West Virginia University University of Nebraska

John A. Daly Barbara G. Cox

Purdue University Pennsylvania State University

Two studies are reported which indicate that high communication apprehensives are perceived as less interpersonally attractive than low communication apprehensives by members of the opposite sex. Effects of physical appearance and attitude similarity on interpersonal attraction, consistent with previous research, were also observed. It was also found that, in general, the more communication apprehensive the subject the less the subject was attracted to other people.

Many explanations have appeared in the scientific literature for way one person is attracted to another (Berscheid & Walster, 1969). Four of the best supported explanations are close physical proximity (Priest & Sawyer, 1967), receipt of personal rewards (Byrne & Griffitt, 1966), attitude similarity (Clore & Baldridge, 1968), and physical appearance (Walster, Aronson, Abrahams, & Rottman, 1966).

Comparatively less attention has been directed toward the effects of communication behavior on interpersonal attraction. Yet it seems intuitively obvious that communication behavior would have a major impact, particularly after initial encounters have passed (Berger & Calabrese, 1975). There is some direct evidence to support this speculation (e.g., Quiggens, 1972; McCroskey, Hamilton & Weiner, 1974; Daly, McCroskey, & Richmond, 1974),l and much of the data indicating that receipt of personal rewards and attitude similarity positively affect interpersonal attraction provide indirect evidence, since communication is a primary means of delivering rewards and exhibiting attitude similarities.

The research that provides direct support for relationship between communication behavior and interpersonal attraction suggests that a certain style of communication behavior is positively associated with interpersonal attraction, particularly social attraction. In their research on communicator style, Norton, Pettegrew, and Land (1975) report that people who are perceived as relaxed and animated are better liked than others and those perceived as relaxed are also judged to be more desirable people with whom to work. The measure of attraction in their study was the Interpersonal Judgement Scale (Byrne, 1971; Byrne & Griffitt, 1973). Similarly, Daly, McCroskey, and Richmond (1974), in a study involving simulated verbal behavior, found that increased verbal activity (up to a moderately high level) was accompanied by increased social attraction. In a study involving live interaction in a small group setting, McCroskey, Hamilton, and Weiner (1974) also reported a positive relationship between increased verbal activity and social attraction. In the latter study it was also observed that the exhibition of tension in the interaction was accompanied by a decrease in social attraction.

Lower levels of vocal activity and increase exhibition of tension in interaction are characteristic behaviors associated with the phenomenon of reticence (Phillips, 1968) or communication apprehension (McCroskey, 1970). Communication apprehension is a broad-based, personality-type characteristic that is characterized by withdrawal from and avoidance of communication, as well as feelings of discomfort and tension while communicating. Based on the results of the Daly et al and McCroskey et al studies, it seems reasonable to hypothesize that people who are high communication apprehensives will be perceived as less interpersonally attractive (particularly on the social attraction dimension) than people who are low communication apprehensives.

This hypothesis, however, must be tempered by an awareness of additional results reported by Daly, McCroskey, and Richmond (1974). When they analyzed their data taking into account the self-reported verbal activity of their Ss, they found an interaction between the verbal activity of the target person and that of the S on levels of social and task attraction. The clear preference for people with high levels of verbal activity (presumably low communication apprehensives), over those with moderate or low levels, with exhibited only by Ss who reported high levels of verbal activity themselves. People reporting moderate levels of vocal activity found others with moderate or high levels to be more attractive than those with low levels, but did not distinguish between the highs and the moderates. Further, those Ss reporting low levels of verbal activity (presumably high communication apprehensives) found people with moderate levels most attractive and did not distinguish between people with high and low levels.

The preceding interaction suggests the possibility of a correlation between similarity of levels of communication apprehension and interpersonal attraction. Personally similarity has been advanced previously as an explanation for interpersonal attraction (Berscheid & Walster, 1969), and communication apprehension can be though of as a personality-type variable. It is broader, however, than most individual personality variables. As noted by Sorensen (1972), communication apprehension is significantly correlated with at least ten other personality variables. Previous researchers investigating personality similarity and interpersonal attraction generally have found no relationship or, at best, only a low positive correlation (Berscheid & Walster, 1969). We may be attracted to people who have personality characteristics similar to ours that we evaluate positively, but we may be repulsed by people who have personality characteristics similar to ours that we evaluate negatively. Considering communication apprehension specifically, there may be a generally more positive orientation toward people who are not apprehensive, but this impression may be less strongly positive for people who themselves are moderately or highly apprehensive.

In the only previous study which directly addressed the question of the impact of communication apprehension on interpersonal attraction, Quiggens (1972) found that source and receiver levels of communication apprehension interacted in the production of social attraction. Low apprehensives were more socially attracted to other low apprehensives than to high apprehensives, but high apprehensives did not discriminate. On the dimension of task attraction, however, both high and low apprehensives were more attracted to low apprehensives than to high apprehensives. Quiggens cautioned against generalizing his results, however, because of his low sample size and the possibility that the sex of his Ss may have confounded his results.

Considering the results of the above studies, we were led to hypothesize both a main effect for the apprehension level of a target persons and an interaction between apprehension levels of both persons in a dyad on perceptions of interpersonal attraction. Keeping in mind the possible confounding effects of the sexual composition of a dyad noted by Quiggens (1972), we advanced the following hypotheses to be tested across two studies with different subject samples:

1. In cross-sex dyads, low communication apprehensives will be perceived as more socially attractive than will high communication apprehensives.

2. In cross-sex dyads, low communication apprehensives will be perceived as more task attractive than will high communication apprehensives.

3. In cross-sex dyads, low communication apprehensives will be perceived as more desirable potential communication partners than will high communication apprehensives.

4. In cross-sex dyads, low communication apprehensives will be perceived as more desirable potential sexual partners than will high communication apprehensives.

5. There will be an interaction of level of communication apprehension of S and target person in the production of interpersonal attraction. This interaction will be produced by the magnitude of the effect for differing levels of subject apprehension. The effects hypothesized in (1) - (4) will apply to all S communication apprehension levels, but will be larger for low communication apprehension Ss than for high communication apprehension Ss.

General Procedure

The hypotheses were tested experimentally. Ss were students enrolled in basic communication courses. Each S completed a self-report measure of communication apprehension. At a subsequent class session, before there was any discussion of communication apprehension, Ss recorded their sex, age, and student number for purposes of later combining these data with the S's communication apprehension score. On this page appeared a description of a person and (in the first study) a photo allegedly of that person. The description was of a person who exhibited either high or low communication apprehensive behavior. On the second page Ss were asked to complete scales designed to measure perceived physical, social, and task attraction of the person described, the desirability of the person as a potential sexual partner. The experiments were administered as class projects and Ss were told they were related to an uncoming unit. The experimental condition to which S was assigned was determined randomly. The results of the study were reported to Ss during the later unit and they were fully debriefed.

Controls

Because of the results of previous research concerning interpersonal attraction, it is known that a number of variables affect the level of interpersonal attraction one person has for another. No single variable should be expected to account for a majority of the variance in interpersonal attraction. Thus, to study any single variable, it is particularly desirable to control or manipulate as many other variables known to influence interpersonal attraction as is feasible. In our effort to study the effects of source and receiver communication apprehension we attempted to control for the effects of physical appearance, sex, age, proximity, attitude similarity, and receipt of personal reward. Since it was not possible to manipulate all these variables in a single study, attempts were made to manipulate some in each study and to control for the others. The means by which each control was attempted is outlined below.

Physical Appearance

A series of photographs of 40 males and 40 females were taken in informal settings at a college other than the locations of these studies. The photographers were instructed to seek wide variety in physical appearance of the people selected for their photos. These photos were initially screened by the researchers to omit any photos that indicated where they were taken or included from distracting information (such as a school letter on a jacket). Photos of people who were clearly older or younger than 19 (the average age of the samples to be studied) or of non-Caucasians (99 percent of the sample to be studied was Caucasian) were also excluded. The remaining photos (14 male, 15 female) were evaluated for physical attractiveness on a seven-point scale by panels of five male and five female graduate assistants, each evaluating members of the opposite sex. On the basis of these ratings, five male and five female photos were selected to represent five a priori categories of physical attractiveness: extremely attractive, moderately attractive, average, moderately unattractive, extremely unattractive. These 10 photos were used subsequently in the first study. No photos were used in the second study, and the procedure employed did not permit a visual perception of the target person.

Sex

These studies were restricted to cross-sex interpersonal attraction. When the experimental materials were distributed, each S was given material including a description of a person with a name characteristic of the opposite sex. In the first study, Ss were also provided a photo of the person.

Age

The person in the photograph was identified as being 19 years of age. Ss recorded their age as part of the study. In the first study there were 44 percent who were also 19, 31 percent who were younger (17-18), and 25 percent who were older (20-26). Seven Ss who were over 26 (30-57) were omitted from all data analyses. In the second study, the percentage breakdown was essentially the same, but there were no Ss older than 29.

Proximity

In the first study, the person in the photograph was identified as a person from the state in which the study was conducted, a student at the same institution and living in a campus dormitory (unspecified). Thus, only fairly remote proximity was assumed to be perceived by Ss. This assumption was not tested because of time and logistical constraints on the administration of the study. In the second study, half of the Ss were informed the person was from the same midwestern state while the other half were informed that the person was from New York City.

Attitude Similarity

In the first study, there was no attempt to manipulate attitude similarity between the photographed person and Ss. The only materials provided that might cause the S to infer attitudes (as opposed to personality orientations) were the following statements: "Her (his) parents own and operate a small business." "...he (she says he (she) likes the [the dorm] very much. However, he (she) hopes to move off campus sometime this year." It was assumed, therefore, that variance in perceived attitude similarity was controlled across experimental conditions.

In the second study, the perception of attitude similarity was manipulated by informing Ss of the alleged attitudes of the target person, purportedly obtained from the person in a survey at the beginning of the semester. Ss were told that the target person had selected "four strong views from a list of 40." Then these four views were listed. A total of six attitude statements were employed. Which four were attributed to the target person for a given S was determined randomly. The six attitude statements used were: 1) Amnesty should be given to Vietnam War draft evaders. 2) The University of Nebraska should provide more parking facilities convenient for undergraduates. 3) A college education is vita in order to attain financial success. 4) Couples should live together at least several months before deciding to marry. 5) The football program at the University of Nebraska should be deemphasized. 6) There should be no amnesty for military deserters.

Receipt of Personal Reward

There was no induction of personal reward in either study. Thus, it was assumed that this variable was controlled (by its absence) across experimental conditions.

Of the variables controlled in the first study, three were controlled in such a manner as to permit statistical analysis of their effects: physical appearance, sex, and age. Proximity, attitude similarity, and personal reward effects could not be tested statistically and were assumed to be randomly distributed across the experimental and control conditions. In the second study, four variables were controlled in ways permitting statistical analysis of their effects: sex, age, attitude similarity, and proximity (region of origin). Physical appearance and personal reward effects were assumed to be randomly distributed (absent).

Experimental Manipulations and Measurement

Apprehension of Target Person

Ss' perceptions of the target person as either a high or a low communication apprehensive were manipulated b including, within the overall description of the person, descriptions of behaviors known to be associated with either high or low communication apprehension. The behaviors chosen were determined by previous research to be clearly associated with communication apprehension. The descriptions were pretested and found to induce perception of high and low apprehension as intended (p>.0001). The portions of the descriptions related to communication apprehension were as follows:

(Low Apprehensive) Bev (Bill) is a very verbal person who participates a lot in class and has a good academic record. She (he) indicates that she (he) would rather work with a group rather than alone; and in fact, when working with a group is a very active participant. She (he) has not as yet decided on a major, but indicates she (he) hopes to go into a profession that requires a lot of direct contact with people. Bev (Bill) lives in a dorm and has a room at the end of the hall right next to the entrance...

(High Apprehensive) Jan (John) is a very quiet person who seldom participates in class, but has a good academic record. She (he) indicates that she (he) would rather work along than with a group; and, in fact, when working with a group seldom says much. She (he) has not as yet decided on a major, but indicates she (he) hopes to go into a profession that requires little direct contact with people. Jan (John) lives in a dorm and has a room at the end of the hall away from the entrance...

Since these inductions had been carefully pretested with a sample from the same population, there was no specific manipulation check included in the first study. In the second study, Ss were asked to indicate on a 10-point scale how "anxious about communication" they perceived the target person to be. The high apprehension condition received significantly (p < .005) higher ratings than the low apprehension condition.

Apprehension of Subject

The extent of communication apprehension of the Ss was measured by means of the Personal Report of Communication Apprehension-College Form (McCroskey, 1970). Ss were divided into three groups based on their PRCA scores. Those with scores beyond one standard deviation above the mean were classified as High Apprehensives; those with scores beyond one standard deviation below the mean were classified as Low Apprehensives; and the remainder were classified as Moderate Apprehensives. The split-halves reliability estimate for the PRCA in the first study was .93; in the second it was .92.

Interpersonal Attraction

Three dimensions of interpersonal attraction (social, task, and physical) were measured by means of scales developed for this purpose (McCroskey & McCain, 1974). The scores on the social and task dimensions were used as dependent variables to test Hypotheses 1, 2, and 5. The scores on the physical dimension were analyzed as a manipulation check for the physical attractiveness manipulation in the first study as a check on the control of this variable in the second. There were no significant effects observed in the second study. Factor analyses (oblique) indicated the presence of the three expected dimensions in both studies with all items loading on the expected factors (>.60) with no secondary loading above .40. As was expected due to the manipulations employed in the first study, the physical dimension was significantly (p < .05) correlated with both the social (r = .40) and task (r = .22) dimensions. In the second study the correlations were nonsignificant (.12 and .09), respectively. The task and social dimensions were significantly correlated in both the first (r = .15) and the second study (r = .17). Internal reliability estimates (split-halves) for the three dimensions in the first study were as follows: social, .90; task, .87; physical, .92. In the second study they were .88, .89, and .91, respectively.

Desirability as Communication Partner

Perceived desirability of the target person as a communication partner was measured by two bipolar scales selected for this purpose. Ss responded on a seven-step continuum bounded by "Very Pleasant" and Very Unpleasant" to the following questions: "How pleasant do you think it would be to talk to this person?" and "How pleasant do you think it would be to listen to this person?" The internal reliability estimates for these two summed scales were .86 in the first study and .81 in the second.

Desirability as a Sexual Partner

Perceived desirability of the target person as a potential sexual partner was measured by two bipolar scales selected for this purpose. Ss responded on a seven-step continuum bounded by "Definitely Yes: and "Definitely No" to the question: "Would you like to date this person?" They also responded on a seven-step continuum bounded by "Very Pleasant" and "Very Unpleasant" to the question: "How pleasant do you think it would be to be married to this person?" The internal reliability estimates for these two summed scales were .80 in the first study and .85 in the second.

Attitude Similarity

Ss' perceptions of attitude similarity-dissimilarity were measured in the second study by means of six bipolar scales developed for this purpose by McCroskey, Richmond, and Daly (1974). The split-halves reliability estimate for these six scales was .91. For purposes of subsequent statistical analyses, Ss were divided at the median score on this measure to form high-similarity and low-similarity classifications.

Statistical Design and Analyses

Study 1: Preliminary Analyses

Since the female (N=341) and male (N=310) Ss (college students in an eastern university) were not exposed to the same photographs, it was both necessary and desirable to treat the sex of the S as a unite of analysis. However, since it could not be assumed that the levels of the photo inductions were equal for the two sexes, it was not possible to analyze the data by simultaneously classifying Ss both by level of photo induction and sex in the same analysis. Further, it was believed that the Ss' perceptions of physical attractiveness, rather than our a priori photo classifications, would provide a better estimate of the relationship between physical attractiveness and our other interpersonal attraction variables. Consequently, the data on subject-perceived physical attraction were subjected to four-factorial analyses of variance, the female and male data being considered separately. The independent variables in the analyses were physical appearance (the five photo levels), age of S (older, same age, and younger), S apprehension (high, moderate, and low), and apprehension behavior level of the target person induction (high or low).

The results of these preliminary analyses indicated that the photo levels did induce significant variance in perceived physical attraction (females, F = 66.30, p < .001; males, F = 51.38, p < .0001) in the ways expected. No other independent variable and no interaction produced significant differences in perceived physical attractiveness. In addition, a t-test of the difference between the female (x = 24.0) and male (x = 23.6) mean scores on perceived physical attractiveness indicated no significant differences (t = 1.21, p < .05). It was possible, therefore, to combine the data from the male and female Ss for subsequent analyses and treat perceived physical attractiveness as a covariate rather than as a classification variable in these analyses.

Since there were four dependent interpersonal attraction variables in this study, a decision had to be made whether to treat these variables as independent from each other (an employ a series of univariate analyses of variance) or consider the four to be part of an interrelated system (and employ a multivariate analysis of variance). The rationale leading to the research hypotheses suggested that the variables may be part of an interrelated system, and predicted the same effects for each variable. Thus, the correlations among the variables were examined. It was found that the social attraction variable was significantly correlated with the other variables (task, r = .15; communication, r = .55; sexual, r =.51). Similarly, task attraction was significantly correlated with the others (communication, r = .31; sexual, r = .25). In addition, attractiveness as a sexual partner were significantly correlated (r = .46).

Since all of the variables were found to be intercorrelated, the possibility that they were operating as a system required that they be subjected to multivariate analysis. However, since the magnitude of the intercorrelations ranged from low (r = .15) to moderate (r = .55), it was also deemed likely that the variables could, in part, be operating independently of the total system. In order to test for such variation, it was also necessary to subject the data to a series of univariate analyses.

Study 1: Main Analyses

The data were subjected to a four-factorial multivariate analysis of variance with one covariate and to a series of four-factorial univariate analyses of variance with one covariate. The covariate in each analysis was perceived physical attractiveness. The other independent variables in the analyses were sex of S (female and male), age of S (older, same age, and younger), S apprehension (high, moderate, and low), and apprehension behavior level of the target person induction (high and low). Hypothesized effects on the individual dependent variables were tested by means of t-tests based on the mean-square error estimate from the univariate analyses of variance. The .05 level of significance was employed for all tests.

Study 2: Preliminary Analyses

Because of the comparatively small sample (N=212), it was not possible to include all of the six independent variables within the same analyses and retain sufficient power to provide adequate tests of higher order interactions. Consequently, a series of three-factorial analyses (multivariate and univariate) were conducted keeping apprehension behavior of target person and apprehension level of S in each analysis, and alternately including sex, age, attitude similarity, and proximity. This procedure permitted testing the research hypotheses against possible alternatives of interactions with other single variables, but did not permit testing the research hypotheses against possible higher order interaction alternatives (e.g., target apprehension by S apprehension by sex by age). Hypothesized effects on the individual dependent variables were tested by means of t-tests based on the mean-square error estimated from the univariate analyses of variance. The .05 level of significance was employed for all tests.

Study 1

Multivariate Analysis

The multivariate analysis indicated the presence of our significant effects. The physical attractiveness covariate (F = 152.46; df = 4/611), apprehension behavior of the target persons (F = 22.28; df = 4/611), subject sex (F = 2.79; d/f = 4/611), and the apprehension behavior of target person by apprehension level of subject interaction (F = 2.05; df = 4/611) effects were significant. In each case a single canonical variable was generated. Table 1 reports the correlations between the dependent variables and the canonical variables. As is indicated by the correlations in Table 1, the four dependent variables did not contribute equally the observed effects found to be significant. Thus, interpretation of these results should be based on the results of the univariate analyses.

Dependent Variable

Effect Producing Canonical Variable Social Task Communication Sexual

Physical Attractiveness (Covariate) .67 .28 .59 .83

Target Apprehension .89 .19 .69 .19

Subject Sex .06 -.06 .14 .97

Target Apprehension X Subject Apprehension .74 -.12 .20 .77

Dependent Variable

Effect Producing Canonical Variable Social Task Communication Sexual

Target Apprehension .96 .07 .53 .57

Subject Apprehension .78 .33 .60 .86

Attitude Similarity .80 .55 .83 .56

Subject Apprehension X Attitude Similarity -.06 .82 .64 .09

Univariate Analyses.

The results of the analysis of social attraction data indicated the presence of three significant effects: the physical attractiveness covariate (F = 271.77; df = 1/614), the apprehension behavior of the target person by apprehension level of subject interaction (F = 5.18; df = 2/614). Table 2 reports the social attraction means reflecting the significant interaction. The results in Table 2 indicate support for both Hypotheses 1 and 5.

Apprehension Behavior of Target Apprehension Level of Subject

Low Moderate High

Social Attraction

Low 27.87a,d 26.32b 25.08c,d

High 22.06a 22.89b 22.92c

Difference 5.81 3.43 2.16

Sexual Attraction

Low 8.10e,g,h 7.10f,g 6.95h

High 6.33e 6.63f 6.81

Difference 1.77 .47 .14

^a-hMeans with same subscript are significantly different, p<.05. No cross-diagonal tests were computed.

The results of the analysis of the task attraction data indicated the presence of two significant effects: the physical attraction covariate (F = 48.78; df = 1/614) and apprehension behavior of the target person (F = 5.34; df - 1/614). The low apprehensive target person was perceived as more task attractive (x = 26.44) than the high apprehensive target person (x = 25.92). While these results provide support for our second hypothesis, it should be noted that the magnitude of the hypothesized difference was very small (d = 0.52) and the variance accounted for was less than two percent of the total variance of the task attraction scores. This analysis provided no support for our fifth hypothesis; there was no significant interaction of target by subject apprehension.

Analysis of the data relating to the attractiveness of the target person as a communication partner yielded three significant main effects: the physical attraction covariate (F = 216.03; df = 1/614), apprehension behavior of the target person (F = 77.13; df = 1/614), and subject apprehension (F = 3.77; df = 2/614). The low apprehensive target person was perceived as a more attractive communication partner (x = 10.68) than the high apprehensive target (x = 9.15). As might have been expected, given our knowledge of the relationship between communication apprehension and the perceived desirability of communication, the low apprehensive desirability of communication, the low apprehensive subjects saw the various target persons as more attractive communication partners (x = 10.52) than did the moderates (x = 9.90) and the highs (x = 9.33). The results indicate support of our third hypothesis but no support for the fifth hypothesis, since there was no significant target by subject apprehension interaction.

Analysis of the sexual attraction scores results in four significant effects: the physical attractiveness covariate (F = 402.99; df = 1/614), apprehension behavior of the target person (F = 5.82; df = 1/614), sex of S (F = 14.34; df 1/614), and the interaction of target person and subject apprehension (F = 5.65, df = 2/614). Male Ss perceived the target persons as more sexually attractive (x = 7.40) than the female Ss (x = 6.60). Table 2 reports the sexual attraction means reflecting the significant interaction. The results in Table 2 indicate partial support for both Hypotheses 4 and 5. While the main effect for target person apprehension behavior predicted in Hypothesis 4 was reflected in the absolute score differences at all levels of subject apprehension, the differences were significant only for low and moderate apprehensive Ss. The magnitude interaction hypothesized in H₅ was present, but the difference expected for high apprehensives, as noted above, was not significant.

Study 2

Multivariate Analysis

Since the only variable other than apprehension behavior of target person and apprehension level of S to be found significant in any of the multivariate or univariate analyses was attitude similarity, only those analyses including the attitude similarity variable will be reported. No significant main effects or interactions involving sex, age, or proximity were observed.

The multivariate analysis of variance involving attitude similarity, target apprehension behavior, and subject apprehension level indicated the presence of four significant effects. The significant effects involved apprehension behavior of the target person (F = 3.50; df = 4/197), subject apprehension (F = 2.10; df = 4./197), attitude similarity (F = 6.26; df = 4/197), and the interaction of subject apprehension and attitude similarity (F = 2.63; df = 4/197). As is noted in Table 1, the four dependent variables did not contribute equally to the effects found to be significant. Thus, as was the case in the first study, the interpretation of these results must be based on the results of the univariate analyses.

Univariate Analyses

The results of the analysis of the social attraction data indicated the presence of three significant main effects: apprehension behavior of the target person (F = 22.87; df = 1/200), subject apprehension (F = 4.53; df = 2/200), and attitude similarity (F = 16.58; df = 1/200). The low apprehensive target person was perceived as more socially attractive (x = 25.02) than the high apprehensive target person (x = 21.80). Low apprehensive Ss found the various target persons more attractive (x = 25.86) than did moderates (x = 23.12) or highs (x = 22.82). Target persons with high attitude similarity (x = 25.26) were perceived as more attractive than target persons with low attitude similarity (x = 21.55). The results on social attraction provide support for our first hypothesis but do not support the fifth hypothesis.

The results of the analysis of task attraction data indicated the presence of two significant effects: attitude similarity (F = 7.80; df = 2/200) and the interaction of attitude similarity and subject apprehension (F = 4.88; df = 2/200). Table 3 reports the task attraction means reflecting the significant interaction. Highly apprehensive Ss were much more task attracted to target persons with similar attitudes than target persons with dissimilar attitudes. Differences for low and moderate apprehensives Ss were in the same direction but were not statistically significant. There was no support for our hypotheses provided by these results.

Analysis of the data relating to the attractiveness of the target person as a communication partner indicated three significant effects: apprehension behavior of the target person (F = 8.13; df = 1/200), attitude similarity (F = 17.43; df = 2/200), and the interaction of subject apprehension level and attitude similarity (F = 3.46; df = 2/200). The low apprehensive target person was viewed as a more attractive communication partner (x = 9.06) than the high apprehensive target person (x = 8.16). The means reflecting the significant interaction are reported in Table 3. The results indicate a magnitude interaction. While Ss with differing apprehension levels did not significantly differ in

Attitude Apprehension Level of Subject

Similarity Level Low Moderate High

Task Attraction

Low 26.22a 24.73 22.74a,b

High 26.58 25.59 27.35b

Difference .36 .86 4.61

Communication Attraction

Low 8.89c 7.62c,d 6.84c,e

High 9.68 9.54d 9.80e

Difference .79 1.92 2.96

^a-eMeans with same subscript are significantly different, p < .05. No cross-diagonal tests were computed.

their perceptions of the target persons who were attitudinally similar, target persons who were attitudinally dissimilar were perceived as increasingly less attractive as communication partners by low apprehensives, moderate apprehensives, and high apprehensives, respectively. The results concerning the apprehension behavior of the target person provide support for our third hypothesis, but no support for our fifth hypothesis was present.

Analysis of the sexual attraction data yielded three significant main effects: apprehension behavior of the target person (F = 4.57; df = 1/200), subject apprehension (F = 4.43; df = 2/200), and attitude similarity (F = 7.99; df = 1/200). Low apprehensive targets were perceived as more sexually attractive (x = 6.56) than high apprehensive targets (x = 6.00). Low apprehensives saw the target persons as more sexually attractive (x = 7.11) than the moderates (x = 6.41) who, in turn, perceived the targets as more sexually attractive than did the high apprehensives (x = 5.21). Ss perceived target persons who were attidudinally similar as more sexually attractive (x = 7.19) than those who were attudinally dissimilar (x = 5.36). The results relating to the apprehension behavior of the target person provide support for our fourth hypothesis, but no support for the fifth hypothesis was present.

We will divide our discussion of these results into two parts, the results bearing on our research hypotheses and other results.

Research Hypotheses

Our first hypothesis predicted that in cross-sex dyads low communication apprehensives will be perceived as more socially attractive than high communication apprehensives. The results of both studies support this hypothesis. In both studies the effect for apprehension behavior of the target person was significant. Moreover, the effect accounted for over 10 percent of the total variance in social attraction in both studies, with additional variance being attributable to the interaction of target and subject apprehension in the first study. From these results we conclude that behaviors characteristic of high communication apprehensives have a significant, meaningful, negative impact on a person's perceived social attractiveness by the opposite sex.

The second hypothesis predicted that in cross-sex dyads low communication apprehensives will be perceived as more task attractive than high communication apprehensives. The results of the first study provided support for this hypothesis; the results of the second did not. We believe these discrepant results can best be explained in terms of the sizes of the two samples studied. The effects observed (magnitude of absolute differences) were essentially the same in the two studies (Study 1, d = 0.52; Study 2, d = 0.43) and accounted for between 1 and 2 percent of the total variance in the scores. The effects, although consistently small, was statistically significant in the first study but not in the second.

We believe that in future studies with large sample sizes (sufficiently powerful, on other words) one should expect this hypothesis to be supported statistically. However, we are led to question the meaningfulness of this observation. The effect size observed, whether statistically significant or not, adds little to our ability to predict a person's task attractiveness. We conclude, therefore, that behaviors characteristic of high communication apprehensives probably have a statistically significant, but not meaningful, negative impact on a person's perceived task attractiveness by the opposite sex.

The third research hypothesis predicted that in cross-sex dyads low communication apprehensives will be perceived as more desirable potential communication partners than will high communication apprehensives. This hypothesis was supported by the results of both studies. In the first study the effect accounted for over 8 percent of the total variance; in the second, over 5 percent. From these results we conclude that behaviors characteristic of high communication apprehensives have a significant, meaningful, negative impact on the degree to which a person is perceived by the opposite sex as an attractive potential communication partner.

Research hypothesis four predicated that in cross-sex dyads low communication apprehensives will be perceived as more desirable potential sexual partners than will high communication apprehensives. This hypothesis was partially supported (for low and moderate apprehensive Ss) in the first study and fully supported in the second study. Although influenced by the interaction of target and subject apprehension in the first study, the effect accounted for approximately 4 percent of the total variance in both studies. We conclude, therefore, that behavior characteristic of high communication apprehensives have a significant, moderately meaningful, negative impact on a person's perceived sexual attractiveness by members of the opposite sex. We restrict this conclusion to moderately meaningful because of the effect sizes observed. It is probably that other variables are much more important predictors of sexual attractiveness. Indeed, both physical attractiveness (37 percent) and attitude similarity (8 percent) were found to be so in these studies.

The final research hypothesis predicted a magnitude interaction between apprehension behavior of a target person and apprehension level of S. This hypothesis received only limited support in the first study and no support in the second. In the first study, there was a significant interaction (of the type predicted) observed in the multivariate analysis results, but only the social and sexual attraction variables were highly associated with the effect. This was also indicated by the results of the univariate analyses. Since even this limited support was absent in the second study, it is clear that we cannot accept this hypothesis based on the present results.

It may be that with very limited information about another person's attitudes; people infere others' attitudes from information about their communication behaviors and that these inferences impact perceptions of attraction. However, the fact that the interaction of subject apprehension and attitude similarity influenced task and communication attraction in the second study while the interaction of subject apprehension and target apprehension impacted social and sexual attraction in the first study is not explained by this speculation. An adequate test of this hypothesis against potentially more viable explanations would involve simultaneous manipulation of physical attractiveness and attitude similarity along with apprehension levels of both target person and S. Thus, we must conclude that interpretation of the results relating to this hypothesis should await future research of the type described above.

Other Results

The outcomes of these studies provide some results replicating earlier research on interpersonal attraction. Physical attractiveness was a strong predictor of all four interpersonal attraction variables in the first study. In terms of effect size, physical attractiveness was the best predictor, by far, of interpersonal attraction of all four types. This was true, however, only when physical attractiveness was operationalized as subject-perceived attractiveness. In other data analyses, not reported in detail here, the physical attractiveness variable was operationalized as the photo levels used as inductions. In these analyses target person apprehension behavior was a superior predictor of social, communication, and sexual attraction for both females and males. While we do not wish to make too large an issue of these results, we believe that future researchers should be cautioned to avoid accepting at fact value the claim of some that physical appearance is the single best predictor of interpersonal attraction. Communication apprehension behavior was a better predictor when physical attractiveness was operationalized by means of judgments of external observers rather than Ss. The concept of "beauty is in the eye of the beholder" will, of course, lead most to prefer basing judgments of physical attractiveness on subject perceptions. This poses a difficult measurement problem, however, because subject judgments of physical, social, and task attraction (and probably communication and sexual attraction as well) have been found to be orthogonal (McCroskey & McCain, 1974). Thus, it is very likely that such operationalizations will result in spuriously high estimates of the relationships, as is likely the case in the present study.

The results of the second study reinforce the finding of previous researchers that attitude similarity is an important predictor of interpersonal attraction. It was a major predictor of all four interpersonal attraction variables. Furthermore, attitude similarity interacted with subject apprehension on both the task and the communication attraction variables. High communication apprehensives were particularly sensitive to differences in attitudes between themselves and others, and were much less attracted to those with different attitudes.

One fairly consistent pattern emerged from these two studies that could not be comfortably hypothesized from the results of previous research, but should not be considered surprising. Generally, the more communication apprehensive the S, the less the S was attracted (on each of the attraction variables) to the target person. This finding is complementary to the general picture of the high communication apprehensive as socially withdrawn and desirous of reduced contact with others.

Taken together with the strong effect, noted previously, for people who engage in such behaviors to be perceived as less attractive (on each of the attraction variables), these results suggest that the highly apprehensive person may have little contact with, or influence upon, others. Neither they nor other people find interpersonal relationships with them as attractive as other alternatives. This may explain why, in data collected as a subsequent part of the second study but not reported in detail here, high apprehensives are perceived to be significantly less sociable, less composed, less competent, less extroverted, less powerful, and less likely to be turned to as an opinion leader than are low apprehensives. All of these perceptions are shared by low, moderate, and high apprehensives alike, though somewhat less severely by high apprehensives.

Considering all the results of these studies, it is clear that there is a strong negative relationship between communication apprehension and interpersonal attraction in cross-sex dyads. We believe it is useful to view the results of these studies in the context of the developmental theory of interpersonal communication set forth by Berger and Calbrese (1975). Portions of two axioms set forth in that theory are particularly relevant here:

1. Given the high level of uncertainty present at the onset of the entry phase [of interpersonal relationships], as the amount of verbal communication between strangers increases, the level of uncertainty for each interactant in the relationship will decrease.

2. Decreases in uncertainty level produce increases in liking.

Essentially, this theory argues that communication reduces the uncertainty people feel about each other, and reducing uncertainty results in increased liking or attraction. If correct, then it follows, as Berger and Calabrese (1975) theorize, that among of communication and liking should be positively related. If we accept the premise that one of the major results of communication apprehension is a reduction in the amount of communication in which the individual engages, the results of these studies provide strong support for the theory. The theory would predict that high apprehensives should be perceived as less attractive by others because they communicate less with them. Conversely, the theory would also predict that high apprehensives would perceive others as less attractive because they communicate less with those others. Both these predictions are supported by the results of these studies.

Although the Berger and Calabrese theory is supported by this research, we would caution against universal application of the theory at this point. In its present form the theory does not consider fully the content of the communication. It seems intuitively obvious that increasing some kinds of communication content will not lead to increased attraction. Negative self-disclosure, display of dissimilarity, and statements of negative regard will all lead to decreased uncertainty, but probably not to increased attraction. However correct this reservation may be, it also may be that such elements of negative content are so comparatively rare withing the totality of interpersonal communication content as to account for a meaningless percentage of the variance in interpersonal attraction. The resolution of this theoretical issue must await future research.

Berger, C. R., & Calabrese, R. J. Some explorations in initial interaction and beyond: Toward a developmental theory of interpersonal communication. Human Communication Research, 1975, 1, 99-112.

Berscheid, E. & Walster, E. H. Interpersonal attraction. Reading, Mass: Addison-Wesley, 1969.

Byrne, D. The attraction paradigm. New York: Academic Press, 1971.

Byrne, D. & Griffitt, W. A developmental investigation of the law of attraction. Journal of Personality and Social Psychology, 1966, 4, 699-703.

Byrne, D. & Griffitt, W. Interpersonal attraction. Annual Review of Psychology, 1973, 11, 317- 336.

Clore, G. L. & Baldridge, B. Interpersonal attraction: The role of agreement and topic interest. Journal of Personality and Social Psychology, 1968, 9, 340-346.

Daly, J. A., McCroskey, J. C., & Richmond, V. P. The relationship between vocal activity and perception of communicators in small group interaction. Paper presented to the convention of the Speech Communication Association, Chicago, 1974.

McCroskey, J. C. Measures of communication-bound anxiety. Speech Monographs, 1970, 37, 269- 277.

McCroskey, J. C., Hamilton, P. R., & Weiner, A. N. The effect of interaction behavior on source credibility, homophily and interpersonal attraction. Human Communication Research, 1974, 1, 42-52.

McCroskey, J. C. & McCain, T. A. The measurement of interpersonal attraction. Speech Monographs, 1974, 41, 261-266.

McCroskey, J. C., Richmond, V. P., & Daly, J. A. Toward the measurement of perceived homophily in interpersonal communication. Paper presented to the convention of the International Communication Association, New Orleans, 1974.

Norton, R. W., Pettegrew, L. S., & Land, S. Communication style correlates of attraction. Paper presented to the convention of the International Communication Association, Chicago, 1975.

Phillips, G. M. Reticence: Pathology of the normal speaker. Speech Monographs, 1968, 35, 39-49.

Priest, R. F. & Sawyer, J. Proximity and peership: Bases of balance in interpersonal attraction. The American Journal of Sociology, 1967, 72, 633-649.

Quiggens, J. G. The effects of high and low communication apprehension on small group member credibility, interpersonal attraction, and interaction. Paper presented to the convention of the Speech Communication Association, Chicago, 1972.

Sorensen, G. A. The use of personality traits and communication apprehension in predicting interaction behavior in small groups. Unpublished masters thesis, Department of Speech Communication, Illinois State University, 1972.

Walster, E., Aronson, V., Abrahams, D., & Rottman, L. Importance of physical attractiveness in dating behavior. Journal of Personality and Social Psychology, 1966, 4, 508-516.

Click Here To Go Back To PERIODICALS