Frequency Concept of Dispositions 1980

8/13/2019 Frequency Concept of Dispositions 1980

http://slidepdf.com/reader/full/frequency-concept-of-dispositions-1980 1/15

The frequency concept of disposition:dominance nd prototypically

dominant acts^

David M Buss and Kenneth H Craik, University o California,

Berkeley

Abstract

Three studies of dominance explore the frequency concept of disposi-

tion, which entails categories of acts that are topographically dissimilar butnonetheless considered to be manifestations of a common disposition. Inthe first study, 100 difierent acts presumably belonging to the category ofdominance were generated through a nomination procedure. In the secondstudy, expert and student panels rated how prototypically dominant eachact is, defined in terms of centrality of membership in the category ofdominant acts. In this manner, an internal structure of the act category wasspecified such that some acts are more prototypically dominant while oth-ers are more peripheral mem bers. Substantial agreement in these ratingsexists within and between panels. The third study found that a m ultiple-act criterion based on prototypically dominant acts is predicted by person-ality scales with significantly greater success than are multiple-act criteriabased on more peripheral acts within the dominance domain. Discussionfocuses on specifying the appropriate act category for other frequency dis-positions and follow-up field studies of them. Implications for altemativenotions of disposition (e.g., purposive-cognitive concepts) are considered.

At the heart of personality research lies the concept of disposi-

tion: the tendency of individuals to behave in certain ways, for

example, to be self-praising, to be subservient in relation to others,

or to adopt a sensation-seeking orientation toward the environment.

The diverse meanings of this basic concept of personality are pres-

1. This study was partially supported by a research grant awarded to David M.Buss from the Psychology Department, University of California, Be rkeley. R equ estsfor reprints should be sent to David M. Buss or Kenneth H. Craik, Institute ofPersonality Assessment and Research, 3657 Tolman Hall, University of California,Berkeley, Califomia 94720.

We are grateful to J. M. Kilkowski for facilitating data collection for the MainStudy, and to the expert panel members: D. S. Butt, L. Cartwright, J. Coyne, L. E.



380 Buss and Craik

ently undergoing reexamination (Alston, 1975; Craik, 1976; Kil-kowski, 1975; W iggins, No te 1). T he assertion M ary is dom inan tcan be understood as a hypothetical proposition (Ryle, 1949) some-wh at akin to a dispositional sta tem ent in phys ics (e.g., T he glassis bri ttle ), taking th e form: it is likely, or a good b et, that the entitywill respond in certain ways {x, y, z to certain circumstances {a,b, c . Alternatively, the concept of personal disposition can be ana-lyzed as a summary statement (Hampshire, 1953) taking the form:so far, the term dominant is the right word to sum ma rize the gen-eral trend in Mary's conduct. The hypothetical proposition focusesupon specific conditional predictions and the role of situational

factors; the summarizing statement emphasizes the relative fre-quency of a specified kind of act over a period of observation.

Both interpretations would appear to be subsumed by Alston's(1975) delineation of what he terms the frequency concept of dis-positions. To paraphrase Alston, when we attribute a disposition ofthis sort to a person, it is part of what we are asserting that, givena representative set of situations (which may be specified broadlyor narrowly), the person will emit a large number of appropriateresponses (relative to the norm for that disposition). Neither per-

sonality scales nor trait ratings constitute the basic measure of thefrequency conc ept of dispos ition; rather, the frequency con ceptinvolves sampling behavior by monitoring the relative frequencyof specific acts within an appropriate response category over anarray of occasions. Alston employs the expression S-R frequencydispo sition, bu t be cau se he later relaxes the situational restric-tions, the expression frequency c once pt of disp ositio n will beused here .

The frequency concept of disposition entails categories of actsthat are topographically dissimilar but nonetheless considered tobe manifestations of a comm on disposition (e.g., some acts co un tas instances of dominance while others do not). The nature of suchdispositional categories, their criteria for membership, and theirstructure constitute fundamental but heretofore relatively unex-amined issues in personality theory.

Our aim is to contribute to this exploration by joining recentformulations of personal disposition to cognitive studies of natural

categories. The disposition toward dominance, one frequently as-sessed in personality research (Butt & Fiske, 1968; 1969), is se-



Dominance 381

1975; 1978; Rosch & Mervis, 1975). Rosch has proposed that suchcategories contain an intem al structure. Th us, m em bers of a naturalcategory differ with resp ect to cen trality of m em be rsh ip. A spar-

row, for example, would be a more prototypical member of thecategory bi rd than wo uld a pen guin , which wo uld be a moreperipheral member. Through the use of prototypicality ratings(Rosch & M ervis, 1975), acts can b e identified wh ich m ost ce ntrallyfit the category or image of the meaning of dominance. One aim ofthe present research is to gauge the reliabili ty with which theseprototypically dominant acts can be identified. The only previousworks to bear even tangentially upon the relation of prototype todispositional concepts (Cantor & Mischel, 1977; 1979) dealt with

trait-descriptive terms and diagnostic labels rather than acts andused different indices for the concept of prototype.

Hampshire's (1953) summarizing notion in particular shifts thevalidation of dispo sitional ass ertions ab out perso ns from single actsto the relative incidence of specified acts over a period of obser-vation. In a com parison of sing le act-single situa tion criteria and m ultip le act-m ultiple situa tion criteria (Fis hb ein & Ajzen, 1974),Jaccard (1974) has indicated that assessments of dominance basedon personality scales correlated around -I-.20 with single acts ofdom inance but around -I- .60 with a compo site m ultiple-act criterion(indexing the number of different acts of dominance out of the 40listed that the person reported having performed). The present re-search was designed to replicate and extend the Jaccard study.Comparison of the single act and multiple-act correlations servesto replicate the Jaccard analysis. In addition, scores on dominancescales are correlated with four composite criteria derived from pro-totypicality ratings. The four composite criteria are arranged on a

gradient from most to least prototypically dominant.

Preliminary Study 1

Method

Subjects

Seventy-five undergraduate students (40 females and 35 males) com-ple ted the p roce du res of Study 1 as an exercise for a psychology class.

Procedure

Each participan t receiv ed the following instructions: Th ink of three of



382 Buss and Craik

own gender on the top right-hand corner of the card. Each participantwas provided with a 3 x 5 index card on which to record the acts.

ResultsThe list of acts generated in this way was supplemented by acts

gleaned from a perusal of dominance scales (Allport & Allport,

1939; Edwards, 1959; Cough, 1957; Jackson, 1967), and was sub-

sequently reduced to 100 by eliminating redundancies and

non-act or general statements such as She tends to monopolize

conversations. Two inventory items, both from the CPI, were in-

cluded in the final list in slightly modified form. The final list of

100 acts was examined for grammatical errors which were then

corrected. Each act was then phrased in a way suitable for perfor-

mance by either sex. For example, the act: He demanded a back-

r ub could also be performed by a female ( She demanded a back-

rub ). Thus, one list consisting of 100 acts had a male (he) as actor

while another list composed of the same acts had a female (she) as

actor.

Prel iminary Study 2

MethodSubjects

Two samples of participants w ere used for the second study. The firstconsisted of 57 undergradu ate volunteers (30 females and 27 males) non eof wh om had pa rticipated in Study 1. Each was paid tw o dollars as a tokenof appreciation for com pleting the procedures. The second sample con-

sisted of 22 expert judges—13 Ph.D.s and 9 graduate students in per-

sonality psychology (8 females and 14 males). The combined n was 79.

Prototypicality RatingsEach participant rated 200 acts (100 with male as actor and 100 with

female as actor) on the extent to which each was prototypically dom inant.The instructions, adapted from Rosch & Mervis (1975) were as follows:

This study has to do with what we have in mind when we use wordswhich refer to categories. Let's take the word red as an example. Closeyour eyes and imagine a true red . Now im agine an orangish red . . . imag inea purple red. Although you migh t still n am e the orange-red or the purple-red with the term red, they are not as good exam ples of red (as clear casesof what red refers to) as the clear tru e red. In short, some reds are

redd er than others . In this specific study you are asked to judge how good an example of



Dominance 383

you feel the act fits very poorly with your idea of what dominance is (oris not a m em be r of that category at all). A 4 me ans you feel the act fitsmoderately well. Use the other numbers of the 7-point scale to indicateintermediate judg em ents . The se instructions were repeated twice, onceeach for male-actor and female-actor statements.

Judgments of social desirability. In addition to the prototypicality rat-ings, approximately half (n = 27) of the undergraduate sample was askedto rate the male-actor statements on social desirability and the other half(n = 30) was asked to rate the female-actor statements on social desirabil-ity. A modification of the Ed w ards (1957) instructional set, pre ced ing thelist of acts, was use d: Belo w you find an exam ple of four things that aperson says that he does. A jud ge , such as yourself, has made an estimateof the degree of desirability or undesirability of these acts. Examples: '2'

to punish your enemies; '5' to read psychological novels; '7' to make ex-cuses to your friends; '8' to go out with your friends.

T he person wh o jud ged these acts believe s that ' to pun ish your ene-mies ' is definitely an undesirable act in others, 'to read psychological nov-els ' is neither desirable nor undesirable, 'to make excuses for your friends'is mo derately d esira ble , and 'to go out with your friends' is quite a desira bleact.

O n a one to nine scale (9 be ing the m ost desirable an d 1 being the leastdesirable) indicate your own judg me nts of the de sirability or u ndesirability

of the acts in the same manner. Remember that you are to judge the actsin terms of whether you consider them desirable or undesirable in others.Be sure to make a jud gm en t about each act. ^

Results and Discussion

Reliability of prototypicality and soci l desirability ratings. Ta-ble 1 presents the alpha reliability coefficients (Cronbach, 1951 sep-arately for the four panels of raters. All are uniformly high, indi-cating that considerable agreement exists among members of each

panel regarding which acts are prototypically dominant and, anal-ogously, socially desirable.Most dominant acts. Below are listed the 15 most dominant

acts as rated by the entire sample.

He (she) issued orders that got the group organized.He (she) managed to control the outcome of the meeting without

the others being aware of it

He (she) took charge of things at the committee meeting.He (she) assigned roles and got the game going.

He (she) readily used the authority of his (her) position.He (she) took command of the situation afier the accident.



384 Buss and Craik

Table 1. Alpha coefficients for prototypicality (PR) and social de-sirability (SDR) ratings.

Male-actor

statements

Female-actor

statements

Pr

.87

.82

.86

.86

SDR

_

-

.97

.96

PR

.89

.84

.89

.82

SDR

—

-

.90

.96

Expert judges (male)

Expert judges (female)

Undergraduate judges (male)

Undergraduate judges (female)

Combined (total sample) 9 97 9 9

He (she) forbade her to leave the room.He (she) set goals for a group.He (she) demanded that he run an errand.He (she) persuaded others to accept his (her) opinion on the issue.On the auto trip, he (she) decided which directions to take when

they got lost.He (she) took the lead in organizing a project.

He (she) persuaded him to do something he didn't want to do.He (she) told her to get off the phone so that he (she) could use it.

It should be noted that these acts involve elements both of con-trol for group goals and of persuasion for self gain.

Correlations between mean panel ratings. Tab le 2 presen ts thecorrelations among the mean panel ratings. All correlations are sig-nificant beyond the .001 level and substantial in magnitude. Thelowest cross-panel agreement is that between undergraduate male

raters and expert female raters, which holds for both the male andfemale actor statements. It should be noted that correlations basedon group means are usually higher than correlations based on pairsof judge s.

Relationship between pro totypicality and social desirability. T helist of means for the prototypicality of each act was correlated withthe list of m eans for social desirability, across each set of 100 item s,separately for each of the four rating panels. Table 3 presents theresults of these analyses. All correlations are significantly positive

with the exception of undergraduate male judges who do not tendto view prototypically dominant acts as socially desirable.



Dominance 385

Table 2. Correlations between mean panel ratings.

Prototypicality ratings

A Expert judges male)

B Expert judges female)

C Undergraduate judges male)

D Undergraduate judges female)

Social desirability ratings

C Undergraduate judges male)

D Underg raduate judges female)

Male-actor

statementsA B C

.82

.65 .48

.87 .80 .74

.97

Femaie-actor

statementsA B C

.87

.64 .44

.74 .64 .89

.93

Note.—All correlations are significant beyond the .001 levei.

on social desirability, high on prototypicality, (c) high on socialdesirability, low on prototypicality, and (d) high on social desir-ability, high on prototypicality. Table 4 presents three exemplarsfrom each of these categories. Petulant and self-centered acts seem

to be rated low on both dimensions. Explicitly directive, yet self-centered acts seem to be rated low on social desirability, yet highon dom inanc e. Lead ership and organizing behaviors for group gainare rated high on both social desirability and prototypicality. Fi-nally, acts in the high socially desirable but low dominance quad-rant seem to be characterized by public involvement without lead-ership or directive connotations.

Main Study

Method

Subjects

Eighty-three undergraduate volunteers (43 females and 40 males) com-pleted the procedures for the Main Study. None of these students hadparticipated in either of the first two studies.

Materials

These were: (1) the dominance scale from the Califomia PsychologicalInventory (Gough, 1957); (2) the dominance scale from the Jackson PRF

(Jackson, 1967); (3) self-rating on a seven-point scale of dominance, andother measures used for a companion study. The Act Report (100 acts



386 Buss and Craik

Table 3.ability.

Relationship between prototypicality and social desir-

SampleMaie-actorstatements

.48***

.64***

- . 0 9

.18*

Female-actorstatements

.38***

.58***

- . 1 3

.46***

Expert judges (male)

Expert judges (female)

Undergraduate judges (male)

Undergraduate judges (female)

• p < .05. p< .01

* p < .001.

act or not. If the answer was ye s, they were reque sted to indicate thefrequency with which they performed the act on a three-point scale— rarely, som etim es, or often.

Procedure

Participants we re adm inistered the first three items listed un de r M a-terials . After a one-w eek interva l, the Act Repo rt was adm inistere d. A

week interval was allowed between the two sessions in order to minimizethe operation of a response set.

Results

Act X Scale correlations. Th e yes /no dichotomy of act per-formance was correlated with each of the two dominance scalesand the self-rating. Similarly, the frequency with which each actwas performed was correlated with the measures. For the CPIDo minan ce Scale, the mean correlations we re .16 (range = - . 2 6 to

54) and .11 (range = -.38 to .57) for act performance and frequencyof performance respectively. For the PRF Dominance Scale, theseresp ectiv e correla tions are .20 (range = - . 1 9 to .64) an d .19(range = -.26 to .63). For the self-rating of dominance, these cor-relations are .10 (range = -.26 to .44) and .13 (range = -.34 to .49).Thus, the prediction of specific acts from inventory scales approx-imates that found by Jaccard (1974).

Multiple-act criteria. Each act was ranked on the basis of its

independently generated prototypicality rating (see PreliminaryStudy 2). This ranked list was then divided into quartiles, each



Dominance

Table 4. Acts from upper and lower quadrants.

387

Low socialdesirability

High social

desirability

Low prototypicality

He she) flattered her inorder to get his her)

way.

He she) hung up the

phone on his her)

lover.

He she) deliberately

arrived late for the

meeting.

He she) asked someone

out on a date.

He she) was highly

involved in a political

campaign.

He she) solicited funds

for a cause in which he

she) was Interested.

High prototypicality

He she) forbade her to leavethe room.

He she) monopolized the

conversation.

He she) directed the

conversation around to

himself herself) and his

her) doings.

He she) took the lead in

livening up a dull party.

He she) took the lead in

organizing a protest.

He she) took command of

the situation after the

accident.

Pred ictors and the mu ltiple-act criteria. Fo r each subjec t, fourscores were generated which represented the number of acts per-formed (the ye s/n o dichotomy ) with in each m ultiple-act crite-rion. The dominance scale scores were then correlated with thesecriterion scores. Table 5 presents these results, separately for eachsex.

In the conceptual framework elaborated earlier, the category ofdominant acts contained an intemal structure such that individualacts differed with respect to prototypicality of membership. It washypothesized that perfonnance of acts central to the category (thatis, prototypically dominant) would be predicted with greater ac-curacy than more peripheral members of the category (acts ratedlow on prototypicality). Inspection of the table columns shows thatfor males on the CPI and PRF scales, the correlations decreaseprogressively in magnitude as the prototypicality of the criteria

decreases. The difference between the correlations with the mostcentral an d m ost pe riph era l category criteria are significant for both



388 Buss and Craik

Table 5. Predictors and multiple-act criteria.

Males {n = 40)

Pro to l

Proto2

Proto3

Proto4

Females (n = 43)

Pro to l

Proto2

Proto3

Proto4

CPI

.48***

.32*

.26*

.14

.31*

.21

.31*

.05

PRF

.67***

.45**

.42**

.33*

.47***

.31*

.48***

.26*

Self-Rating

.25

.26

.04

.29*

.31*

.15

.27*

.12

*p < .05.** < 01.

***p < .001.

dieted than the second with both scales. Again, the most centraland most peripheral differ significantly {p < .01 for the CPI; p <

.05 for the PRF). For the dominance self-rating, the hypothesizedpattern is obtained for the female sample, with the difference inmagnitude of correlations for Protol and Proto4 significant {p <.01). The pattern of correlations is slightly lower than that obtainedby Jaccard (1974) but in no case significantly so.^ To summarize,the hypothesis that prototypically dominant acts (using a multiple-act criterion) will be pred icte d with grea ter accuracy than acts morepe riph eral to the category, is by and large confirmed by the pre sen tdata.

Discussion

The frequency concept of personal disposition entails a multiple-act criterion a nd a me ans of specifying the appro priate act category.The finding of a gradient of validity correlations associated withprototypicality of acts lends support to the hypothesis that dispo-sitional concepts function through natural categories of acts (Rosch

3. Th e focus of the CP I Do scale upon asse ssing constructive forms of dom inanc e(Gough, 1968) is reflected in a correlation of +.62 between the social desirability



Dominance 389

& Mervis, 1975), structured so that some acts serve as prototypicalwhile others fall along a continuum toward the periphery of thecategory boundaries. Furthermore, the evidence indicates that, for

the co ncept of dom inance at least, highly reliable judg m ents of theprototypicality of acts can be made, even among a set previouslyand ind ep end ently nom inated as dom inant, and that prototypicalityjudgments show striking agreement between panels of persono-logical experts and nonexperts. Prototypicality ratings offer a directand convenient means of examining category structure. However,conve rgent evid enc e from other typicality indices shou ld be sought(e.g., verification times for category membership; probability ofitem output in membership nomination tasks; expectations gener-

ated by the category name) (Rosch, Simpson, & Miller, 1976).The frequency concept of personal disposition generates an al-

ternative to traditional approaches in the analysis of personality.Through the variables of the Califomia Psychological Inventory(CPI), Cough (1957; 1968) has sought to assess folk concepts ofpersonality, dimensions of individual functioning that are universalamong cultures, are often calibrated with sociological continua andserve to forecast how persons will be described and appreciated byothers and how they will fare regarding significant societal out-comes (e.g., graduating from medical school; being diagnosed asalcoholic; having one's works and performances deemed creative).Alternatively, Rlock (1977) has differentiated four classes of per-sonality data: R-data (based on observer reports), S-data (based onself-reports), L-data (based upon personal and societal life out-comes), and T-data (based upon laboratory test situations), and hasemphasized the demonstration of coherence among them. Atten-tion to act-data (based upon naturally occurring behavior in the

individual's everyday ecology) is not incompatible with either ofthese two orientations to personality research. Its neglect, however,leaves the behavioral linkages that presumably establish the oftenobserved statistical coherence among sets of R-, S-, and L-data assomething of a mystery. An act approach, as an addition to tradi-tional means of personality research, could clarify these behaviorallinkages.

The frequency concept of disposition is, however, logically dis-tinct from notions of personal disposition which Alston terms pur-

posive-cognitive (PC) concepts, and prefers on theoretical grounds(Alston, 1975). In the PC approach, theoretically postulated deter-



390 Buss and Craik

need for close relations with others; but due to lack of interest andfear of rejection, respectively, he may rarely or never exercise hisability or seek to satisfy his need. Furthermore, consistency in the

same kind of overt acts, e.g., trying to dominate others, may beaccounted for by different PC concepts, e.g., a fear of what mighthappen if someone else were in control versus a strong desire forreassurance of one's own worth plus a belief that dominating othersprovide s a good chan ce of secu ring that reass uran ce (Alston, 1975).Thus , it is understandable that some of our personological experts,committed to PC concepts of personality (e.g., the Murray needsystem), found the making of prototypicality ratings subjectivelydifficult, for each act could be viewed as an expression of a varietyof PC concepts. By way of contrast, in the strict frequency ap-proach, dispositions are not considered causes (Wiggins, Note 1);rather, coherence among acts embodies regularities in individualbehavior that may possess fundamental descriptive and predictivevalue, but that call for rather than afford explanation.

The replication of Jaccard's (1974) findings regarding single ver-sus multiple-act criteria for validational appraisal of personalityscales also illustrates the rather straightforward means available for

linking act-data with S-data, and in analogous fashion, with R-data.Comparison of single-act and multiple-act criteria is analogous tothe comparison of items and scales: thus, higher correlations forthe latter are to be expected on psychometric grounds. The sub-stantive point of the Jaccard analysis and this replication is thatmultiple-act criteria constitute the appropriate criteria in person-ality research from the perspective of a summary or frequency in-terpretation of the concept of personal disposition.

The present analysis links one form of S-data (scale scores) withanother (self-reported acts). In addition, the reliabilities for the pro-totypicality ratings are based upon verbal presentation of acts rath-er than directly observed acts. One evident next step in this lineof research calls for field studies, entailing prototypicality ratingsof acts observed in situ. Video-taping the everyday behavior ofpersons can be expected not only to reveal stable differential baserates (or batting averages) for the already specified and nominatedacts of dominance but also to highlight the artistry of dominance

and yield more subtle exemplars of the multiple-act category. Suchnaturalistic field studies, providing fuller elaboration and docu-



Dominance 391

him to . . . versus she ordered him to . . . ). Second, in assessingthe strength of a disposition, what are the implications of using th etotal frequency with which any acts falling within the dispositional

category are manifested over the period of observation in contrastto using the number of different acts within the category? Third,it is possible that the population of core acts for a given dispositionis quite limited and readily specifiable, while the population ofperipheral but pe rtinen t acts is broader and more difficult to iden-tify fully. How likely is it then that an individual who manifests adisposition (e.g., dominance) by a moderate frequency of core actswould be more accurately assessed than an equivalently domi-nant individual who displays the disposition via a higher rate of

peripheral acts, which may be less completely and thoroughlyidentified and monitored? These issues can be clarified by dem-onstration field studies.

Finally, what about the acts that are not prototypically dominant?They may simply be peripheral mem bers of the category, but morelikely they share attributes and membership with other disposi-tions. The examination of prototypical acts for an array of disposi-tional categories is obviously an important item on the researchagenda. A guide to this form of inquiry is offered by the circumplexmodel of trait-descriptive terms within the interpersonal domain,recently presented by Wiggins (1979). A procedure of act nomi-nations and prototypicality ratings for each of the sixteen-markercategories of that model would test it empirically: acts peripheralto a given category should also display membership within adja-cent, rather than remote categories along the circumplex. Our strat-egy offers a means of exploring domains such as those describedby the circumplex, by focusing upon acts, the categories that con-

tain them, and the internal structure of these dispositional con-structs.

Reference Note

1. Wiggins, J. S. In defense of traits. Vancouver, B.G.: University of British Co-lumbia, unpublished report, 1974.

References

Allport, G. W., & Allport, F. H. The A-S reaction study. Rev. Edition. Boston:Houghton Mifflin, 1939.

Alston, W. P. Traits, consistency, and co nceptual altem atives for personality theory.Journal for the Theory of Social Behaviour, 1975, 5 , 1 7 ^ 8 .



392 Buss and Craik

Butt, D. S., & Fiske, D. W. Differential correlates of dominance scales. Journal ofPersonality, 1969, 37, 415-428.

Cantor, N., & Mischel, W. Traits as prototypes: Effects on recognition memory.Journal of Personality and Social Psychology, 1977, 35, 38-48.

Cantor, N., & Mischel, W. Prototypes in person perception. In L. Berkowitz (Ed.),Advances in experimental social psychology. New York: Academic Press, 1979.

Craik, K. H. The personality research paradigm in environmental psychology. InS. Wapner, S. B. Cohen, and B. Kaplan (Eds.), Experiencing environm ents.New York; Plenum, 1976, pp. 55-79.

Gronbach, L. J. Coefficient alpha and the intemal structure of tests. Psychometrika,1951, 16, 297-334.

Edwards, A. L. The social desirability variable in personality assessment and re-search. New York: Holt, Rinehart and Winston, 1957.

Edwards, A. L. Manu al for the Edwards Personal Preference Schedule, Rev. Edi-tion. New York: Psychological Corporation, 1959.

Fishbein, M., & Ajzen, I. Attitudes toward objects as predictors of single and mul-tiple behavioral criteria. Psychological Review, 1974, 8 1 , 59-74.Gough, H. G. M anua l for the California Psycholog ical Inventory, Rev. Edition.

Palo Alto, Calif.: Gonsulting Psychologists Press, 1964.Gough, H. G. An interpreter's syllabus for the Galifomia Psychological Inventory.

In P. McReynolds (Ed.), Advances in psychological assessment. Vol. 1. PaJoAlto, Galif.: Science and Behavior Books, 1968, pp. 55-79.

Hampshire, S. Dispositions. Analysis, 1953 , 14 , 5—11.Jaccard, J. J. Predicting social behavior from personality traits. Journal of Research

in Personality, 1974, 7 , 358^67.Jackson, D. N. Personality research form manua l. Goshen, N.Y.: Research Psy-

chologists Press, 1967.

Kilkowski, J. M. Traits, situations, and the logic of explanation in psychology: Aselective review of the issues and evidence. ORI Research Monograph, 1975,15, No. 3.

Megargee, E. L. The California Psychological Inventory Handbook. San Francisco,Galif.: Jossey-Bass, 1972.

Rosch, E. Gognitive reference points. Cognitive Psychology, 1975, 7, 532-547.Rosch, E. Prin ciples of categorization. In E. Rosch and B. B. Lloyd (Eds.), Cognition

and categorization. Hillsda le, N.J.: Erlbaum , 1978.Rosch, E., & Mervis, G. B. Family resemblances: Studies in the intemal structure

of categories. Cognitive Psychology, 1975, 7, 573-605.Rosch, E., Simpson, C, & Miller, R. S. Stm ctural bases of typicality effects. Journal

of Experimental Psychology: Hum an Perception and Performance, 1976, 2,491-502.Ryle, G. The concept of mind. New York: Bames & Noble, 1949.W iggins, J. S. A psychological taxonomy of trait-descriptive term s: I. T he interp er-

sonal domain. Journal of Personality and Social Psychology, 1979, 37, 395-412.



Documents

Frequency Concept of Dispositions 1980