|
|
 |
 |
 |
 |
 |
 |
 |
 |
 |
 |
 |
JAC
Volume 5 |
 |
|
Editor: Download PDF |
Experimental Study of Written Communication Apprehension and Language Choice in a Business Setting Don W. Stacks and Robert W. Boozer One of the major fears that people have is the fear of communication. Although there has been an interest in the area of oral communication apprehension for some time, only recently have educators begun to express interest in the fear or apprehension of written communication. Research in this area, however, has taken a unidimensional approach to the identification of who is apprehensive about writing, an approach which may not adequately operationalize writing apprehension. The purpose of this study is to provide an initial test of a multidimensional model of written communication apprehension as the construct relates to syntactic language use. Writing Apprehension and Language Choice The instrument used most often to ascertain written communication apprehension has been that developed by Daly and Miller.1 This instrument consists of 20 statements reacted to on a Likert-like scale. It has a high reported reliability estimate (usually .80 and above) and has been used rather extensively. However, Boozer, Lally, and Stacks and others have recently questioned the construct integrity of the measure.2 Specifically, they note that writing apprehension may be the result of a number of factors, to include both attitudes toward writing and the more behavioral manifestations. Boozer, Lally, and Stacks present a multidimensional instrument which contains many of Daly and Miller's original statements and others. They report a six-factor solution and argue that it better represents writing apprehension than the unidimensional instrument. What constitutes writing apprehension and what aspects the writer contributes to writing apprehension have been investigated. The results, however, have not led to any major pedagogical change in the teaching of writing. In most cases, the research has been generalized; that is, the subjects were enrolled in freshman composition courses and not representative of a single group. It would be of interest to the business communication educator, for example, to establish how the business communication student perceives and behaves according to his/her apprehension toward writing. What we do know of writing apprehension in general comes mainly from the work of Daly. Daly and his associates have found that writing apprehension is predictive of occupational and academic choice, to include perceptions of occupation and course of study desirability, enrollment in writing courses, writing style in terms of intensity of language used, and a number of satisfaction indices.3 Hence, writing apprehension has a demonstrated effect on general perceptions of importance to the writer. What we know about the structure of writing suggests that high apprehensive writers write differently than low apprehensive writers. In this regard, low apprehensives score higher on indices of grammar, mechanics, and general skill.4 Low apprehensives use more words in their writing, more qualification, and choose to write with higher language intensity than do high apprehensives.5 Faigley, Daly and Witte found that indices measuring syntactic development were also affected by writing apprehension.6 Specifically, they found significant differences in syntactic "fluency" or "maturity" as measured by the length of T-units and clause length between high and low apprehensive writers.7 Significance was almost obtained for two other indices of syntactic fluency: the ratios of T-units with final nonrestrictive modifiers of total T-units and number of words in final nonrestrictive modifiers to total number of words (p < .07). However, these findings were highly dependent upon the type of message written. For an argumentative essay none of the syntactic indices were significant; for a personal narrative/descriptive essay, however, significant results were obtained for number of words per T-unit, number of words in final nonrestrictive modifiers, and T-units with final nonrestrictive modifiers. This body of research suggests that syntactic language choice may be a significant predictor of writing apprehension. If it is and it can be demonstrated that low and high writing apprehensives differ in their syntactic choice, then pedagogical tools could be derived to reduce writing apprehension. In the business world, for example, many written communications tend toward the narrative, simple response to some stimulus that do not bring into account the argumentative form of discourse or, where argumentative, policy and procedures yield a more descriptive writing style. Given this, it follows that syntactic choice might predict writing apprehension. This study is a further exploration of writing competency (as measured by actual syntactic choice) in a business situation. Further, it tests the multidimensional measure of writing apprehension developed by Boozer, Lally, and Stacks.8 Instrument Development The writing apprehension instrument developed by Boozer, Lally, and Stacks consists of 64 items, including 26 items from Daly and Miller Writing Apprehension Test (WAT) and 38 other statements, covering a number of attitudes toward business communication. The instrument asks each subject to respond to the items on a strongly agree to strongly disagree, 5- point, Likert-like scale. All items were randomly listed and presented to 428 undergraduate students enrolled in a number of courses and disciplines at the University of South Alabama. The sample crossed both college and discipline and was representative of students just enrolling to those who have been in college for a number of years. Thus, a valid cross-sampling of student was obtained. Data were then submitted to a principle components factor analysis
with both varimax and oblique rotations. An eigen- value of 1.0 was
the criterion used as the cut-off for rotation and only those factors
that loaded above .60 and which had no secondary loadings greater than
.40 were considered for inclusion in the instrument. Comparison of the
oblique and varimax rotations yielded identical results: six factors
meeting the loading criteria with only modest interfactor correlations.
Consequently, the six factors were treated as orthogonal.9 Subjects Subjects were 59 undergraduate students enrolled in business communication courses at the University of South Alabama. Subjects completed the instrument, other forms, and tasks as a regularly scheduled part of the courses they were enrolled in. Procedures Each subject completed the writing apprehension instrument during the first week of the quarter. After approximately one week's time, subjects were given a written assignment to complete. The assignment asked that each write a business letter responding to a problem. Each subject was provided with all the necessary background data for writing the letter. After the letters were collected they were coded by two upper-division English majors and were submitted to computer analysis. The computer program used, Syntactic Language Choice Analysis (SLCA III), is the third generation of a system that provides profiles of language reflecting basic psychological and cognitive states.10 The program analyzes the grammatical characteristics of amessage in terms of eight qualities of language: social perception, sensation, existence, motion, disposition, time, symmetry, and conditionality. Operationalization of each quality is based on three categories of language behavior: information units (nouns), qualitative-quantitative units (adverbs and adjectives), and relations (verbs). Relative densities are then calculated for each category as ratios of the number of particular units divided by the total number of units.11 SLCA III operationalizes each of the eight language qualities based on several sub-measures within each quality. Social perception is composed of five indices which measure inanimate perception, audience perception, self perception, and two forms of other perception (generalized and authority). Sensation is composed of four indices which measure sensed and unsensed information and sensed and unsensed qualifiers. Existence consists of six indices: positive and negative information, positive and negative qualification, and positive and negative relation. Motion is composed of two indices (non-motion and motion); disposition, two indices (disposition and assertion language); and symmetry, two indices (symmetric and asymmetric relations). Time is measured as being past, present, or future- oriented. Finally, conditionality consists of four indices: qualified and unqualified information and qualified and unqualified relations. Additionally, a total cognitive activity score is also generated (perceptual cognitive activity) which measures the total frequency of occurrence of information units, relational units, and modifiers in the message corpus. Variables The dependent variables in this research were operationalized as the six factors obtained from the Boozer, Lally, and Stacks writing apprehension instrument. Additionally, because of the moderate inter-factor correlations, the total test score was also computed and used as a dependent variable. Because the inter-factor correlations were moderate, and, because the varimax and oblique rotations yielded the same basic structure, the factors were treated as independent. Scores for each subject were treated as a state variable. In this way the entire range of scores were used in regression analysis.12 Most research has taken a more trait approach, which uses as its basic distinction the mean apprehension score and then those scoring greater than one standard deviation above or below the mean are scored as high and low writing apprehensive, respectively. The independent variables in this research were the 36 syntactic choice variables as identified by the SLCA program. Each was entered as a density and ranged in value from 0.00 to a possible 1.00. Data Analysis The data were analyzed via a multiple regression procedure, the Maximum R2 Improvement technique, developed by Goodnight.13 This technique is superior to the normal stepwise procedure in that all variables in the equation are reexamined at each step. In this way a variable's contribution to the statistical model is constantly being checked as new variables are entered. To select the best multiple regression model, the overall F due to regression had to be significant (p < .05) and a minimum increase of 1% in R2 had to occur by the addition of an additional item. All analyses of the relationship between dependent and independent variable were accomplished by analyzing the beta weights of each predictor in the model.14 To summarize, the multiple regression models regressed the 36 syntactic
language choice measures on the six dimensions of writing apprehension
and the total summed measure. Because of the exploratory nature of this
research, no specific entry order was included; the independent variable
with the highest F-ratio was entered first. Also, because of the exploratory
nature of the study, trends (p <.10) were also analyzed.
Reliability of Measures Reliabilities for all dependent variables were computed via Chronbach's Coefficient-Alpha.15 Reliability coefficients for the six factors (using the total sample of 400+ respondents) were: .86 for Factor 1, .87 for Factor 2, .79 for Factor 3, .62 for Factor 4, .66 for Factor 5, and .52 for Factor 6. The low reliabilities for Factors 4, 5, and 6 may be in part a function of the number of items on each factor; two items were used to compute Factors 4 and 6, while three items were used to compute Factor 5. Since reliability is in part a function of the number of items in the measure and their contributions to explained variance, those factors with two or three items would be expected to produce lower reliability coefficients.16 The reliability coefficient for the total (summed) measure was .83.
Predictive Models Table1 summarizes the data and presents the means and standard deviations for each of the factors and the total summed scale.
Means and Standard Deviations
Total Summed 59 53.39 10.51 24 73
Factor 1 59 18.36 2.68 12 23
Factor 2 59 15.92 2.28 11 21
Factor 3 59 10.81 2.71 5 18
Factor 4 59 5.10 1.57 2 8
Factor 5 59 11.34 2.18 5 15
Factor 6 59 3.11 1.18 2 6
An examination of Table 2 indicates the relationships of each factor to the total test and to each other. The negative relationships to the total summed scale are due to the scoring of the Likert-like statements on a (1) strongly agree to (5) strongly disagree basis. Factors 1 and 5 must be interpreted as the lower the score, the more apprehension or the more negative the attitude. Table2 Correlation Matrix: Total Summed and Factors
Summed Fact 1 Fact 2 Fact 3 Fact 4 Fact 5 Fact 6 Total
Summed 1.000 .750 -.584 .667 .028** .719 -.367
Factor 1 1.000 -.472 -.439 .021** .652 -.227*
Factor 2 1.000 .342 .142** -.311 .422
Factor 3 1.000 .060** -.475 .148**
Factor 4 1.000 -.045** .230*
Factor 5 1.000 -.258
Factor 6 1.000
Correlations with no star are significant at or beyond the .05 level. *p<.10 **p>.25
Based on the items loading on each factor from the factor analysis, tentative names are provided to help in the analysis. Factor 1 seemed to represent a more behavioral manifestation of apprehension, "Blank Page Paralysis." Factor 2 seemed to refer to a more general attitude toward writing, "General Affect toward Writing." Factor 3 looked more at the writing from a business perspective and was titled, "Positive/Negative Business Affect." Factor 4, "Alternative Modes," dealt with alternative modes of communicating (face-to-face, via telephone). Factor 5 seemed to stress writing competency or "Attitudes toward Writing Competence." Finally, Factor 6 represented an instrumental attitude, "Attitude toward Professional Writing Skills." The total instrument, the sum of all items found in each subscale is called simply, "Business Writing Apprehension." The multiple regression procedure previously discussed was used to obtain a predictive model of apprehension or attitude toward writing on each factor and the total summed scale. For Blank Page Paralysis, the best model had 12 variables that accounted for 35% of the variance (F = 2.09; df = 12/46; p < .05; intercept [INT] = 12.53). The following summarizes obtained significant and near significant (p < .10) variables and their relationship to Blank Page Paralysis. (Note: For this factor, the lower the score, the greater the Blank Page Paralysis.) The variables are listed in the order of their entry into the multiple regression model. Variable Relationship 1. Motion Density Linear (B =.529, p <.05). As the relative frequency of verbs and
verb phrases not of the form "to be" increased, blank page
paralysis decreased. 2. Negative Generalized Other Perception Linear (B = .292; p < .05). As the frequency of nouns or pronouns
which refer to negative unspecific other persons increased, blank page
paralysis decreased . 3. Pos. Audience Perception Linear (B = -.299; p < .05). As the frequency of proper nouns which
refer to specific other positive persons or groups of persons increased,
blank page paralysis increased. 4. Positive Existential Linear (B = .394; p < .05). As the frequency of subjects and objects
of verbs that have no negation increased, blank page paralysis decreased. 5. Sensed Information Linear (B = - .309; p < .05). As the frequency of subjects and objects
that refer to things, persons, places, that can be sensed increased,
blank page paralysis increased. 6. Positive Self Perception Linear (B = -.243; p < .10). As the frequency of subjects and objects of verbs that first-person positive pronouns increased, blank page paralysis increased. For General Affect Toward Writing the best model had six variables that accounted for 25% of the variance (F = 2.90, df = 6/52; p < .05; INT = 23.53). The following summarizes the variables meeting the significance criteria and their relationship to attitudes toward writing in general. (For this factor, the higher the score, the more negative the attitude is toward writing.) Variable Relationship 1. Negative Qualification Linear (B = .270; p < .05). As the frequency of qualifiers associated
with information units and by the use of "no" or "not”
increased, the less positive the attitude toward writing. 2. Motion Density Linear (B = -.434; p < .05). As the frequency of verbs and verb
phrases increased, the more positive the attitude toward writing. 3. Unsensed Qualifier Density Linear (B = -.324; p < .05). As the frequency of modifiers that cannot be sensed increased, the more positive the attitude toward writing.
For Positive/Negative Business Affect the best model had nine variables that account for 50% of the variance (F = 2.90, df = 6/52; p < .05; INT = 23.53). The following summarizes the variables meeting the significance criteria and their relationships to writing in a business context. (For this factor, the higher the score, the more negative the attitude toward business writing.) Variable Relationship
1. Negative Qualification Linear (B = 4.39; p < .05). As the frequency of qualifiers associated with "no" or "not" increased, the more positive the attitude toward business writing.
2. Positive General-Other Perception Linear (B = -.266; p < .05). As the frequency of nouns and pronouns which refer to positive unspecific other persons or groups of persons increased, the more positive the attitude toward business writing.
3. Positive Relational Density Linear (B = -.411; p < .05). As the frequency of verbs which do not have negative indications in the verb phrase increased, the more positive the attitude toward business writing.
4. Positive Self Perception Linear (B = .276; p < .05). As the frequency of subjects and objects of verbs that are first-person positive personal pronouns increased, the more negative the attitude toward business writing.
5. Non-Motion Language Linear (B = .294; p < .05). As the frequency of verbs or verb phrases of the form "to be" increased, the more negative the attitude toward business writing. No significant multiple regression was obtained for Factor 4, Alternative Modes (F = 2.42, df=1/57, p >.05, R2 =.04). For Attitude toward Writing Competence the best model had eleven variables
that accounted for 35% of the variance (F = 2.32; df = 11/47; p <
.05; INT = 4.22). The following summarizes the variables meeting the
significance criteria and their relationships to attitudes toward writing
competence. (For this factor, the higher the score, the more competent
is the perception of writing ability.) Linear (B = .221; p < .05). As the total amount of informative units,
qualitative-quantitative units, and relations increased, so too did
perceptions of writing competence. Linear (B = -.262; p < .05). As the frequency of proper nouns which
refer to specific other positive persons or groups of persons increased,
perception of writing competence decreased. Linear (B = -.307; p < .05). As the frequency of simple past tense verbs or verb phrases increased, perception of writing competence decreased. For Attitude toward Professional Writing Skills the best model had seven variables that accounted for 30% of the variance (F = 3.16; df = 7/5; p < .05; INT = 4.26). The following summarizes the variables meeting the significance criteria and their relationships toward professional writing skills. (For this factor, the higher the score, the more negative the attitude toward professional writing skills.) Variable Relationship 1. Neg. Audience Perception Linear (B = -.332; p <.05). As the frequency of subjects and objects
of verbs which are second-person in nature and perceived negatively
increased, so too did attitudes toward professional writing skills. 2. Negative Relational Density Linear (B = .339; p < .05). As the frequency of verbs having "not"
or other negating prefixes increased, attitude toward professional writing
skills decreased 3. Positive Self Perception Linear (B = -.229; p < .10). As the frequency of subjects and objects
of verbs that are first-person positive personal pronouns increased,
attitude toward professional writing skills increased. 4. Sensed Qualifiers Linear (B = -.256; p < .05). As the frequency of modifiers referring to qualities or quantities which can be sensed increased, attitude toward professional writing skills increased. For general business writing apprehension the best model had nine variables that accounted for 37% of the variance (F = 3.16; df = 9/49; p < .05; INT = 53.52). The following summarizes the variables meeting the significance criteria and their relationships toward business writing apprehension. (For this analysis, the higher the score, the more apprehension.) Variable Relationship 1. Negative Qualification Linear (B = -.338; p < .05). As the frequency of qualifiers associated
with information units and relations by the use of "no" or
"not" increased, reported writing apprehension decreased. 2. Motion Density Linear (B = -.453; p < .05). As the relative frequency of verbs
phrases not of the form "to be" increased, reported writing
apprehension decreased. 3. Past Time Density Linear (B = .405; p < .05). As the frequency of simple past tense
verbs and verb phrases increased, reported writing apprehension increased. 4. Conditional Density Linear (B = -.291; p < .05). As the frequency of verbs that are
of the subjective mood or are in the sentence form of a question increased,
reported writing apprehension decreased. 5. Neg. Authority Perception Linear (B = -.222; p < .10). As the frequency of proper nouns which refer to specific other negative persons or groups of persons increased, reported writing apprehension decreased.
Discussion
The results of this investigation offer strong support for both a multidimensional model of writing apprehension and the impact that syntactic choice has on an individual's perceptions of writing apprehension. People who use passive voice, rather than active voice, for example, are more apprehensive about writing across a number of factors. But most important, as an initial test of the multidimensional model, is the finding that the instrument identifies those who are writing apprehensive and provides diagnostic areas for further examination. Obviously, some refinement is still needed, but the amount of variance accounted for by the syntactive choices ranged from .30 to .50, fairly large portions of the explained variance. An examination of the instrument suggests that not all of the factors were perceived similarly. The factor dealing with alternative approaches to writing failed to yield a significant multiple regression model. This should have been expected, given the nature of the writing assignment which was a business letter and provided no other outlet for communication. Also, the type of the letter, the "no" letter, may have provided some interesting results in of itself. As noted by Faigley, Daly, and Witte, the type of writing assignment influences both apprehension and syntactic choice.17 In this case, however, it seems that writers who have less apprehension toward business writing (as measured by Factor 3), who have a positive attitude toward their professional writing skills (as measured by Factor 6), and who score low on the total writing apprehension instrument use more negative qualification. This finding is interesting since the assignment should have produced less negation and are more positive qualification. (Or, it may be that less apprehensive writers worry less about following directions because they feel their writing skills will be able to carry them through any writing assignment.) An examination of the predictor variables and their clustering by major qualities of language indicates that different factors produce different groupings. Of the eight language qualities (social perception, sensation, existence, motion, disposition, time, symmetry, and conditionality), seven yielded one or more significant predictor variables. As might be expected for the total summed instrument, the five variables predicting general writing apprehension came from five different qualities (existence, social perception, motion, time, and conditionality). For the subscales, however, the predictors seemed to group themselves. For Blank Page Paralysis (Factor 1), three of the six predictors grouped on measures of social perception. For Positive/Negative Business Affect (Factor 3) two of the five predictors grouped on measures of existence and two on social perception. Social Perception again was the grouping area for Attitudes toward Professional Writing Skills (Factor 6). For General Affect toward Writing (Factor 2) and Attitude toward Writing Competence (Factor 5), the predictors did not group but were found across qualities. This, like that of the total summed instrument, may reflect the more general nature of the areas while the more specific factors produced grouping. The importance of these findings underlies the general nature of the development of the instrument. Concern at the time of creation was divided in two areas: the identification of writing apprehension and the problem that both logic and other research seemed to suggest a multidimensional rather than unidimensional treatment of the construct.18 The results of this investigation lend strong support for the multidimensional treatment of writing apprehension. That the total instrument (1) predicted writing apprehension as related to different syntactic qualities, (2) that it correlated highly with the factors obtained, and (3) a specific dimension for "business writing" was obtained providing both strong identification and diagnostic capabilities. Based on this bifurcation, it may be possible to assess an individual's total writing apprehension and then, based on the subscale scores, diagnose where attention should be placed. It may be that one or two subscales yield the problem areas; exercises to attack both the behavioral (instrumental) and cognitive (attitudinal) problems can then be administered. Although the results of this research are encouraging, they also point out that further research and refinement is needed. Specifically, the addition of statements on those factors possessing only two or three items may increase their reliability and make interpretation easier.19 There needs to be research undertaken that further refines the relationship between oral and written apprehension. And, finally, self-concept and the discrepancy between idealized and actualized self-concept needs to be examined in more detail. As Stacks and Stone noted, the selfconcept and the discrepancy between actual and idealized selfconcept plays a major factor in predicting oral communication apprehension.20 It should also play a major factor in writing apprehension; some of the predictor variables found in this study seem to imply that writer self-concept may play a major function in both the assessment and treatment of writing apprehension. In summary, this research has lent strong support to a multidimensional representation of writing apprehension. It has also supported the earlier research on the effect of syntactic choice and apprehension and extended that research significantly. In all cases the relationships between writing apprehension and individual factors were linear, suggesting that treatments should examine ways to either decrease or increase particular syntactic choice and use. These findings suggest that writing apprehension can be identified and diagnosed. They further suggest that syntactic language choice is one predictor across at least five of the six dimensions found to correlate with writing apprehension. Future research need to refine the instrument and expand its use to other types of writing (reports, speeches, etc.) and examine its relationship to other related constructs. University of Alabama Memphis State University Notes 1 John A. Daly and Michael D. Miller, "The Emperical Development of an Instrument to Measure Writing Apprehension," Research in the Teaching of English, 9 (1975), 242-249. 2 Robert W. Boozer, Tim D. P. Lally, and Don W. Stacks, "Written Communication Apprehension I: What Is It and How Can It be Measured?" submitted to Journal of Business Communication; Judee K. Burgoon and Jerold L. Hale, "A Research Note on the Dimensions of Communication Reticence," TS, Department of Communication, Michigan State University, 1983; Judee K. Burgoon and Jerold L. Hale, "Dimensions of Communication Avoidance and their Impact on Verbal Encoding," International Communication Association, Minneapolis, MN, May 1981; Judee K. Burgoon and Randall J. Koper, "Communication Reticence and Relational Message Behavior," International Communication Association, Dallas, TX, 27 May, 1983. 3 Daly and Miller, John A. Daly and Michael D. Miller, "Further Studies in Writing Apprehension: SAT Scores, Success Expectations, Willing ness to Take Advanced Courses, and Sex Differences," Research in the Teaching of English, 9 (1975), 250-256; John A. Daly and W. Shamo, "Writing Apprehension and Occupational Choice,” Journal of Occupational Psychology, 49 (1976), 55-56; John A. Daly and W. Shamo, "Academic Decisions as a Function of Writing Apprehension,” Research in the Teaching of English, 12 (1978), 119-126; Michael D. Scott and Lawrence R. Wheeless, "Communication Apprehension, Student Attitudes, and Levels of Satisfaction," Western Journal of Speech Communication, 17 (1977), 188-198. 4 John A. Daly, "Writing Apprehension and Writing Competency," Journal of Educational Research, 72 (1978), 1-14. 5 Lester Faigley, John A. Daly, and Stephen P. Witte, “The Role of Writing Apprehension in Writing Performance and Writing Competency,” Journal of Educational Research, 75 (1981), 16-21. 6 Faigley, Daly, and Witte. 7 See: K. W. Hunt, "Syntactic Maturity in School Children and Adults," Monographs of the Society for Research in Child Development, 35 (1970), No. 134. 8 Boozer, Lally, and Stacks. 9 The six factor instrument is available from the authors. See also, Boozer, Lally, and Stacks. 10 H. Wayland Cummings and Steven L. Renshaw, "SLCA III: A Meta-Theoretic Approach to the Study of Language," Human Communication Research, 5 (1979), 291-300. 11 H. Wayland Cummings, "Computerized Language Analysis," Speech Communication Association, Louisville, KY, 5 November 1982. 12 Most research in communication apprehension, oral or written, has taken a distinctly trait approach in which the basic distinction between apprehensives and nonapprehensives is measured by an artificial difference in instrument scores. Usually the standard deviation is the delimiter whereby those scoring one standard deviation above the mean response are classified as apprehensive, those one standard deviation below the mean as nonapprehensive, and those in the middle are not considered. This format does not allow for the naturally occurring apprehension we all have when faced with a writing assignment, which is more state-like. 13 James H. Goodnight, "Stepwise," in Kathryn A. Helwig, eds., SAS Users Guide, 1979, 391-392. The small sample size is indicative of the exploratory nature of this research 14 Norman R. Drapier and Harry Smith, Applied Regression Analysis (New York: Wiley-Interscience, 1966), pp. 86-103; Fred N. Kerlinger and Elazer J. Pedhazur, Multiple Regression in Behavioral Research (New York: Holt, Rinehart, and Winston, 1973), pp. 63-65. 15 L. J. Cronback, “Coefficient Alpha and the Internal Structure of Tests," Psychometrika, 16 (1951), 297-334. 16 Cronback. 17 Faigley, Daly, and Witte. 18 Burgoon and Hale, "A Research Note"; Burgoon and Hale, "Dimensions"; Burgoon and Koper. 19 c.f., Boozer, Lally, and Stacks. 20 Don W. Stacks and John D. Stone, "The Effect of Self-Concept, Self-Disclosure, and Type of Basic Speech Course on Communication Apprehension," Communication Journal of the Pacific, (1983, in press); Don W .Stacks and John D. Stone, "An Examination of the Effect of Basic Speech Courses, Self-Concept, and Self-Disclosure on Communication Apprehension," Communication Education, (1983, in press). |
||
|
 |
 |
|||||||||
|
|||||||||
