This article was downloaded by: [York University Libraries] On: 18 November 2014, At: 20:13 Publisher: Routledge Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK

The Journal of Psychology: Interdisciplinary and Applied Publication details, including instructions for authors and subscription information: http://www.tandfonline.com/loi/vjrl20

The Relationship of Personality Traits as Measured in the Questionnaire Medium and by Self-Ratings a

Charles A. Bukdsal Jr. & Steven A. Schwartz

a

a

Wichita State University , USA Published online: 02 Jul 2010.

To cite this article: Charles A. Bukdsal Jr. & Steven A. Schwartz (1975) The Relationship of Personality Traits as Measured in the Questionnaire Medium and by Self-Ratings, The Journal of Psychology: Interdisciplinary and Applied, 91:2, 173-182, DOI: 10.1080/00223980.1975.9923939 To link to this article: http://dx.doi.org/10.1080/00223980.1975.9923939

PLEASE SCROLL DOWN FOR ARTICLE Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”) contained in the publications on our platform. However, Taylor & Francis, our agents, and our licensors make no representations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose of the Content. Any opinions and views expressed in this publication are the opinions and views of the authors, and are not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be relied upon and should be independently verified with primary sources of information. Taylor and Francis shall not be liable for any losses, actions, claims, proceedings, demands, costs, expenses, damages, and other liabilities whatsoever or howsoever caused arising directly or indirectly in connection with, in relation to or arising out of the use of the Content.

Downloaded by [York University Libraries] at 20:13 18 November 2014

This article may be used for research, teaching, and private study purposes. Any substantial or systematic reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any form to anyone is expressly forbidden. Terms & Conditions of access and use can be found at http://www.tandfonline.com/page/terms-and-conditions

Published as a separate and in The Journal oj Psychology, 1975, 91, 173-182.

THE RELATIONSHIP OF PERSONALITY TRAITS AS MEASURED IN THE QUESTIONNAIRE MEDIUM AND BY SELF-RATINGS* Wichita State University CHARLES

A. BURDSAL, JR., AND STEVEN A. SCHWARTZ

Downloaded by [York University Libraries] at 20:13 18 November 2014

SUMMARY

The general purpose of the study was to determine (a) whether individuals tend to have an accurate perception of their personality, and (b) the manner in which an individual's personality (as measured by a personality questionnaire) affects self-ratings on the same traits. The Clinical Analysis Questionnaire (CAQ), Part 1, was administered to 98 male and female undergraduates. One week later, 5s were asked to self-rate themselves on the 16 personality traits measures by the CAQ, using a modified version of the I6PF Test Profile. High correlations between measured and self-ratings indicated that the instruments were measuring similar, but not identical, traits. Following a factor analysis of the intercorrelations, 12 factors were identified, representing the interrelationship between perception of self and measured personality traits.

A.

INTRODUCTION

In the last decade, relatively few reported studies have dealt with comparisons of personality trait scores obtained from 'standardized tests versus those obtained from direct self-ratings on the same traits. To the present time, however, some of the traits examined in this manner include "drive" (8), "anxiety" (1, 8, 10, 11), and "extraversion" (10, 14). The results obtained in these studies indicate that 5s tend to perceive themselves in a similar manner on both questionnaire and self-rating data in estimating broad, second-factors, such as anxiety and extraversion. However, the question remains whether similar results might be obtained in estimating primary source traits as well. One attempt to examine such relationships was that of Wetsel, as reported in Peterson (13). He administered the 16 Personality Factor Questionnaire [16PF (6)] to a group of undergraduates and obtained peer and parent *Received in the Editorial Office on June 19, 1975, and published immediately at Provincetown, Massachusetts. Copyright by The Journal Press.

173

Downloaded by [York University Libraries] at 20:13 18 November 2014

174

JOURNAL OF PSYCHOLOGY

ratings on each student on the same dimensions. The relationships between the sets of ratings were essentially nonsignificant for the primary factors, although some significance was found among the second-order factors. These results promoted the conclusion that broad group factors are generally more efficient and stable than narrow primary factors (13). Such a conclusion implies that peer-parent ratings, in fact, are the criteria for measuring personality dimensions. Questionnaire data, by definition, are a type of selfreport. It would thus appear that a better criterion for such an examination would be how the individual rated himself (on the basis of the description of the primary factors). Another issue often ignored involves assuming validity of the personality factors measured by a questionnaire. Once such an assumption has been made, one can then ask what the relative influence of existing personality traits is upon the S's perception of the same traits; i. e., what sort of individual sees himself in what manner? Pedersen (12) approached this question to an extent in reporting personality correlates of a high self-concept. For the male Ss, the correlates included emotional stability, cooperativeness, extraversion, self-acceptance, low manifest anxiety, and low neuroticism. For the females, traits included self-acceptance and low manifest anxiety. Such a combined approach of examining the correlation of self-ratings with measured ratings and examining the interrelation of the two sets of data may appear at first to be self-defeating. It would seem that if one were to find exceedingly high correlations (+.8 or better), one would then expect selfratings and measured personality traits to be unidimensional. Thus, there is the dual implication that (a) one could get as good an estimate of personality traits from self-rating as one could get from questionnaire data, and (b) selfrating variables would pattern with corresponding questionnaire variables in a factor analysis. It might be a more reasonable position to assume that if questionnairederived traits are valid, they will (a) show significant but not extremely high correlation with their corresponding self-rating traits, and (b) have similar but not exactly the same side correlations as do the measured traits. If this is the case, one might expect differential patterning of the self-rating variable with the personality traits as measured in the questionnaire medium. Thus, the present study proposed to examine two central issues: (a) Do individuals tend to have an accurate perception of themselves (using a personality questionnaire as the instrument for comparison)? (b) In what manner are measured personality traits related to self-ratings of the same traits?

Downloaded by [York University Libraries] at 20:13 18 November 2014

CHARLES A. BURDSAL, JR., AND STEVEN A. SCHWARTZ

175

B. METHOD The 98 Ss (42 males and 56 females) who served in the present study were volunteers solicited mainly from introductory psychology classes, but including about 15 from experimental psychology classes and a business administration class. The two instruments utilized were (a) Part 1 of the Clinical Analysis Questionnaire [CAQ (7)] and (b) a retyped version of the 16-PF Test Profile sheet. The CAQ is a two-part multiple choice test which yields results on 16 bipolar personality factor scales (Part 1) and 12 pathological scales (Part 2). For the present study only the information obtained from Part 1 (a shortened version of the 16PF) was included in the analysis. The 16PF Test Profile is a standard form for plotting and recording sten scores on the individual personality factors for each S. On the left side of the form is a series of "Low Score Descriptions" composed of various adjectives and phrases which are descriptive of certain personality attributes. Each group of "Low Score Descriptions" is followed by a series of 10 dots (for recording the converted sten scores). Following each is a "High Score Description," representing the opposite pole of the personality trait. The major differences between the original profile and the rewritten version were (a) the rewritten version was not identified as being the 16PF Test Profile, and (b) the psychological-psychiatric term for each factor (such as "sizothymia") was omitted. The CAQ and the 16PF Test Profile were administered on two separate occasions one week apart. Immediately prior to the administration of the 16PF Test Profile the Ss were presented the following instructions: On the sheet before you are 16 different groups of descriptive adjectives or phrases. For each individual item there is a "Low Score Description" (on the left) which is followed by a row of x's numbered from one to 10, and then a "High Score Description" (to the right). You are to read each pair of descriptive phrases and determine where you fall between them. For example, in the first item if you feel that you are an extremely reserved and aloof individual, you might choose a rating of 1 or 2. If, on the other hand, you feel yourself to be a bit warmhearted and outgoing, you might pick a rating of 7 or 8. Circle the one x which has the appropriate number above it for each pair of descriptions on this test. As it is noted on the sheet, a choice of 5 or 6 indicates an "average" response, indicating that you feel yourself to be right between the two descriptions.

Pearson product-moment correlation coefficients were calculated on all pairwise combinations of the 32 variables (16 from the CAQ, 16 correspond-

176

JOURNAL OF PSYCHOLOGY

TABLE 1 INTERCORRELATIONS BETWEEN THE MEASURED PERSONALITY TRAITS AND THE SELF-RATINGS (N 98)

=

Scale A

B C E

Downloaded by [York University Libraries] at 20:13 18 November 2014

F G H I L M N

0

QI Q2 Qa Q.

Correlation

p

.263 .141 .474 .403 .362 .665 .593 .131 .151 .206 -.121 .425 .268 .295 .286 .299

.01

NS .001 .001 .001 .001 .001

NS NS .05

NS .001 .01 .01 .01 .01

ing variables from the 16PF Test Profile). A factor analysis was then done as described below. C.

RESULTS AND DISCUSSION l

The intercorrelations between the CAQ traits and the self-ratings may be found in Table 1. Eigenvalues were computed for the correlation matrix. The Scree test (3) indicated 12 factors. An iterative principal axis solution was applied to the correlation matrix until the communalities stabilized in the third decimal place. A Kaiser Varimax Orthogonal Rotation (9) was applied to the resulting factor matrix followed by a Maxplane oblique rotation (5), 13 graphical oblique rotations (4), and finally a Maxplane cleanup rotation, yielding a 74.2 percent .10 hyperplane. The salient loadings for the 12 factors (absolute value greater than .25) may be found in Table 2. 1.

Self-Ratings vs. Questionnaire Measured Traits

The obtained correlations between the measured and self-report traits indicated a moderately high relationship between the two. However, since the mean correlation was moderate, +.302, one can assume that self-ratings and the questionnaire measures are not measuring entirely the same thing. In examining the nonsignificant relationships one must observe the differ1 Matrices such as the Ro ' Vrs' L, etc. may be obtained from Charles Burdsal at the address shown at the end of the article.

177

CHARLES A. BURDSAL, JR., AND STEVEN A. SCHWARTZ

TABLE 2 Vip OF QUESTIONNAIRE AND SELF-RATING VARIABLES

Scale

I

II

III

IV

Downloaded by [York University Libraries] at 20:13 18 November 2014

VIII

+95 +43

-39

-37 +34

+37

-52 +65

+49 -37

+60 +36

-63

+34

-45

-47 -35 +50 -60 +35

+28

+35 +46 -31 -52

+53 +85

+61 +28

-29 -32

-26

+54

+80

+87 -35 +45

+40

+58 +63 -60

M'

-100 +92

N'

-51

-67

-27

-40 +43

-59

+38

+49

-68 -29

-31

-95

Q2'

Qa' Q4'

-28

+66

L'

0' Q1 '

XII

+77

Q2

C' E' F' G' H' I'

XI

-30

+71

Q1

B'

x

-33

M N

Qa Q4 A'

IX

-75

I L

o

Factor VI VII

-30

A

B C E F G H

V

+58

+26

+26

-38

-60

Note: 'indicates self-rating; all values rounded to two demical places with the decimal assumed; alI values less than an absolute .25 omitted.

ences between the means of the self-report and the measured traits (see Table 3). For factor B (intelligence) the mean for the self-rating was significantly higher than the mean for the measured trait. The possibility of the intrusion of social desirability upon the self-rating seems quite likely. Such an explanation might also be applicable for scales L (trusting vs. suspicious) and I (tough-minded, self-reliant, realistic vs. tenderminded, sensitive, clinging). In both cases the self-rating mean was in the direction (significantly) of the socially desirable side of the factor. It is interesting to note that the one scale (N) whose t value was not

178

JOURNAL OF PSYCHOLOGY

TABLE 3 MEAN DIFFERENCES OF MEASURED AND SELF-REPORTED PERSONALlTY TRAITS THAT WERE NOT SIGNIFICANTLY CORRELATED

Scale

CAQ mean

Self-rating mean

B I L N

5.45 5.31 5.49 5.00

7.07 4.42 4.56 5.18

Downloaded by [York University Libraries] at 20:13 18 November 2014

Note: CAQ

p 7.04 3.27 3.57 .57

.001 .01 .001 NS

= Clinical Analysis Questionnaire.

significant and which did not appear to have social desirability involved was also found by Burdsal and Vaughn (2) to be the weakest psychometrically of all the scales on the 16PF. Thus, it seems reasonable to conclude that the lack of significant correlation of the self-rating with factor N is further evidence of psychometric imperfection of this scale. 2.

Factor Analysis of the Self-Rating and Questionnaire Measured Personality Traits The factor analysis lent further support to the proposition that measured personality traits (by questionnaire) and self-reports, while related, are not the same thing. As demonstrated by Table 2, the self-report variable patterned in some cases with the measured traits and in some cases did not. This, then, leads to the question of just what are the factors found in the analysis. In general, it appears that the factors represented the interrelationship between how one sees oneself and how one is as measured by questionnaire data. Factor I: Structural Stability vs. Structural Instability. This factor was dominated primarily by factor C (emotional maturity vs. emotional immaturity). Persons high on this factor were emotionally mature (C+) and saw They appeared to be both secure (0-) and themselves that way (C' relaxed (Q4-), were somewhat used to getting their way (E+), and had considerable self-control (Q3+). They had an accurate perception of both their emotional maturity (C'+) and aggressiveness (E'+). They liked to see themselves as somewhat outgoing (A'+) and liberal (Ql'+) though this might or might not be true. This end of the factor seemed in general to represent an overall structural stability. The opposite end of the factor seemed to describe an anxious, passive, self-deprecating individual. Factor II: Traditional vs. Liberal Values. Superego (G) and radicalism (Qi ) dominate this factor. Stronger superego strength coupled with conservatism of temperament, both upon the questionnaire (G+, Ql -) and in Qt'-). Both perceived and measured self-control the self-report (G'

+).

+,

Downloaded by [York University Libraries] at 20:13 18 November 2014

CHARLES A. BURDSAL, JR., AND STEVEN A. SCHWARTZ

179

(Qs, Qs') were also to some extent related to the factor, with high control (Qs+) going with stronger superego. It is interesting to note that the selfperception of control was notably higher than the measured trait. Persons high on this factor were thus conscientious (G+), traditional (Ql-), selfcontrolled (Qs+), and perceived themselves that way. Notably they tended to be a bit emotionally less stable (C-) and saw themselves as somewhat shy (H'-) and tender-minded (1'+). This pattern, when seen with the opposite end of the factor (disregards rules, liberal, low self-control, and viewing self that way; being of moderate emotional stability, while viewing oneself as being bold and self-reliant), appeared to be one of more traditional values, tempered by free thinking, and possibly resulting in some discomfort. This is indicated by lower ego strength and higher self-control vs. a newer set of values with a relaxed stability, perhaps to the point of looseness. Factor III: Surgent Boldness vs. Sober Timidity. This factor, dominated by Surgency (F) and Parmia (II), demonstrated major differences in selfperception of individuals who are impulsive (F +) and socially bold (H+) vs. those who are more sober (F-) and shy (H-). The factor indicated an accurate self-perception by the individual as to his impulsiveness and shyness. However, the impulsive individual saw himself as outgoing (A' +), trusting (L'-), self-assured (0'-), and a bit controlled (Qs'+). On the other hand, the shy, sober individual saw himself as being a bit reserved, suspicious, guilt-prone, and low in self-control. It is interesting to note that the selfperceptions do not tend to be related with the questionnaire measurement of of the traits (at least for A, L, 0, and Qs). It should be observed that, except for the lack of factors C and Q4 and the presence of factor F, the selfperceptions fit the usual 16PF anxiety pattern (C-, H-, L+, 0+, Qs-, Q4+)' Thus, one might conclude that the socially bold, impulsive individual sees himself as being less anxious than does the sober, shy individual. Factor IV: Imaginative Artlessness vs. Practical Shrewdness. Self-perception of level of intelligence marked this factor (B /). Individuals who measured to be imaginative (M+) and forthright (N-) tended to perceive their intelligence as being high (B' +) and saw themselves as being liberal (Q/+ ) , while practical individuals estimated their intelligence, at least in terms of abstract thinking, as being low and saw themselves as somewhat conservative. It appears logical that a shrewd person estimates his intelligence as being low whether or not that is the case. Factor V: Perceived Sensitive Imagination vs. Perceived Realistic Practicality. This factor was dominated by the self-perception variables. An individual who saw himself as sensitive (1'+) also saw himself as imaginative

Downloaded by [York University Libraries] at 20:13 18 November 2014

180

JOURNAL OF PSYCHOLOGY

(M' +), while the individual who, on the other hand, had a self-perception of practicality also saw himself as realistic. The person high on this factor (self-perceptions of sensitivity and imagination) tended to be reserved (A-) and actually somewhat sensitive (I+), but not necessarily to the degree of the self-perception. The low end of the factor involved realism and outgoingness as well as the perceptions of down-to-earth practicality. Factor VI: Shrewd Defenselessness vs. Naive Suspiciousness. This factor represented the degree to which an individual reported himself as being trusting (L'- ), naive (N'- ), passive (E'- ), and unstable (C'- ). The persons who did this (high on the factor) tended to measure as shrewd (N+), sensitive (I+), somewhat guilt-prone (0+), and actually a bit passive (E-). Essentially, an individual who was shrewd, sensitive, and somewhat guilt-prone appeared to wish to obtain maximum advantage from an appearance of an almost naive helplessness, possibly with the goal of giving himself the advantage by having other people make assumptions as to his sophistication. The person very low on this factor tended to be straightforward, dominant, and relaxed and wanted others to see him as fairly hardnosed in terms of being suspicious, shrewd, dominant, and stable. Factor VII: Intelligence. This factor was dominated by B on the 16PF. The brighter, in terms of abstract thinking, the individual appeared to be, the more he tended to be impulsive (F +) and somewhat forthright (N - ) , but perceived himself as a bit shrewd (N'+) and conservative (Ql'-). Factor VIII: Straightforward Realism vs. Shrewd Sensitivity. Individuals high on this factor tended to be forthright (N - ), realistic (I - ), and emotionally stable (C+). They tended to see themselves as impulsive (F' liberal (Qt' +), and controlled (Qa' On the other end of the factor, individuals tended to be shrewd, sensitive, and emotionally unstable and saw themselves as sober, conservative, and having low self-control. Factor IX: Perceived Level of Group Adherence. The level of the individual's perception of his group dependence (Q2') marked this factor. Individuals who perceived themselves as being joiners and sound followers tended also to measure on the questionnaire as being group dependent (Q2-), but not nearly to the level of their self-perception. They also saw themselves as being outgoing (A+). They tended to be, as measured on the 16PF, somewhat happy-go-lucky (F+). It appeared that this factor was mainly involved with an individual's perception of whether he was introverted or extraverted. The self-perception scales that loaded this factor are those that most obviously are popularly associated with extraversion-introversion.

+).

+),

CHARLES A. BURDSAL, JR., AND STEVEN A. SCHWARTZ

181

Factor X: Suspicious Dejensiueness vs. Trusting Self-Deprecation. This factor seemed to be the level of an individual's reaction to his degree of protension. The more suspicious an individual was (L+), the more controlled he was (Qa+) and the more likely he was to report emotional stability (C'+), although he did tend to report his level of suspiciousness accurately (L'+ ). Conversely,· the more trusting he was, the less controlled he was and the more he was likely to report some emotional instability. Factor XI: Detached Looseness vs, Outgoing Self-Control. This factor individual's degree of aloofness (A) and selfwas primarily defined by control (Qa), on both the measured traits and the self-perceptions. On the high end of the factor, individuals tended to be reserved (A-) and had low self-control (Qa-) in terms of both measured traits and self-perceptions. They tended to be radical (Ql+) and independent (Q2+) and saw themselves as being forthright (N'- ). The low end of the factor involved outgoingness, high self-control, conservativeness, and group dependency, with self-perceptions of outgoingness, highself-control, and a degree of shrewdness. Factor XII: Perceived Internal Comfort vs. Perceived Internal Discomfort. This factor involved three of the four scales (C, Qt, Q4' I) that best differentiate neurotics from normals (6). However, the scales were. involved only in the self-perception. Persons high on this factor saw themselves as generally emotionally stable (C'+), self-assured (0'-), relaxed (Q/-), and somewhat traditional (Ql'- ). They measured to be somewhat reserved (A-) and surgent (F+), although these last two loadings were moderately low. On the other hand, persons low on the factor saw themselves as guiltprone, emotionally unstable, tense, and somewhat radical, while measuring to be outgoing-and serious. This factor, as did Factor X, seemed to contain a strong element of self-deprecation. Persons low on this factor were willing to report negative aspects of their personality, while persons high on this factor wished to give a report of general comfort. The only measured differences seem to be small loadings on A and F, with the self-deprecating end associated with being reserved and a bit impulsive.

Downloaded by [York University Libraries] at 20:13 18 November 2014

an

D.

CONCLUSIONS

The primary conclusions of the research were that some evidence was provided for the validity of measured personality traits and that differential patterning for perceived vs. measured traits exists, indicating some interacting systematic relationship of personality traits and the perception of the traits. Perhaps further research will help uncover these systematic factors

182

JOURNAL OF PSYCHOLOGY

which have been called perturbation factors by Cattell. It might be interesting, for example, to factor the differences between the self-perceptions and measured traits. Finally, the existence of the factors described might, after replication to demonstrate their stability, result in further understanding of self-ratings in systematic patterns.

Downloaded by [York University Libraries] at 20:13 18 November 2014

REFERENCES 1. ACKER, M., & McREYNOLDS, P. On the assessment of anxiety: III. By self-ratings. Psychol. Rep. 1966, 19. 251-254. 2. BURDSAL, C. A., & VAUGHN, D. S. A contrast of the personality structure of college students found in the questionnaire medium by items as compared to parcels. J. Genet. Psychol., 1974, 125, 219-224. J. CATTELL, R. B. The scree test for the number of factors. Multivar. Beha. Res., 1966, 1, 245-276. 4. CATTELL, R. B., & FOSTER, T. The rotoplot program for multiple single plane visually guide rotation. Behau, Sci., 196J, 8, 156-165. 5. CATTELL, R. B., & MUERLE, J. C. The maxplane program for factor rotation to oblique simple structure. Educ, & Psychol, Meas., 1960, 24, J-JO. 6. CATTELL, R. B., EBER, H. W., & TATSUOKA, M. M. Handbook for the Sixteen Personality Factor Questionnaire (16PF). Champaign, Ill.: Inst. Personal. & AbiI. Test., 1970. 7. DELHEES, K. H., & CATTELL, R. B. Manual for the Clinical Analysis Questionnaire. Champaign, III.: Inst. Personal. & Abil. Test., 1971. 8. FARLEY, F. H. Global self-ratings, the independence of questionnaire drive and anxiety, and social desirability variance. Acta Psychol., 1968, 28, J87-397. 9. HARMON, H. H. Modern Factor Analysis. Chicago: Univ. Chicago Press, 1967. 10. HARRISON, N. W., & McLAUGHLIN, R. D. Self-rating validation of the Eysenck Personality Inventory. Brit. J. Soc. & Clin, Psychol., 1969, 8, 55-58. 11. McREYNOLDS, P., & ACKER, M. On the assessment of anxiety: II. By self-report inventory. Psychol. Rep., 1966, 19, 2J1-237. 12. PEDERSEN, D. M. Evaluation of self and others and some personality correlates. J. of Psvchol., 1969, 71, 225-244. B. PETERSON, D. R. Scope and generality of verbally defined personality factors. Psychol. Reu., 1965, 72,(1), 48-59. 14. VINGOE, F. D. Validity of the Eysenck Extraversion Scale as determined by selfratings in normals. Brit. J. Soc. & Clin. Psychol., 1966, 5, 89-91.

Department of Psychology Wichita State University Wichita, Kansas 67208

The relationship of personality traits as measured in the questionnaire medium and by self-ratings.

The general purpose of the study was to determine (a)whether individuals tend to have an accurate perception of their personality, and (b)the manner i...
818KB Sizes 0 Downloads 0 Views