.

C-::(I J )

COMMUNICATION APPREHENSION ANDSHYNESS: CONCEPTUALANDOPERATIONAL DISTINCTIONS

James C. McCroskey and Virginia P. Richmond

FOR half a century communicationscholars have been concerned. with

identifying and helping people who areanxious about communication. The

early research generally employed theconstructs of "speech fright" and "stagefright" and was focused on the anxietyexperienced by public speakers andactors.l As the field of communicationevolved, more .cholars directed attentionto communication in contexts other than

public speaking. With this evolutioncame an awareness that many peopleexperience anxiety in settings that donot involve the fonnal presentation ofspeeches, such as communicating in meet-ings, communication in small groups,and communicating with one other in-dividual. With the advancement of theconstruct of "reticence,":! the field'sconcern with communication anxietymoved from the narrow context of

public speaking to the broader contextof communication in general.

The construct of "communication ap-prehension"' evolved from the earlier re-ticence conceptUalization.a The original

conceptualization, which has. remainedunchanged, 4 viewed communication ap-prehension as the "fear or anxiety asso-ciated with either real or anticipatedcommunication with another person orpersons.";} Two elements of this con-ceptUalization are noteworthy. First,there is no reference to the context of

the communication. Presumably, then,communication apprehension could beexperienced in any context, not just apublic speaking context. Second, in theoriginal enunciation of the construct itwas unclear whether communication

apprehension was viewed as a trait of anindividual or a response to a specificcommunication encounter.6 However,

the measure reported at that time (the. Personal Report of CommunicationApprehension, PRCA) clearly focused ona presumed trait-like response. In laterwritings, this aspect of the conceptualiza-tion was made explicit. Communicationapprehension can be viewed either asan individual's trait or as an individual's

response to a given sitUation.7

graphs, 37 (1970),269.77.. . 4 See James C. ~IcCroskey and Virginia P.

James C. McCroskey is Professor and Chal~- Richmond. The Quiet Ones: Communicationper;on of the Depa7:,,~ent ?f Sp'eech Com.m~n~' A.pprehension and Shyness (Dubuque, Iowa:catIOn at West VIrginIa Umverslty, and VirginIa Gorsuch Scarisbrick, 1980) and James C. Mc.P. Richmond is A.ssociate Professor in the De- Croskey, "Oral Communication Apprehension: Apartment of SPeech Communication at West Reconceptualization," in Communication YearVirginia University. book 6, ed. Michael Burgorn (Beverly Hills.

1 Theodore Clevenger, Jr.. "A Synthesis of California: SAGE, 1982).Experimencal Rlsearch in Stage Fright," Quar- ;}James C. McCroskey, "Oral Communicationterly Journal of Speech, 45 (1959), 134.45. Apprehension: "A Summary of Rec~nt. Theory

:! Gerald ~L Phillips. "Reticence: Pathology and Research. Human CommunicatIon Re-of rhe Normal Speaker," Speech Monographs, 35 search, 4 (197i), 78-96.(1968), 39.49. 6 :\(cCroskey, "~Ieasures. . . ."

a James C. McCroskev, "Measures of Com. 7 ~IcCroskey, "Oral Communic:ltion Ap.mllnic:ltion-Bound Anxietv." Speech Mono- prehension: A Summary. . . ."

CE~TRAL STATES SPEECH JOUR~AL, Volume 33. Fall 1982

CO:'IMUN1CA TION APPREHENSION A.'olD SHYNESS

Communication apprehension, wheth-er viewed as a trait orientation ofan individual or as the individual's

reaction to a particular situation,is conceptUalized as an internalized,affectively experienced response of theperson experiencing it. No specificphysiological characteristic or behavioris presumed to be perfectly correlatedwith this internal state, since differentpeople may have different behavioral orphysiological manifestations of theiraffective states.S Xevertheless, certain be-

havioral tendencies are theoreticallyassociated with communication appre-hension. )\"umerous studies have con-firmed hypotheses based on this theory.9Thus, both theoretically and empirically,communication apprehension is viewedas an internal, affective response withexternal, observable impact on com-munication behavior.

In recent years the work of psycho-logists under the rubric of "shyness" hasreceived increased attention. Much of

this work has paralleled the work in thecommunication apprehension area, bothin terms of effects and treatment. This

rese3.rch and clinical parallel has led toconfusion of the two constructs, with

some going so far as to consider com-munication apprehension and shyness tobe "conceptual twins".l0 Much of thisconfusion seems to have stemmed from

the lack of a clear conceptualization ofwhat is meant bv shyness. As Zimbardo, ,

puts it, "shyness is a fuzzy concept")1

8 James C. McCroskev, "Validitv of the PRCAas an Index of Oral Communication Apprehen-sion." paper presented at the Speech Communi-cation .-\ssociation convention. Houston. 1975.

9 James C. :.rcCroskey, "Validitv of the PRCAas an Index of Oral Communication Apprehen-sion", Communication Monographs, 45 (1918).192-203.

10 Malcolm R. Parks. "A Test of the Cross-SitUational Consistency of Communication Ap-prehension", Communication ,'I,[onographs, 47(1980). 220-32.

11"Phillip G. Zimbardo, Shyness: What It Is,What 10 Do About [t (Reading, :'Iassachusetts:Addison. Wesley, 1977).

159

Zimbardo, one of the leading writersin the area of shyness, carefully avoidsproviding a constituent definition of theconstruct and prefers to employ only anoperational definition-a single question:"Do you presently consider yourself tobe a shy person?" He then asked hissubjects who answered yes to indicatewhat kinds of people and situationsmake them shy. I:!Strangers, members ofthe opposite sex, and authority figureswere the people most of the shys saidcaused them a problem. A majority ofthe shys found the following situationsproblematic: giving a speech, largegroups, having lower status than others

. in the situation, social situations ingeneral, new situations in general,sitUations requiring assertiveness, sitUa-tions involving evaluation such as aninterview, and situations in which theperson is a focus of attention, such asa small group setting. Clearly, Zim-bardo's data indicate that shyness has animpact across a wide variety of communi-cation settings. much like communica-tion apprehension. Also, shyness seems tobe stimulated only by the possibility ofcommunication. It does not seem to berelated to non-social situations.

CONCEPTUAL DISTINCTIONS

If communication apprehension andshyness seem to be produced by the samesituations. and they seem to have similareffects. it becomes increasingly importantto clarify the conceptualization of shy-ness to determine, if, in fact, the con-structs are different in any meaningfulway.

The closest Zimbardo comes to pro-viding a constituent definition of shynessis to quote the Oxford English dictionaryand suggest that to be shy is to be "diffi-cult of approach. owing to timidity,GlUtion or distrust," and the shy person

I:! Zimbaruo.

460 CENTRAL STATES SPEECH JOURNAL

is "cautiously averse in encountering orhaving to do with some specified personor thing" and "wary in speech or action,shrinking from self-assertion, sensitivitytimid" and the shy individual may be"retiring or reserved from diffidence".13To continue in the mold of Zimbardo,the Thorndike-Barnhart dictionarv de-

I

fines shy as "uncomfortable in company,bashful, easily frightened away, timid,cautious, wary." Similarly we find inRoget's Thesarus that shy is taken tomean "reserved, unsocial, reticent" and

"unaffectionate, uneffusive, unrespon-sive." Running through these non-technical definitions of "shy" are thefollowing components: discomfort insocial-communicative situations, a ten-

dency to withdraw, a tendency to betimid and lack confidence, and a ten-dency to be quiet, neither assertive norresponsive to others.

Pilkonis, a former student of Zim-

bardo, provides a definition of shynesswhich seems consistent with the non-technical definitions. He indicates that

shyness is "a tendency to avoid otherpeople, to fail to respond appropriatelyto them. . . , and to feel nervous and

anxious during interactions with them.."Behaviorally, he notes, "shy people arecharacterized by avoidance of socialinteraction, and when this is impossible,by inhibition and an inability to respondin an engaging way; they are reluctantto talk, to make eye contact, to gesture,and to smile."u Girodo, another majorwriter in the area of shyness, while fail-ing to provide a specific definition ofshyness, indicates that shyness is com-posed of three ele~ents: undevelopedsocial skills, social anxiety, and low socialself-esteem. He also indicates that oneor more of these elements will be

1;\ Ihid.14 Paul A. Pilkonis. Carol Hcape. and Robert

H. Klein. "Tre:ltin~ Shyness and Other Relation-ship Difficulties ill Psychiatric Outpatients."r:v;T!/lIlinication Education. 29 (1980). ~50.

dominant for any given shy person.lll Keyto this conceptualization is the idea thatthe behaviors of all shys will be similar,but the reason(s) for those behaviors maydiffer sharply.

In his recent book on shyness, Phillipsalso fails to provide a specific definitionof shyness.16 While noting' that shypeople talk less than others and aresocially ineffective, Phillips stresses theview that shy people do not have apsychological problem but are the waythey are because of inadequate skills. Heindicates a preference for the term"reticent" rather than "shy" because thefr-rmer suggests that the person has achoice to speak or be silent while thelatter suggests the person has some kindof disability.

How, then, may we distinguish be-tween shyness and communication ap-prehension? To begin, we may not beable to distinguish between the con-structs on the basis of projected be-havior. Both would predict withdrawaland reduced communicative output.Specifically, both would predict less talk-ing. The critical distinction seems to bein terms of immediate causes of the be-

havior predicted. The communicationapprehension construct. predicts thebehavior from a single cause-fear oranxiety. Shyness, on the other hand, sug-gests the behavior may be the product ofsocial anxiety, low social skills (notknowing how to behave), or low socialself-esteem (e.g. expecting to fail in thesituation). Thus, we conclude, communi-cation apprehension and shyness are notparallel constructs, nor are theyisomorphic constructs. Rather, they forma genus-specie relationship. The genusis shyness-the tendency to be timid, reoserved, and most specifically, talk less.

1:>::\lichel Gir_.lo, Shy! (New York: PocketBooks. 1978).

16 Gerald'::\<I. Phillips. Help for Shy People(Englewood Cliffs. New Jersey: Prentice-Hall.1981).

COMMUNICATION APPREHENSION AND SHYNESS 461

One of the species is communicationapprehension-the tendency to behave ina shy manner (talk less) because of fearor anxiety. Another of the species isreticence-the tendency to behave in ashy manner (talk less) because of a lackof communication skill.

OPER.-\TIO:'-l.-\L DrSTI:'-ICTIONS

.-\[though other measures have beendeveloped recently,I7 the 20-itemI8 and25-iteml!1 PRCA have been the opera-tional definitions of communication

apprehension in the overwhelmingmajority of the over 200 studies in-volved with this construct which have

been reported to date. A strong case hasbeen advanced for the validity of theseinstruments.~o

Within the area of shyness, no con-sensus measure has emerged. The mea-sure employed by Zimbardo and hisassociates, as noted previously, is asingle-item, forced-choice scale.:!1 Themetric qualities of the scale are notdiscussed by Zimbardo, but the findingsgenerated by use of the scale are sug-gestive of its validity. More recently,two new instruments have been advanced

to measure shyness. Cheek and Buss haveprovided a nine-item, Likert-type scale

which they report as. unidimensional.:!:!:\-Iost of the items on this scale suggest

17See James C. McCroskey. Janis F. Andersen.Virginia P. Richmond. and Lawrence R."'heeless. "Communication Apprehension ofElementarv and SecondarY Students andTeachers:' Communication Education. 30 (19Hl).122.32: James C. McCroskey, An Introduction toRhetorical Communication. 4th ed. (EnglewoodCliffs. New Jersey: Prentice-Hall. 1982). ch. 2.

IS ~[CCroskev, ":'vIeasures. . . .n19 :-'fcCroskey, "Validitv. . . :':!OJohn A. Daly. "Th'e Assessment of Social-

Communicative Anxiety Via Self-Reports: AComparison of Measurers", Communication.HollOgraphs. 45 (19i8), 204-18; McCroskey,"Validitv. . . .n

:!1 Zimh:lrdo.:!2J. Cheek amI .-\.rnold H. Buss. "Scales of

Shvness. Sociabilitv. and Self-Esteem and Cor-rc1ations .-\.mong Them." unpublished researchrepon. Cniversity of Texas. 19i9.

nervousness or discomfort; thus. the

measure may tap communication appre-hension better than shyness. McCroskey.Andersen, Richmond, and Wheeless

have provided. a 14-item, Likert-type scale which they have labeled a"shyness scale.":;:;;They report the scaleto be unidimensional and. factorallydistinct from. though correlated with,communication apprehension. There isno substantial data base from which to

argue the validity or invalidity of eitherthe Cheek and Buss scale or the Mc-

Croskey et at. scale.

RESEARCH QUESTIONS AND HYPOTHESES

The primary purpose of the presentinvestigation was to determine whetherthe conceptual distinction between com-munication apprehension and shynesswas amenable to empirical opera-tionalization. The first research question.therefore, was:

Q,. Are measurers of communication apprehen.sian and shyness which are isomorphic withconceptualizations of these constructs em-pirically distinct from each other?

Within the trait framework employed inthis study, it would be expected thatmeasures of communication apprehen-sion and shyness would be correlated.However, the correlation should not beso high as to suggest interchangeabilityof the measures or a single responsestructure. Therefore, two hypotheseswere advanced:

HI: A measure of communication apprehensionand a measure of shyness will be moderatelycorrelated.

H.: With oblique factor analysis. the items on ameasure of communication apprehensionand the items on a measure of shyness willform separate factors.

Support of these two hypotheses wouldprovide an empirical foundation for the

:!3 McCroskev,Wheeless. '

Andersen, Richmond. and

462 CE~TRAL STATES SPEECH JOURNAL

conceptual distinction between com-munication apprehension and shynessoutlined earlier. The question wouldremain, however, whether the behavioralimplications of the distinction are mean-ingful. If, as theorized, shyness isproduced by multiple causes whilecommunication apprehension stems frombut one of these causes, observations of.

behavior should be better predictors ofshyness than of communication appre-hension. Thus. we advanced a thirdhypothesis:

H3: Observer reportS of shyness are more highlycorrelated with self-reports of shyness thanthey are with self.reportS of communic:ltionapprehension.

Support for this hypothesis. in conjunc-tion with support for the previoushypotheses, would provide substantial\"alidation for the conceptual distinctionadvanced here. Failure to support thishypothesis would indicate weakness in'the conceptualization. the operationalmeasures, or both.

METHOD

Self-Report l.y[easures

The 25-item PRCA was selected as

our measure of communication appre-hension.2-t As noted previously, this isthe most conunonly employed measureof this construct, and the instrument has

a good record in terms of both reliabilityand validity. In our preliminary study,discussed below, we also employed asecond measure of conununication ap-prehension, the Personal Report ofCommunication Fear (FEAR).25 Thisinstrument is reported to have highreliability and to be highly correlatedwith the PRCA, which" establishes con-current validity.

The 14-item Shyness Scale (SHY) was

24McCroskey."Validity. . . :'25 McCroskev. Andersen. Richmond.

Wheeless. .

selected as our measure of shyness.26 Al-though the Zimbardo scale has been usedin more research and generated a bettercase for validity, it was rejected for thisstudy because of its single-item naturewhich would result in major problemsin terms of data analysis.27' The Cheekand Buss scale was rejected because itwas insufficiently isomorphic withthe shyness conceptualization advancedabove.28 The SHY instrument, althoughreceiving only limited previous use, hasbeen found to be highly reliable and tobe factorally distinct from at least onemeasure of communication apprehen-sion.29 In addition, since all of the Lemson the scale refer directly either to shy-ness or to talkativeness, the scale is

highly isomorphic with our shynessconceptualization.

0 bserver-Evaluation i\t[easures

The design of this study required useof untrained observers providing evalua-tions based on observations of different

subjects in different communicationsituations across varying lengths of time.Because of this variability in observationconditions, it was necessary to generaterating scales for communication appre-hension and shyness that were generalin nature and applicable across bothsubjects and observers. In order to fulfillthis objective, the PRCA and SHY in-struments were reworded to reflect an

observer's rating rather than a self-report. This was accomplished by remov-ing personal pronouns and substitutinggrammatically appropriate versions of"this person" in their place. Post-studyinterviews indicated the observers had

little difficulty completing the measures.

and

26 IcCroskey. Andersen. Richmond. and'Wheeless.

27'Zimbarao.28 Cheek and Buss.29 McCroskey. Andersen. Richmond. and

Wheeless.

COM~ICNICATION APPREHENSION AND SHYNESS 4.63

The Preliminary Study enrolled in graduate classes in instruc-

A preliminarv studv was conducted tional communication. The remainingto explore our r~search' question and test ~ubJe~ts (n = 295) were friends of theseour first two hypotheses. A sample of mdIviduals. The teacher-subjects were606 college students completed the assigned random identification numbersPRCA, FEAR, and SHY scales at the for use by both themselves and a friend.

beginning of the semester in which they Each of these subjects was asked to com-were enrolled in basic communication plete the PRCA and SHY measures onclasses. No discussion of communication him or herself and to complete the

apprehension or shyness preceded the revised versions of these scales on adata collection. Although this study "friend you know well and who knowspermitted us to test our first two hy- ~ou wel1.". After completing a~d return-po theses, the primary purpose of the mg these mstruments. the subjects werestudy was to ensure that the measures given an identical set and asked to havewe intended to employ in the major t~e selected friend complete them over-study were, as hypothesized, correlated nIght. seal the completed measures in anwith each other but factorally distinct. envelope provided. and return them theWere this found not to be the case, next day. Thirty-two subjects (from an

additional measurement development original sample of 327) were unable towould be required before an appropriate secure .data from the selected friend, duetest of Hypothesis 3 could be conducted. to theIr unavailability or unwillingness

In previous work, as noted above, the to participate. These subjects were ex-FEAR and SHY scales were found to eluded from all data analyses.

be correlated but to form separate Th~ procedure just described yieldedfactors in an oblique factor rotation. data m the form of self-reports from 590Even though the FEAR scale had been subjects and data in the form of observerfound to correlate highly with the ratings from the same number ofPRCA, it could not be assumed that a subjects. It should be recognized that

simi~ar factoral relationship would hold these observers must be classified asfor PRCA and SHY. Particularly trouble- completely untrained. There was no in-some to such an inference is the fact that struction given as to what to look for

one of the items on the PRCA specifical- or what to consider. Even with thisly refers to shyness: "I ta~k less because limitation, we believe that our choiceI'm shy." In order for us to claim sup- of f~ie~ds as obsc:rvers is the best that isport for our second hypothesis, therefore, realIstIcally possIble. Only friends canit was decided that, not only would we share the natural environment of anrequire that the items on the SHY individual with minimal impact onmeasure load on a factor distinct from tha.t individual's behavior. More highlycommunication apprehension items, but tramed observers would alter behavioralso that this item from the PRCA would by their very presence. In addition, the

be required to load with the SHY items friends' observations came prior to beingrather than with the other PRCA items. aware they were to provide data, which

mitigated against biased observation.We believe the best observation of be-havior across situations, across timecomes from observers who. in the course

of their natural lives, are present acrosssituations. across time. The only people

The Major Study

The major study involved 590 subjectsand observers. Half of the subjects (n =295) were elementary and secondaryschool teachers (ages 23-64) voluntarily

464 CEXTRAL STATES SPEECH JOURXAL

who meet this qualification are friendsor spouses.

Data Analyses

The Preliminary Study. Internal re-liability estimates (split-half) were com-puted for each of the three measures.The correlations among the raw,summated scores also were computed.

The main analysis was a factor analysiswith oblique rotation. The SAS promaxoption was employed. One, two, andthree factor solutions were examined,although the third factor accounted foran increase of less than five percent ofthe total variance. The one-factor solu-

tion was rejected because it generatedlow (below .30) communality estimatesfor many of the items. The three-factor solution was rejected because itproduced two communication apprehen-sion factors, one representing positivelyworded items and the other representingnegatively worded items, an otherw'isenon-interpretable result frequently ob-tained in previous research.3O Therefore,the two-factor solution was retained fordiscussion.

The iYIajoT Study. Internal reliabilityestimates (split-half) were computed forthe PRCA and SHY self-reports and forthe observers' ratings on the PRCA andSHY measures. Correlations between the

raw, summated self-report scores andbetween the raw, summated observer

report scores were also computed.

The self-report and observer-reportdata were submitted to separate factoranalyses. The two-factor solutions weresubmitted to oblique rotation employingthe SAS promax option. These analysespermitted us to test our first two hy-potheses and respond to the researchquestion. A t-test for significance be-tween dependent correlations (observerSHY! self SHY vs. observer SHY/self

30 ~rcCroskcy. "~[e:lsurcs. - . ."

PRCA) was computed to test hypothesis3.

REsULTS

The Preliminary Study

The internal reliability estimates forthe three measures were as follows:PRC.-\. .95; FEAR .90; SHY .92. Thecorrelations among the raw scores ob-tained were: PRCA/FEAR .85; PRCA/SHY .57; FEAR/SHY .57.

The results of the oblique factorrotation are reported in Table 1. Theobtained inter-factor correlation was .52.This indicates that the raw score cor-relations between SHY and either

measure of communication apprehensionshare approximately five percent morevariance (32%) than scores based onfactor weights (27%). As indicated inTable 1, this is partly a function of oneitem on the PRCA loading with theitems from the SHY scale-the item

which specifically mentions shyness.The results of this study provided

support for our first hypothesis, themeasures of communication apprehen-sion were each moderately correlatedwith the measure of shyness, sharing ap-proximately 30 percent of the variance.The results also support our secondhypothesis, the measures of communica-tion apprehension fonn one factor whilethe items from the shyness measure forma second factor, the only exception beingthe shy item from the PRC.-\. whichloads, as it should, with the other shynessitems. We may, therefore, tentativelyprovide an affinnative answer to ourresearch question: measures of communi-cation apprehension and shyness whichare isomorphic with conceptualizationsof these constructs are empirically dis-tinct.

The positive results of this studypermitted us to conduct the major studyas planned. However. since the one itemon the PRC.-\. appears to be out-of-place

I Ie III

Preliminary Studyl'actor liactor

I (CA) 2 (SHY)

Major SiudySelf-Report

Factor FactorI (CA) ~ (SHY)

Major StllllyObserver. Report

liaclor liaclorI (CA) 2 (SHY)

TABLE I

ROTAT£I) FACTOR LOADINGS

I'RCAJ. While participating in a conversation with a new acquaintance I feel 8.

very nervous. .50* ,43 .48* .47 .62* .452. I have no fear of facing an audience. .71* .35 .!i:;* .:!:! .59* .413. I lalk less because I'm shy. .57 .67* ;; . .54 .62*4. I luok forward to expressmg my opinions at meetings. .65* .46 .1,4* 55 .58* .54:>. I alii afraitl 10 express myself in a group. .61* .47 .(j,(* .'17 .59* .46 &;6. I luok foreward to an opporlunity to speak in public. .72* .37 .tiG* ,10 .64* .44 o-j7. I Iilid the prospect of speaking mildly pleasant. .35* .20 AS* .37 .55* .32 08. When communicaling. my posture feels strained and unnatural. .50* .29 .48* .32 .49* .30 Z9. I am lense and nervous while participating in group discussions. .70* ,46 ./i5* :17 .67* ,41 >

10. Although I' talk fluently with fi'iends. I am at a loss for wonls on the .74* .30 .fi7* .28 'tj

pialforlll. .69* .30 'tj

II. I have no fear about expressing myself in a group. .63* AI .60* ,48 .53* ,46:0t>1

12. My hands tremble when I try to bandle objects on the platform. .59* .23 .58* .20 .56* .15 :I:13. I always avoid speaking in public if possible. .71* .35 .71* .37 .72* .35 t>1

14. I feel Ihat I am more fluent when talking to people than most other en

people arc. .50* ,46 .57* .511 .52* .50 015. I am feadul and tense all the while I am speaking before a group of Z

I'eupk. .70* .34 .72* .3.1 .71* .84 >Hi. Illy thoughts become confused and jumbled when I speak before an

,IIHlielife. . .67* .27 .70* .29 .65* .30 t:I

17. I like to get involved in group discussions. .62* .57 .57* .58 .Ii2* .55 en

18. Although I am nervous just before getting up. I soon forget my fcars:I:

anll enjuy the experience. .59* .32 .62* .29 ,47* .2419. Cunversiug with people who hold positions of authority causes me F;1

III be fea rful and lense. .53* .21 .43* .31 .55* .24 en

20. I dislike to use my body and voice expressively. .52* .42 .32* .22 .47* .8421. I feel relaxed and comfortable wbile speaking. .65* AI .67* .40 .71* .3922. I feci self-consciolls when I am called upon to answer a question or give

an opinion in class. .56* .37 ,49* .3-1 .60* .3023. I face Ihe prospect of making a speech with complete confidence. .64* .2'1 .68* .23 .71* .392.1. I'm afraid to speak up in conversalions. .55* 042 .54* .48 .56* .45

25. I wonld enjoy presenting a speech on a local television show. .61* .29 .60* .2ti .61* .43

....0>U.

TABLE I - (Continued)ROTATED FACTOR LOADINGS

~

SHY

I. I am a shy person.2. Othcr peoplc think I talk a lot.3. I am a vcry talkative person.4. Othcr pcople think I am shy.5. I talk a lot.6. I tcnd to be very quiet in class.7. I don't talk much.8. I talk more than most people.9. I am a quict person.

10. I talk more in a small group (3-6 people)II. Most people talk marc than I do. .

12. Other people think I am very quiet.I~. I talk morc in class than most people do.14. Most pcople are more shy than I am.I'EAR

1. Talking with someone new scares me.2. I look forward to talking in. class.3. I like standing up and talking to a 1?"°up of people.4. I like to talk when the whole class hstens.5. Stauding up to talk in front of other people scares me.6. I likc talking to teachers.7. I am scaned to talk to people.8. I like it when it is my turn to talk in class.9. I like to talk to new people.

10. Whcn somcone asks me a question, it scares me.II. Thcre are a lot of people I am scared to talk to.12. I like to talk 10 people I haven't met before.13. I like it when I don't have to talk.14. Talking to teachers scares me..Primary loading. Items renected to equal polarity before analysis.

than other people do.

,56 .7'!J* .54 .74* .63 .71*.15 .63* .20 .73* .22 .70*.31 .76* .32 .81* .30 .78*.42 .66* .39 .70* .52 .74*.21 .69* .33 .78* .29 .82*.47 .54* .51 .56* .53 .54*.38 .77* .40 .81* .40 .80*.33 .68* .33 .74* .40 .74*

Q.41 .71i* .37 .82* .46 .84*.27 .52* .33 .47* .25 .38* Z.41 .70* .46 .76* .44 .75*.43 .71- .37 .79* .42 .77* ':d

;J..43 .52* .48 .56* .48 .61* t"<.32 .53* .37 .54* .37 .46* en

.52* .44

.68* .52 en

.76* .40 "d

.73* .42

.72* .36 n

.49* .30 :J:

.49* .43 .....0

.70* .48 (j

.46* .44

.55* .33

.53* .32;J.t"<

.43* .33

.52* .51

.53* .28

COM~ICNICATION APPREHENSION AND SHYNESS 467

on that scale. separate analyses, scoringthe PRCA with and without this item,

were performed in the major study.

T~le Major Study

The internal reliability estimates forthe self-reports on both PRCA and SHYwere .94. For the observer ratings. theestimate for the PRCA was .95 and forthe SHY the estimate was .94. The cor-

relations between the self-report rawscores (PRC.-\.jSHY) was .59. The cor-rebtion between the observer-report rawscores was .63. These correlations are not

significantly different (Z = .91,P > .05).Including the PRCA item which men-tions shyness increa~ed the self-reportcorrelation to .61 and the observercorrelation to .65, both non-significantchanges.

The results of the oblique factorrotation for the self-report data are re-ported in Table 1. The obtained inter-factOr correlation was .49. This indicatesthat the raw score correlations between

PRC.-\. and SHY share approximately 11percent more variance (36%) thanscores based on factor weights (25%).This discrepancy is reduced by omittingthe PRC.-\. item which mentions shyness,as suggested by the results of thepreliminary study, but only by abouttwo percent.

The results of the oblique factorrotation for the observer-report data alsoare reported in Table 1. The obtainedinter.factor correlation was .52. This in-dicates that the raw score correlationsbetween PRCA and SHY share approxi-mately 13 percent more variance (40%)than scores based on factor weights(27%). As with the self-reports, omittingthe shyness item from the PRCA sc,?resreduces the correlation somewhat.

These results provide support for ourfirst hypothesis. The measures ofcommunication apprehension were cor-related with the measures of shyness for

both the self-report and the observer-report data and shared approximately 30percent of the variance.

The results also support our secondhypothesis. The items from the com-munication apprehension measure formone factor while the items from the shy-ness measure form another factOr, for

both self-reports and obserVer-reports.'When the analyses included the shynessitem on the PRCA. that item loadedwith the items from the shyness scale.The remaining loadings were alteredno more than .01 either direction by

including this item or omitting it.

Support for our first two hypothesessuggests an affinnative answer to ourresearch question: measures of communi-cation apprehension and shyness whichare isomorphic with conceptualizationsof these constructs, although meaning-fully related as conceptualized, areempirically distinct. The fact that theresults for self-report and observer-report data were consistent strengthensthis affinnation.

Our final hypothesis (H3) also wassupported. Observer reports of shynesswere more highly correlated with self-reports of shyness (r = .53) than theywere with self-reports of communicationapprehension (r = .37; t = 5.07. P<.001). In contrast, the correlations ofobserver reports of communication ap-prehension with self-reports of com-munication apprehension (r = .46) andself-reports of shyness (r = .44) werevirtUally identical (t < 1). These resultsreinforce the conceptual distinction be.tween communication apprehension andshyness advanced earlier. Since theformer is an internal trait and is but oneof the causes of shy behavior. it is moredifficult for observers to identify than isthe shy behavior itself.

DISCUSSION

The purpose of this research was to

468 CDITRAL STATES SFf.ECH JOl'~'" u.

Jc:tw a conceptual distinction betWeencommunication apprehension and shy-ness and to determine whether measures

reflecting the distinction would provideempirical validation of the conceptuali-zations advanced. Our first hypothesiswas that a measure of communication

apprehension and a measure of shynesswill be moderately con-elated. Thishypothesis was supported by the self-report results of both the major studyand the preliminary study as well as bythe observer-report results of the majorstudy. The second hypothesis suggestedthat the items on a communication ap-prehension measure would form a factorseparate from the items on a shynessmeasure. This hypothesis was also sup-ported by the self-report results of boththe major study and the preliminarystudy as well as the observer-report re-sults of the major study. Finally, thethird hypothesis was supported. Ob-servers' reports of shyness were morehighly associated with self-reports of shy-ness than self-reports of communicationapprehension in the major stUdy. Insum, all thre hypotheses were supported.

Our research question was concernedwith whether communication appre-hension and shyness measures areempirically distinct from each other. Thesupport obtained. for our three. hy-potheses strongly indicltes the appro-

. priateness of an affirmative response tothis question. Not only are thesemeasures distinct when employed asself-reports. they are also distinct whenemployed as instruments for observerreports. Of course, this conclusion ap-plies only to the measures employed inthis investigation, PRCA. FEAR, andSHY. Other operationalizations may noty;~ld empirical distinctions. For ex-ample, we would not expect the Cheekand Buss shyness scale to be distinct fromthe PRCA or FEAR scales, since manyof . the items on the scales are highlysimilar.31 In this reg:1rd. however, itshould be noted that the item refer-

encing shyness that app,:ars on thePRCA shold not be included in future

research where an empirical distinctionbetween communication apprehensionand shyness is of importance.

31 Cheek and Buss.