .29 .34 .25 .36 Note. Each entry i s the mean i n t e r c o r r e l a t i o n among 2 to 6 ratings of the same in d i v i d u a l (depending on the size of the group and the number i n attendance each week). Correlations exceeding .19 are s i g n i f i c a n t at p=.05; correlations exceeding .25 are s i g n i f i c a n t at p=.01; correlations exceeding .32 are s i g n i f i c a n t at p=.001. E = extraversion; A = agreeableness; C = conscientiousness; N = neuroticism; O = openness 41 Table 8 Prediction of Corresponding Peer Ratings from  Self-Reports Over Time E A C N 0 PA LM MEAN Week 1 35 .01 .33 .25 .27 .45 .07 25 Week 4 .28 25 .25 10 35 52 .05 26 Week 7 40 27 29 .18 31 51 .18 31 Overall .41 23 .36 21 36 54 .12 32 Note. Correlations exceeding .19 are s i g n i f i c a n t at p=.05; correlations exceeding .25 are s i g n i f i c a n t at p=.01; correlations exceeding .32 are s i g n i f i c a n t at p=.001. The o v e r a l l index refers to the sum of three ratings (Weeks 1, 4, and 7). E = NEO-FFI Extraversion; A = NEO-FFI Agreeableness; C = NEO-FFI Conscientiousness; N = NEO-FFI Neuroticism; O = NEO-FFI Openness to Experience; PA and LM = IAS-R subscales corresponding to dominance and warmth. Table 9 Prediction of F i n a l Peer Ratings from Self-Reports Peer Ratings Self-Reports E A C N 0 TT Big Five T r a i t s Extraversion .40 .37 .22 -.34 .21 .41 Agreeableness -.02 .27 .27 .01 .10 .10 Conscientiousness .27 .06 .29 -.17 .20 . 37 Neuroticism -.33 -.12 -.20 .18 -.14 -.32 Openness .31 .22 .28 -.31 .31 .28 Circumplex T r a i t s PA .51 .16 .10 -.37 .15 .51 LM -.06 .18 .25 -.02 .16 .05 Note. Correlations exceeding .19 are s i g n i f i c a n t at p=.0 correlations exceeding .25 are s i g n i f i c a n t at p=.01; correlations exceeding .32 are s i g n i f i c a n t at p=.001. PA = IAS-R dominant/assertive; LM = IAS-R warm/agreeable 43 Table 10 Prediction of Peer Ratings from Corresponding Self-Reports After  Controlling for Time Talking Self-Report T r a i t s Big Five Circumplex E A C N 0 PA LM MEAN Week 1 .17 -.07 .23 .18 .05 .18 .01 11 Week 4 05 .24 .11 -.09 .12 .18 .02 . 09 Week 7 09 .20 .13 .04 .21 .15 .14 14 Overall 10 28 28 -.01 21 20 .18 .18 Note. Correlations exceeding .19 are s i g n i f i c a n t at p=.05; correlations exceeding .25 are s i g n i f i c a n t at p=.01; correlations exceeding .32 are s i g n i f i c a n t at p=.001. Numerals 1,4,7 refe r to weeks during which peers rated each other on the Big Five. Overall index refers to the sum of three ratings (Weeks 1, 4, and 7). E = NEO-FFI Extraversion; A = NEO-FFI Agreeableness; C = NEO-FFI Conscientiousness; N = NEO-FFI Neuroticism; O = NEO-FFI Openness to Experience; PA and LM = IAS-R subscales corresponding to dominance and nurturance. Table 11 Intercorrelations of Peer Ratings Over Time Week 1 E A C N 0 E A .61 C .67 .55 — Mean=.44 N .17 .14 .09 O .75 .64 .60 .20 Week 4 E A C N O E A .37 C .68 .37 — Mean=.38 N .21 .22 -.02 O .60 .47 .64 .22 Week 7 E A C N O E A .54 C .57 .38 — Mean=.41 N .19 .15 .19 O .66 .55 .75 .09 Note. Correlations exceeding .19 are s i g n i f i c a n t at p=.05; correlations exceeding .25 are s i g n i f i c a n t at p=.01; correlations exceeding .32 are s i g n i f i c a n t at p=.001. E = extraversion; A = agreeableness; C = conscientiousness N = neuroticism; O = openness. 45 Table 12 Mean Ratings of Free Descriptions by Expert Raters Rater E A C S O Mean Groups 1-8 Rater 1 4.30 4.63 4.41 4.35 4.33 4.40 (1.32) (-77) (-57) (-69) (.54) Rater 2 4.33 4.58 4.44 4.10 4.30 4.35 (.81) (-64) (.62) (-71) (.46) Groups 9-16 Rater 3 4.50 4.90 4.48 4.29 4.63 4.56 (1.36) (.79) (-69) (.83) (.78) Rater 4 4.41 4.96 4.56 4.20 4.55 4.54 (1.39) (.81) (-82) (1.03) (.93) Mean 4.39 4.77 4.47 4.24 4.45 4.46 Note. No differences among raters are s i g n i f i c a n t . A l l means are on seven-point L i k e r t scales; numbers i n parentheses are standard deviations. Groups 1-8 : n = 43; groups 9-16: n = 45. E = extraversion; A = agreeableness; C = conscientiousness; S = emotional s t a b i l i t y ; 0 = openness. 46 Table 13 Inter-Rater Agreements of the Expert Ratings Extraversion Agreeable Conscientious Neuroticism Openness Rater 1 & 2 .84 .88 .86 .76 .74 3 & 4 .93 .85 .87 .83 .86 Note. A l l correlations are s i g n i f i c a n t at p < .001. No differences between sets of raters are s i g n i f i c a n t . 47 Table 14 Intercorrelations of Big Five Expert Ratings E A C N E — A . 13 — C .32 .13 — N -.67 -.16 -.32 — 0 .31 .44 .39 -.35 o Note. E = extraversion; A = agreeableness; C = conscientiousness; N = neuroticism; 0 = openness. Correlations exceeding .25 are s i g n i f i c a n t at p=.01; correlations exceeding .32 are s i g n i f i c a n t at p=.001. Table 15 Intercorrelations of Expert Ratings Blocked into Low and  High Verbal P a r t i c i p a t i o n Low (n=43) E A C N E — A .13 — C . 27 . 26 N -.67 -.21 -.24 — 0 .23 .49 .48 -.21 High (n=45) E A C N E — A .34 — C .03 .09 N -.31 -.19 -.18 — 0 .28 .43 .29 -.39 Note. E = extraversion; A = agreeableness; C = conscientiousness; N = neuroticism; 0 = openness. Correlations exceeding .26 are s i g n i f i c a n t at p=.05; correlations exceeding .34 are s i g n i f i c a n t at p=.01; correlations exceeding .43 are s i g n i f i c a n t at p=.001. 49 Table 16 Predictions of Expert Ratings from Self-Reports Blocked  on Verbal P a r t i c i p a t i o n E A C N 0 Mean Low (n=43) .16 .47 .13 0 .19 .19 High (N=45) .29 .24 .08 .16 .44 .24 Note. E = extraversion; A = agreeableness; C = conscientiousness; N = neuroticism; 0 = openness. Correlations exceeding .26 are s i g n i f i c a n t at p=.05; correlations exceeding .34 are s i g n i f i c a n t at p=.01; correlations exceeding .43 are s i g n i f i c a n t at p=.001. Table 17 Relative V a l i d i t i e s of Peer Ratings, Ranked Peer Ratings,  and Expert Ratings of Free Descriptions E A C N 0 Mean Peer Ratings .40 .27 .29 .18 .31 .29 Ranked Peer .33 .29 .25 .27 .28 .28 Ratings Free Descriptions .38 .32 .20 .19 .37 .29 Mean .39 .30 .25 .19 .34 Note. E = extraversion; A = agreeableness; C = conscientiousness; N = neuroticism; 0 = openness. A l l correlations s i g n i f i c a n t at p < .05. 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
https://iiif.library.ubc.ca/presentation/dsp.831.1-0098680/manifest

Comment

Related Items