Vol. 3 N9. 1 Fall 1976

06"1-

THE RELATIONSHIP BETWEEN COMMUNICATIONAPPREHENSION AND ACADEl\'1IC ACHIEVEMENT

AMONG COLLEGE STUDENTS

JAMES C. McCROSKEY and JANIS F. ANDERSEN

West Virginia University

A series of studies are reported which indicate that high communication apprehensiveshave lower academic achievement in traditional interaction-oriented educational sys-tems than low communication apprehensives. but that no similar relationship exists in acommunication-restricted educational system. Data are also reported indicating thathigh communication apprehensives prefer mass lecture classes over small classes whilemoderate and low communication apprehensives' preferences are the reverse. Theimplications of these results for choosing or designing instructional systems are discus-sed. .

Communication between teacher and student

plays a major role in traditional learning environ-ments. Teachers must communicate with students

for students to achieve maximum learning. It isthrough the process of communication that teacherstranslate the course content into a symbolic codewhich can be decoded and interpreted by students.Teachers who fail in this communication fail in theirresponsibility to educate students.

Student achievement is also partly determined bythe student's communication behaviors. First,classroom questioning aids the student in conceptclarification and concept integration. Second, stu-dents participate in determining the pace of contentpresentation through verbal and/or nonverbal feed-back that communicates to the teacher whetherlearning has occurred. On the basis of this informa-tion, the teacher may decide to review old material,offer additional information to explain old con-cepts, or introduce a new idea. Third, in someclassrooms, student communication is p~rceived tobe so important that it is directly linked to evaluationof Sourceperformance. In these classrooms, a classparticipation evaluation is part of the course designandcommunication performance is directly influen-tial on course grade.

Communication between teachers and students is

not always adequate. There are two major barrierstosufficient teacher-student interaction. Large class

size is the most frequently noted barrier. The morestudents enrolled in a class, the less time for indi-vidual students to express their ideas in class.Nationwide, elementary and secondary teachers areciting reduced class size as an important concern innegotiation and bargaining meetings with schoolboards. Teachers argue that large classes do notallow time to communicate with all students and,thus, inhibit student learning. The several-hundred-student college lecture class has also beenextensively criticized as being impersonal and inef-fective. Some educators have argued for its elimina-tion on the grounds that it does not allow for theamount of student-teacher interaction necessary foradequate student questioning, feedback, and learn-ing. Most educators agree that increased class sizeresults in decreased communication potential be-tween students and teachers. Many further agreethat large classes are detrimental to learning due tothis decr~ased interaction potential.

The second major barrier to classroom interactionmanifests its effect primarily in smaller classes.Even if there is ample opportunity for student-teacher interaction, some students are functionallyunable to communicate because of communication

apprehension. These students are so anxious aboutcommunication with teachers and other studentsthat their anxiety interferes with their communica-tion efforts. Based on a series of recent research

74 McCroske)' and Andersen

studies, McCroskey (1975) has concluded that atleast 15to 20 per cent of all students suffer fromdebilitatim! oral communication apprehension.- -These students may be severely handicapped insmall classes .because they do not ask questions,give feedback. or participate in class discussions.Apprehensive students may learn less because theydo not attempt to restructure the classroom presenta-tion of information to meet their specific needs.They may also receive lower evaluations due totheir failure to participate in classroom interactions.

RATIONALE AND HYPOTHESES

The amount of success in a learning environmentwhich can be attributed to student initiated com-munication has not been empirically determined.Previous research, however, has isolated four majorpredictors of student success in learning environ-ments. First, measures of intelligence and aptitudeare predictive of grade point average (GPA). Al-though intelligence tests are currently under seriousattack. many investigations lead to the conclusionthat the correlation of intelligence test scores andschool grades is substantial (Thorndike & Hagen,1969;Binder, Jones & Strowig, 1970). Correlationsof .50 and .60 between intelligence scores andschool grades are typical.

Prior achievement, as measured by standardizedachievement tests, is a second major predictor ofGPA. The validity of the American College Test(ACT) scores and the College Entrance Examina-tion Board (CEEB) testing program have beenclearly established (Hoyt & Munday, 1968;Schrader & Stewart, 1971). The correlations be-tween achievement test scores and GPA generallyvary from about .33 to .65, depending on the natureof the academic institution used and the homo-

geneity of ability in the group studied. Past grades inschool are also predictive of future GPA. The corre-lations between high school grades and collegefreshmen grades are usually in the area of .50(Bloom, 1964).

A third major predictor of academic success is thepossession of certain nonintellectual personalityvariables. Cattell. Sealy, and Sweeney (1966) re-

port that 25 per cent of the variance in GPA isassociated with personality differ~nces. When per-sonality predictors are combined with achievementscores the predictive power can be increased toaccount for 33 per cent of the variance (Mandryk &Schuerger, 1974). .

A final predictor of student success which weneed to consider is expectation. Research indicatesthat teachers' expectations of their students' per-formance are highly predictive of the students' ac-tUal performance (Dusek. 1975), particularly intutorial or semitutorial learning situations (Beez,1968).

The various intercorrelations between factorswhich have been isolated to predict student perfor-mance contribute to some confusion in attempting

to predict student GPA. When all previously iso-lated factors are combined in a predictive modelthere is at least 25 per cent of the variance in GPAleft unexplained. More often, over 50 per cent is notexplained. This study attempted to isolate one addi-tional potential predictor of stUdent achievement,stUdent's level of communication apprehension.

As noted previously, intelligence, priorachievement, various dimensions of personality.and teacher expectations have been found to predictstudent achievement. Recent research has estab-lished a relationship between some of these predic-tors and communication apprehension, particularlypersonality and teacher expectations.

Communication apprehension is substantially as-sociated with certain personality variables.McCroskey, Daly, and Sorensen (1975) found thatcommunication apprehension is significantly corre-lated with many of the personality variables asses-sed by Cattell's Sixteen Personality Factor Ques-tionnaire, namely cyclothmia, emotional maturity,dominance, surgency, character. adventurousness,trustfulness, confidence, self-control, and generalanxiety. Several of these personality characteristicshave been found previously to be predictive of stu-dent achievement, although the causal link bet\Veenthese variables and achievement has been elusive

(Cattell, Sealy & Sweeney, 1966). Because of!he in-trinsic role played by communication in traditionaleducation, we reasoned that these personality vari-ables may attain their predictive power through their

APPREHENSION AND ACADEMIC ACHIEVEMENT 75

association with communication apprehension,which is the causal agent producing reduced com-munication, hence reduced learning in the tradi-tional educational system. Thus, we hypothesizedthat high communication apprehensives will havelower achievement in a traditional educational sys-tem than will low communication apprehensives.

Communication apprehension has also beenfound to be associated with negative teacher expec-tations. (McCroskey and Daly (1976) found thatteachers project greater achievement for a childidentified with behaviors typical of a low communi-cation apprehensive than a child with characteristicbehaviors of a high apprehensive. Children iden-titied with behaviors typical of high communicationapprehension were projected to perform less well inall areas of the school environment. Since teacher

expectations have been found to be linked withstudent achievement, as previously noted, we wereagain led to pose our hypothesis that high communi-cation apprehensives wiil have lower achievementin a traditional educational system than will lowcommunication apprehensives.

If communication apprehension were found tohave a substantial negative correlation with intelli-gence or high school achievement, we would haveeven more reason to consider our hypothesis tena-ble. Such associations, however, have not beenestablished. McCroskey, Daly, and Sorensen(1976), although finding numerous significant rela-tionships between communication apprehensionand personality variables, found no significant rela-tionship between communication apprehension andintelligence among college students. Bashore(1971) found a slight negative relationship in astudy of high school seniors, but the correlation wassignificant only for female students and his sampleof 75 senior students in a specialized universitylaboratory school precludes generalization withoutextreme reservation. Notably, Bashore (1971)failed to find a significant correlation betweencommunication apprehension and GPA in that samestudy.

The only previous research that presents evidencebearing directly on our hypothesis is that of Bashore(!971), parts of which are noted above. Bashore

found communication apprehension to be signifi-cantly negatively correlated with scores on theAmerican College Test (ACT), the Illinois HighSchool Test (IHST), and the verbal portions of theCollege Entran.ceExamination Board (CEEB) testand the Preliminary Scholastic Aptitude Test(PSAT). These findings should be considered tenta-tive, of course, because of the problems with thesubject sample noted above. However, they do pro-vide a modicum of support for the rationale underly-ing our primary hypothesis.

To this point our rationale has focused on thetraditional educational system, the single teacherwith a moderate number of students in a potentiallyinteractive setting. Other teaching-learning systemshave existed for many years, and the number ofalternative methods appears to be increasing. Thesealternative systems range from the commonplacecollege lecture system. with upwards of 100 stu-dents receiving instruction simultaneously. tosemitutorial, personalized systems of instruction.While our hypothesis sho'lld be expected to holdtrue for general academic achievement in college orhigh school, because most instruction in mosthigh schools and colleges still employs the tradi-tional system: whether it would hold for other sys-tems would depend on the nature and extent ofcommunication demands of the particular alterna-tive system under consideration.

In order to examine the potential generalizabilityof our hypothesis beyond the traditional system, weconsidered achievement in one additional type ofinstructional system, the mass-lecture system. Be-cause student-to-teacher communication is highlyrestricted in this type of system, nothing in ourprevious rationale would lead us to expect thatcommunication apprehension should impactachievement.

Therefore, based on the rationale that communi-cation apprehension will reduce student-initiatedcommunication, and hence student achievement, in

. instructional systems that rely on student-teacherinteraction but not in those that restrict such interac-tion. the following hypotheses were tested:

HI: High communication apprehensives will

76 McCroskey and Andersen

have lower achievement in a traditional edu-cational s-ystemthan will low communica-tion apprehensives.

H2: High and low communication apprehensiveswill not differ in achievement in a

communication-restricted system involvinginstruction primarily employing the mass-lecture method.

It is important that our second hypothesis isframed as a null hypothesis and thus is not amenableto statistical confirmation. If it can not be rejected,as suggested by our rationale, the crucial concern inthe interpretation of the obtained results is the powerof the design and analysis employed to detect differ-ences, if indeed any existed. We will address thisissue in the results section below.

If the previous hypothesized relationships arefound to be correct, it would follow that studentsmay be intuitively aware of the impact differentinstructional systems have on their learning, at leaston an affective level. While it is a virtual truism inhigher education that students dislike mass-lecturecourses, projection of student attitudes based on ourknowledge of previous research on communicationapprehension would suggest this attitudinal orienta-tion would not be held by highly apprehensive stu-dents. They should find mass-lecture courses lessthreatening than, and thus preferable to, small clas-ses which permit considerable interaction. For lowand moderate apprehensives, those who seek com-munication and the rewards it can bring, the mass-lecture is probably too restrictive. Thus. these stu-dents should find small classes preferable to mass-lecture courses. Based upon this rationale, we posedthe following interaction hypothesis:

H3: High communication apprehensives will ex-press more positive attitudes toward mass-lecture classes than toward small classes thatpermit interaction, but low and moderatecommunication apprehensives will expressmore positive attitudes toward small classesthat permit interaction than toward mass-lecture classes.

METHOD

In order to test our hypotheses, studies based onfour samples of college students were conducted. Ineach study Ss completed the Personal Report ofCommunication Apprehension, College Form(PRCA, McCroskey, 1970). Scores on the PRCAserved as the operational definition of communica-tion apprehension in all four studies. Althoughcommunication apprehension has been found to be amultidimensional construct (Wheeless, 1975; Daly& Miller, 1975), the PRCA was selected for thesestudies because its focus is on oral communication

apprehension, the type of communication whichunderlies the rationale which led to our hypotheses.

The PRCA has been employed extensively inprevious research involving communication ap-prehension and has consistently produced internalreliability estimates above. 90 and test-retest relia-bility above .80 (McCroskey, 1970). On the basis ofa survey of over 20 studies which employed thePRCA, McCroskey (1975) evaluated the reliabilityand the predictive validity of the instrument. Heconcluded that researchers "can employ the PRCAas an index of oral communication apprehensionwith confidence in both its reliability and its valid-ity. " In the present studies the internal reliabilityestimates for the PRCA ranged from. 93 to .95.

The achievement data required for the first threestudies were retrieved from university records.Coded student numbers were employed in this pro-cess to insure the maintenance of confidentiality ofstudent records. The particular achievement dataobtained for each study are noted below.

Study 1

The first study was designed to test our firsthypothesis. Achievement in a traditional educa-tional system was operationalized as scores on theAmerican College Test (ACT) taken by the Ss dur-ing their final year of high school. The ACT is oneof the most widely administered measures 'ofacademic achievement and is employed to screenstudents for admission by many colleges and uni-versities. At the institution whe~e this study was

APPREHENSION AND ACADEMIC ACHIEVEMENT 77

conducted, students are requested to submit ACTscores with their admission applications, but stu-dents are not excluded from admission on the basisofthe scores. The scores available included not onlytheoverall or composite score but also subscores forEnglish, social science, natural science, andmathematics.

5s in this study were 825 students enrolled in abasic course in communication during the 1974-75academic year. Approximately half of the studentswere male, half female. The breakdown byacademic rank was approximately 30 percentfreshmen, 45 percent sophomores, 15 percentjuniors, and 10 percent seniors.

Stud 2

The second study was also designed to test ourfirst hypothesis. Achievement in a traditional edu-cational system was operationalized as the Ss' gradepoint averages through and including the springsemester of the 1974-75 academic year. Ss were1454 students enrolled in two basic courses in

communication during the 1974-75academic year.The sex and academic rank breakdowns were the

same as for Study 1. The GPAs available repre-sented a minimum of one full academic year and upto a maximum of four years. Because of the codingsystems to maintain confidentiality, it was not pos-sible to identify Ss' sex or rank in the final data setsfor any of these studies. Thus, these classificationscould not be employed to reduce error variance inthe data analyses.

Study 3

The third study was designed to test our secondhypothesis. Achievement in a communication re-stricted system was operationalized as the summedscore from two 15-item, one 20-item, and one 50-

item multiple-choice (5-foil) examinations adminis-tered as regular exams in a mass lecture course incommunication. The 709 5s who took all four testswere enrolled in the course during the fall semesterof the 1975-76 academic year. The sex and

academic rank of the Ss were approximately thesame as those of the Ss in the first two studies.

Our second hypothesis focuses on achievement ina communication restricted system involving in-struction primarily employing the mass-lecturemethod. The course in which the study was con-ducted was ideally suited as the operationalizationof that instructional system. Students in this classreceive all of the content through mass lectures(there is no textbook) and there are no recitation orlaboratory sessions associated with the class. Con-tact with' the instructor is discouraged, questions

during lectures are not permitted, and less than onepercent of the students ever talk to the instructorabout course content outside the classroom.

The test items had been tested previously andfound to discriminate. The estimated reliability ofthe combined score for the 100 items on the fourtests was. 79.

Study 4

The fourth study was designed to test our third

hypothesis. Attitudes toward mass-lecture andsmall classes were operationalized as the responsesof S5 to two 5-point, Likert-type statements. Thescales, which permitted a response option fromstrongly agree to strongly disagree, were as fol-lows: "I prefer lecture courses to any other kind"and "r like small classes where we can have a gooddeal of discussion." These scales were includedamong 75 items in an attitude questionnaire to avoidcalling attention to the research purpose.

The 775 Ss in the study were concurrently enrol-led in two basic communication classes, one a two-credit class in interpersonal communicati'on withenrollment per section limited to 25, the other aone-credit mass-lecture class in communicationtheory with enrollments per section ranging from350 to 400. These Ss were selected because of thisdual enrollment in the belief that this selection

would mitigate against possible bias in response dueto the type of class the S was currently taking.

The data were collected during the fall semesterof the 1975-76 academic year. The sex and

78 McCroskey and Andersen

academic rank of the Ss were approximately thesame as in the previous studies.

Statistical Analyses

Ss in each study were classified as high, moder-ate, or low communication apprehensives on thebasis of their PRCA scores. Two analyses wereperformed on the data from each of the first threestudies. For the first analysis in each study, Ssscoring more than one standard deviation above themean were classified as high apprehensives; Ssscoring more than one standard deviation below themean were classified as low apprehensives; the re-maining Ss were classified as moderate apprehen-sives. The second analysis in each study was con-ducted to examine the impact of extremes in ap-prehension (the points where the greatest impactwas expected). In these analyses Ss scoring morethan one and a half standard deviations above or

below the mean were classified as extremely highapprehensives and extremely low apprehensives,respectively. High and low apprehensives in theseanalyses were the same as in the first analyses,except that the extremes were removed. A secon-dary analysis was not performed for the fourthstudy. The limited sample size would have madesuch an analysis questionable because of the veryfew Ss who would have constituted the extremeclassifications.

Each of the dependent variables in the first threestudies was submitted to single classificationanalysis of variance employing the two classifica-tion systems noted above. Although our hypothesesdid not include references to moderate apprehen-sives, they were included in these analyses for pur-poses of comparison. Supplementary analyses in-cluding only high and low apprehensives and

. analyses including only extremely high and ex-tremely low apprehensives were also conducted inorder to determine effect sizes related to our specifichypotheses. The 0.05 level was set for significanceof all of these tests.

The data from the fourth study were subjected to atwo-factor analysis of variance with one repeatedmeasure. The three apprehension levels served as

one factor and the type of class to which resppnded(the repeated measure) served as the other. Since asignificant interaction was observed, correlatedt-tests were computed to test the effects predicted byour third hypothesis. The 0.05 level was employedfor these tests.

Since single scales were employed as the depen-dent variable in this study, the question of theirreliability should be considered. No direct estimateof their reliability, of course, was possible. Becausethe two scales were used as a measure of the same

thing (attitude toward instructional systems), how-ever, it was possible to examine them together as ifthey were a combined instrument. The obtainedcorrelation between the scales was - .49(thenega-tive association, of course, was expected). Correct-ing this correlation for length would result in aninternal reliability estimate of the two scales to-gether of .66.

RESULTS

Study I

The analyses of the data from the first studyindicated support for our first hypothesis: as shownin Table 1, the effects attributable to communica-tion apprehension were significant for the ACTcomposite score and all of the ACT subscores. Thelargest effects were found on the composite scoreand the social science subscore, the smallest on themathematics subscore. Given the common methods

by which social science and mathematics are taught,these findings lend credence to our rationale for thefirst hypothesis. Social science classes in highschools are typically much more interaction-oriented than mathematics classes. It would follow,therefore, that level of communication apprehen-sion would have greater impact in the former thanthe latter.

Study 2

The results obtained from the second study alsoprovide support for our first hypothesis (Table I).The GPAs of high communication apprehensiveswere found to be significantly lower than those of

APPREHENSION AND ACADEMIC ACHIEVEMENT 79

low communication apprehensives. The differencebetween the extreme highs and extreme lows wasparticularly striking, almost half a grade point.

Study 3

Results of the third study indicate that our secondhypothesis (the null hypothesis) can not be rejected.Obtained F-ratios, even when extreme apprehen-sion levels were considered, did not even approachsignificance (see Table I).

Because of the statistical impossibility of con-firming a null hypothesis, we conducted a poweranalysis to determine the likelihood of our failing toreject this hypothesis because of low power. Thisanalysis indicated that, with alpha set at 0.05, thepower of this test to find a significant effect if it werepresent, even if it were very small (f = .10), was.73

and if it were only as large as f = .15, thepowerwas.99 (Cohen, 1969). Such an effect size in this study,even if statistically significant, would account forless than one percent of the variance of the depen-dent variable. Consequently, low power must berejected as an explanation for the lack of observedstatistical significance.

Study 4

. The analysis of variance of the data from the -fourth study indicated a significant interaction be-tween level of communication apprehension andinstructional system on the subjects' attitudes.Table 2 reports the observed means and the resultsof the tests related to our third hypothesis. As indi-cated in the table, the results provide support for thehypothesis. Not only were the differences in the

TABLE 1

Mean Scores, F-Ratios, and Effect Sizeson Achievement Variables in Studies 1-3

Dependent Variables

ACTCommunication

Apprehension Natural Social CourseClassification Composite English Science Mathematics Science GPA Test

Iean Scores--Extreme High 21.12 19.50 22.92 21.69 19.81 2.46 84.39

High (minus

extremes) 21.49 20.03 22.70 21.93 20.84 2.53 78.99

High 21.28 19.88 22.56 21.73 20.46 2.52 79.48

!odera te 21.73 19.88 23.00 22.56 20.95 2.56 79.59

Low 23.03 20.94 24.36 23.38 22.82 2.69 79.25

Low (minusextremes) 22.74 20.81 23.97 23.19 22.39 2.66 78.92

Extreme Low 24.80 22.15 26.55 24.85 24.85 2.91 82.25

F-Rados and Alpha

Excluding Extremes

F-Ratio 6.88 3.91 4.61 3.32 7.09 4.73 0.03

Alpha '".002 < .02 '" . 01 <.05 <.002 <'.01 NSD

Including Extremes

F-Ratio 4.10 2.65 2.90 1.79 3.84 3.54 0.36

Alpha <.005 <:.05 <.03 NSD <:.005 <..01 NSD

Effect Size (Percent of Variance)

Regular High-Low .04 .02 .02 .02 .04 .02 .00

Extreme High-Low .21 .14 .15 .08 .19 .10 .00

80 McCroskey and Andersen

TABLE 2Mean Attitudes of Students

Toward Instructional Systems*

*Higher scores indicate more positive attitude, possible range of 1.0 to 5~0

with 3.0 the presume neutral position.

direction predicted, and statistically significant, butthey also indicated that on an absolute basis highapprehensives were favorable toward mass lecturecourses and unfavorable toward small classes. This

interpretation is based on the presumed neutral pointbeing 3.0, the number assigned to undecided re-sponses. Moderate and low apprehensives recordedmeans in just the opposite direction. The dataanalysis indicated that 22 percent of the variance inthe dependent variable was attributable to ap-prehension level.

DISCUSSION

The theory upon which our hypotheses werebased was that communication between student andteacher is a valuable component of many instruc-tional systems, but that some students are muchmore likely to seek this communication while othersare more likely to avoid it. Such student behavior isnot unique to the learning environment, but rather ischaracteristic behavior of the individual because of

the individual's level of communication apprehen-sion. The results of these studies support both ourhypotheses and our rationale leading to those hy-potheses. When the instructional system studiedpermitted student-initiated interaction with theteacher, significant differences in achievementwere observed between high and low apprehen-sives, but in a communication-restricted system, nosuch differences were observed.

The lack of statistical significance in our thirdstudy. although expected. requires special consid-eration. While we cannot specify the precise proba-

bility that the null hypothesis is confirmed, in ~hisinstance we have good reason to believe that oursecond (null) hypothesis is substantially correct.Since we did not expect to find a reliable differencein this study, we carefully designed the study toachieve as much power as was feasible so that anyreal effect present would not escape our attention.The power analysis noted above indicates that hadan effect been present that would have accounted foras little as one percent of the variance, it would havebeen statistically significant. Such, of course, wasnot the case. Consequently, while we cannotexclude the possibility that there may be some vel}'small relationship between communication ap-prehension and achievement in a communication-restricted instructional system, there is good reasonto believe that no meaningful relationship exists. Inthe absence of additional research indicating other-wise, decisions concerning choices of instructionalsystems should be made on the basis that no rela-tionship exists.

The results of these studies raise an importantquestion concerning traditional methods of class-room instruction as well as some of the newer in-structional systems involving voluntary tutorial in-structi0n. Communication from student to teacher ishighly valued in both approaches, but these studiessuggest that r~liance on instructional systems em-phasizing voluntary, student-initiated interactionwith teachers penalizes a large number of studentssuffering from communication apprehension. Theyare placed at a competitive disadvantage becausethey are too apprehensive to engage in the behaviorsrequired to achieve success.

ApprehensionInstructional System

Level N Mass Lecture Small Class Difference t p

High 44 3.16 2.80 .36 6.51 '.0001

loderate 186 2.52 3.74 1.22 22.06 "'.0001

Low 45 2.09 4.40 2.31 41.77 <.0001

APPREHENSION AND ACADEMIC ACHIEVEMENT

On the other side of the coin, the instructional

system which consistently receives the most criti-cism from educators, the mass-lecture system, wasnot found to impose' a penalty on the highly ap-prehensive student. This suggests, of course, thatunder ideal circumstances the instructional systememployed should be tailored to the needs of theindividual student-a communication-restricted

system for high apprehensives and acommunication-emphasized system for other stu-dents.

Such tailoring of instructional systems is likely tobe received favorably by students. Their expressedattitudes. as indicated in our fourth study, indicaretheir preferences are in line with the decisions oninstructional system recommended by the results ofour studies on achievement.

While there is a need for further research on the

impact of communication apprehension on learningin the various instructional systems (e.g., casestudy. PSI, fully mediated, independent study,CAI), it appears clear from this series of studies thatno one instructional system will be superior for allstudents. One of the best predictors of the value of agiven instructional system appears to be the interac-tion of communication apprehension level of thestudent and communication requirements of the sys-tem, the consideration of which has not previouslyentered into the construction or evaluation of new

instructional systems.

REFERENCES

BASHORE. D.N. Relationships among speech anxiety,IQ, and high school achievement. Unpublished mas-ters thesis. Illinois State University, 1971.

BEEZ. W. V. Influence of biased psychological reportson teacherbehaviorand pupilperformance.Proceed- .

ings of the 76th Annual Com'ention of the AmericanPsychological Association, 1968, 3, 605-606.

BLOOM. B.S. Stability and change in human charac-teristics. New York: John Wiley and Sons, 1964.

BINDER. D.M.. JONES. J.G.. & STROWIG, R.W.Non-intellective or self-report variables as predictorsof scholastic achievement. Journal of EducationalResearch. 1970. 63. 364-366.

81

CATTELL. R.B.. SEALY. A.P.. & SWEENEY, A.B.What can personality and motivation source trait mea-surements add to the prediction of school achieve-ment'? British Journal of Educational Psychology.1966,36. 280-295. .

COHEN, J. Statistical power analysis for the behw'ioralsciences. New York: Academic Press. 1969.

DALY. J.S.. & MILLER. M.D. The empirical develop-ment of an instrument to measure writing apprehen-sion. Research in the Teaching of English. 1975. 9.242-249.

DUSEK. J.B. Do teachers bias children's learning? Re-\'iew of Educational Research, 1975.45.661-68"+.

HOYT, D.P.. & MUNDAY. L.A. Your collegefreshmen: 11Iterpreti\'eguide to ACT research ser\'icesfor higher education. Iowa City. Iowa: ACH Researchand Development Division. 1968.

MANDRYK, T.R.. & SCHUERGER. J.M. Cross-validation of the HSPQ as a predictor for high schoolgrades. Educational and Psychological Measure-ment. 197"+,34, 449-454.

McCROSKEY, J.c. Measures of communication-boundanxiety. Speech Monographs. 1970. 37, 269-277.

McCROSKEY, J.C. Validity of the PRCA as an index oforal communication apprehension. Paper presented atthe annual convention of the Speech CommunicationAssociation. Houston. December. 1975.

McCROSKEY. J.c.. & DALY. J.A. Teachers' expecta-tions of the communication apprehensive child in theelementary school. Human Communication Re-search. 1976. 3.67-72.

McCROSKEY,J.C.. DALY.J.S.. &SORENSEN, G.A.Personality correlates of communication apprehen-sion. Human Communication Research, 1976, 2.376 - 380.

SCHRADER. W.B.. & STEWART. E.E. Descriptivestatistics on college board candidates and other refer-ence groups. In W.H. Angoff (Ed.). The collegeboard admissions testing program: A technical reporton resea~chand development activities relating to thescholastic aptiwde test and achieveme1lt tests. NewYork: College Entrance Examination Board. 1971.79-115.

THORNDIKE. R.L.. & HAGAN. E. Measurement and

evaluation in psychology and education. New York:John Wiley and Sons. 1969.

WHEELESS. L.R. An investigation of receiver appre-hension and social content dimensions of communi-cation apprehension. Speech Teacher, 1975. 2"+.261-268.