recurrent personality factors based on trait ratings - Defense Technical [PDF]

IUNCLASSIFI]ED "",e%" ,II '•--a--•

AD ,FO,

267

778

SERVICES TECHNICAL INFORMATION AGENCY ARLINGTON HALL STATION ARLINGTON 12, VIRGINIA

C%

UT N CL ASFH E D

NOTIlCE: Men Mov-rumat or other dz'viWg, spedIications or ot~her dst^ um. used for my prpos.o other Can in conn-ction wi'th a definitely related goverment pzociu~nt operit Ion, the U. S. Gove~mwnt thereby incars no responsibiit~y, nor any obl~igatirnn wbattoever; and the fact that toe Government may havco foiuiated, furnished, or In axzy %ay ruppl.ied the said dmilupx, spedlticatirns, or other data is not to~ bo reprided by Jim~1cation or othervise as ia any maner 11cerslag the bolder or aI'! otaer person or corporatioo, or conveying ak, . J.t. or pemwiss~om to wwattetgre. us* or cell any patentei invention tbat may to a?'y vay be related t~lereto.

TECHNICAL

ASD-TR-61-97REPORT

Cq

Ma~y 196,1

Recurrent Personality Factors Based

¶

on Trait Ratings

!

(,

S~ C30

(*

Ernest C, "ruipes Rasmond E'. (Christal

PEIRSONNEL,

ILABORATOR

Y

AEItONAIITICAI, SYSTEMS DIVISION AIt FORCE SYS'I'I'EMS COMMA NI) LA(,0.AND Alit FORM,* BASE, TKXAS

Fred E. Hoidreg., Cal USAF ChiefA

c-~

Pe.oenn.1 Labotatorip ""Aoronouu

Sy"1606emDivision (A FSC)

NOTICES Ween Government drawings, Specdzjcý;rijns% Or other data cre used for purposf other than in Connection an.ý with a defimn.W) related Government procurement operation, the United Stotes Covernmer, thereby Incurs no resprigisibility not any obligation whatrooever; ennd the faCt that tho' Government may have, fmwaulated. Iuruis'~ed, of in any way supplied the -said dtcwinq~, sezacaan date, is not to be reqde o t sior Spycifications otherws ' mne ies tb. holder or any othe, peFdrbson mo~ricortpavoratilonl,eotr soenvoesynqaanyy rnaqnhnssolr permtinjsiotomanufactute, use Ot se1l any patented InVent-on !hat may .. any way be eltd thereto.

Quclitied tv-4iiestors may obtain copies of this report frow. the Afted Servac's Technictid Information Age~cy, Documents Service Center. Aflinqtlon S.':na Depament of Defense contractors may obtain unclassified docujment,~ on request by stating their offi-ial frort- kSTIA need and citmnq the U.I,. 5 catrr mnvolve'd.

This feport has ben released to the Office of Technical Services Ur.S earaitof Commerce. W tthavo 1'5. D.Z. &ImSal* to the qeneta. r1: .1c.

Cootes Of ASO T~~cc Repowts an4 Toctica.o Notes, sionutd to tIe AwattSyje~mv Lhsvjsio Vovo such telurn is rpUmi'tlenot be #1~tr c~n1ctw obliaqeuons at notice an a specific rlocamnet.by 2cu4vrs

TECHNICAL REPORT ASD-TR-61.97 Mey I94I

RECURRENT PFERSONA.LITY FACTORS BASED CM4TPAIT RAI GS

80 Ernest C. Tu~s Roymmnd E. Ch1•istel

Pejotoc 7717. Task 17110

Peoneel Lete"e~y AEROKAUTKA'AL SYSTEMS DIV¢4M AIR FORCE SYSTEMS COMAND UNITEO STATES AIR FORCE Llo,,d Ab Pa Veto. Teem

ABSTRACT Intausair~jtions among ratings on 35 personality traits, selected as representative of the personality domain, were obtained for eight samplei. These samples differed in length of acquaint an.-eship from three days to more than a yews; in L~ind of acquaintancerhip from assessment program~s in a military training COUrs to a fr~temity house =ituatimi; in type of subject from airmen with only a high school education to male and female, undergraduate students to f ifst-yeas graduate students; mid in type of rater from very naive persons to clinical psychologists and psychiatrists with years of experience in the evaluation o: personality. Centroid or multiple-group factors were extracted and rotated orthogonally to simple structure. For one study, an independent solution was obtained in which analytic rotations were accomplished an an IBIA 6%0O computer using Kaiser's normal vasimax criterion. Five fairly strong and recurrent factors emerged from each analysi!- labeled as (1) Surqency. (2) Agreeableness, (3) Dependability. (1) Emotia .r I Stability, and (5) Culture.

It

TABLE 01 CONTENTS

lntrnduction

........................................

.. .......................

Page I

.....................

2

Method ..................................................

De3,a tiption of the Eight Studies .................................................. Study Study Study Study Study

A. B. C. D. E.

OCS 790-Case Sample .............................................. OCS 3-Day Aszessnient Sample ...................................... OCS End-of-Traininq Assessment Sample ............................. Command and Staff School Sample ................................... Cattell's Male University Sample ...................................

Study F. CattelPs Female University Sample

3 3 4 4 4

.................................

5

Study G. Fiske's Teammates Ratings of Graduate Students ...................... Study H. Fiske's Staff Members Ratinsq of Graduate Students .................... Resu'ts ........................................................................ Factor Factor ractar Factor Factor

U: UI: IV; V:

Surgency ......................................................... Aq eablen ss .................................................... Dependability ................................... Emoticoal Stability .............................................. Culture .........................................................

3

5 5 5 6 7

.............

a 9 10

Discussion .....................................................................

11

Summ ry .......................................................................

13

ReferenLces .. ................................................................

14

Appendis A: Study A Factor Analysis ....

16

.......................................

Table Al: Inte.cowelations Amorng Trait •atings lice Officer Candidates ............. ........................................ Table A2: Loadings as Nine Ceatioid factors from 710 Male Officer Candidates ............. ................................. Table Al: Loadings ca Nine Roilard Factors troee 79 Male Office Candidates ............................................. Appendix B: Study B Factor Aislysit

Table 81: late•--nlatsioas Am...1 I-r•ry Trait Retllas from Officer Cuididates .... ..... ...................................... Table 82: Loadils an Five Othloqaoillsmied Mltiplt-Grp med Thm Ceatmid Factors from -Day Oficr Candidate Ratings ................. Table P'1 Loadigqs as Ciqht Rolated |lecior fro. )-Day Officw Candidate R tinau ............. .................

Table Cl: ltercarreletias Amoeg Truit Ratings Obtaiod Now rnd of OCS ........................... ........ v

17 i

19

........................................

Appeadix C: Sthdy C Factor Analysi .......................

16

.................

................

19 20 21 2

22

Table of Contents (Conffnue) Page Table C2: Loadings on Five Orthoqonaliztd Multiple-Croup and Three Controid Factors from Ratings Obtained Late in OCS ...................... Table C3: Loadings on Eight Rotated Factors from Ratings Late in OCS ....................................................... Appendix D: Study D Factor Analysis ............................................... Tabie DI: Interconelaticas Among Trait Ratings from 499 Command mad Staff School Officer ....................................... Table D2: Loadings on Five Ortbogonalized Multiple-Group Factarx from 499 Command and Staff School Officers ........................... Table D3: Loadinsq on Five Rotated Factors from 499 Command and Staff School Officers .......................................... Appendix E: Study E Factd

Analysis ..............................................

Table El: Intercorrelations Among Trait Ratings of ............................ Cattell's Wale !oll.q* Students ................. Table E2: Loadings on Five (ftbogoealized Multiple-Group and Three Centroid Factors from Cattell's Male College Students ................... Table U.3: Loadings on Five Rotated Multiple-Group Factors and Three Additional Centroid Factors from Catten's Male College Students ............... Table E4: Loadings on Eight Final Rotated Factors Lom Cattell's Male Colleqe Students ............................................. Appendix F: 3tudy F Factor Analysis

..............................................

Table Fl: Intercarrelatioms Among Trait Ratings of Cattell's Female College Students .......................................... Table F2: Loadings an Five Orthoganalized Multiple-Group and Seven Ceatroid Factors from Cattea Femaleoe conee Students ....................... Table Fl: LoLwqs am 12 Rotated Factors from Cattell's Feasle Coll*eg Students .................................................. Appendix G: Study G Facter Anaiysis ............................................... Tqble GI: iatwe'cv.iatioss Ame" T.-t Ratings fram .... .................................. Fiske's Temmetes ............... Tabe G2: Loeadis an Five Otbogqaelised Muttiple-Group and One Ce•traid Fact has Fiske's Teemmaes Rating .s.................... Tae G3: Loadi* am Si- Rotated Factors Ira& ............................. ................. Fiske's Temmetes Retia Aep meliu It: Study II Fsctor A lysm

..............................................

Table HI: lstercarreiatims Among Trait Ratings from .................... ..................... .......... Fiske's Staff Rati Tible H2: Loeadis8e as riv Othoaemeaiisd Msltiple Gro and One Ceatmid Frctm iorw Fiske's Stuff Retiqg ..............................

Table mn Loadis as Six Rotated rctera from Fisk0 8 Staff Rating .s .......................

'S

..............................

23 24 25 25 26 27 28 28 29 30 31 32 32 33 34 35 35 16 17 38 10 35 40

LIST OF TABLES Table

Pap

1

Loadings on Recurrent Factor I (rom Eight Analyses

............................

6

2

Loadings on Recurrent Factor 1Ufrom Eight Analyses ............................

7

3

Loadings on Recurrent Factor IU from Eight Analyses ...........................

8

4

Loadings on Recurrent Factor IV from Eight Analyses ............................

9

5

Loadings on Recurrent Factor V from Eight Analyzes ............................

10

6

Comparison of Normal Vmirmax Solutlon with that Obtained Using Graphic Rotations ...................................................

13

RECUIRENT P~ERSONALITY FACTORS BASCD OW T."A1T RATINGS' INTRODUCTION Thie measurement of personality by means of trait ratings has a history of at least fifty dating back to the investi~jations of Heymans & Wiersma (1909) in which 400 physicians rated o.ver 2S00 indivie.aals. After a brief surge of interest in the area in the 1920's. few to-t-arch studies were carried out on trait ratings until fairly recently. Several early investigators reported findioq-t In4icatinq that ratings of personality traits were quite, unreliable, and thus not ':er, useful measurements. At the same time. the incaeasing popularity ofthGsalan !)yzamic Psychology theories with their view that man must be studied as a whole person in "a~ion did little to stimulat* interest in trait "'easuraemet. However, the concept of the "whole, person." *lth',ugh it may ultimately lead to greatest accuracy of description and prediction. is unmanage~able fram a measurement point of view and will likely remain so for same, tine to coste. rurtL-ermare,. as Cattell (1946) has pointed out, the trait calicept does not Preclude the concept of :he whale person. since my person can be rAiquely and adequately described by a combhination of a number of sndependenot traits or factors. Although early studies indicated p~ersonality trait ratings to he unreliable (low rater agreemienti, unstabl (specific to the rating situation). and .ontaminated by a lI"g general foct'x (balo%. Symoeds (1931) end Allpart (1937) concluded that such def iciencies might he overcome by the .me of fairly large groups of rotem who have observed the subjects' day-to-day behavior Paer a considarablie period of tine, ad by requirin the Maing of several st-bjects on one trait at a time, rather than the riaig of each subject on several traits at a ti-te. ?Clrsr,

Several ni-cet studies hore horn out Symonds' and Allpart's conclusions. Results fromn the Veterans Admizszsration Research Program on the Selection of Clinical Psychologists -s reported by Kelly & Fiske (1951) tentatively itiicated that ratings on personality tralts wr predle tiv* of ftwore behavior. Topes (19S7. 1959) in studies using Air Faux officer cmadidftes and seal.r Air Force officers as subjects found wer ratings on personality traits to het predctive of later performance as sec.n lleiteriests is the case, of the officer candidates. cad to be reluled to concurtent but iftdeF~z!es itmeoW*3 of offiarm performanCe in thie latter groupi. Pitaethee. the profiles of the correlotions betwee, personality traits an officer performance were, quite similar in the two groups. All three of these itudies. as well as one by Mays (19%4). inicated4 thak even though personalityV trait ratings by a single raer may be quite warellahl (ahout .2 to M3)ratins obtained from a group Of rater (10 tIQ 20 raters) when summed iied~ad scorses whose r4Iabl93e were quite satisfactory (.4 to .9). Io generl. it w*my be concluded that raings on paersreIty traits wre vSfCA predictors of fture wbhavwo Ond that sucht ratings yield sufficiently reliable Individual ,fiffretivgs to bot wieies in themseslves, either lat the study of isdivl~oal differences in personA-lity or as -istere against8 whichi oher types of persafnelity mveaer-8 (for **ample. paper-m4.-peacil tests) may !wcValidated. As .adicale by Eyseock (19%31 anid rrench (1%) man --tndis hvef been reported cerningMa the fector %treciare undeirling persatielity trait rating vortemos. In sp"t of ths facd the domain hes sot bees vt all clearly de'inad. Cottell (194S 1947. 194) bas puhbliod two lector aselyses of men Ond are 0 ofumeit, each besed G"~setiq of 1%persopolity Irnis selacted to reprnsent the entire Persmoelity 6.zeo. Is Weach he foud I i of 12 fectars Wkich he Iao ,danOfie as similar ia the three anITy"e. Fra mean at ths** factors.owe. the factor loadings me so stmall thatom fect.t analysts wou4ld eiwateXto, try to iW4 1rprejt Won at all. Fiske (1949) emelTZ04 rattas of 22 of the

.~

~~No6

of-i...d *,s. I-..S

p~b.5.. An

ASti 1.osS..s4

6%mla A k,rsl

1W1

same or highly shsiarl variables using begiannin graduate students in clinical psycol~ooqy for his sample. He obtained about the saime factorial structure from ratings of the studems by themaclv--s (self-ratiags). by their peers, cand by clinical p:.ychoJogasts&. P~owev~r, a colaparison of the fectnrs isolated by Fiske with those defined by Cattoll is quite, difficult, in spite of the fadt that the variables used by Fisk* in the main corresponded quite closely with those used by Cattell. Same. similarities can be moted between the Cattell aod Fiak* factors. but it is difficult to tell whether the diferences observed ore a function of divergent extraction and rotational phbilosophies, the natare of the samples rated, the nature of the rater groups, or the omission of 13 of the trait vmuiable* from the Fisk., study. Attempts to compae. the results of either the Fiske or Cattell analyses with those found by other investigators ame generally futile. since it is rarely possible to determnine from the studies whether all, some, or for that matter, any of the variables used ore similar from one study to mother. When what might he -Prureat factors ame found (&.g.. extroversion-introvorsion, exiotionality-stability. and confomnity-indepeedence). differences in the nature of variables identifyinq theme factors ate such as to makeo impossible any but subjective Judgments as to their possible -limilaritios. The present study was designed to help ciarify the personality ..ait-ratinq doad!. Te 4;oal was to isolate meaningful and relativr ly independent trait-rating factors which ore universal enoug to app - in a variety at samples. manwhich me not undulv sensi'ive to the rating cooditioas or situations.

ItCT11)D Eight intercorrelation miatricesi were fainterd and roated orthogonally to approximate simple structure. 2 The matrices were selected in such a way that differeocts dos, to the ttaiha tuted would be minimized. while differences in the type of subjects. raters, and situations would be nmaximi~d. Mon* of the analyses were carried out "blind" (without identification of the vormables). nor were they made isdepoodeetly of one another. The goal was to rotate the separate faictor matrices anto similar structure while at the ame time following accepted prisci,-!ss of rolatios and arriving at siampl. structure. For comiporson purposes. one of the solutions w-- redetrmised in a completely objective mumme by sublectiag the contrei fuctors to a varimex rotational 1 -9g-anonanIM 610 computer. The trait variobles eaturagintoW *a*c analysis were on" the IS developd by 'tel(1947). wko *sed as a basis the comprehensive list of adjectives originally ideelif led by Allpost & Odher (1961Q as describing heom bhabvior, Coch trait is bipolar. with each pal definesd by a sabrt ,eo of .idjettives or phrozes. Thase begst orv believed cspeciaily appropriate for as iawestigetiom of the trait-rating domeai since the methodt boywhich they were developed lod to omw acommne that Othy 4"e4rereetive ut the entire perunaslity arme. The bWpoliff s~of these traits appear in Telile I-& below. for their defeisig edkectew,. cc phru** Ohe rede to Mrefrd to Ca~ttal (19471. risk* (1941. or T"a (1917). Three W1the istercervelet Weam .eticwaowe aeseo Ai# Force '3fIerk Coadidire School sub-. Owts who ruled eacls other is varboas sited q, -p-j One analysis is based on Air Fiur, fiel grade offices (Malors and a4. tootiotenewn PleseIs) who r-.% each etber while meadeows at the Ar orace Coman un Staff School. Too analyses we reetoosins of analyses published by Cattell (1947. JIM S is which the subjecs are mal*eand ..Ideml coewlq el~tuscet. The two 4" awyevys we !-*to on two of rli&ea (UM4) .Mrcue itse ues -4 rslimg of first-yae gradute students is c'.lical payhoegyi . tv the twoe cf ohe.. Mains" weft oomtelendk peers. in the second,. rwatus ware obtair*d ben emporieftod cismIce p*yclilegists Ond psychiatriats. aSW.o Aw.. soot?*" ovv.w em

eat

"*we vei

2

w..

Ows....

aesi.

All groups of subjects oDd raters are dearibed below. Briefly, they difftr in length of acquaintanceship from three days to a year or more; in kind of acquaintanceship from assessment programs to a military trainin:j cnurse to a fraternity house situation; in type of subject from airmen with only a high Rrhool education to niale and female underqraduate students to first-year graduate students; and in type of rater from very naive persons to clinical psychologists or psychiatrists with years of experience in th,: evaluation of personality. It would appear that any factors common to all of these groups would hove a wide range of generality both in terms of type of subject and type of rating situation. DESCRIPTION OF THE EIGHT STUDIES STUDY A.

OCS 790-CASE SAMPLE

The subjects were 790 male graduates of OCS Classes 49B, 50A, 50B, SOC, SiB, and SID. The earliest class, 49B, was graduated in December 1949; the latest, 51D, was graduated in December 1951. All candidates in each class had been selected from a much larger number of applicants (selection ratio about ten applicants for each vacancy) on the basis of a board interview, a biographical inventory designed to measure leadership characteristics, and differential credit for completion of more than the required minimum of two years of college. For applicants on active duty in an enlisted status, an evaiuation form completed by the applicant's commanding officer was also considered. The a •erage age was 23.6 years, with a standard deviation of 1.5 and a range of from 20.5 to 26.5 years. The average education was 3.6 years of college, with a standard deviation of 0.6 and a range of from 2 to 6 years. Distributions on both variables were decidedly skewed toward the lower end. Slightly over half of each class came from an enlisted status, with the others selected for OCS directly from civilian life. Each OCS class was divided at the start of training into flights of from 25 to 30 candidates each. Each flight lived together in one dormitory, ate as a flight, and attended classes and drill as a flight. In fact, nearly all of each candidate's time was spent with his flight, and he soon became intimately acquainted with each of his fellow flight members. It was the well-organized OCS flight which constituted the rating group in the present study. Each candidate rated all his fellow flight members and was in turn rated b7 all his fellow flight members on 30 of the 35 Cattell traits. Each rater was required to pick one-third of the group as best described by the definition at each end of each bipolar trait. Lengths of acquaintanceship at time of rating varied from as little as three weeks for one class to one year in another (this class rated each other six months after graduation from OCS at the end of an on-the-job training period at Lackland Air Force Base). Product-moment lntercorrelati4,n matrices of the 30 traits were computed for each class separately. A final matrix was then obtained by taking the median correlation between each pair of traits in the separate class matrices. Eight factors were extracted from this matrix using the complete centroid method, and rotated to orthogonal simple structure. STUDY B. OCS3-DAY ASSESSMENT SAMPLE

The subjects were 125 male officer caaididates in OCS Class 55B, whose ages ranged from 20V2 through 27. A little more than half had no college training; about a fifth had some college; and about a fifth were college graduates. All had some previous Air Force enlisted service rangIng from one year to seven, with a mcdlan of 21,1 years. The majority weie planning on an Air Force career and all had been required to sign a contract for three years of commisslo,-ed service after

graduation from OCS. All had been screened on a measure of general learning ability-the Officer Quality composite of the Air Force Officer Qualifying Test. Eighty-five per cent of the class }•ad OQ scores as high as the upper 10% of the genoral population of young males and as the upper 40% of college freshmen. 3

0

,i

Ratings were obtained at the end of a three-day assessment program just prior to the start of OCS. Rating groups consisted of 12 candidates, six of whom had observed each other in an intensive series of group and individual performance testr, and six of whom had only shared a barrack floor anc dining table with the other six. Each rater was required to pi:k the four subjects who were best described by each end of the- bipolar trait. Five multiple group factors (corresponding to the five found in Study A) were extracted, along with three centroids. All sight factors were rotated to simple structure. STUDY C.

OCS END-OF-TRAINING ASSESSMENT SAMPLE

These are the same subjects who were used in Study B. At the end of the assessment, the groups were re-formed into OCS flights of from 15 to 20 candidates each. No two flight members had been members of the same assessment group. Near the end of the six-months OCS course, members of each flighL rated each other on the 30 traits. Raters wee asked to pick the third who were best described by each end of each bipolar trait. These ratings, although based on the s-rme subjects, were entirely independent of the ratings amalyed in Study B.

Five multiple-group and three centroid f.-tors were extracted from these data and rotated to simple structure. STUDY D.

COMMAND AND STAFF SCHOOL SAMFLE

The subjects were 500 students in the Air Force Command and Staff School Class of 1958. These officers had been screened originally or. about the same basis as the OCS samples. However, at the time the trait ratings were obtained the average Command and Staff School officer was about 15 years older and had approximately 15 more yeats of military experience than the average OCS subject. Nearly all of the officers rated held the rank of major, although the sample included a few holding the rank of lieutenant colonel. Hatings on 30 of the bipolar traits were obtained on these subjects after they had been in attendance at the Command and Staff School about two months. Each rating group was composed of from 12 to 14 officers who attended all classes as a unit. Only a third of each group served as raters; these rated all members of the q. oup by selecting the four subjects in their seminar group who were best described by each pole of each trait. Only five multiple-group factors (and no centroids) wsre extracted and rotated to simple structure. However, at a later time factors were extracted from the intercorrelation matrix by the complete centroid method and rotatedorn rm IBM 650 computer using the varimax program. sTUDY E. CATTELL'S MALE UNIVEAS-TY

SAMPLE

Suiects were 133 male university students with an average age of 20 yeas. Some were returning veterans. Ratings on 35 bipolar traits were obtained in groups of 17 men, all of whom lived together in fraternity houses or dormitories. Each rater rated all members of his group on each trait as below average, average, or above average on each trait, with a suggested distribution of '4X ., and '/. foz the three categories. For a complete description of this sample see Cattell (1947). The intercorrelatio,* mctrix 3 was factored and rotated twice. In one instance, eight centroid factors were extracted and rotated; In the second case, five multiple-group factors and three controid factors extracted from the resulting residual matrix were rotated. Only the latter solution is reported, since .he solution based upon the centrold extractions was discovered to contain errors introduced by the graphic rotational procens.

3 Mad* av.~ltbl* throuqh the courtesy of Prof e~r Table El, with his pormiemlun.

4

•qtymqnoed 8. Cattill esd regr4eaed in the Appeadix.

STUDY F. 'CATTELL'S FEMALE UNIVCHSITY SAMPLE

The sub~ects were 140 female uriversity students. Ratings on these stu~ents were collected from their peers (nil women) at the -ames time and in the same mcmnner as in Study E. This sample is more completely described by Cattell (1948). Since this was the only femal,ý sample studied, it was considered desirable to in'!!udp several moie factors in the analysis than might reasonably be expected to exist. Therefore 12 fr4aors were rotated to simple structure; five~ of these were ortnoqonalized multiple-qrroup, factors and the other 4 seven were centroids extracted fram the residual matrix. STUDY

G.

FISKE'S TEAMMATES' RATINGS OF GRADUATE STUDENTS

The subjects were 128 male graduate students in clinical psychology who participated in an intensive ass.essment program during the summer before they started their graduate training. Their median age was 26, near'y all were veterans, and neari:' al! had World War 11eisraiat~ce as military psychologists. During the week-long cssessment, they ate, roomed, and took their recreation toaether. 'wenity-f our trainees were assessed each week and were sriI # -itrarily into groups of four who parv ticivated in a series of situational tests. At the end of the week, each subject raed himself and the other three members of his aroup on a series of Variables, including 22 bipoltu personality traits. Rating= were made nm an eight-point scale. The three ratings made, on each subject by his three teammates werv summed to obtain the rating scores used in this study. For a complete description of the sample, the variables, and the rating procedure, see Fiske (1949). Five multiple-group factors and three centroid factors were extracted from the correlatlon matrix. STUDY It. FISKE'S STAFF MEMBERS' RATINGS OF GRADUATE STUDENTS

These subjects were the some as those of Sti~dy G. The same ratinq variables and rating scales were usod except that staff members were asked to normally distribute theair ratings on each trait. The raters wor' three assessment staff members assigned to each group of four subjects, and the raing scores were, the num of the ratings made by these individuals. Lach Ltatf member was a clinical psychologist (a few were psychiutrists) with years of experivoce. The raters had not only intensively obsorved each subject during a period of one week but In addition-haid the results of ten objective tests, four prolective tests, a biographical Inventory, an autobiography, and the write-as of three intervieweirs. The staff ratings were made without knowledge of the taeatmates' ratings. Five multiple-grop cad three centroid factors were extracted and - 1,"P4 to o~wohlmatia orthogona I simple structure. RESULTS In each analysis five fairly strong roated factwo emorged. In tue four studies involviin Air Force somples and the two risk* sacnpib.al.ol but bhefive it-ong I actors were reuidusitted (e.g.. none of the residualized fackorz hod loadings am any trait wibble ~ve at arbitrary :30 level). In the analysis of Cattell's mzlv 3ample. two factors were residualized. and a webk factor lseelving primarily an intelligence test was defined. In Cattell'A female aMple. What had bee. ideatified at. the fifth factor in the other nnalyses split into two foctors. Both of thoea factors are reported for comparisoa pmrpose..

41It o theoulk'hit* .pwu.- Tt.i kItoit.c low" ois she .MII-eI 8,leteve tale the :01064R4.4 prowse.. fe"C" of s.0 it'eI se 0ls.4sqte.tly v...s..d Ike effect "ot.*wyt eft. *I-* 0. toeeeeeby eeed 614010 It *1**e 1@4 1* escoaftbvtly. Melth as.inrsGs 0.4tbel tofu r"01 "4 othe efmwse M114eted

lobe glegisi *oVtrwtr*.

ir

tho

tat:.%

The five factors are reported in Tables 1 through 5. To save space and to make comparisors easier. each factor loading is presented to only one decimal space (e.g., .6).s In each table, the trait rating variables are listed in the first column, grouped togeth~er in accordance with thi. fuctoi to which they make the highest contribution. Thus the first group are those variables defininq Factor 1, the~ second group are those defining Factor 11, etc. Each of the other columns shows the loadings of each variable an the appropriate factor in each of the eight studies. These columns are Pr-1 beled A through H to correspond with the letters assiqned to each sample above. Traits not rulted in any particular study are so indicated by ..m X in the appropriate column. FACTOR 1: SURGENCY

Factor I appears to he that labeled by Cattell (1947. 1948) and French (1953) as Surgency and by many other investigators as Extroversion. It is hii-1 defined by the traitst Tillkat'veress, Frenikness, Advrenturone..nss, Assertiveness, Sociability, Energetic, Composed, Interest in. Opposite Sex, and Cheerfulness. It appears to be a 'rue bipolar factor with negative loadings (.1 or greater) obtained for the traits of Emotional Maturity, Mildness, Kindliness, Conventionality, and Cnlr Paess. TABLE 1. Leading& on Rgecoreet Fatter Trult Vw~ie~ble he. 4..

from Elgue Analysts

J5

Analyis

k

.L

_L

A.

5 5 .5

.5

.5

.9

.9

."

.7

.7 .7

.s 7

.5 A

.7 .5 .7 X X X

.4 .7 .5 .6 .5 .5

.5 7 .9 .7 .4 .6

.7 X .5 X .4 .7

.1 x

.31 .2 13 .- 1 .4 .0 13 3 S5 .. SI

.1 _. -. 1 -. 2 .4 .3 .2

.,J X X X .2 .3.3 X

X X X 1. .0 .3 X

.7

-5

.6

.. s

X

14 25 16 I 29 7 33 35 32

Silent vs Talkatve Sectetlv. weFrank routi~uo. w. Advviuroran Submissinve we Assertive S.u-centahz,.d vs Sociable Langui~d. Slow to Frm..qe&c shy. Bashulweaoempoa* SIjqbl we h4-sked 3st~rv..t i" OpuPei,. Srz Depressed we Cheerful

.8 .0 .5 .7 .7 .7 X X X

.7 .7 .1 .7 .7 .7 X X X

.7 .5 .5 .8 .4 .7 X X X

IV 20 32 13 I 9 21 17 5 33

Spiteul vs Gooda.tiareJ J*6iouswe NeitSo DemandingvwieEmtionally Unto,.e Sell-wi&Ued we Mild Obstructive ws Costierotive fuspicloue we Tfustlut Alaid vs A'44pleae Hord. 31*m weKindly Coal. Aloel vs Altoeitaws to People Atgootie. Gatti eq we Sell-sauficisaI

.1 .. .. 1 .1 .2 . .2 I1 $1 X

.0 -.. -.5 -. 5

.1 .1 .. 1 .. 2 . 13 .2 .4 -.641 .3 X

-k .. 1 -. 4 ..; -. 2 13 1 1 .. .4 X

1S 4 3S IS a!

Male3i.u.4Loidelet estvnalten2ivyOtd~ti rtviiii we ltespinaat3 Unscruapulous we Coancl~entisws Quitting we Per*...l,01 Usceswe"164001 weCoftweanune3

..j .C .4 .31 .4

.. 1 .0

-. 2

..Y .1 -.2 .31 -.4

is

Is

we

.3 .So

NOW14~ we lsteL.1iiveol. Cultured Looking AIsec roeting we reshsoI r..ttice3. iet..u.imtaginative Clumsy. Awbwos4 W,,Pubasbe" emetafe -. 1.4edtne

"Oo~.

-A roll entry .i Xrdeno~tes

see-Sd soAppeodlt

~..2

.2 .. S

e0

-. 3 .3 -.4

.*I

Weuavoticvetw e So. Worrying. Anslows vo Placid 6 Costly Upeol we Poised. Tough 32 lypoehbssdutetelvwe Sol I3I tao.felmovC3 we C41 I Chaeng.eable "W. .. v..es. 17 Dependent wesoeli..o16ihdtt

Is 24

a 27 34

.3 .2 .S .4 X

Feitc-t odio.

6

.2 . 130 .4 ý2 .5 .1..

.e X

K

.0 .0 K

1a .. 1 x

1

.0

.2

.0

.6

.6

4

w.*6e.o -@1 wood in atwp.

A-0.

.

5 5 .3 .0 X . .3 -3I K

I3

.. I 13

..

.. 3)

.4 .3 .b .1 .0 0 .4

~..3

.. I .0 .2 .3I .- 1 .0 .1 .2 -.3I .4 .0 X

.2 .3I

a0 e0 .-

.0

.0

3

.-

K -. 4

-. 4 X K .. 1 .0 K 3I -.. I .2 X .I .0

3

Hi

.7 X .4 .5

X X

~..)

-. 3 X x K .3 .3 K K -..2 .2 .2 X .I -. 1

FACTOR 11: AGREEABLENESS

This factor corresponds quite closely to that ca' led Agreeableness by French (1953). It, too, t- a bipolar factor, defined on the positive end by the variables Goodnatured, Not Jealous, Emaotionally Mature, Mildness. Cooperativeness, Trustfulness, Adaptability, Kindliness, Attentiveness to People, and Self-sufficiency (as opposed to Attention-Cettinq). Traits loaded negatively on Factor II include Assertiveness and to a lesser extent Talkativeness and Orderliness.

TABLE 2. Trait Variablle Me. None

Loadings on Recurrent Fact.,

A-

a

C

D

F

6

H

.0

-. 2

..!

-. 2

-.

-. 1

..2

.1

.1 .1

.1 .0

.3 .1

.0 -. 2

-2 .0

.2

.3 .2

-. 4

-. 4

-. 3

-. 4

.2

.1

.2

.1

-.2, .1 -.6 .0

.0 X X X

.0 X K K

.0 X X K

.0 X X X

.2 -. 1 .0 .3

.1 .0 ..1 .4

Spiteful vs Goodnaturea Jealousvamet So De...ndia vs Emaztlaal.Uy Msatur Salt-wilwlwed vKIM obstructive vs Comaerative Ss~itCleve vs Trustful Rigid vs Adaptable 1 ird. St..m vs Kindly 'lAloof wsAtesatv. toPePtle AltetaSoa Get9inag 6& Seli-asiiieet

.3 .6 .6 .7 .7 .4 .6 .6 .7 X

.7

.8

.7

.7

.5

.7

.8

.6

.6

.6

.0. .6 .1 .S .4 .4 .5 x

.1 .7 .7 .7 .7 .5 .6 K

.6 .6 .6 .6 .4

.7 .6 .6 .6 .6

.6 .6 .6 .7 .6

.3

.7

.7

K

.5 K

.3 .4

.7 .6

.4 X

.6 X X K .7 .7 .7 x .3 X

Rte''am.dIndoenatvswsslsstoly Orderly Frit-elous we Respasibie tVasc.up"ie. I vs Conscenetious ,it jswepersevering "Jneoaventieasi vs Cjmv~aeftuii

.. ] .3 S3 .3 .2

-. 2 .1 . .2 .3

.23 .4 .6

..1 .2 .4

-. 1 .3 .4

..1 .4 .6

X .0 .3

X .2 .4

.4

.0

.. 1

.0

X

X

.4

.3

.2

.3

X

X

.3 .3 .1 .4 .4 .1 X

.1 .3 .. 1 .3 .1 .6 x

.4 .S .2 .4 .9 Ar K

.3 .1 .0 .2 -J .3 K

.2 .0 .0 .31 .1 .2 X

-$ .2 -..a .4 .4 .6 X

x A.4 -I 3 x K .4 .2

K .3 x X .2 .0

.2 1&~ K -.. .1

.2 .0 x .2 .4

.3 .0 K .2 .1

.1 X .0 3 4

.1 .6 .3 3 .6

Analyss

14 26

Silent vs Talkaotive Secretive vsFrank

II 3 .9 7 33 35 32

Cautious vs Adventurous 3Subaissive weAssertive Sall-cmutaimed vs Socks).!. Languid. Slow vs EnergaUc Shy. BashfulvweComposed Slight vs Marked Interest in Opposite See ZEjpressed vs Ch~edui

10 20 22 13 I 9 21 17 S 31 is I is I$ 23 2S 34 a Ij II

27 .4

11fremi Eight Analyses

Neurelti vs Not S. Is~wyksa Ansjuus vs Placid 1,06. L,Upset wePalmed. 740"It H1p. toh0*dtleta vs HtI S £Emtional vswe a 2Ck..a.40abllvs I'sat~stietey Stable 37Dopadow ve SaiI-stii sat 8401a4ih vs "nteilortwol. Cuohared Lerikint Artistic Feeling vs 3sthwA'coi~j r'rsstlsa. LogiCal *a la0q440SUVO 38 lensy. Awkwo#4 ". Pollak" lo&.aursettfý*ade*.sieal-mladed

rest,

Waft.-A cell entry of X ilsoee. amabolo not wood lon Atudfy.

7

E

.1

.3 -

. .1 .1

:1 .~.4 X .0 X .0 .3 .7 X X X .6 .6 .4

.0 X .3 .1 -.4

-. 4 X .1 X .2 .4

.1 K . .1 .

FACTOR 11:

DEPENDABILITrY.

The primary definers of this factor are Ordelianesn, Responsibility, Conscientiousness, Perseverance, and Conventionality, with several other variables (Cooperativeness, Mildness, and Emotional Stability) having positive loading* above .3. Practically all definers of Factor I are loaded negatively on this factor, as are Trustfulness, and Imaginative. The factor in many respects is like that labe!.. 4 bhyFvench (1953) as Dependability or by Fiske (1949) as Conformity. It appewrs to be quite similar to the old "w" or Will factor found by Webb (1915).

TABLE 3. LoedIags on Rocuwow PFecto III from Eight Analyses Trit Vmri-%Ie me. NaW

A

B

C

D

E

P

G

H

14 28 16 3 21 7 33 3S 32

Silent v Talkative SecreUve ve Frank Coatioas vo Adventurous Submisivvov8 AssorUve Sll-cntmaunod vw Socibleo Lamguid, Slow v Ennorgetic Sky, Bashi:,l ws Composed SL!qbt we fMaked Inteeoot in Oppogdto Sez Dopressod vm Chqodul

-. 2 -. 2 -. 4 -. 1 -. 4 .3 X X X

-. 3 -. 3 -. 1 -. 3 -. 4 .0 X X X

-. 2 .0

-. 3 '.1 -. 4 .. 1 -. 4 -. 2 -. 3 -. 2 -. 4

-. 3 -*2 "., .0 -. 3 .0 -.2 -. 3 .. 3

,,0 .0 -. 4 .0 X -. 1 X -. 4 ..1

-,1

-. 2 -. 2 .1 X X X

-. 2 -. 1 -. 2 -. 1 -.4 -. 1 X X X

10 20 22 13 I I 21 17 S 31

Spiteful we Goodneturod Jealous wv Not So 'nomendinq vy Emotionally Maist stil-willed vs Mild Oi.tnaegivo v CooperatUve Suspllcous vs Trustlul RlqitA we Adoptbleo Hurt. Stem wv Kindly Cool, Alool vs Attentivo Re Poeplo Attoatimon Getting w, Soll-aullicies'

.0 .0 .2 .2 .4 .1 -. 3 .0 .X

.0 .0 .1 .1 .5 .2 -. 2 .2 -1 X

.3 .2 .3 .4 .5 .1 -. 2 .2 .I X

.0 .1 .2 .3 .2 .2 -. 2 .1 .1 X

.1 .0 .2 .3 .3 .0 .. 3

-. 2 .0 .2 .1 .3 .0 -. 1 -. 1 .0 1

.0 X X X .2 .1 .. 1 X .2 X

.4 x X X .3 .3 .1 X .I X

It 4 2S IS 23

Rolosed. ndoleoot Vs Insistently Orderly Flvelon ss Vs Rospunsiblo UnsclvpuiouS vs Coesci.tioun QuittpUq vw Poeoodvrtq Unconventional vs ConvoeUnsol

.7 .6 .5 . .6

.4 .4 .5 .3 .4

.5 .6 .6 .4 .6

.S .6 .6 -S .4

.6 -6 .4 .4 .,

.7

X .7 .6

X .7 .7

x

x

X

X

.1 1. .0 ..2 :I

.0 .0 -.. .1 1 .3 I

.3 -. -.A .0 .1 .3 K

.1 ". .0 .0 .3 -1 I

.0 -. 2 -.. .1 .1 .Ia

.1 ".1 .1 .0 4 4 x

X -0 .1 X X .4 -:

.2 .2 x .1 .0

.0 .2 x .2 .0

.,

.

.3 .0 .. 1

. . .. $ .3 .2

.3 I x .1 .3 .1

A

26 NHos•.,c vo Nol So Z4 Wffo.sq. AnninomawoPlacid 6 Easily UpsIeve Pesi*. Toug I2 I ypockoodrioecl vs Nt So II L•otluel we Cota 2 Chageoablo e Emelieaoolny S1obi 17 Dopeedoot we vs oll-0slliel! 1 27 34 19 30

116es"1" v*a stolle,*N.el. rolutovd Lckinq Artitoic roollag ,s etuoi~cosi, rt.oniaco. Logeiel svo Cloa=y. Abwowd vs Polimold Ins.6toqo v wsosaolM dd

.

roaludivos

M1..--A e.oil entry of X doselos verieble oei usod &a%tudy.

5

.,?

.2 X ..2

A .4 .2 .

*k

.0 .4

..

.0

.o

.3 .7 .5

.2 ,1 .1 X .0 X -. 4 .0

x -.

.I X x .4 .2 .0 x .0 .1 .2

ERIATUM Tupes, E.C. & Christol, R.E. Recurrent patsonaJity Ictor=z bod on trait ratings. Licklaid A;r Force Base, Texas: Personnel Laboratoty. Aeronautical Systems Division. May 1961. (Technical Report ASD-TR-61-T7) Page 8, line 5 For

. . . loadcd negatively on this factor, as are Trustfulness. and Imaginative.

Read

. - loaded negatively on 'his factnr. as at, Adaptability and Imaginative.

FACTOI 17: EMOTIONAL STABILITY

Tthe inverse of this factor seems to be that listed by French (1953) as Emotionality. It .s loaded highest by Not Neurotic, Placid, Poised, Not Hypochondriacal, Calm, Emotionaily Stable, and Self-sufficient (as opposed to Dependent). Seconaary definers of the factor cte Lack of Jealousy, Emotional Maturity, Cooperativeness, Trustfulness, Adaptability, Res!winnsibility, Perseveranc-ý, 7!nd Indapendent-Mindedness. Kindliness has a significant negative lo~ading on this factor.

TABLE 4. Trait Variable me. NMeM

Leandings on Recurrent Factor IV from Eight Analyses

Analysts

A

a

C

D

E

P

G

H

14 26 10 3 29 7 33 35 32

Silent vsTalltotive Secretive,we rank Cautious vs Adventurous Submissive vs Assertive Self-contained vs Sociable Lanq'.id, Slow vs Energetic Shy. Bashful vs Composed Slight vs Marked Interest in Opposite Sesl Deipressed vs Cheerful

-. 2 .1 .2 .2 -. 1 .3 X X X

-. 1 .1 .4 .4 .1 .2 X X X

-. 4 -. 1 .1 .1 -. 2 .1 X X X

-. 1 .0 .1 .2 .. 1 .1 X X X

.0 -. 1 .1 .1 ..1 .4 .2 ..1 .1

-. 2 .0 .2 .2 -. 3 .1 .2 -. 1 .0

.2 .0 .2 .1 X -. 2 X .2 .3

.0 .0 .2 .2 X -. 4 X .3 .4

10 20 22 13 I 9 21 17 5

Spitela' vs Goodnatured Jealous VSNot SO Demanding vs Emotionally Mature Sell-willed vs Mild Obstructive vs Cooperative Suspicious vs Trustful Rivid va Aduptable Mrrd, Stern wo X,,d~j Coal, Aloof vs Attentive to People

.2 .5 .4 .3 .3 .6 .4 ..4 .1

.2 .3 .3 .1 .4 .5 .4 -. 3 .0

.2

.1

.0

-. 2

.1

.1

.4

.3

.4

.1

X

X

.2 .2 .3 .4 .3 ..4 ..-

.2 .0 .1 .5 .0 -. 4

.3 .1 .1 .4 .2 -. 3

.1 .. 2 ..1 .3 .1 -.5 .1

X X .1 .1 .3 X .0

X X .1 .3 .1 X .1

31

Attenticm Gettarg vn Seli-suiiici,..t

x

x

X

x

.4

.1

X

y!

Iii 4 IS IS 23

Relaxed, Indoleinz vs rinsistlortly Orderly F~rivolous vs Rtesponsible Unscrupulous vs, Cons~cientious Quitting vs Perssv ;ring Unconventional vs Conventional

.0 .5 .2 .4 .1

.0 .2 .2 .5 -. 2

.0 .4 .2 .4 .0

-. 1 .2 .1 .2 -. 1

.3 .3 .0 .4

.6

.0 .1 -.23 .2 -. 4

X .0 .0 X X

X .0 ..1 X X

16 Nourotic vs Not So 24 Worrying, Anxious vs Placid 6 Easily Upset vs Poised. Toulih 12 "71ipschundrfiecel vs Not 36 I I Emotional vit Calm 2 Changesable vskrmetionally Stable 3? flepead....t vs Sell-sufficients

.7 .7 .7 .7

.7 .7 .5 .6

X .7 .7 X

X .8 .6 X

.11 .6

.1

I 21 34 19 30

.2 .1 X .7 .5

3!

141"orish v. Intellectual. CIultued Lacking Artistic rooting ws Estheticully Fast~dt..,Aa Ptna-ttrel. Legical va Imaginetfive Clumsy. Awkward vs Polished Immature vs Indeopendent-mindeld

Met*

A

.II

0-,v

..i

%'.40-10.

v..rftki

.4

not uned in s.1tov.

9

.1 .1 X .1 .4

.1

-. 1

..

S5 .6 .6 .6 .6 .6 X

.6 .7 .S .5

.7 .6 .6 .S

.S .6 .7 .5

.4

.4

.5

.4 X

.4 3!

.4 X

.2 .4

.4 .6

.2 .0 X .2 .5

.2 X .1 .2 .4

.3 .,1 ..3 .1 .5

.1 .0 -. 4 .0 .4

.2 X .0 .3 .3

.1 X .0 .6 .

X

X

FACTOR Vs

CULTUIRE

Factor V is the Itast clear of the five factots identified by the elqht analyses. It appears to be similar to the factor labeled by French (1953) as Culture and by Fiske (1949) as the Inquiring Intellect. It is defined by the variables, Cultured, Esthetically Fastidious, Imaginative, Socially Polished, and Indopendent.Uinded, with secondary loadings by Energetic, Poise, Emotional Stability, and all the variables in Factor II. It will be noted that loadings for two factors are shown under Column F. This is the analysis of the female colleqe students, and in this sample only, Factor V split into two quite distinct subfactors. The first of these has a pattern of loadings quite similar to the Factor V found recurring throughout the studies. The second of these is defined by the variables Esthetically Fastidious, Socially Polished, and Interest in the Opposite Sex.

TABLE S. Loadings en Recurrent FPoct. V from Eitig

Anelyses

Trait Variable

NAnalysis 14 28 16 3 29 7 33

l

32

SeIen we TalkatiUve Secreuveve Frank Cautious vs Adventurous Submissive vs Assertive SeUl-contained vs Sociable Lviguld, Slow vw Energetic Shy. Beshful vs Composed Slight vs Marked Interest in Opposite See Depressed vs Cheerful

10 20 22 13 I 9 21 17 S 31

Spiteful vs Goodnntured Jealous vs Not So Demamdlng ve Emotionally Mature Self-wl3ed v8 Mild Obstructive vs Cooperative Suspicious vs Trustful Riqid vs Adoptable Herd. Storn vw Kindly Cool, Aloof vs Attentive to People Attention Getting vs Self-sufficient

.0 .0 .1 .0 -i .2 .0 -. 2 .2 X

.0 .0 .0 .1 .3 .2 .. 1 .0 .0 X

.2 .2

-.,

.1

.0

.0

.0

-. 2

.1

.0

.2 .2 .3 .4 .4 .3 .3 X

.0 -. 1 .2 .0 -. 1 .. .1 X

.1 .1 .3 .0 -. 2 -.. .4 -. 2

.1 .0 .2 .0 -. 2 .0 -. 1 .0

1I Relexed. Indolaet vs Insistently Orderly 4 Frivolous vs ftesponsim3e 25 Unserupulous vs Conscientious I5 Quitting vs Persevering 23 Unmenventional vs Conventional

.3 .3 .1 .4

.6 .4 .4 .2

.6 .3 .4 .4

.3 .4 .3 ,5

.4 .2 .3 .4

.-

.2

.3

.0

-.3

.1

.2 .1 .3 .

.1 .0 .4 '-. 2

-.2

YS

26 24 6

Nowetic vn Not So Worrying, Ansious vs Plecid Eaosly Upset we Poised. Tough Itypohandrineelv Not So

Dependent ve Sell-salicient

27

Loeehio

34 Is 10

Autistic Feeling we Isethoticalny restdiouti

Proctireel, Logi1cal te ImagnativeU Clemsy, Awkwerd we PolisheJ lemobnt we IftogdslenIn-Milnded

Not..-A cell entry of A denotes veriable not u#e9

-. 2 -. 2 .. 2 .1 -. 1 .2 X X X

-. 2 .0 ,1 .0 .0 .4 X X X

.1 .I .4 .1

32 II Eotionalvn we .m 2 Chengeable vs Em.loonelir Stable 37

.0 .1 .1 .3 .1 .5 X X X

-.1 .2

.2 .0 .1 .3 .0 .4 X X X

3

.0 .2

.3

.-

.2 X

.1 X

.3 X

.

6O

.6

.

X .1 .5

in study.

X .$ .4

X .8 .4

.0 .2 -. 1 .2 .1 .3 .4

.0

.1

.0 .0

-.1 .I

X

.S X

I

.X

.S .6 .6

-..1 -. .0 -. 1 .0 . .3 .1 -. 2 .2 .I .1 .3 .2

. .6 -S

.,2

.4

-. 2

.2

£

i

.2 .2 .4 .2 X

.1 .1 .1 .2 X .1 X

.-

X

.2 .3

-..

-. 1 -. 1 .1 .2 .2 .0 -. 2 -. 1

.2 X X X .2 .2 .5 X .3 X

.0 X X X .0 -. 2 .4 X .0 X

.2 .0 .S .. 1 .3 -. 2 .5 -. 1 -., .0

X .0 .2 X X

X .2 .1 X X

-. 2 .1) .4 -. 1

.0 .1 .1 .0

X . .1 X

.I .3

X .0 .3 X

.I .0

X -.2

X -.3

X

X

.4

,3

-

.6% 2 .2 - .5 .S .1 .. 1

.0

X

. 7

,7 .4•

.4

.

D SCUSSION The results of these analyses clearly indicate that difrerences in samples, sit;,atins, raters, and iengths and kinds of acquaintanceship have little effect on the factor structure i.n,,-Lying ratinq•s of personality traits. Statistical tests are not needed to indicate the similarity of cwxresporiding factors from one analysis to another. There can be no doubt tnat the five factor.; found througbout all eight analyses are recurrent. In evaluatin3 the results of a series of factor solutions, such as thns' pres,:a' ed ir Tab .1esj through 5, it is natural for the rea,i,-r to wonder to what exten': the results .-.iqht ef •!-t b gses• T_ e 2 oe'ry part of the authors. There is little doubt but that the words "simple structure" ha,- ,0._ ýý j be fa-.ed e&•nyn loosely by many analysts, and it is also undoui.'.dly true thrt.,a preconceived .•-, through a little "forcing" durinq the rotational . -.cess. The first factors rotated were those from 'he .:"J-case OCS sampie described i.i Stu Wy , K. Wi these rotations were not made bli.,;, they were riau with no preconceived notions us to I o.. the ftnal soiution should appear.. Even so, there wore certuin "choice points" during t eot,.L .u;: il pr.,ce-.s at which somewhat arbitrary decisions were ,-r,3de. These are the same types of decision: v. ..ch ac,-; familiar to all who have participated in orthogonal graphical rotations. In the main they -e If two types: (a) those concerning final positioning of reference axes when there was a choice f f ivorinc one or the other of two factors or of balancing the two; and (b) those concerning 'ihether t 3 t: i*enpt the buildup or residualization of weak factors introduced into the rotational systirm. The ui -s rimple structure do not provide clear guidance in either event, and the rotator is generally left w'th the job of imposing some subjectivity in deciding which alternatives best fit the criteria. The choice on final positioning of the reference axe:, is probably not too c 'itcca;, .;I. c'. 4A generally affects oniv the relative magnitude of the loadings on the two factors :onsIdere. a' .d doe" not greatly affect the pattern of factor definers. The dec. ion concerning the b- ildup or ic .. uali2ation of weak factors is considerably more serious, and whether the cioice goes one way v. ti e othe'r can aiffect both the number and nature of factors reported. Individuals seem to .u.ive at their decisions ir m,any different ways. Gc, erailly "he '.'nc' ptn'tioning of reference axes is subjective, although it many times is tempered wit I reascri. L. rtgard 1-i the rotation of weak factors, some preier to be guided by one or more of the tw !nty ccd'. mui ic -natic:•! criteria which propose to estimate the true rank of the original intercorrelation Ytatrix. Unf vr unately, the various criteria often do not aqree, even when the beginning communality cstiniates are ic.entical. Other individuals prefer to over extract and fight the battle on the rotation boa:d. If a weal f ictor can be built up into something they interpret as meaninaful they accept it; otherw:se they make a strong attempt at residualization. In the current study the final positioning of rclerence axe- . Study A wis made arbit! inil'i witin the general bounds of acceptable simple structure. Once then " :cisions ha'. been made, k1 .endency was to make choices in the same direction in i,:tu- ana!yses-still staying wi hin the bound. f simplc cI' on ana structure. A variety of criteria were considercd 'n making decisions concern-wiff 'he iw rotation of weak factors. These included sever.,I statistical criteria relating to . 1.IN " k, t ie results reported by past irvesti'•ators anulyzing tne same data, the results of attempted rui. -- ý -iid residualizations of such factors, and, admitted)-,, a little subjective judgment. The actual i u abe" oi factors rotated varied from only five in the Con'rmand and Staff School analysis to 12 in the r nrc,'.•ion of the Cattell women's sample. In every sarr.ple except one t!..,r appeared to be five relati%.1", stronq and recurrent personality factors and .othing r1ore of any consequence. In the Cattel vomer,:,. -aniple, the fifth factor appears to ' ive splil, into two related factors. Subsequent to completion of all eight analyses, j. program became a'.ailabl, for accoml lishinq analytic rotations by means of the IBM 550 computer %.singthe normal varimax cri irion (Kai: c:r, 1958). There are good indications that this complet-.iy objective analytical rotati , procedu v will

•Co9

Best w~a~

o'

ot only save ma~ny hours of labor, but wil! bring considerably moie rigcor to what has thus far beer. '2 it! er loose area. Perhaps the most encou.raging note is that the nricrral varirnux solution appears Ll' !.,nvraiant ur~der changes in t~ie rompos~tirm of L.test I attery. Thus submitting one or more of the 1.11a yses in the current paper for -'ulyi-Ic rutution usi::g- tbhe ro~rma vaijmax criterion would serve at i'a--t two purposes: (al it wo:-!d remove (or .:onfirn') a-,y doiubts rhe reader (or authors) might have r, iz.erning the biases i"! in 'ne reported snhilicns, und (b) it would produce factors likely to he iva.iar~t under chanaý.. i..) in-~ :c: nosition ý,f .1!c ..aii-raflng battery. The Command and tn. t samPi: ;`3tudy M))was selected for onalytic rotation because in 4wt 91cctos'jdmn " most subject to criticism. This is Lecause all the factors were rac~ed by the mult ipie-r- oup n"ethod and oniy five factors weire introduccd into the rotational proces S. if.i ;rder to maximize the imlr-~p-ndence o! 'he new solution, only the .i.tercorrelation matrix was c* .e statistH,ýad servico.; s-ectiua, with instruc~tions to extract eight c,ýrntrcid factors and obtain ojo.%rariinax solution. "he vý7riables in this matrix wiere not identified. It was th- decision of

`f,* cc'n!- iltonts in the statistical services section to rotate only six factors, tr~e List of which was n-Ad.,dl~irrd by the analytic procedure. 6 The fi.'re identifiable factors are reported in Table 6 along v i-h t .c -or-.esponding solution obtained via graphic rotations. 111can be seen that, the two solutions c. - fo. a I practical purposes identical. In every instaatce the lu.adings f-ýr the def iring variables are (:,.Gctly Tie same or differ by only .1. No loadinq differs by mcre than .2. even among the nondefining riabkIn wany ways it seems remarkable that such stability should be found tr an area which to date .S gritnt'-d anything but consisteýnt results. Undoubtedi- 'he consisteacy has always been there, 1!t it itaý been hidden by inconsistency of factorial teefiniques and philosophies, the lack of replicaon u-.ni- identical variables, an3 disr'Treement e.-,ong analysts as to factor titles. None of the fac* :sidiri inthi stdy iie re ew.The hae ten identified many times in previous analyses

,Ilthouc.. hey have not always been called by the samne name%. F -. so. it might surpri.;ru some to tim! ft1e Zaine factors emerging from such a wide variety of ;ample:: ,nd conditions. One interpretation is thcA there are only five fundamental concepts running :h.-nuih ic 35 trait names used ii.- these :;ttt'jies. If the common variance in these 35 bipolar traits cfle-ý -.;Y five fundamental meaning c'inicepts, then it is reasonable to expect these concepts to -3rýp -!to the factors identified in cay -,ample to -luicb the 35 trsits are applied. It si -ruld be noted that there may cAsst little relationship between the magnitude of intercorrelaLons oo-'- ned amnong trait-rating vart-Al'es and he level of inter-ratar agreement concerning which 1.aits appiy to given individuais being rated. Thus it would be possible to identify very strong trait* itig fazt~rs having no practical utility. As indicated above, however, trait ratings bc.3ed on the -tariable;: ncluded in this study n,.t only grant satisfactory Inter-rater agreenien coefficients, but are "liated to '1.tet meaningful criteria. it : s unlikely that the five luctors identified are the orly fundamental per onality factors. There :r.quite I kely other fundamental roncepts involved among the Allport-Odbert *djectives on which the r- ables t sed in the present study were based. The 35 traits (or more accurc ely trait clusters) used ;a the prt:sent study represent the distillate drawn by Cattell fi-jm the interre tionships among some 175 trait -Yhich In turn were selet tejd as representative of the Allport-Odbert idjectives. The comrnun-itRies o' the trait-rating variables ir. the various samples studied are on the whole quite sizable (avcviging .EC to .85); however, for some traits they are as low as .4 or_... Thus many of the traits have spe-.ýic vrieiances greatly in. excess of their rommon variance. In many cases these specific variances wo.!d hec~me common variances were other wyriables to be included In the analyses. Thus it is likely thLt other Lindamental factors may be identified in future studies.

6

0Ont variable 'had a tooat~r. nof .24 on the sLixth factor, ail other variables had loading& below .20.

TABLE 6. Comparison of Normal Virlimx Solution with that Obtained using Graphic Rotations

Trait No.. Nae

_uiab I V G

14 28 16 3 29 7

Silent vs Talkative Secretive vs Frank Cautious vs Adventurous Submissive vs Assertive Self-contained vs Sociable Lcouquid, Slow vs Enerqetic

10 20 22 13 1 9 21 17 5

Spop,'l vs Goodnatured Jeolois vs Not So Demunding vs Emotionally Mature Self-wilied vs Mild Obstru•tive vs Cooperative Suspicious vs Trustful Riqid vs Adaptable Hard, Stern vs Kindly Cool, Aloof vs Attentive to People

-.1 -? -.S -.6 -.3 .0 .0 -.3 .3

.1 -.1 -.4 -.5 -.2 .1 .1 -.3 .4

18 Rela:.:ed, Indolent vs Insistently Orderly 4 Frivolous vs .Responsible 25 Unscrupulous v: Conscientious 15 Quittinq vs Persevering 21 Xlnconvertional v, Conventional

-.1 -.2 -.2 .1 -.5

-.2 -.1 -.2 .1 -.4

26 24 6 12 11 2

.1 .2 .0 .1 .3 .4 .1 .2 -.6 -.5 -.2 -.2

Neurotic vs Not So WaMrying, Anxious vs Placid Easily Upset vs Poised, Touqh Hypochondriacal vs Not So Emotional v Calm Changeable vs Emotionally Stnble

.9 .8 .? .7 .6

.8 .8 .8 .7 .5

.

8

Boorish vs Intellectual. Cultred

.2 .1

-. 1

19 30

Clumsy, Awkward vs Polished Iramature vs Indepeudeat-Mloided

,f, .4

.0 .4

34

Practical, Logical vs lawinactiv*

N..t.-V

.1

JifIL._ .. V G V . -. 1 -.2 .1 .0 -.1 -.2 -.4 -.4 .2 .1 .0 .0 .8 .6 .5 .5 .7 .5 .5 .6 .6.5

.7 .6 .6 .6 .6 .6 1 .4 "1 .51

IV V G

V V.9

.0 -.2 .0 -. 1 -.1 -.2 -.4 -.4 1 .1

.0 ..1 .1 .0 .2 .1 .120 . -.1 -. 1 .2 .1

.1 .0 .0 .232 .1 .3

.0 .0 .1 .1 .1 .2 .3 .3 .1 .2 .1 .2 -.4 -.2 .0 .1 .1 .1

.2 .1 .4 .3 .3 .2 .0 J) .1 .1 .5 .5 .1 .0 -.3 -.4 .0 -.1

.0 .1 .1 -. 1 .3 .2 .1 -.2 .0

-. 1 .0 .0 -. 1 .2 .0 -. 1 -. 2 .1

.7 .6

-.3 -.1 .2 .2 .1 .1 .1 .2 -.1 -. 1

.4 .3 .2 .4 .0

.3 .4 .3 .5 .0

-. 1 -.1 .2 .2 .3 .4 .0 .0 .2 .3

.. 4

.2 .0 -.1 .2 .2 .2

.1 .1 .0 .2 .3 .3

.1 .1 -.1 -. 1 .1 .0 .0 .0 .2 .3 .3 .3

.0

.1

.3

.2 .0

.0

.1

.2

.2 .3 .. 2 -. 1

.2 .3

.1 .1

.2 .3

.2 .4

.0

.0

.6.6 .6.5

.1

. .7

lS

.7 .l

6.5

.3 .4 1.4.4 .1!

.2 .0 .1 4 .0 .4

.1 .1 .2 .0 .4 .4 -..1 -. 2 .2 .1 .5 .5 5-

Vurima- Solution. G - Graphic Rotations. SUMMARY

"4Th, present study was •esiqgnd to help clarify the personality trait-- .Am" domain. The aoal was to isulae meaninqful WmK relatively independent, It-ratiag foctors at .ch ame mivel Ieole to eppt.r in a eisty of samples, and which m not unduly senitive to the ra•tiq conditions or tutal ef 3%personality traits were selected as represeMtotive of the personality demate. Isate.t mas amo"q thest traits were obtained to eiqht samples. These solos differed in lenth of 'tw.rantant- %hpfroc three days to a yew or awe: in kind of acqloimteaecohip from essessms t

I1

programs in a military training course to a fraternity house situation; in type of rnabjcct from airmnen with only a high school education to male and female undergradu~ate students to first-year graduate students; and in type of cater from very naive persons to clinicclI psychologists and psychiatrists with years of experience in the evaluation of personality. Centroid at nsuitiiile-group factors were extracted from the intercorrelat ions and rotated orthogonally to simple structure. For one of tite studies an independent solution was obtained in which analytic rotations were accomplished bv an IBM 650 computer using Kaiser's normal vnzimax criterion. In all solutions except one there app:*.:ctl~ to he five relatively strong and recurrent factors cand nothinig more of any coasequenc In one solution, based upon data from undergraduate women, the fifth fxctot split into two hiqhlj lated factors. The solution obtained by analytic rotations usin. the normal vartmax criterioi was for ail practical purposes identical to the corresponding so~ution obtained via rqik~ic rotations to the simple structure crikerion. The five recurrent factors wert labeled as (1) Surgeacy. (2) Agreeableness. (3) Dependability. (4) Emotional Stability. and (5) Culture. While no claim is made bV the authors that the five factors identifiedl are the only personality dimensions, reasons are given in suppart of tboru fundamiental nature and probable invariance. REFERENCES Ailpert, G.W. Personality: A psychological mnterpretation. New York: Henry Holt, 1937. Alipimet, G.W. & Odbert, H.S. Trait-names, a psyche-lexical study. Psychol. Monsoqr.. 1936. 47. No. 1. (Whole No. 211) Coftell. U.S. The description of personality: Principles and findings in a factor analysis. Amer. I. Psycho)., 194%, 68. 0M-q. Ceweall, R.S. Def-crptioa and meajsurement of personality. New York: World Book. 1946&

Cewell, U.S. Confirmation and clarification of primary personality factors. Psychometnka. 1947. 12. 197-220. Ceeseti, U.S. The prizoay personality fetters in women compared with thoe Posychol., 13".M1. 114-130. E 7 .eorh, N.J.

710 setrucItu

in men. Brit 1.)

of humon personality. London: Mothava. New York: Wiley. 19S3.

risk*. D.V. Cea.sisteecy of the lactorial structure of persoality raisgs frau dufferet Souces. I. aboor'.. soc. Psychol.. I' 19. U4. 3215.144. Pseerk, i.1. The description of Permweeity measurements in tewms a) r610104 Sectors. Princeta., N. J.: tdecotbaoal Testing !arxvic*. Merck 1911. Me, s., G. & Vieroems 9. Beitreq aw spextelle. Fisychologie eel Gremed eimer Mosseavlersuch"ag Zeschr. L. Psychol.. 1909. 51. 1-72. Kaiser. N.F. The vorlues cv'tehta for ewIlyuk wateiou .a !kcor anslysix. Psyc"Momtefn.. '9M. 23. 187-200161ly. E.L. & Fishe, O.W. The predciom of perhemence &aciahacd girychole". Ann Arbn.: Usiverity of IOCL940 Preo". 19i1.me"s. R.J. Reietaaships Wefwee ln0t0 of ocV"Otafimce W4.4% ralre of froso roef .6 4 af nteemeaf between totesa. Locklead Air rat: Be,". Toeess: Ast loarce Ptrsomeel end Trsuutan Hipsoorch toemus.M.A

D, gomsiae potsmeisfy 4o4m

a vedf.

how Yarit: Appuio.eeuv

1,.

Tues, E.C. Relationship between behavior trait ratings by peers cAd later officer p&lotUmUc* 0! USAF Officer CandilcteSchooI Graduates. Lackland Air Force Base, Texas: Air Force Personnel and Training Research Center, October 1957. (Researc Heport AFPTRC-TN-57-123, ASTIA Document No. AD-134 257) Topes, E.C. Perso, fity ,raits related to e. -rfiv osS of junior and senior Air Force officers. Lackland Air Force Base, Texas: Personnel Laboratory, Wright Air Developmen: Center, November 1959. (Technical Note, WADC-TN-59-198.) Tpes, E C. & Christel, R.E. Stability of personality trait rating factors obtained ,cder diverse

conditions. Lackland Air Force Base, Tezas; Persente L oratory, Wnqbt Air Development Center, May 19F8. (Technical Note WADC-TM-.5-I. ASTIA Docmuat No. AD-ISI 041.) Webb, E. Character and intelligence. Brit. I. Psychol., Manor. Supp.. 1915. 1, No. 3.

OVp.q

*o*,t.cu..

.4-tn.

""fl-flynn

0n

*

*fi*

.*vc

-

n

9ne

C.S*fl

An,..

..

O* eqs

.. n

.e.t! .te..o..o

fl!O

aye

Q.nO

7::;::!;

a 2

310S m.

**

.. fly

4 Cl

6

t...

C-

0*

II'

*;

:

s4 ;:: I

4

.

4-

:;;; C-.

a

4

A11? I

t*

r:

ii!I

*

te?

I -

--

ir.

fI

JJ atii 3: 'IJk

'I

nil

fits

i3I

-

f I U

j

~~iIj

~

UN~fCC

11"

or-

0-

go00-

"A~

r.

WA

no

9o

C-Y

d

l-4,S OD.

71

3I

-4

.01

q4

9 C?0,

U k4

toLd HaCl3

II

('.-0o 0

-4

X

Z

-'7

LAiT rl U"y

3j

O

"N

'O-4m

0,4

('UUN,4'S

I~

~

rI.r-q-~eecc1-

~ ~

9f

2 N'ft4~

:3gE

,

S

MAIM4*Q

O .

5 l..-f-.o4

~c

.s

6

C.n

A.JS

a-

P

N4

1

T

T

2-.

23

22j

19a

sui~i

i

t~

-

W

O

.

C

.

0

0

'-4'r

-4-4

ti,

ý4C

ý,HOU 0-.0

~

~

00

4

0.

C30

-441C

00.~-i

Cj.4

.IO

5

000

00,CHA

-U0

0\o

C-,O

001

-'I

.4~C~5

~

C'

.i')4

0

0-

('4*O

,-dN

NZO~SoC"'9

4 -0

'tCJCI

9

C

A

~ ~~

3.

)

~

~ 40~10~

t-.1 a~

U,.1s

,l

~ ~

J

CY

' 10C

IC

'

T0--

)ia.

~

V'

A-

00

cm -

C.

*0

O

w

0CA

r. n

- ýN

3

4-

1O9,

sA

9

00

ON

bm V it~

21

u

.-..

,.

..

e nfnff

1c

7-.

9

7a-

~

i-t,.c a

a,

~

3~3~3

ner

2

;;j~

Zý

-

an

V33 R,

ar

-.

~

n

7 n...

c

ne

Tv...

ry

-~c

ti

nT.

Jo~u

cloga~c

-ci-

*ec

NIZ

Pig N~-fflc0 9.-

00.

~9

0

mv ll:;I

4.-g

u

Vo

s

I

VIl

~

.

jt...

""I

44-

0E

A 4

II

.~-

.

Nl S

I

gal

?M

-

C-"

ARfHM

"

-t% 4iHMmpo

A

.

*

-~~

I

-

0

-4I~.10..

C'

ý

4.4V.ý

"

got;J~

,.I

C-..0

0-~'

~--"I24

C.-A'O

1O

k

I

0 0

'

V~?V4.

NS144O0 Vc .' N vV

O

.'... ...

-

.04 7.

~

ft~

0O0'.V' 0N vCI~,

0

t0

0

ONO.

'...

0,0?'.A N 0 .' 1 w N.M-.

NNW..e .P

O AC'1n4.. .. M V

-0 A 'm0

2IZ

~

NOOC'0

tO.M.j

V VM424

- -

'N

0

O'-Nf

Nr.0I....?o4N

.0 00

..1 N

- -

oo '

070 o

?1?,m. -?'? re, -00 - 1?1vn.

.e mr... v..O VHN.M4 miO

~~~~~ eo no

*

r.

0... co

L

. .n

0

o

2

0~.0

N o .0

4

a

ON

-

.i'0 .-..C

44. mtae

~

i

..

0 0

'4

0.N4NO

N .1

0.Nlo

0~ 0

en

u .

e~

inrv .0-tV

N

4 cX

c 1

5

5oc

V...'.

'0.

"m-

.n4 oc0rvN-nST r.' ..

544I0

?'wiov

0941v ...041' t¶119N

-

O

'Or--0,'

.

~~~O (4

o.

'

s

..

N.0

1" N

..

m a.0c V.I00

NO

CU .f~O'.0

N

er-

N

N.N0e ,norw.,~. N~1..O

.

ae.

w'4

64.

~

.1N v.* e~

1

oa

9Uo

oN

tO

e,,vwl.,.

, cc' I~n

.ti m,40-0

0001

tIV 1

v0.

0 I~~~~~~~~ rv,..er.

N

e..e~'

Id

0t0

OH~~~~~

Ii~

0.0 d.(

N

~~~~~~~~ O.. 0'1 e...., ~~~~I 1 10g 1.,--.. N

'55~

Io

0

0.Cm

ne

e0

-'.

04. C

oive.

-

.

vV.2m Ceom

v~..

I

.O'

.0.

-w.. . - -4Ot-r. -~~fwr..1

0

I.', w x

0.MV

m---

~V

vw

c.O7N O

co

-4pN

A C2

-fj

,Cj

f)

noac

sLIr

FO ro

I

C in

U

N

n

r-i4

-*1

C-o

4"

I

0

o

\,

114

-Os C1.C'J

(

,4l

c

'o

''ON

-

j-:I~

c 4

5.

3

H

HC'ici

-.;1P

AU

a 4

C\-0C'.AD, H

o'

e4

H~ H\.n C - ^1

0"_

C4

ý

H .r4, l

(5aC

4-,

U ,

ao.

-

CV0

0.

0

9-.

o

H1

0I

0a1

r

YC,

c

ný

CLL

o(

0. 0

0'

S

0

4

Hu

T1 !-?

1 M

0

4

3

C",.0

61

0

r

13

:-:fII.

b

tt m v

.-

.~~U

,

4'0

ri-

In IN

) S0

:.

4 641 41

0~ H

H

'A i

l, I.,

tm4 D..

44

0

HI

0I

-l

r4 1 o Ti

0HC

-4~.4r It,

C

~

(,a~~

-

*y

10V%-

IIo 1

fn

*q

0c

a

,

'

-4 -4

CY ul

1

.a

4

0Lo0

AH

-

r

C..

4..3

4-4

f6

417

.ý4

0, H

•" Wita

l

•ti! aIt

i

iii

ii t eWi aati i

*

2lit

3ii !

:z!!i14 i. .

l

"I 3

i

ll

Ii

..i

n

-.-

-

Il

!3 ! i

!

i

9*

i

-fai*

-

1Iiii

*

"

li.....

b

i

--

i

IS ;A-3i;i~i 1

.I

till

, i

i i '* t i 0.

'

,-

0 V

0

*

5

5

5

I

R

t.'

Op

SIS

o.41 -

toS.

l,3

~dQ'12

-I

%73ý6

~'r

rSm-

~

2

S,

S.N

IV

0 4

.

fta

f

N.~

gA~f

r r~ * o eI~.

x

ft VA

.

M,

9

4f6-

~f ,1

P~z

2

c

AIR

-R,'O~a. 00 w

IA

-

-9t' 8

C-f

H~~t

c-.I-.'

1

-

0-

IIN 1"11,-

'

"

^

0,7

-1

.4

1i

10

s

gy

0.

'nU

~

--10'~i O~.O

2

N o oN-

0% d~4 0

IV

0ý

H

la *0ýP.ý1 o*O. 4.4

.0 '.0

!

ý

-"3

4'A .4P'o 10 4

A

U'0

4 r4.

,4-:1't fn4

rn

do, 0

y

-M *f"m1

-*c

H -*,

HI

Yu

00

C?

0

n.ff

"'-4;qna

8

4

..----..

e

cm co4c.

o"

"

m

c

9c

0

0

I

C

lb 44)

GV

I

C4

31

Hi'O 0ý 2c

1,3,-)

Z -

41 td

n '

I

oy'"u

1

0)

,- l

'A 0

-1''ir cm4

-

"+I

To'Iqy°

++ " + +

P+ m

-... * *++ -" 1;¶-I++ + + + + +

"*

++++-

1o

f

It :l+ I+ ;

.

.°

IL

+

+ :+

.: ; *'

++

.

!

----

++.

-

.

-

-...

;"3:" +

I:l +

-

I

+!..

-

-o+:..~:°t+:

-I

".*--

-

+"

...

,-

-

.* .+

'lI

Cn. ..

.

++ m

+: : :

.+ +. ; + +

: +o

j++

+ +e : + .

IL+

*

1,

I

II

!+'++'+j

£ ~ --- *

- Lz,

-

~~

~

~

'

d 0 rO

,4

- "

-

"

I I

-I

I0 SA

W,I

a

S,

KA

2

0

IleU

I

I

e

31U

m 'A4A' ot

N

• N.--4 oo

4

r

*

~~QN~ 44.4

9

•A-g -

S•"•

.-4 4ýO s'0

4

4

Q-M9

'i

,•"A ý -5iq !

-

0.

1

;

!

-4

it 4i

33

I01~

Ai

-I

U

&,

-a1 4g

•

=

t , '14 I1 = 9

:d~tf'II

t ii tidi; rihjlftni

a

=l 34 ~

;lddil ll

i~f

A,

l

ii•.:

I

,

~

MMM

-M100.

11

0

E V.

INNO

Cc,

gýo

Z

C0

N

00. MI

0

V

0

0A

M OM

9.0.

Nvg

0

.'

CM0 .0

N1

0000 -

.

a

.-. -I1MMW

MO

0

M

0r.

NI0

X

"

r71.' 00.0

Cc

Er-1t

00C 1 1"

o"MM 0.

...

IN 01M

Oo.

00

NY

!

0,ow1

1!C-MM cM

c

In

p'S

MP...

-

u

U

o41

N 0*1 NO. ca

*N

0

cc w

N.

0

-c

I

N

IL *

~

~

~ O

0 2- '1 E0

7:-..M

..

0. -0

E

o" .0

t a.

CCMEAI

COCIS~~o

0

M

M

0C

.

a.

I

4A II

z

z

.

n

0

C *

0

cv-uI

*I,

tR

m U1

a

~9 II

04

0

CD

0

A

u

MH

0

o

IA-A

0

.4

I~

^'0

M

U%00Of. C4~I

o%1

U

094 ,j

41-G S.4

P

0

to

oil

.M

6

r6-vU1

m4

P C4

37

0

A*w

opW

A

vil~7

0

;; 0

al00

00

44

0~~lv

W)

DN

10-0i.

IV 20

om

or

A

*n N

00

-0000'

rt

m4

ur~r

~

a-

1

w

0.

ow r

4C

01

0.

CLC -(.4

(014 ,-..

-

~'0 ei 110

7

au

u

W,

~S o

.2

*I

COP41

0

LA

i

to

~

~I18

u

Wo

wC

Lut

cc,

.0 'Aý S

A

f

-

-

U

'I

UN

9

09 k0 0

92

0.

0

I.

r

'I

.

54

V)

M

. §

§

?

0ýf

~ ~

'

-.oa

~4

*

ba C2

4 c4. I

441

O

Ow

aai a

U In

-l

^N

5

1

U

N;

Zc

54u

I. p,

390

>1 [ýý -R N ý4

0

~

(V 4

8

-4

0

0

u~

-H

c

'

24PfN'

COI Lm o,

0

Al

D

490

cu

,

0

U

.4f

?41

01

va

0

Sou

z

0

-0

6

0 -p

-,1

o

40ao

~

E

-

U0

.

a

A

c

u

0 .04 0. OS

*..

-

i

.

4.

.. -*

am

v

*-

;

f

.

...

c

otU

6

o

L6 U0

cc

c

2

u

..

L

a.U

Ir

nw

0

u

w0

InU

Li

u

w.

zz

-

Uc

c

A

Uc C

'Ie~ u

0w-*'t~

~~;'

~w