INTERNATIONAL

JUNG TYPEINDICATORtechnical manual

ONTENTS1 THEORETICAL OVERVIEW

2 THE JUNGIAN TYPES

3 SCALE CONSTRUCTION

4 THE PSYCHOMETRIC PROPERTIES OF THE JTI

5 RESOURCE MATERIALS ON PSYCHOLOGICAL TYPE

APPENDICES

I ADMINISTRATION AND SCORING INSTRUCTIONS FOR THE JTI

II PSYCHOMETRIC DATA FOR THE ORIGINAL VERSION OF THE JTI

III REFERENCES

SUPPLEMENT

I SUPPLEMENT TO RELIABILITY AND VALIDITY EVIDENCE

ASSOCIATED FOR JTI RELEASE 2

c

2

31.2.3.4.5.6.7.8.9.10.11.12.13.14.15.16.17.18.19.20.21.22.23.24.25.26.27.28.29.30.31.32.33.34.35.36.37.38.39.40.

LIST OF TABLESMBTI® Intercorrelation Matrix (n=229)Reliability (alpha coefficients) for the JTIShort and Long Term Test-Retest coefficients for the JTIStandard Error of MeasurementIntercorrelations between the JTI subscalesFactor structure of the JTI itemsCorrelations between the original and revised versions of the JTICorrelations between the JTI and MBTI® subscalesFactor structure for the (revised) JTI and MBTI® (sample 1)

Factor structure for the (revised) JTI and MBTI® (sample 2)Correlations between JTI and NEO (Ffi)Factor Analysis of JTI & NEO scale scoresCorrelations between JTI and 16PF-5Correlations between the JTI and 15FQ subscalesCorrelations between the JTI and EPQCorrelations between the JTI and OPP Correlations between the JTI and VMI Correlations between the JTI and OIP JTI subscale mean differences and standard deviations for men and womenAlpha coefficients for JTI subscales by sexCorrelations between the JTI subscales and ageJTI subscale Mean differences and standard deviations, for white and ethnic groupsCoefficient alpha for each of the JTI subscales on an ethnic minority sampleCorrelations between the JTI and 15FQ subscales on an ethnic minority sampleCorrelations between the JTI and OPP on an ethnic minority sampleJTI scale internal consistencies and mean ITC for various samplesShort & Long Term Test-Retest Coefficients for the JTIJTI Intercorrelation Matrix (N=5575)Uncorrected correlations with the MBTI®JTI Corrected correlations with the MBTI®Correlations between JTI and NEOCorrelations Between the JTI & 16PF (N=210)Correlations between JTI and 16PF-5Correlations between JTI and 15FQ (N=5575)Correlations Between OPP & JTI (N=158)Correlations between JTI and OIP (n=1971)Correlations between JTI, EPQR and I7 (N=221)Correlations between JTI & Kline’s PPQ (N=193)Correlations Between JTI & Gordon’s SIV (N=160)Factor Structure of the JTI and MBTI®

4

1THEORETICALOVERVIEWTheories of human typology can betraced back at least as far as to theGreek physician Hippocrates (ca.460-377 b.c.).

● INTRODUCTION TO HUMAN TYPOLOGY

● THE PIONEERING WORK OF MYERS

AND BRIGGS

● THE TWO BASIC ATTITUDES

● JUNG’S THEORY OF THE BASIC

PSYCHOLOGICAL PROCESSES

● THE FOUR PSYCHOLOGICAL

FUNCTIONS

● DOMINANT AND AUXILIARY

FUNCTIONS

● SECONDARY PROCESSES

6Hippocrates postulated four basictemperamental types: sanguine; cho-leric; melancholic and phlegmatic.This typology persisted for manycenturies; for example the philoso-pher Immanual Kant (1798) basedmuch of his psychology on thistheory of four psychological types.The importance of this early fourfoldtypology is that it contains within itthe seeds of modern typologicalthinking. Although this fourfoldtypology is explicitly categorical (i.e.each person can only be one type),the concept of behavioural traits isnonetheless implicit in this theory,with each type being defined by ref-erence to a number of traits which,when taken together, make up a par-ticular personality type. For exam-ple, Kant attributes the traits ofcarefreeness, hopefulness, sociability,and being easily bored etc. to theSanguine temperament. Thus psy-chometric typological theories, whichstress that a personality type is ahigher-order construct obtained froma cluster of behavioural traits, can beseen to have been foreshadowed bythese early typological theories.

Wilhelm Wundt (1903), theacknowledged founder of experi-mental psychology, demonstratedhow a fourfold typology can arisefrom a two-dimensional theory.Wundt postulated two dimensions,one reflecting the strength of aperson’s emotions and the other thespeed at which such emotionschange. Given this two-dimensionalsystem Wundt showed that fourtemperamental types could bedefined by the behavioural traitscontained within the four quadrantsof this system.

The next development inpersonality typology was carried outby C G Jung (1921), who proposed atypological theory in whichindividuals could be assigned to oneof eight types. (As the JTI has beendeveloped to assess Jung’s theory afull description of this typology willbe saved until later in this chapter.)Moreover, as is the case with theearlier typologists, while Jung’stheory describes people in terms ofcategories, Jung himself made itclear that he viewed personalitytypes, such as Extraversion versusIntroversion, as being dimensionalrather than consisting of discretecategories.

Perhaps the most experimentallyrigorous personality typology is thatdeveloped by H J Eysenck (1960,1969), who provides a threedimensional type descriptionconsisting of:Extraversion/Introversion;Emotional Stability/EmotionalInstability, and Psychoticism (aTough-minded/Tender-mindedcontinuum). In line with the ideasnoted above, Eysenck has shown,from an extensive body of research(based on numerous factor analyticstudies), that a typology is purely ahigher-order organisation ofpersonality traits. Thus, Eysenck(1947) proposes four levels ofpersonality description. At the lowestlevel are the person’s specificresponses to specific situations. Thesecond level of description consists ofhabitual responses; those responseswe typically make when similarsituations occur. At the third levelthese habitual responses areorganised into personality traits,these are the primary factors thatemerge from a factor analysis ofpersonality test items. The highestlevel consists of the organisation ofthese personality traits into a generaltype. In factor analytic terms theseare the second-order factors whichemerge when the primary factors arethemselves submitted to factoranalysis.

INTRODUCTION TO HUMAN TYPOLOGY

7After formulating his theory ofpersonality types in 1921, Jung didlittle further work in this area. Theseideas were then taken up by themother and daughter team ElizabethMyers and Catherine Briggs, whorecognised the intrinsic value ofJung’s personality typology. Realisingthat if the theory was to be of anypractical value, it was necessary todevelop a method for assessing aperson’s type, they set aboutdeveloping a questionnaire measureof psychological type. Thispioneering work led to thedevelopment of the Myers BriggsType Indicator®, which is now one ofthe most widely referencedassessment tools.

The Jung Type Indicator (JTI) is afurther development that aims toprovide a modern, psychometricallysound measure of Jung’s typology.Thus, while the JTI is firmly rootedin Jung’s psychological theory, it hasbeen developed using modernpsychometric techniques (describedin detail in Chapters 3 & 4) to ensurethat the questionnaire provides areliable and valid measure of aperson’s Jungian (psychological)Type.

Table 1: MBTI® Intercorrelation Matrix (n=229)

E I S N T F J P

ExtraversionIntroversionSensingiNtuitionThinkingFeelingJudgementPerception

1 -.94 -.19 .14 -.02 .04 -.14 .10-.94 1 .20 -.13 -.01 -.04 .13 -.08-.19 .20 1 -.89 .09 -.07 .42 .44.14 -.13 -.89 1 -.15 .14 -.36 .39

-.02 -.01 .09 -.15 1 -.80 .27 -.27.04 -.04 -.07 .14 -.80 1 -.27 .28

-.14 .13 .42 -.36 .27 -.27 1 -.96.10 -.08 -.44 .39 -.27 .28 -.96 1

THE PIONEERING WORK OF ELIZABETHMYERS AND CATHERINE BRIGGS

As well as being rooted in Jungiantypological theory the JTI also incor-porates some of the modifications toJung’s theory, suggested by ElisabethMyers (1962), which have beenincluded in the Myers-Briggs TypeIndicator. Principal among these isthe inclusion of the Judging/Perceiving types, which although notexplicitly stated as part of Jung’stypology, are nevertheless implicit inhis theory of types. In addition, theJTI also adopts the method of ascer-taining the dominant function usedby the MBTI® (see discussion of thesuperior function later in this chap-ter). Despite these similaritiesbetween the JTI and the MBTI®there are, nonetheless, some funda-mental and important differencesbetween these two instruments interms of their psychometric proper-ties. This results from the JTI havingbeen constructed using modern psy-chometric test theory. The moststriking difference between these twoinstruments is that the JTI, in keep-ing with modern type theory (seeabove), construes psychologicaltypes as being best described bypoints on a continuum, rather thanby discrete categories. Thus, unlikethe MBTI®, the JTI views the twobasic attitudes (Extraversion versusIntroversion), the four psychologicalfunctions (Thinking versus Feeling

and Sensing versus Intuiting) andthe two secondary processes(Judging versus Perceiving)described by Jung as being continu-ous variables, rather than discretecategories. This is supported by thefact that the evidence demonstratinga discrete categorical view of psycho-logical types is very scant. In partic-ular, statistical analysis of MBTI®data does not reveal the bimodal dis-tributions which would be expectedif this instrument were assessing dis-crete categories.

Moreover, examination of the cor-relations between the separatelyscored pairs of MBTI® attitudes,functions and secondary processes(presented in Table 1) indicates thatthese are effectively measuring oppo-site ends of the same dimension.Therefore the JTI has been devel-oped to assess bipolar continuousconstructs, with each type beingdefined by those personality traitswhich cluster at one end of the typedimension with the type boundariesset in the middle of the scale. AsEysenck (1969) notes, this is a truetypology in the modern sense of thatterm, as ‘…the widespread notionthat typologies imply discontinuities,bimodal distributions, and the like,does not accurately represent thewritings and views of modern typolo-gists.’

THE JTI AND THE MBTI®

8

THE TWO BASIC ATTITUDESThe two basic attitudes that Jungidentified function as opposites, withone of these attitudes being domi-nant and conscious, while the otheris auxiliary and unconscious.

THE EXTRAVERTED ATTITUDE

Extraversion is oriented towards theexternal, outer world. Thus, peoplewho prefer this attitude would ratherto spend time interacting with theoutside world, than dealing with theinner world of subjective experiencesand mental events. Thus extravertsenjoy spending time in the companyof other people and enjoy translatingtheir ideas and interests into overtbehaviour and action.

THE INTROVERTED ATTITUDE

Introversion is oriented towards theinner, subjective world. Thus peoplewho prefer this attitude like to spendtime in quite contemplation andreflection. Thus introverts are quiet,introspective individuals who do notfeel the constant need for contactwith the outer world of people andevents.

JUNG’S THEORY OF THEBASIC PSYCHOLOGICALPROCESSES

In his personality theory Jung identified two core mental activities: ● receiving, or taking in, information; which he termed Perceiving, and ● processing that information (e.g. organising the information and

coming to conclusions form it), which he termed Judging. He further identified two alternative ways of Perceiving information,

which he termed Sensing and Intuiting, and two alternative ways of Judginginformation, which he termed Thinking and Feeling. Moreover, Jung notedthat these four mental process can be directed either at the external world ofpeople and things, or at the internal world of subjective experience. Thus theexistence of these four basic psychological process, which can be used eitherin the internal or external world, means that people can use their mind in oneof eight ways; thus resulting in eight psychological types.

9As described above, Jung identifiedfour basic psychological functions,with a person’s preferred functionbeing dominant and conscious, whilethe non-preferred function is auxil-iary and unconscious.

THINKING

Thinking involves the logical analy-sis of information in terms of thestrict principles of cause and effect.Thus people who prefer this form ofJudging approach life in a rational,analytical way; searching for logicalrelationships between events andideas.

FEELING

Feeling involves identifying the emo-tional value that is attached toobjects or events. Thus people whoprefer this form of Judging are moreconcerned with what they feel abouta person or event, rather than, withwhat they can learn about it throughlogical, rational analysis.

SENSING

Sensing involves directly receivinginformation through the senses.Thus people who prefer this form ofperceiving tend to focus on the factsin a given situation, and on harddata.

INTUITING

Intuiting involves going beyond theinformation provided by the senses todiscover possibilities which might notbe immediately obvious from sensorydata. Thus people who prefer thisform of perceiving have a preferencefor focusing on theoretical issues andhidden patterns of meaning.

THE FOURPSYCHOLOGICALFUNCTIONS

bkDOMINANT ANDAUXILIARY FUNCTIONS

Jung further noted that people’spreferences for each of these func-tions vary, with these differencesgiving rise to the individual differ-ences we observe between people.Moreover, Jung observed that thedominant function is used in the pre-ferred world (i.e. the external worldfor Extraverts and the internal worldfor Introverts) while the auxiliaryfunction is used in the non-preferredworld (i.e. the inner world forExtraverts and the outer world forIntroverts).

In addition, Jung realised that asThinking and Feeling are both ratio-nal functions they tend to opposeeach other, as do the non-rational

functions of Sensing and Intuiting.Thus neither member of an opposingpair of functions can have an auxil-iary role for the dominant function.In this way, if a rational function isdominant then one of the non-ratio-nal functions must provide the auxil-iary function and vice versa. The JTIadopts the approach developed byElizabeth Myers (1962) for identify-ing which of the four functions isdominant. This consists of using anindividual’s secondary process ofJudging or Perceiving (see next sec-tion) as an indicator of which of thefour functions will be dominant andwhich will be auxiliary.

SECONDARY PROCESSESAs noted above one of ElizabethMyers (1962) great insights was torecognise that the secondaryprocesses of Judging and Perceivingcould be used to identify whichpsychological function will bedominant for each person.

JUDGING

Judging is concerned with organisingand processing information. Thuspeople who prefer this mentalprocess for receiving informationabout the world like to plan aheadand organise, as they order and reg-ulate their mental events.

PERCEIVING

Perceiving is concerned with directlyreceiving information without evalu-ation. Thus people who prefer thismental process like to put off deci-sion making in order to gain as muchinformation as possible. Moreover,when they do decide to act they willdo so in an unstructured and flexiblemanner, without detailed prior cate-gorising of events.

The two basic attitudes, whencombined with the four functions,produce the eight Jungian Types.These eight types can be furtherdifferentiated in terms of theauxiliary function.

● INTROVERTED-THINKING

● EXTRAVERTED-THINKING

● INTROVERTED-FEELING

● EXTRAVERTED-FEELING

● INTROVERTED-SENSING

● EXTRAVERTED-SENSING

● INTROVERTED-INTUITING

● EXTRAVERTED-INTUITING

2THE JUNGIAN TYPES

bmViewing life from a detached intel-lectual perspective, Introverted-Thinking types are objective clearthinking and rational. Quietly intro-spective and abstract thinking, somepeople may view them as being a

little cool and reserved, or evensomewhat distant, in their dealingswith others. Rational, analytical andlogical, they are inclined to be quietobservers of life.

INTROVERTED-THINKING

EXTRAVERTED-THINKINGLogical, analytical and rational,Extraverted-Thinking types arequick to challenge ideas that are notbased on sound logic. Direct and tothe point in their dealings with

others, they are lively energetic andtalkative. They actively participatein all that is going on around them,greatly enjoying social contact.

INTROVERTED-FEELINGQuietly thoughtful, sensitive andconsiderate, Introverted-Feelingtypes are deep introspective individ-uals. Keenly perceptive observers oflife, they have a genuine empathicconcern for others.

While they have a great capacityfor warmth, they may not displaythis with people they do not knowwell. Their reliance on feeling meansthat they hold strong values that arecentral to their personal identity.

EXTRAVERTED-FEELINGWarm, friendly and sociable,Extraverted-Feeling types show agenuine interest in the people aroundthem. Disliking conflict and discord,they will actively try to seek com-promise and consensus. Thoughtful

empathic and understanding, otherswill be quick to warm to them.Valuing genuiness in interpersonalrelationships, they will be open andexpressive in their dealings withothers

bnIntrospective, down-to-earth andrealistic, Introverted-Sensing typeshave a quiet, matter-of-fact interper-sonal style. Focusing on practical,achievable goals, they will wish toget things done with a minimum of

fuss and discussion. Deeply privateindividuals, they are not quick toexpress their inner most thoughtsand feelings and, as a result, othersmay see them as being somewhatreserved.

Outgoing sociable and lively,Extraverted-Sensing types are actionorientated individuals. Pragmatic,down-to-earth and realistic, they liketo achieve immediate, practicalresults. Goal orientated and problem

focused, they have little interest inthe theoretical nuances of a problem,preferring instead to focus on factsand hard data. Friendly and enthusi-astic, they will be popular teammembers

INTROVERTED-INTUITINGQuietly observing life from anabstract intellectual perspective,Introverted-Intuiting types are deepindividuals. Often engrossed in theirown theoretical musings, others whoare less abstract minded than them

may sometimes view them as being alittle cool, distant or even somewhatreserved. Deeply private individuals,they may hold back from expressingtheir deeply held personal views andfeelings.

INTROVERTED-SENSING

EXTRAVERTED-SENSING

EXTRAVERTED-INTUITINGLively, talkative and outgoing,Extraverted-Intuiting types are quickto enthuse others with their abstract,theoretical ideas. They greatly enjoyparticipating in theoretical debatesand discussions, and are quick to

express their personal insights andunderstandings. Greatly enjoyingsocial contact, they will participateenthusiastically in all that is going onaround them.

bo

SCALECONSTRUCTIONThe JTI subscales were constructedfollowing the principles of classicaltest theory (see Kline 1986 for adetailed exposition of test theory).

● ITEM CONSTRUCTION

● ITEM TRIALING

3

ITEM CONSTRUCTION

An initial item set was constructedby three psychologists who areexperienced in the theory ofpsychological type and itsassessment. Each psychologistindependently generated a set ofitems designed to assess the corecharacteristics of each type. Whendeveloping these items eachpsychologist made extensivereference to Jung’s original work onpsychological type, in addition toreferring to more recent publishedwork in this area. Once an initialitem pool had been generated,consensus was sought among thepsychologists on the wording of eachitem. Minor alterations were thenmade to item wording untilconsensus was achieved, and thoseitems on which consensus was notachieved were rejected.

ITEM TRIALING

The items were then trialed on threeseparate samples, two of which alsocompleted the Myers-Briggs TypeIndicator®. Items were selected forinclusion in the final instrument ifthey met the following criteria:

1. Each item correlated substantially(0.3 or greater) with the targetMBTI scales.

2. Each item did not correlatesubstantially (0.2 or less) withnon-target MBTI scales.

3. The items combined to formhomogeneous item sets acrosseach of the three samples(corrected item total correlationsexceeding 0.3)

4. Removing any item from the itemset did not reduce the scale’salpha coefficient.

5. When more than 15 items met theabove criteria, those with thelowest item-total correlations wereremoved from the item set on aniterative basis.

bq

THE PSYCHOMETRICPROPERTIESOF THE JTIThis chapter presents detailsconcerning the psychometricproperties of the Jung Type Indicatorand demonstrates that thisinstrument meets the technicalrequirements of a modernpsychometric measure.

● INTRODUCTION

● RELIABILITY OF THE JTI

● VALIDITY

● BIAS

4

bsRELIABILITY

An important technical requirementfor a psychometrically sound test isthat the measurements obtainedfrom that test should be reliable.This assesses the extent to whichvariation in test scores reflect truedifferences between people (on thecharacteristic being measured)rather than reflecting random mea-surement error. The reliability coeffi-cient indicates the proportion ofvariance (variability) in a scale thatis attributable to the underlyingcharacteristic, rather than randommeasurement error. Measures ofinterests, attitudes and values areconsidered to meet acceptable levelsof reliability if they have reliabilitycoefficients above 0.7.

Reliability is generally assessedusing one of three methods. The firstmethod assesses the internal consis-tency, or homogeneity, of the scaleitems. (That is to say, the extent towhich the scale items all assess thesame underlying construct.) Thesecond method assesses the extent towhich parallel forms of the same testare correlated with each other, andcan thus be considered to be measur-ing the same underlying trait orcharacteristic. The third methodassesses the stability of scale scoresover time.

Reliability: Internal ConsistencyAlso known as scale homogeneity,internal consistency measures of reli-ability assess the extent to which allof the items in a scale are measuringthe same underlying construct ortrait. A scale is said to be internallyconsistent if all the items are corre-lated with each other, which in turnimplies that the items are all assess-ing the same trait or characteristic.

The most common measure ofinternal consistency is Cronbach’sAlpha coefficient. If the items in ascale are highly correlated with eachother, and with the total scale score,then coefficient alpha will be high.Thus coefficient alpha indicates theextent to which all the items in thescale are measuring the same con-struct.

Reliability: Parallel FormThis assesses the extent to which twoequivalent forms of the scale are cor-related with each other. The Pearsoncorrelation coefficient between thetwo forms is used as the parallelform reliability coefficient.

This reliability coefficient is rarelyreported in the literature because ofthe difficulty in creating parallelforms. However, this can be a usefulmeasure of reliability when such par-allel forms exist.

INTRODUCTION

btReliability: StabilityAlso known as test-retest reliability,this measure of reliability assessesthe extent to which a psychometricinstrument produces consistentscores when used on two differentoccasions. The occasions may beeither a few hours, days, months oreven years apart. The Pearson corre-lation coefficient between the (same)scale scores, obtained on two differ-ent occasions, is used as the test-retest reliability coefficient.

Test-retest reliability coefficientsprovide a useful indicator of a scale’sstability over time. If these coeffi-cients are low (less than 0.6) thenthis suggests either that the behav-iours/attitudes being measured arevolatile and/or are situationally spe-cific, or that scale scores are influ-enced by high levels of random error(i.e. they vary randomly over time).However the fact that a scale hashigh internal consistency & stabilitycoefficients only guarantees that it ismeasuring something consistently. Itprovides no guarantee that the scaleis actually measuring the construct itpurports to measure. Questions con-cerning what a test actually mea-sures, and its relevance to particularassessment situations, are dealt withby looking at the test’s validity.Reliability is generally investigatedbefore validity as the reliability of ascale places an upper limit on thescale’s validity. In this regard it canbe mathematically demonstratedthat a validity coefficient for a par-ticular scale can not exceed thatscale’s reliability coefficient.

VALIDITY

This assesses the degree to which ascale measures what it claims tomeasure.

Two key aspects of validity areknown as criterion validity and con-struct validity.

Validity: Construct ValidityConstruct validity assesses whetherthe characteristic which a scale isactually measuring is psychological-ly meaningful and consistent withthe scale’s definition.

The construct validity of a scale isassessed by demonstrating that thescale is correlated with otherpsychometric instruments whichmeasure similar constructs, and isnot correlated with tests whichmeasure different constructs.

Validity: Criterion ValidityThe extent to which the scale canpredict a real world (measurable)criterion. The criterion validity of atest is demonstrated by showing thatscale scores are related in meaning-ful ways to an external criterion.Criterion validity can take the formof either predictive or concurrentvalidity.

Validity: Predictive ValidityPredictive validity assesses whether atest is capable of predicting a criteri-on that occurs at a time after the testwas completed (e.g. that a selectiontest can predict the likelihood ofsomeone successfully completing atraining course).

Validity: Concurrent ValidityConcurrent validity assesses whetherthe scores on a test can be used topredict a criterion which is assessedat the same time as the scale is com-pleted (e.g. that a selection test canpredict current job performance).

ck RELIABILITY OF THE JTIData on the internal consistency andstability of the JTI subscales are pre-sented below.

INTERNAL CONSISTENCY

Table 2 indicates that the Jung TypeIndicator subscales have a high levelof internal consistency (reliability)across a number of different sam-ples. Inspection of this table demon-strates that these subscales haveparticularly high levels of reliability(internal consistency) for such short(15 item) scales.

STABILITY (TEST-RETESTRELIABILITY)

The data presented in Table 3 indi-cate that the JTI subscales have ahigh level of stability (test-retest reli-ability) when completed by the samerespondents over a weel, or a threemonth, period. Most significantly,these subscales demonstrate veryhigh levels of test-retest reliability(stability) for measures of personali-ty and/or preference.

STANDARD ERROR OFMEASUREMENT (SEM)

Table 4 presents estimates of thestandard error of measurement ofthe JTI subscales (presented in stenunits). When interpreting any psy-chometric test score it is important totake into account the fact that if arespondent were to be repeatedlytested the score they obtained wouldbe expected to fluctuate due torandom measurement error. Indeedtheir true score on the test would beexpected to fall within a range of ± 1SEm from their obtained score 68%of the time.

Table 2: Reliability (alpha coefficients) for the JTI

Sample Sample Sample Sample Sample Sample1 2 3 4 5 6

EISNTFJP

.83 .88 .85 .87 .84 .85

.78 .86 .84 .83 .81 .80

.75 .78 .80 .81 .84 .79

.82 .79 .75 .83 .82 .85

Sample 1 = Business Studies undergraduates (n=40)Sample 2 = Technician applicants (n=107)Sample 3 = European Business School undergraduates (n=112)Sample 4 = Personnel Professionals (n=158)Sample 5 = Volunteer Sample (n=117)Sample 6 = Undergraduate Psychology Students (n=222)

Table 3: Short & Long Term Test-Retest reliability coefficients for the JTI

JTIScales

Three month Long-termSEm

EISNTFJP

cl

SEm

EI 0.72SN 0.82TF 0.87JP 0.82

Sample = total Sample (N=787)

Table 4: Standard Error ofMeasurement

JTISubscale

Sample 1 = Training course attendees N=101Sample 2 = Psychology undergraduates N=222

0.920.880.790.86

0.860.800.850.80

One week Short termSEm

0.60.660.920.82

0.750.890.770.89

Sample 2Sample 1

cmTHE INTERNAL STRUCTUREOF THE JTI

The intercorrelations between theJTI subscales, presented in Table 5,are all small and thus indicate thatthese subscales are measuring rela-tively independent characteristics.The highest correlation(0.27) isobserved between the SN and TFsubscales. It should be noted that inthe MBTI® the equivalent correla-tions between J / P and S / N rangefrom 0.36 to 0.44, suggesting aslightly higher degree of overlapbetween the scales than occurs in theJTI.

FACTOR STRUCTURE OF THEJTI ITEMS

The strongest test of the internalstructure of the JTI is provided byfactoring analysing (principal axisfactoring with oblimin rotation) theJTI items. The scree test indicatedthat the data were best described bya four factor solution, which is pre-sented in Table 6. (Factor weightssmaller than 0.3 have been excludedfrom this table to aid interpretation.)Inspection of this table indicates thateach of the JTI items load on theappropriate factor, demonstratingthe good internal structure of theseitems.

EQUIVALENCE BETWEEN THEORIGINAL AND REVISED VERSIONS OF THE JTI

Table 7 presents uncorrected correla-tion’s between the subscales of theoriginal and revised versions of theJTI. All these correlations are large,and indicate that the original andrevised JTI subscales are substantial-ly equivalent to each other. Of par-ticular note is the observation thatthe weakest of these correlations isbetween the Thinking-Feeling sub-scales in the original and revised ver-sions of the JTI. This reflects the factthat a major revision was undertak-en of this subscale, with the intentionof removing this subscale’s previous-ly high loading on emotionality. Thevalidity data, (presented later in thisChapter), clearly indicates that thisrevision was successful in that therevised TF subscale now more close-ly maps onto the original Jungianconcept of the Thinking -Feelingpsychological function.

The above correlations thus sug-gest that, while the original andrevised forms of this instrument arenot strictly interchangeable, they dononetheless measure similar con-structs. A considerable quantity ofdata has been collected on the origi-nal version of the JTI, which demon-strates the validity of this measure(see Appendix I). Thus the revisedJTI can be viewed as building uponthe extensive body of evidence thatdemonstrates the validity of the orig-inal measure.

VALIDITY

Table 6: Factor structure of the JTI items

Factor 1

Item 1 .40Item 5 -.70Item 9 -.68Item 13 .67Item 17 .77Item 21 .30Item 25 .71Item 29 -.51Item 33 -.32Item 37 .67Item 41 .61Item 45 -.69Item 49 .31Item 53 -.64Item 57 -.61Item 2 .65Item 6 -.46Item 10 .50Item 14 -.51Item 18 -.32Item 22 .52Item 26 .51Item 30 .39Item 34 -.52Item 38 -.50Item 42 -.31Item 46 .33Item 50 .55Item 54 .52Item 58 .47Item 3 .33Item 7 -.36Item 11 -.48Item 15 .46Item 19 -.52Item 23 .39Item 27 -.57Item 31 .54Item 35 .44Item 39 -.31Item 43 .54Item 47 -.43Item 51 .59Item 55 .50Item 59 -.43Item 4 .59Item 8 -.57Item 12 -.35Item 16 -.30Item 20 -.58Item 24 .31Item 28 .49Item 32 -.35Item 36 .47Item 40 .50Item 44 .83Item 48 .47Item 52 -.50Item 56 -.65Item 60 .82

Factor 2 Factor 3 Factor 4

Sample = total sample n=778

Table 7 Correlations between the originaland revised versions of the JTI

ScaleIntercorrelations

EI .82SN .77TF .60JP .69

Sample = Personnel Professionals (n=40)

cnEI SN TF

Table 5: Intercorrelations between the JTIsubscales

EI –SN .05 –TF -.26 .27 –JP -.04 .25 .10

Sample = total Sample (n=787)

coTHE RELATIONSHIP BETWEENTHE JTI AND THE MBTI®

The correlations between the MBTI®and JTI subscales presented in Table8, indicate that the JTI subscales are,for all practical purposes, measuringdimensions that are identical to thoseassessed by the MBTI®. This isclearly demonstrated by the fact thatall the corrected correlationsapproach unity. Moreover, the size ofthe uncorrected correlations betweenthe respective MBTI® and JTI sub-scales are in the order of those thatwould be expected to be foundbetween parallel forms of the sametest.

However the strongest test of theJTI’s concurrent validity is achievedby factor analysing the MBTI® andJTI subscales. Table 9 and Table 10present the results of such factoranalyses (with factor weights of lessthan 0.3 not being reported in orderto aid the interpretation of this factorstructure).These factor analyses(principal axis factoring withoblimin rotation) produced clearfour factor solutions on two indepen-dent samples.

The factor analyses presented inTables 9 and 10 demonstrate thatthe JTI subscales are clearly measur-ing four independent factors, whichmap closely onto the MBTI® sub-scales. Most importantly, each of theJTI subscales weight on only onefactor, with this factor being clearlydefined by the relevant MBTI® sub-scales.

RELATIONSHIP BETWEEN THEJTI AND THE NEO

The JTI subscales were correlatedwith the NEO short form to assessthe JTI’s construct validity. Thesecorrelations are reported in Table 11.

As expected, the NEO Extraversionsubscale correlated highly with theJTI Extraversion-Introversion sub-scale, providing strong support forthe validity of this JTI subscale.Similarly, a high correlation wasobserved between the JTI Sensing-Intuiting subscale and the NEOOpenness subscale. This is consistentwith these scales’ definitions, as boththe NEO – O and the JTI – SN sub-scales measure an interest in theworld of ideas and abstract thought.

The JTI Judging-Perceiving sub-scale correlated significantly with theNEO Conscientiousness subscale,with this being attributable to theNEO – C subscale assessing a prefer-ence for structure and order in dailyactivities. Moreover, the JTIThinking-Feeling subscale correlatedsignificantly with the NEOAgreeableness subscale, with thisbeing attributable to the NEO – Asubscale assessing an empathic con-cern for others; which features as acomponent of the Jungian concept ofthe Feeling type. Moreover, the factthat these latter two correlations(NEO – C with JTI – JP; NEO – Awith JTI – TF) are relatively modestsuggests, as would be predicted, thatthese two JTI subscales are measur-ing dimensions that are not fullyaccounted for by the NEO. Finally,as anticipated, none of the JTI sub-scales correlated highly with theNEO Neuroticism subscale,indicat-ing that this is not a construct that isdirectly measured by the JTI.

A scale factor analysis was under-taken on the same data-set to exam-ine, in more detail, the convergentand discriminant validity of the JTI.Factors were extracted using princi-pal axis factoring and were rotated tosimple structure via normalised

Table 8: Correlations between the JTI and MBTI® subscales

JTISubscale

MBTI® Subscale

CorrectedCorrelation

UncorrectedCorrelation

EI E 1 .87I -1 -.90

SN S .96 .75N -.94 -.71

TF T .97 .75F -.96 -.70

JP J .93 .76P -.97 -.80

Business Studies undergraduates (n=40)

Table 9: Factor structure for the JTI and MBTI® (sample 1)

Factor 1 Factor 2 Factor 3 Factor 4

MBTI – J -.93MBTI – P .94JTI – JP .71MBTI – E -.94MBTI – I .93JTI – EI .81MBTI – T -.90MBTI – F .91JTI – TF .71MBTI – S -.85MBTI – N .88JTI – SN .72

Sample 1 = European Business School undergraduates (n=112)

Table 10: Factor structure for the JTI and MBTI® (sample 2)

Factor 1 Factor 2 Factor 3 Factor 4

MBTI – J -.94MBTI – P .94JTI2 – JP .85MBTI – E -.96MBTI – I .97JTI2 – EI .89MBTI – T -.95MBTI – F .95JTI – TF .83MBTI – S -.87MBTI – N .87JTI2 – SN .88

Sample 2 = Psychology Undergraduates (n=96)

cp

cqvarimax rotation. Table 12 presentsfactor loadings for this factor solu-tion. (Loadings below 0.3 have beenexcluded from the table to aid clarityof interpretation.) Inspection of Table11 demonstrates that this factor solu-tion has good levels of convergentand discriminant validity, with allfour of the JTI subscales clearly load-ing substantially on one four of thebig-five factors. The fifth factor,Neuroticism, constitutes a factor inits own right, as would be predicted.The results of this factor analysisthus provide further evidence indica-tors that the JTI is measuring fourdistinct psychological dimensions,each covering a different area of thepersonality spectrum.

RELATIONSHIP BETWEEN THEJTI AND THE 16PF-5

Table 13 presents correlationsbetween the 16PF (version 5) and theJTI. As expected, the JTIExtraversion-Introversion subscalecorrelated highly with all the 16PF-5extraversion factors; most notablywith factors H (Social Boldness) andF (Liveliness), and to a lesser extentwith factors N (Directness), Q2(Group Orientation) and A (Warmth).This provides strong support for thevalidity of the JTI Extraversion-Introversion subscale. Moreover, aswould be expected, the JTI subscaleSensing-Intuiting correlated substan-tially with the 16PF-5 factors M(Abstractness) and Q1 (Openness toIdeas), both of which assess an orien-tation towards the world of creativethought and ideas. Thus this findingprovides strong support for the validi-ty of the JTI Sensing-Intuiting sub-scale.

While somewhat smaller, the corre-lations between the JTI Thinking-Feeling subscale and the 16PF-5factors A (Warmth) and I (Intuitive),were nonetheless significant andpsychologically meaningful. As wouldbe expected these correlations indi-cate that Thinking types tend to have

less concern about other people’s feel-ings, than do feeling types, and areless in touch with their own feelingsthan are Feeling types. This findingthus provides further suport for thevalidity of the JTI Thinking-Feelingsubscale. Finally the Judging-Perceiving subscale of the JTI wasfound to correlate substantially withthe 16PF-5 Factor Q3(Perfectionism), in line with expecta-tion; as Judging types have apreference to be perfectionistic andorderly in their work and daily lives.Thus this finding provides strongsupport for the validity of the JTIsubscale Judging-Perceiving.However, somewhat surprisingly, the16PF-5 factor G (Rule-Boundness)was not found to be correlated withthe JTI Judging-Perceiving subscale .This lack of correlation between thesetwo subscales may reflect the fact thatthis 16PF-5 factor assesses a narrowconcern with rules, as opposed to thebroader concept of Conscientiousness,that is assessed by earlier form of the16PF.

RELATIONSHIP BETWEEN THEJTI & 15FQ (n=43)

Table 14 presents correlationsbetween the 15FQ and the JTI. Asexpected, the JTI Extraversion-Introversion subscale correlatedhighly with all the 15FQ extraver-sion factors; most notably with fac-tors FH (Social Boldness), FF(Liveliness), FA (Warmth) and FQ2(Group Oriented). This providesstrong support for the validity of theJTI Extaversion-Introversion sub-scale. Moreover, as would be expect-ed, the JTI subscale Sensing-Intuiting correlated substantiallywith the 15FQ factors M(Imaginative) and Q1(Radical), bothof which assess a creative and imagi-native orientation towards the worldof ideas. Thus this finding providesstrong support for the validity of theJTI Sensing-Intuiting subscale.

Variable

Table 12: Factor Analysis of the JTI & NEO scale scores

NEO_N .90NEO_E -.79NEO_O .94NEO_A .84NEO_C -.84 -.34 JTI2_EI .95JTI2_SN .78 .37JTI2_TF .82 .38 JTI2_JP .86

European Business School undergraduates (n=112)

Factor 1 Factor 2 Factor 3 Factor 4 Factor 5

Table 13: Correlations between the JTI and 16PF-5

EI SN TF JP

PF5_A -.47 -.11 .37 .05PF5_B -.03 .25 .06 .20PF5_C -.15 -.07 -.05 -.03PF5_E -.20 .23 -.51 .08PF5_F -.73 .29 .37 .38PF5_G .20 -.37 -.11 -.35PF5_H -.78 -.03 .11 .23PF5_I -.01 .24 .33 .01PF5_L .01 .02 -.10 .05PF5_M -.01 .64 .13 .31PF5_N .59 -.12 -.20 -.20PF5_O .25 -.21 .42 -.13PF5_Q1 -.21 .65 .08 .34PF5_Q2 .49 .25 -.31 -.08PF5_Q3 .38 -.50 -.26 -.72PF5_Q4 -.07 .15 .04 .14PF5_IM -.06 -.22 -.06 .02PF5_INF .02 .03 -.18 .08

Course Delegates (N=43)Table 14: Correlations between the JTI and15FQ subscales

FA -.62 .22 .38 .28FC .10 -.07 -.36 -.25FE -.09 .44 -.16 .22FF -.69 .46 .15 .33FG .31 -.46 -.26 -.76FH -.81 .22 .11 .35FI .21 .39 .18 .08FL .08 -.09 -.12 -.15FM -.10 .78 .16 .43FN .29 -.38 .01 -.38FO .35 -.16 .38 -.11FQ1 -.16 .62 -.03 .48FQ2 .61 .22 -.31 .02FQ3 .25 -.41 -.20 -.37FQ4 .13 -.03 .29 .05FMD -.10 .10 -.15 -.05CENT .02 .02 -.07 .07INF .11 .04 -.28 -.07

Course Delegates (N=43)

EI SN TF JP

Table 11: Correlations between the JTI and NEO (Ffi)

Variable EI SN TF JP

NEO_N .20 .30 .33 -.05 NEO_E -.73 .37 .23 .06 NEO_O -.22 .66 .05 .22 NEO_A -.28 .06 .44 -.02 NEO_C -.25 -.10 -.10 -.49

European Business School undergraduates (n=112)

cr

csThe Judging-Perceiving subscale

of the JTI was found to correlatesubstantially with the 15FQ subscaleFG (Conscientiousness), as would beexpected, with this 15FQ factorassessing a desire for order andstructure in daily activities. Thus thisfinding provides further support forthe validity of the Judging-Perceiving subscale of the JTI. TheJTI Thinking-Feeling subscale didnot correlate substantially with anyof the 15FQ subscales, suggestingthat this JTI subscale is assessing aconstruct that is distinct from thoseassessed by the 15FQ.

RELATIONSHIP BETWEEN THEJTI & EPQ (n=220)

Table 15 presents the correlationsbetween the JTI and the EysenckPersonality Questionnaire(EPQ).The most substantial correlation wasobserved between the JTIExtraversion Introversion scale andthe EPQ Extraversion scale. This isconsistent with each of these scale’sdefinitions, and provides furthersupport for the validity of this JTIsubscale. A modest correlation wasalso found between the EPQPsychoticism scale and JTIThinking-Feeling subscale. Thisreflects the observation thatThinking Types tend to be rather lesscaring and empathic in attitude thanare Feeling Types. Thus, this resultprovides further support for thevalidity of the Thinking-Feeling JTIsubscale. In addition a small, butmeaningful, relationship wasobserved between the JTI Judging-Perceiving subscale and the EPQ Liescale; demonstrating that there is asmall tendency for Judging Types towish to present themselves in a moresocially acceptable light. Finally, fur-ther support for the discriminant

validity of JTI is provided by the lowcorrelation found between the JTIThinking-Feeling subscale and theEPQ Neuroticism scale. This pro-vides further evidence that thisrevised JTI subscale is no longerassessing aspects of ‘emotionality’.

RELATIONSHIP BETWEEN THEJTI & OPP (n=43)

Table 16 presents Correlationsbetween the JTI and OPP subscales.As would be expected, the OPPGregarious – Reserved subscale cor-related highly with the JTIExtraversion-Introversion subscale,providing strong support for thevalidity of this JTI subscale.Similarly, a high correlation wasobserved between the JTI Sensing-Intuiting subscale and the OPPPragmatic – Imaginative subscale.This is consistent with these scales’definitions, as both of these subscalesmeasure an interest in the world ofideas and abstract thought. Thusthis strong correlation between thesetwo subscales provides further sup-port for the validity of the JTISensing -Intuiting subscale.

The JTI Judging-Perceiving sub-scale correlated significantly with theOPP subscale Flexible-Detail-Conscious, with this being attribut-able to both of these subscalesassessing a preference for structureand order in daily life. Thus thisstrong correlation provides furthersupport for the validity of theJudging-Perceiving subscale. The JTIThinking-Feeling subscale was notfound to be substantially correlatedwith any of the OPP subscales,reflecting the fact that this JTI sub-scale is measuring a characteristicthat is not fully accounted for by theOPP.

Table 15: Correlations between the JTI and EPQ

EPQ_E -.73 .12 .24 -.11EPQ_N .25 .02 .14 .05EPQ_P .19 .17 -.35 .23EPQ_L .00 .02 .03 -.28

Undergraduate Students (n-221)

EI SN TF JP

EI SN TF JP

ASSERTIVE -.29 .01 -.17 .06FLEXIBLE -.01 .43 .11 .45TRUSTING -.04 .09 .18 .02PHLEGMATIC -.12 .14 -.32 .17GREGARIOUS -.62 -.11 .15 -.03PERSUASIVE -.50 .34 .02 .18CONTESTING -.11 .04 .08 .06EXTERNAL .16 -.23 .12 -.24PRAGATIC -.11 -.66 -.24 -.25CONFORMING .02 .07 .12 .02CENTRAL -.11 .03 .06 -.11

Course Delegates (n=95)

Table 16: Correlations between the JTI and OPP

ct

dkRELATIONSHIP BETWEEN THEJTI & VMI (n=26)

Table 17 presents correlations(values lower that 0.3 have beenexcluded from this table to aid inter-pretation) between the JTI and theValues and Motives Inventory (VMI).As would be expected, the JTIExtraversion-Introversion subscalecorrelated substantially with thosevalues that relate to a need for inter-personal contact (Independence andAffection), providing further supportfor the validity of the Extraversion-Introversion subscale. Moreover, fur-ther support for the validity of thethe JTI Sensing-Intuiting subscalewas provided by the substantial cor-relation between this subscale andthe Aesthetic values subscale of theVMI, with this correlation reflectingthe fact that Intuiting types express apreference for the abstract, aestheticand unknown, rather than for evi-dence and facts.

As would be predicted, the JTIThinking-Feeling subscale wassubstantially correlated with thevalue of Affection, with thisreflecting the empathic interpersonalsensitivity of Feeling types. Thus thissubstantial correlation providesstrong support for the validity of theJTI Thinking-Feeling subscale.Moreover, further support for thevalidity of the JTI Judging-Perceiving subscale was provided bythe substantial correlation betweenthis subscale and the VMI Moralvalues subscale, reflecting thepreference that Judging Types havefor order and control in their dailylives.

RELATIONSHIP BETWEEN THEJTI & OIP (n=45)

Table 18 presents correlations(values lower that 0.3 have beenexcluded from this table to aid inter-pretation) between the JTI and theOccupational Interest Profile (OIP).As would be expected the JTIExtraversion-Introversion subscalewas found to correlate substantiallywith an interest in occupations thatinvolve working with other people;including an interest in sales jobs(Persuasive). This provides furthersupport for the validity of the JTIExtraversion-Introversion subscale.Similarly, support for the validity ofthe JTI Sensing-Intuiting subscalewas provided by the substantial cor-relation between this subscale andthe OIP subscale assessing interest inArtistic (aesthetic) occupations. Thisreflects the creative and imaginativeorientation towards the world ofideas that is characteristic ofIntuiting Types

The JTI Judging-Perceiving sub-scale was found to correlate substan-tially with an interest in occupationsthat involve following set systemsand procedures (Structure). Thisreflects the preference that JudgingTypes have for order and control intheir daily lives, and thus providesfurther support for the validity of theJudging-Perceiving subscale.Furthermore, the JTI Thinking-Feeling subscale correlated substan-tially with an interest in occupationsthat involve caring for others(Nurturing). This reflects the factthat an empathic concern for othersis a core characteristic of the FeelingType. Thus, this result provides fur-ther support for the validity of theJTI Thinking-Feeling subscale.

Table 17: Correlations between the JTI and VMI

VMI (scale)

Traditional -.41Moral -.25 -.33 -.58Independent .62 .42EthicalAltruistic .35 -.34Affiliative -.45 -.34 .44Affection -.61 .61 -.39Nach -.49 -.48Financial -.40Safety .30 -.49Aesthetic .38 .66 .31

Undergraduates (n=23)

JTI-EI JTI-SN JTI2-TF JTI-JP

Table 18: Correlations between the JTI and OIP

VarietyStability -.34 -.41Structure .30 .81People -.83 .36Control -.36 -.40 -.33Persuasive -.56ScientificPracticalAdministrative -.37Nurturing .54Artistic .70 .34Logical -.34

European Business School Students (n=45)

JTI2_EI JTI2_SN JTI2_TF JTI2_JP

dl

BIASdmSEX DIFFERENCES

Table 19 presents data on sexdifferences on the JTI subscales.Inspection of this table indicatesthat, for the current sample, the onlysubscale which shows significantmean differences between men andwomen is Thinking-Feeling. Thistable further indicates that, incomparison to the men in the presentsample, women demonstrate a smallmean preference for Feeling overThinking. This sex differnce has alsobeen reported for the MBTI (bothForm G and Step 1). Thus thisobserved sex difference may reflect areal difference between men andwomen, rather than reflecting testbias

Table 20 presents alphacoefficients for each of the JTIsubscales by sex. Inspection of thistable indicates that each of thesubscales has comparable levels ofinternal consistency for men andwomen.

AGE DIFFERENCES

Table 21 presents correlationsbetween age and each of the JTIsubscales. All these correlations areclose to zero, indicating that age isnot associated with JTI scores.

RACE DIFFERENCES

Table 22 presents data on racedifferences on the JTI subscales,based on samples of undergraduatesfrom white and ethnic minoritybackgrounds. Inspection of this tableindicates that there are no significantmean differences between thesesamples.

THE EFFECT OF RACE ON THERELIABILITY OF THE JTISUBSCALES

In order to adequately demonstratethat a test is not biased againstcertain racial groups it is insufficientsimply to show that there are nosignificant mean differences betweenthe scores obtained by differentracial groups. Rather, it is necessaryto demonstrate that the test is asvalid and reliable when used on thespecified subgroup as it is when usedon the population on which it wasinitially developed. To this end Table23 presents the alpha coefficient foreach of the JTI subscales, on asample of ethnic minority under-graduates. Inspection of this tableindicates that all these alphacoefficients are high, demonstratingthat the JTI subscales continue to bereliable when completed by ethnicminorities.

Table 19: JTI subscale means and standard deviations for men and women

MeanWomen

Mean Men

pvalue

SDWomen

SDMen

EI 26.2 28.2 .102 10.6 9.9SN 34.1 34.5 .801 9.1 9.3TF 40.0 36.1 .005 7.1 7.8JP 29.1 29.4 .854 8.2 9.2

Sample = personnel and training professionals (men = 63 women =63)

Table 20: Alpha coefficientsfor JTI subscales by sex

Men Women

EI .88 .86SN .81 .80TF .83 .79JP .83 .84

Sample = combined sample(men = 274 women =421)

Table 21: Correlations betweenthe JTI subscales and age

JTI Subscale

EI -.06SN -.00TF .02JP .14

Sample = personnel & trainingprofessionals (n = 158)

AGE

Table 22: JTI subscale means and standard deviations, for white and ethnic groups

MeanWhite

Mean Ethnic

p value

nWhite

nEthnic

SD White

SD Ethnic

JTISubscale

EI 24.01 21.18 .133 71 51 10.75 9.43SN 37.06 34.94 .120 71 51 7.31 7.43TF 41.68 39.73 .144 71 51 7.65 6.56JP 26.62 25.06 .394 71 51 10.59 8.94

Sample = undergraduates (n = 122)

dn

AlphaCoefficient

EI .85SN .76TF .77JP .83

Sample = undergraduates (n=62) from ethnic minority groups

Table 23: Coefficient alpha for each of the JTI subscales on anethnic minority sample

do THE EFFECT OF RACE ON THE CONSTRUCT VALIDITY OF THE JTI

Table 24 presents correlationsbetween the 15FQ and the JTI sub-scales for a sample of undergradu-ates from ethnic minoritybackgrounds. As would be predicted,the JTI Extraversion-Introversionsubscale correlates highly with allthe 15FQ extraversion factors; mostnotably with factors FH (SocialBoldness), FF (Liveliness), FA(Warmth) and FQ2 (GroupOriented). This provides strong sup-port for the validity of the JTIExtraversion-Introversion subscalewhen completed by ethnic minoritygroups. Moreover, as would beexpected, the JTI subscale Sensing-Intuiting correlated substantiallywith the 15FQ factors M(Imaginative) and I (Intuitive), bothof which assess a creative, imagina-tive intuitive orientation towards theworld of ideas. Thus, this findingprovides strong support for thevalidity of the JTI subscale Sensing-Intuiting when completed by ethnicminority groups.

The Judging-Perceiving subscaleof the JTI was found to correlatesubstantially with the 15FQ subscaleFG (Conscientiousness), as would beexpected, with this 15FQ factorassessing a desire for order andstructure in daily activities.Similarly, there was a modest corre-lation between this JTI subscale andthe 15FQ factor Q3 (Disciplined),reflecting the fact that both of thesesubscales assess a preference fororder, discipline and self-control indaily life. Thus, this finding providesfurther support for the validity of theJudging-Perceiving subscale of theJTI when completed by ethnic

minority groups. The JTI Thinking-Feeling subscale did not correlatesubstantially with any of the 15FQsubscales, suggesting that this JTIsubscale is assessing a construct thatis distinct from those assessed by the15FQ.

Finally it is noteworthy that thepattern of correlations obtainedbetween the JTI subscales and the15FQ factors, on the present sampleof undergraduates from ethnicminority backgrounds, was broadlysimilar to the pattern of correlationsobtained on a sample of white euro-pean course delegates (reported inTable 14). This supports the con-struct validity of the JTI when com-pleted by ethnic minority groups.

Table 25 presents correlationsbetween the JTI and the OPP sub-scales on a sample of undergraduatesfrom ethnic minority backgrounds.As would be expected, the OPPGregarious – Reserved subscale cor-related highly with the JTIExtraversion-Introversion subscale,providing strong support for thevalidity of this JTI subscale whencompleted by an ethinc minoritysample. Similarly, a high correlationwas observed between the JTISensing-Intuiting subscale and theOPP Pragmatic – Imaginative sub-scale. This is consistent with thesescales' definitions, as both of thesesubscales measure an interest in theworld of ideas and abstract thought.Thus this strong correlation betweenthese two subscales provides supportfor the validity of the JTI Sensing -Intuiting subscale when completedby ethnic minority groups.

dpEI SN TF JP

FA -.53 -.18 .06 .26FC .34 -.20 -.14 -.46FE -.38 .10 -.21 -.00FF -.66 .10 .02 .06FG .07 -.46 -.19 -.63FH -.71 .22 .14 -.13FI .04 .45 .36 .30FL .31 -.24 -.16 .14FM .08 .64 .34 .43FN .36 -.18 -.02 -.24FO .28 .05 .01 .17FQ1 .01 .17 -.05 .26FQ2 .58 .11 -.13 .08FQ3 -.07 -.25 -.12 -.46FQ4 .14 -.05 .08 .20FMD -.20 .05 .10 -.30CENT -.14 .03 -.19 .21INF -.12 -.08 .08 .04

Sample = undergraduates (n = 62)from ethnic minority groups

Table 24: Correlations between the JTI and 15FQsubscales on an ethnic minority sample

EI SN TF JP

ASSERTIVE -.34 .09 -.23 -.21FLEXIBLE .09 .16 .13 .54TRUSTING -.28 .21 .33 -.03PHLEGMATIC -.24 -.15 -.07 -.28GREGARIOUS -.57 -.24 .21 -.26PERSUASIVE -.46 .25 -.00 .09CONTESTING -.05 .11 .07 .13EXTERNAL .23 .16 .15 .28PRAGATIC .00 -.66 -.36 -.24CONFORMING -.07 .08 .13 -.30CENTRAL .08 -.03 .06 .13

Sample = undergraduates (n=62)from ethnic minority groups

Table 25: Correlations between the JTI and OPP on anethnic minority sample

dq The JTI Judging-Perceiving sub-

scale correlated significantly with theOPP subscale Flexible – Detail-Conscious, with this being attribut-able to both of these subscalesassessing a person’s preference forstructure and order in their dailylives. Thus this strong correlationprovides support for the validity ofthe JTI Judging-Perceiving subscalewhen cmpleted by ethnic minoritygroups. Finally it is noteworthy thatthe JTI Thinking-Feeling subscalewas not found to be substantiallycorrelated with any of the OPP sub-scales. This clearly reflects the factthat this JTI subscale is measuring acharacteristic that is not assessed bythe OPP.

Finally it is noteworthy that thepattern of correlations obtainedbetween the JTI and OPP subscales,on the present sample of undergrad-uates drawn from ethnic minoritybackgrounds, was broadly similar tothe pattern of correlations obtainedon a sample of white europeancourse delegates (reported in Table16). This further supports the con-struct validity of the JTI when com-pleted by ethnic minority groups.

RESOURCEMATERIALS ONPSYCHOLOGICALTYPE5

BOOKS

Please Understand Me, An Essayon Temperament Styles by DavidKeirsey and Marilyn Bates.Prometheus Nemesis BookCompany, P.O. Box 2748, DelMar, CA 92014 One of the more widely knownbooks describing the MBTI®.

Portraits of Temperament, DavidKeirsey. Prometheus NemesisBook Company, P.O. Box 2748,Del Mar, CA 92014

Gifts Differing, Isabel Briggs-Myers(with Peter Myers). ConsultingPsychologists Press, 1980ISBN 0-89106-011-1 (pb)0-89106-015-4 (hb).

Manual: A Guide to theDevelopment and Use of theMyers Briggs Type Indicator, byIsabel Briggs-Myers and Mary H.McCaulley. ConsultingPsychologists Press, 1985.

LifeTypes, by Sandra Hirsh andJean Kummerow, ISBN 0-446-38823-8 USA and ISBN0-446-38824-6 Canada. WarnerBooks, Inc., 1989.

Facing Your Type, George J.Schemel and James A. Borbely.Published by Typofile Press,Church Road, Box 223,Wernersville, PA 19565.

Type Talk. Otto Kroeger and JanetM. Thuesen. Bantam DoubledayDell Publishing Group, Inc.(Tilden Press also mentioned.)ISBN 0-385-29828-59.An easy-to-read book that givesprofiles for all sixteen personalitytypes.

Type Talk at Work. Otto Kroegerand Janet M. Thuesen. ISBN 0-385-30174-X.

Type Watch. Otto Kroeger and JanetM. Thuesen.

The Leadership Equation. LeeBarr and Norma Barr. EakingPress, Austin, Texas. 1989.

Using the Myers-Briggs TypeIndicator in Organizations.Sandra Krebs Hirsh. ConsultingPsychological Press, Inc., PaloAlto, CA. 1985.

People Types and Tiger Stripes.Gordon Lawrence. Available fromCenter for Application ofPsychological Type, Gainesville,Florida. ISBN 0-935652-08-6.This book is written primarily tohelp teachers counsel students,but it applicable for other relateduses.

Working Together. Olaf Isachsenand Linda Berens. New WorldManagement Press, Coronado,CA. 1988.

ds

Psychological Types, C.G. Jung,H.G. Baynes (translator).Bollingen Series, Princeton U.P.,1971 ISBN 0-691-01813-8 (pb)0-691-09770-4 (hb).

An Introduction To Theories ofPersonality, B.R. Hergenhahn.Prentice-Hall, New Jersey, 1990.

An Empirical Investigation of theJungian Typology, by LeonGorlow, Norman R. Simonson,and Herbert Krauss. In Theoriesof Personality, Primary Sourcesand Research, editors: GardnerLindzey, Calvin S. Hall, MartinManosevitz, Robert E. KriegerPublishing Company, Florida,1988.

The Measurement of LearningStyle A Critique of FourAssessment Tools Timothy J.Sewall, University of Wisconsin,1986.

Dichotomies of the Mind: ASystem Science Model of theMind and Personality,WalterLowen (with Lawrence Miike).John Wiley, 1982 ISBN 0-471-08331-3.

Introduction to Type: by IsabelBriggs Myers, published byCenter for Applications ofPsychological Type, Gainesville,Florida 32601.

PERIODICAL LITERATURE

The Type Reporter. Susan Scanlon,Editor. For Subscription informa-tion, mail to: 524 North PaxtonStreet, Alexandria, VA 22304.

Journal of Psychological Type.The official research journal ofthe Association for PsychologicalType, 9140 Ward Parkway,Kansas City, MO 64114.

dt

ek

ADMINISTRATIONAND SCORINGINSTRUCTIONS FORTHE JTIIA

PPE

ND

IX

ADMINISTRATION

The Jung Type Indicator (JTI) isadministered in a self-scoring (paper& pencil) questionnaire.

SCORING

The JTI is self-scoring. Scoringinstructions are provided inside thequestionnaire booklet, by tearing offthe top (flimsy) sheet.

ADMINISTRATIONINSTRUCTIONS FORTHE JTIBEGIN BY

● Introducing yourself and explaining your role

● Explaining the purpose of thesession

● Emphasising confidentiality● Describing when and how feed-

back will be given – Answeringany questions

Continue by reading aloud the fol-lowing instructions exactly as givenbelow. Say:

From now on, please do not talkamong yourselves, but ask me ifanything is not clear. You willbe completing the Jung TypeIndicator. While you arecompleting the questionnaire Ishall be checking to make sureyou are not making anyaccidental mistakes when fillingin your answers on thequestionnaire. I will not bechecking your answers.

WARNING: It is mostimportant that questionnairesdo not go astray. They shouldbe counted out at the beginningof the assessment and anyquestions that have not beencompleted should be returned.Remember to collect all the top(flimsy) sheets that contain theJTI questions. The respondentscan however keep the profilechart and capsule typedescriptions.

DISTRIBUTE THEQUESTIONNAIRES

Then ask:

Has everyone got two sharppencils and a questionnaire.

Rectify any omissions, then say:

Please open the questionnaireand print your first name andfamily name, in the spacesprovided. Please also indicateyour title, age, sex andeducational level in the spacesprovided. Do not forget to writedown the highest educationalqualification you have obtained(e.g. BTech Engineering) andyour occupation (studentsshould state their field of study).

Allow time for completion of the bio-graphical information and then say:

Please follow the instructions asI read them to you.

This is a questionnaireconcerning your interests,preferences and attitudes abouta range of things. There is notime limit, however most peopletake about 10 minutes tocomplete thequestionnaire.

Answer each question byfilling in the box that bestdescribes you.

em

Indicate the example response on theanswer sheet. Now say:

When answering the questionsplease remember the following:

Do not spend too much timepondering over the answer toeach question. The informationgiven in a question may not beas full as you might wish, butplease answer the question asbest as you can.

Please try to avoid the middle(in between) answer whereverpossible.

Try to be as truthful as you can.Don’t give an answer justbecause it seems to be the rightthing to say.

Make sure you answer everyquestion, even those which donot seem to apply directly toyou.

Do not erase an answer. If youwish to change an answer, putan X through the incorrectresponse and fill in the correctresponse.

Please note that the questionsand answers run across thepage (indicate this torespondents by holding up aquestionnaire and pointing)and not down the page.

Indicate how the questions andanswers run across the page.

Then say very clearly:

Is everyone clear about how tocomplete this questionnaire

Deal with any questions, appropri-ately, and then say:

Please begin

Answer only questions relating toprocedure at this stage, but enter intheAdministrator’s Test Record anyother problems which occur. Walkaround the room at appropriateintervals to check for any potentialproblems that may occur.

When everyone has completed thequestionnaire then say:

Thank you for completing the Jung Type Indicator

PROCEED TO SCORINGINSTRUCTIONS

en

The JTI is administered in a self-scoring booklet. The test admin-istrator should ensure that thequestionnaire has been correctly beencompleted before scoring the JTI.Respondents can score their own JTIby following the simple scoringinstructions inside the test booklet.

The instructions for scoring theJTI, the scoring key and Profile chartare exposed by tearing the question-naire along the perforation. The totalraw score for each JTI subscale isobtained by adding the numbers ineach column that have been marked.(Note: the subscales are scored bycolumn and each subscale scoreshould be within the range of 0 to60. column). Each of the four rawsubscale total scores should then beentered into the appropriate box at

the bottom of each column. Theseraw total scores are converted intoprofile scores by referring to theScore Conversion table (which isindicated by a bold arrow). The fourprofile subscale scores are thenmarked on the JTI Profile Chart.(Note: do not enter raw subscalescores on the JTI Profile).

The four letter JTI code can beidentified by referring to the JTIprofile. Profile scores 1-4 areclassified as E, S, T and J, whileprofile scores 7-10 are classified as I,N F and P. Middle values (5 and 6)are treated as borderline preferences.When borderline preferences areindicated, additional JTI codes canbe generated by using both ends ofeach subscale.

SCORING INSTRUCTIONSFOR THE JTIeo

PSYCHOMETRIC DATAFOR THE ORIGINALVERSION OF THE JTIIIAPPENDIX

● RELIABILITY

● VALIDITY

● FACTOR STRUCTURE FOR JTI

INTERNAL CONSISTENCY

Internal consistency reliabilities(Cronbach’s Alpha) were computedon samples 1 & 5. The coefficientsare computed over both the com-bined males & females for each rele-vant sample. Table 26 belowprovides these coefficients separatelyfor each sample and for the entire setof respondents. In addition, themean ‘corrected’ item-total correla-tions for each scale are also reportedfor each of these samples. Theseparameters index the mean associa-tion between the constituent itemswithin a scale and the scale scoreitself, each individual item-scalescore coefficient is corrected for theinflation of the coefficient due to theitems inclusion in the scale score.Table 26 also provides informationconcerning the total number of itemswhich make up each scale of the JTI.

Table 26 shows that all the JTIdimensions have reliability coeffi-cients above .7indicating that thetest meets a high standard of relia-bility. The reliability of these scalescompares favourably with the relia-bility coefficients reported in the usermanual for the Myers-Briggs TypeIndicator.

STABILITY

Table 27 displays temporal consis-tency data for the JTI scales. Bothshort-term (2 weeks)and long-term(3 months) reliability coefficients arereported based on a sample of 80and 83 respectively. As can be seenthe JTI is highly reliable in terms oftemporal stability over the periodsreported.

The reliability of the JTI over thetime periods indicated is extremelyhigh and compares very favourablywith test-retest coefficients reportedfor other tests.

RELIABILITYeq

Table 27: Short & Long Term Test-Retest Coefficients for the JTI

JTI ScaleShort-termTest-retest

Long TermTest-retest

Short-termSEm

Long-termSEm

EI .91 .86 0.6 0.75SN .89 .80 0.66 0.89TF .79 .85 0.92 0.77JP .83 .80 0.82 0.89

Table 26: JTI scale internal consistencies and mean ITC for various samples

Scale

Sample 1 (N=618)

Sample 5(N= 456)

Total Sample(N=5577)

No. ofItems Alpha ITC Alpha ITC Alpha ITC

EI 10 0.82 0.31 0.78 0.27 0.80 0.29SN 15 0.70 0.24 0.72 0.20 0.74 0.21TF 15 0.72 0.15 0.76 0.18 0.75 0.16JP 15 0.73 0.15 0.74 0.16 0.74 0.16

er

VALIDITYAs was said in the introduction oncewe have ascertained the reliability ofa test we must address its validity. Itis important to know that the con-structs we are measuring are valid,that it is indeed measuring the char-acteristic it purports to measure.This section of the manual providesconsiderable evidence to demon-strate that the dimensions of the JTIare consistent with similar measures.

THE INTERNAL STRUCTUREOF THE JTI

Table 28 is based on the total JTIstandardisation sample describedearlier and indicates that thecorrelations between the JTIdimensions are fairly modest in size.This demonstrates that the fourdimensions measured by the testassess different personalitycharacteristics. For purposes ofcomparison the inter-correlationmatrix for the MBTI is presented inTable 28 opposite. This is based on asample of 229 UK undergraduates.It can be observed that theseparately scored pairs of MBTIattitudes and functions areeffectively opposite ends of the samescale, providing support for JTIsingle dimensions.

THE RELATIONSHIP BETWEENTHE JTI AND THE MBTI

A sample of 131 undergraduatevolunteers completed both the JTIand the MBTI as part of a testvalidation exercise. Table 29opposite displays the significantcorrelations between the various JTIdimensions and the MBTI.

As can be seen from the abovetable the JTI dimensions correlatehighly with their MBTI counterparts.All of the correlations are above 0.65showing that there is an excellent-match between the correspondingconstructs of the two tests.

Corrections for unreliability asso-ciated with each of the two instru-ments, would result in the followingcorrelations assuming no measure-ment error.

Table 30 opposite demonstratesthat there is most congruencebetween JTI and MBTI Extraversion-Introversion, with all other matchedpair registering corrected values of.80 or better.

es

Table 28: JTI Intercorrelation Matrix(N=5575)

EI 1.00 -.02 .05 -.21SN -.02 1.00 .33 .46TF -.05 .33 1.00 .20JP -.21 .46 .20 1.00

EI SN TF JP

Table 29: Uncorrected correlations with the MBTI®

MBTI scales

Extraversion .82 -.03 .08 -.23Introversion -.80 .03 -.13 .19Sensing -.22 .65 .09 .45iNtuiting .14 -.66 -.16 -.42Thinking -.02 .18 .66 .21Feeling .10 -.15 -.65 -.10Judging -.30 .35 .18 .70Perceiving .27 -.38 -.18 -.69

EI SN TF JP

MBTI scales

Extraversion 1.0Introversion 1.0Sensing .81iNtuiting -.83Thinking .83Feeling -.81Judging .88Perceiving -.86

EI SN TF JP

Table 30: JTI Corrected correlations with the MBTI®

et

RELATIONSHIP BETWEENTHE JTI AND THE NEO

As part of an occupational assess-ment training course 108 personnelprofessions completed both the JTIand the NEO Short form.

The strongest relationship wasobserved between JTI Extraversion-Introversion and its NEO equivalent(-. 72). This would tend to suggestthat to all intents and purposes theseare virtually interchangeable.Assuming no measurement error thecorrected correlation would increaseto .90, suggesting over 80% commonvariance. The JTI Sensing-Intuitingscale correlates most highly withNEO Openness to change andExperience, whereas both the JTIThinking-Feeling and Judging-Perceiving do not register as strongrelationships with the NEO. Thosethat are observed are meaningful,however. Thinking-Feeling correlateswith Neuroticism, Openness,Agreeableness and inversely withConscientiousness, suggesting thatFeeling types tend to be more emo-tional, open to experience and ideas,empathic and less concerned withstructures. Finally JTI Judging-Perceiving relates primarily withOpenness and marginally (inversely)with Conscientiousness, suggestingthat high scorers on the Perceivingend of the J-P continuum are moreopen to ideas and less conscientious.

RELATIONSHIP BETWEENTHE JTI AND THE 16PF

Table 32 presents the correlationsbetween JTI and 16PF based on asample of 210 delegates primarily inthe Personnel and Training disci-plines attending courses organisedby Psytech International. The resultsdemonstrate some interesting, andmeaningful correlations between theJTI dimensions and the scales of the16PF.

The JTI Extraversion/Introversiondimension shows strong correlationswith the extraversion factors of the16PF – reserved/outgoing,restrained/spontaneous, timid/ven-turesome and group-oriented/self-sufficient. The Sensation/ Intuitiondimension of the JTI shows logicalcorrelations with such 16PF factorsas conservative/ experimenting, sub-missive/dominant and tough-minded/tender-minded.Judgement/Perception on the JTIstrongly correlates with the 16PFfactor of expedient/conscientious, acorrelation you would expect giventhe respective definitions of thesetwo dimensions. The Thinking/Feeling dimension of the JTI corre-lates primarily with 16PF anxietyfactors which might be expectedgiven the emotional sensitivity of theFeeling type.

fk

16PF Scales EI SN TF JP

A Warm .45C Stable .35 .31E Dominant .38 -.32F Impulsive .53G Conscientious -.30 .65H Venturesome .73I Tenderminded -.36L Suspicious -.33M ImaginativeN Diplomatic -.47 .32O Apprehensive -.48Q1 Experimenting -.43Q2 Self-reliant -.54Q3 Self-disciplinedQ4 Tense-driven -.37FG Fake Good .50

Table 32: Correlations Between the JTI & 16PF (N=210)

Neo Scales

N .29 .02 .42 -.04E -.72 .11 .09 .20O -.32 .53 .34 .37A -.14 .24 .31 .17C -.03 -.29 -.36 -.29

N NeuroticismE ExtraversionO Openness to ExperienceA AgreeablenessC Conscientiousness

EI SN TF JP

Table 31: Correlations between JTI and NEO

fl

RELATIONSHIP BETWEENTHE JTI AND 16PF-5

A sample of 84 delegates on aPsytech training course completedboth the JTI and the16PF version 5as part of the course requirement.The results yield extremely strongrelationships between JTI and 16PF-5 in a number of areas, namely, E-I,SN and J-P, with less significant, butnonetheless meaningful relationshipsbetween TF and relevant 16PFfactors.

The JTI E-I scale registers amongthe most prominent correlations with16PF-5. Those high on the E-I con-tinuum, tend to be Reserved,Serious, Shy, Private and Self-reliant.The JTI S-N scale correlatesmost highly with Abstractedness on16PF-5 but also notable correlationsare observed with Sensitivity,Openness to Change, andExpediency. The JTIT-F scale fails toregister the very high correlationswith 16PF-5, although this is fairly-consistent with other findings e.g.16PF vs. MBTI and 15FQ and JTI.Those that are elevated, suggest thatthose high on the T-F scale tend tobe Warm, Sensitive andAbstract butare also inclined to register a degreeof anxiety in terms of Emotionalityand Apprehension. Finally, JTI scaleJ-P correlates primarily with 16PFFactor G:Expediency (-.67) but alsowith Openness (Q1), Abstractedness(M) and Tolerance of Disorder (Q3) .Perhaps a little surprising is thefairly elevated correlation with H:SocialBoldness.

RELATIONSHIP BETWEENTHE JTI & 15FQ

Table 34 presents the correlationsbetween JTI and 15FQ based of5575 individuals which constitutesthe entire standardisation sample ofJTI described in section 4.2.

The results highlight a number ofmeaningful correlations between thepersonality dimensions of the 15FQand the JTI scales. It will be seenthat the JTI EI scale correlates quitestrongly with the extraversiondimensions of the 15FQ such asOutgoing, SociallyBold etc. The SNscale correlates with those 15FQdimensions – Intuitive, Conceptual&Radical – that one would expectgiven the nature of the SN dimen-sion. The JP scale, with its emphasison organisation and planning, corre-lates most strongly with just thosedimensions on the 15FQ – Detailconscious, Radical & Disciplinedthat one would expect. Thinking-Feeling (TF) tends to correlate moststrongly with the 15FQ anxietydimensions such as Calm-stable,Self-doubting etc., again this is whatone would expect given the emotion-al sensitivity of those people on thefeeling end of this dimension.

fm

15FQ Scale EI SN TF JP

Table 34: Correlations between JTI and 15FQ (N=5575)

Outgoing -.81Calm-stable -.27 -.46Assertive -.37 -.28Enthusiastic -.68 .26 .38Detail conscious -.24 -.58Socially Bold -.80Intuitive .41 .59SuspiciousConceptual .72 .23 .26Restrained .35 -.27 .24 -.43Self-doubting .31 .55 -.23Radical .70 .60Self-sufficient .69Disciplined -.49 -.74Tense-driven .49Distortion -.28

16PF-5 Factors EI SN TF JP

Table 33: Correlations between JTI and 16PF-5

A Warmth -.52 .43 .31B ReasoningC Emotional StabilityE Dominance -.33F Liveliness -.71G Rule-Consciousness -.46 -.67H Social-Boldness -.72 .40I Sensitivity .53 .52 .31L Vigilance -.35M Abstractedness .75 .42 .45N Privateness .55O Apprehension .45 -.30Q1 Openness to Change .50 .47Q2 Self-reliance .68Q3 Perfectionism -.31 -.53Q4 TensionIM Impression Mgt

fn

THE RELATIONSHIP BETWEENJTI AND OPP

A sample of 158 volunteers complet-ed both the OPP and the JTI as partof a test validation exercise. Thesecomprised primarily personnel andtraining delegates attending Psytechcourses but also a group of under-graduate Business Study studentswho completed both tests as a partcourse on Personnel Selection.The table opposite suggests goodoverlap with OPP dimension and JTIwith simple correlations with each ofthe JTI scales and a single OPPdimension ranging from .57 to.73.The most notable relationships existbetween Gregarious and EI (-.64),Pragmaticand SN (-.61), Phlegmaticand TF (-.57) and Flexible with JP(.73). In addition to being linked toanxiety, TF is also related to empa-thy and imagination. With the levelsof overlap reported here, the OPPcould provide fairly accurate esti-mates of type.

RELATIONSHIP BETWEEN JTIAND OCCUPATIONAL INTERESTPROFILE

A total sample of 1971 completedboth JTI and OIP as part of selectionand assessment procedures with adiverse number of organisations. Ascan be seen from Table 36 oppositeExtraversion on the JTI correlatedquite strongly with the OIP needs forvariety, change and people exactlythose needs which one would predictfor Extraverts. Extraverts showed aninterest in Persuasive roles involvingthe communication of information aswell as in those activities centred onother people. Intuitives, as indicatedon the SN scale, showed the need forVariety and Change which would beexpected given the questioning,exploring nature of those peopleplaced at the N end of this dimen-sion. Intuitives were also highlyinterested in Artistic pursuits, againthis is in line with the definition ofIntuitives. Feeling types tended to below scorers on Stability, a measure ofemotional resilience & self-confi-dence, while scoring highly on needfor Change, which would indicatesome lack of emotional self-control.There was also a fairly high correla-tion between Feeling types andArtistic interests, an expression ofthe sensitive emotionality of Feelingtypes. Finally Perceptive types withtheir emphasis on spontaneity andflexibility tend to score highly on theOIP needs for Variety and Change.As would be expected of the sponta-neous Perceptives they showed a dis-like of Administrative tasks such asclerical and financial work.

fo

OPP Scale EI SN TF JP

Table 35: Correlations Between OPP & JTI (N=158)

Assertive -.31 -.37Flexible .42 .73TrustingPhlegmatic -.32 -.57Gregarious -.64Persuasive -.54ContestingPessimistic .37 -.27Pragmatic -.61 -.32 -.29Distortion

Table 36: Correlations between JTI and OIP (n=1971)

OIP Scale EI SN TF JP

Variety -.21 .16 .22Change -.23 -.30Structure .15 .22 .24 .35People -.57Control -.24Persuasive -.38 .17Scientific .13Practical - .13Administrative -.15Nurturing -.25 .15Artistic -.18 .40 .24 .13Logical .10 -.10

OIP-Pers Persuasive Interests OIP-Nur Caring InterestsOIP-Sci Scientific Interests OIP-Art Creative InterestsOIP-Prac Practical Interests OIP-Log Logical InterestsOIP-Admin Administrative Interests

fp

EI SN TF JPEPQR Scale

Table 37: Correlations between JTI, EPQRand I7 (N=221)

Psych .30 .45Extra -.70 .27Neurot .51LieImpul -.23 .29Vent -.27 -.30 .33Emp .56

EPQR-Psych PsychoticismEPQR-Extra ExtraversionEPQR-Neurot NeuroticismEPQR-Lie Social-DesirabilityI7-Impul ImpulsivityI7-Vent VenturesomenessI7-Emp Empathy-Sensitivity

THE RELATIONSHIP BETWEENTHE JTI AND EPQR & I7

A sample of 221 volunteers complet-ed both the JTI and the EPQR andI7 as part of a test validation exer-cise. Table 37, below, providesdetails of the most significant corre-lations found. The JTI Extraversiondimension correlates highly with thecorresponding Eysenck scale, it alsocorrelates with the I7 scales ofimpulsivity and venturesome, aswould be expected. The JTI TFdimension correlates well with the I7scale of empathy and with the EPQNeuroticism scale. Both of these cor-relations are in line with the defini-tion of the TF dimensions. The JTIJP dimension correlates with I7Venturesome and Impulsivity,reflecting the opposite ends of plan-ning & spontaneity of the JP dimen-sion.

RELATIONSHIP BETWEENTHE JTI AND KLINE’S PPQ

Table 38 below gives the correlationsbetween the JTI and Kline’s PPQobtained from a sample of 193 vol-unteers. It can be seen that the JTIEI scale correlates most highly with-the PPQ Extraversion dimension.The SN dimension correlates fairlyhighly with the PPQUnconventionality scale, reflectingthe Intuitives desire to go beyond theobvious. SN also correlates with thePPQ Tenderminded andConscientiousness scales, whichreflects the sensitive and adaptablenature of the Intuitive type. JTIThinking/Feeling correlates highlywith the PPQ Tenderminded scalewhich is to be expected given theemotional sensitivity of the Feelingtype. The JTI JP scale, as one wouldexpect, correlates most highly withthe PPQ Conscientiousness andUnconventionality scales. This isconsistent with the definition givenfor the conscientious, conventionalJudgmental type.

fq

Table 39: Correlations Between JTI &Gordon’s SIV (N=160)

SIV Scales EI SN TF JP

Support .30Confor -.36 -.67Recogn -.20Indepen .26 .38Benevo .27 .28Leader -.29 -.33

SIV-Support SupportSIV-Confor ConformitySIV-Recogn RecognitionSIV-Indepen IndependenceSIV-Benevol Benevolence

PPQ Scales EI SN TF JP

Insecure .29 .23Tender .21 .48Extra -.21Conscient -.27 -.41Unconv .43 .38

PPQ Insecure InsecurityPPQ Tend TendermindednessPPQ Extrav ExtraversionPPQ Conscient ConscientiousnessPPQ Unconv Unconventionality

Table 38: Correlations between JTI & Kline’sPPQ (N=193)

A sample of 160 volunteers complet-ed both the JTI and the Gordon’sSurvey of Interpersonal Values aspart of a test validation exercise. Ascan be seen from the table below anumber of interesting correlationswere found. The JTI E-I dimensioncorrelates quite well with the SIVLeadership scale, suggesting thatleadership is particularly valued byextroverts. E-I also correlates withSIVRecognition and Support reflect-ing the extroverts desire for socialrecognition and the support of otherpeople. The JTI S-N dimension cor-relates most strongly with theSIVConformity scale which is anindication of the solidity and conven-tionality of JTI Sensation types. S-Nalso correlates with SIV Benevolenceand Independence reflectingtheIntuitive’s basic friendliness alongwith their desire for freedom fromrestrictions. Thinking/Feeling on the

JTI correlates with the SIVLeadership and Support. Thelogical,rational Thinking types tended tovalue leadership qualities more thanthe emotional, sharing Feelingstypes. The relationship with supportis a reflection of the need on the partof Feeling types for the support andencouragement of other people. TheJudgmental/Perceptive scale dis-played a high relationship with theSIV Conformity scale. This would beexpected given the desire of theJudgmental type to live an orderedlife, conforming to social values andexpectations. JP also correlateshighly with SIV Independence,reflecting the independent nature ofthe Perceptive type, who likes to befree to act in a spontaneous,unplanned way.

THE RELATIONSHIP BETWEEN JTI AND SURVEY OF INTERPERSONAL VALUES

fr

Table 40: Factor Structure of the JTI and MBTI

I II III IV

MBTI E -.970MBTI I .962MBTI S .906MBTI N -.921MBTI T .912MBTI F -.897MBTI J .938MBTI P -.930JTI EI -.887JTI SN .809JTI TF .841JTI JP .732

Normalized Factor Loadings(Normalized Varimax Rotation)

FACTOR STRUCTUREOF THE JTI

Table 40 displays the results of car-rying out a factor analysis on dataobtained from a sample of 131undergraduate subjects who com-pleted both the JTI and the MBTI.Principal Components extractionwith Normalised Varimax Rotationwas carried out on all the variables.Loadings below 0.4 have beenexcluded for the purpose of clarity.

As can be observed from Table 40below, both the JTI and the MBTIhave extremely clean factor struc-tures with each of the four Jungiantypes emerging as separate factors.

This is clear evidence that the fourscales measured by the JTI, and bythe MBTI, are independent of eachother and that each scale is measur-ing a separate psychological entity.The results of this factor analysissuggest that there is every reason tobelieve that the JTI does measurefour distinct psychological dimen-sions, each covering a different areaof the personality spectrum.

fs

REFERENCES

Agnew M (1997) GeneSys™Assessment & Profiling System Manual.Letchworth: Psytech International Ltd.

Budd R. J. (1992) 15FQ TechnicalManual. Letchworth: PsytechInternational Ltd.

Budd R. J. (1991) The OPP TechnicalManual. Letchworth: PsytechInternational Ltd.

Cronbach, L.J. (1951). Coefficientalpha and the internal structure oftests. Psychometrica, 16, 297-334

Costa P & McCrae R (1992) NEO-PI-RProfessional Manual, Oddessa Florida:PAR Inc.

Darnell B (1992) The OIP TechnicalManual. Letchworth: PsytechInternational Ltd.

Eysenck, H.J. (1960). The Structure ofHuman Personality. London:Routledge& Keegan Paul.

Eysenck, H.J. and Eysenck, S.B.G.(1969). Personality Structure andMeasurement. London:Routledge &Keegan Paul.

Eysenck, H.J. and Eysenck, S.B.G.(1992). Manual to the EysenckPersonality Questionnaire and I7 scales.London:Routledge & Keegan Paul.

Gordon, L.V. (1985) Manual for theSurvey of Interpersonal Values. Henley,UK: Science Research Associates.

Jung, C.G. (1921). PsychologicalTypes, Vol. 6, the collected works ofC.G. Jung. London: RKP

Kline, P. (1984) The Handbook of TestConstruction. London: Routledge andKeenen Paul.

Kline, P. (1993) Personality: ThePsychometric View. London: Routledge.

Kline, P. and Lapham S. (1991) TheProfessional Personality Questionnaire.London: Psychometric Systems Ltd.

Myers, Briggs I. (1962) Manual for theMyers Briggs Type IndicatorConsulting Psychologists Press,California 94306

Russell M, Karol D, 1994, (1994) 16PFFifth Edition Administrator’s ManualInstitute for Personality and AbilityTesting Inc. Champaign, Illinois

IIIAPPENDIX

gk

SUPPLEMENT TORELIABILITY ANDVALIDITY EVIDENCEASSOCIATED FOR JTIRELEASE 21 EQUIVALENCE BETWEEN THE ORIGINAL AND REVISED VERSIONS OF THE JTI

2 RELIABILITY OF THE (REVISED) JTI SUB-SCALES

3 STRUCTURE OF THE (REVISED) JTI SUB-SCALES

4 EQUIVALENCE WITH THE MBTI

Tables1 CORRELATIONS BETWEEN VERSIONS 1 & 2 OF THE JTI (N=40)2 RELIABILITY (ALPHA COEFFICIENTS) FOR THE JTI (REVISED) SUB-SCALES

3 STANDARD ERROR OF MEASUREMENT FOR JTI SUB-SCALES IN STEN UNITS

4 INTER-CORRELATIONS BETWEEN THE JTI SUB-SCALES

5 CORRELATIONS BETWEEN THE JTI AND MBTI SUB-SCALES

6 FACTOR STRUCTURE FOR THE (REVISED) JTI AND MBTI

ISUPP

LE

ME

NT

gmEQUIVALENCE BETWEEN THEORIGINAL & REVISEDVERSIONS OF THE JTI

The substantial correlations present-ed in Table 1 indicate that the twoforms of the JTI are substantiallyequivalent (all above 0.6). Of partic-ular note is the observation that theweakest of these correlations isbetween the Thinking -Feeling sub-scales in the original, and revisedversions, of the JTI. This reflects thefact that a major revision was under-taken of the original TF sub-scale,with the intention of removing thissub-scale’s previously high loadingon emotionality. The validity data,which is presented later in themanual, clearly indicates that thisrevision was successful in that therevised TF sub-scales now moreclosely maps onto the originalJungian concept of the Thinking -Feeling mental function.

A considerable quantity of validi-ty evidence was collected for theoriginal version of the Jung TypeIndicator. This demonstrated thatthe scales of the JTI were measuringthe characteristics that they set outto measure. The above correlations,would suggest that while the formsare not interchangeable, version 2 ofthe JTI builds upon this body of evi-dence and should demonstrate assimilar, if not better construct validi-ty as further data is collected.

RELIABILITY OF THE (REVISED)JTI SUB-SCALES

The data presented in Table 2:Reliability (alpha coefficients) forthe JTI (revised) sub-scales indi-cate that the Jung Type Indicatorsub-scales have a high level of relia-bility across a number of distinctsamples. Most significantly, thesesub-scales demonstrate extremelyhigh levels of reliability for suchshort (15 item) scales.

The Standard Error ofMeasurement (SEm) for JTI sub-scales is presented in Table 3 below.This value provides the 68% confi-dence interval for JTI sten scoreswhich are used on the profile chart.That is for example, an individualobtaining a sten score of 8 on the EIscale, would be expected to scorewithin the range of 7.25 to 8.75 on68% of occasions. This demonstratesthat the JTI scales provide fairlyrobust measurements.

STRUCTURE OF THE (REVISED)JTI SUB-SCALES

The inter-correlations between theJTI sub-scales, presented in Table 4,are all small, and thus indicate thatthese sub-scales are measuring rela-tively independent characteristics.

The highest correlation, betweenSN and JP compares favourably withthe correlation obtained betweenequivalent scales on the MBTI (0.42and 0.39).

gnScale

Intercorrelations

EI .82SN .77TF .60JP .69

Table 1: Correlations between versions 1 & 2of the JTI (n=40)

sample 1 sample 2 sample 3 sample 4

EI .83 .88 .85 .82SN .78 .86 .84 .84TF .75 .75 .80 .78JP .82 .79 .75 .76

1 Business Studies undergraduates (n=40)2 Technician applicants (n=107)3 European Business School undergraduates (n=112)4 Personnel Professionals (N=40)

Table 2: Reliability (alpha coefficients) for the JTI (revised) sub-scales)

SEM

EI .75SN .77TF .94JP .96

Table 3: Standard Error of Measurement forJTI sub-scales in sten units

EI SN TF

EI –SN .15 –TF -.14 .21 –JP .01 .30 .12

Table 4: Inter-correlations betweenthe JTI sub-scales

goEQUIVALENCE WITH THE MBTI

The correlations presented in Table 4 indicate that the JTI sub-scales are, forall practical purposes, measuring dimensions that are identical to thoseassessed by the MBTI. This is clearly demonstrated by the fact that all thecorrelations (corrected for attenuation due to measurement error) approachunity. Moreover, the size of the uncorrected correlations between the MBTIand JTI sub-scales are in the order of those that would be expected to befound between parallel forms.

The strongest test of the JTI’s equivalence to MBTI is achieved by factoranalysing the MBTI and JTI sub-scales. The factor analysis (principal axisfactoring with varimax rotation) of these sub-scales produced a clear fourfactor solution. Table 6 presents the results of this factor analysis (with factorweights of less than .3 not being reported in order to aid the interpretation ofthis factor structure). These results demonstrate that the JTI sub-scales areclearly measuring four independent factors, which map closely onto theMBTI sub-scales. Most importantly, each of the JTI sub-scales weight on onlyone factor, with this factor being clearly defined by the relevant MBTI sub-scales.

gpJTI MBTI Corrected UncorrectedSub-scale Sub-scale correlation correlation

EI Extraversion 1 .87Introversion -1 -.90

SN Sensing .96 .75Intuition -.94 -.71

TF Thinking .97 .75Feeling -.96 -.70

JP Judging .93 .76Perceiving -.97 -.80

Business Studies undergraduates (n=40)

Table 5: Correlations between the JTI and MBTI sub-scales

factor 1 factor 2 factor 3 factor 4

MBTI - J -.93MBTI - P .94JTI - JP .71MBTI - E -.94MBTI - I .93JTI - EI . 81MBTI - T -.90MBTI - F .91JTI - TF .71MBTI - S -.85MBTI - N .88JTI - SN .72

European Business School undergraduates (n=112)

Table 6: Factor structure for the (revised) JTI and MBTI