John A. Daly, James C. McCroskey, and Virginia P. Richmond

A large amount of research has indicated that vocal activity, or how much an individual talks, is a major, and often ignored variable in interpersonal interactions. In the past, vocal activity has been related strongly to rankings of leadership, influence, and liking. The present study sought to examine the relationships between vocal activity and three other, less often examined variables: the attribution of quality contributions, perceived listening, and perceived understanding. Correlational analysis revealed a significant positive relationship between activity and attribution of quality contributions. Significant negative relationships were observed between vocal activity and perceived listening and understanding.

Considerable research has indicated a strong linear relationship between what may be termed vocal activity and leadership-influence rankings. Vocal activity is the frequency and durations of an individual's speech, or simply, how much an individual talks. Terms such as talkativeness,1 interaction rate2 and productivity3 all have been used to define vocal activity in the past. The research on the relationship between activity and leadership-influence rankings is, according to Willard and Strodtbeck "one of the most consistent findings in the small group literature."4 Reports of this relationship have been offered by a number of other researchers.5 A similar, significant, but curvilinear relationship between vocal activity and liking has also been often reported.6

Yet comparatively few studies have examined the relationship between the extent of a person's vocal activity and the perceived quality of the person's contributions. However, available research points to a significant positive relationship between vocal activity and perceived quality. That is, people who talk more are more likely to be perceived as offering the correct conclusions and as suggesting the best decision than do people who talk less, regardless of whether such perception shave any base in objective analyses of the interaction.7 In short, it appears that judgements of the quality of interaction are more dependent on the quantity of interaction than on the content of the interaction.

Virtually no research has examined the degree to which extent of vocal activity is related to perceptions of how understanding an individual is or how good a listener the person is; yet these two perceptions can be very important in determining the nature of interpersonal exchanges. Individuals will seldom communication with one who they believe will not or cannot listen or try to understand them.

The present study sought to examine the relationships between vocal activity and three interpersonal judgements: perceived quality of contributions, perceived listening, and perceived understandingness. It was hypothesized that there would be a significant positive relationship

John A. Daly (M.A., West Virginia University, 1974) is a doctoral student at Purdue University; James C. McCroskey (Ed.D., Pennsylvania State University, 1966) is Professor and Chairman in the Department of Speech Communication at West Virginia University; Virginia P. Richmond (M.A., West Virginia University, 1974) is a doctoral student at the University of Nebraska.

between vocal activity and perceived quality of contributions, and significant negative relationships


One hundred thirty undergraduate students enrolled in an introductory communication course at West Virginia University during the Fall semester, 1973, served as subjects in this research. Participation was part of a general research requirement of all students enrolled.

Each subject was provided with a five-page booklet. The first page asked Ss to indicate their sex and provided general instructions for completing the remaining pages. Subjects were told:

The following is a series of scales on which we will ask you to evaluate people about whom you have very little information. Try to evaluate as best you can within this very extreme limitation.

At the top of the following four pages Ss were asked to "evaluate a man (woman) who talked _____ percent of the time in a small group." Pages two and four asked for evaluations of a man, while pages three and five asked for evaluations of a woman. Percentages of talk were systematically varied from 0% to 95% with no subject being asked to evaluate hypothetical individuals that were less than 15% apart in their alleged vocal activity level. Thus each subject evaluated four individuals, two males and two females, each of whom was described only on the basis of talking different amounts of time.

Among the several questions asked were three relevant to this research report. Ss were asked to answer on nine-point scales the following three questions which were interspersed among other non-related questions:

1. How good a listener do you believe the person was in the group?

Very poor: 1 2 3 4 5 6 7 8 9: Very good

2. To what degree do you think this person tried to understand your point of view?

Very little: 1 2 3 4 5 6 7 8 9: Very much

3. What was the quality of this person's contributions to the group?

Very low: 1 2 3 4 5 6 7 8 9: Very high

Initially there was some doubt about the reliability of using descriptions of hypothetical individuals rather than having Ss actually engage in or observe real interactions. However, past research provided support fro the decision to employ the hypothetic procedure. For example, Hayes and Meltzer8 in a four-part study compared interpersonal evaluations based on a full knowledge of all verbal, vocal, and nonverbal cues with those based strictly on knowledge of vocal activity. In the first study, observers watched either a video-tape of a small group interaction or a panel of lights which represented the conversational patterns of the group members. In this latter condition, a light flashed every time a group member talked. The researchers found virtually no differences between the rankings of the individual group members made by observers in the video-tape conditions and the light-board condition. The second study utilized a different, more salient topic and different rating scales. The results were quite similar to those obtained in the first study. The correlation between rankings of small group members made by video-tape observers and light-board observers was extremely high. The third study had the same observers watch both the video-tape and light boards. Again the correlation between actually observing a group discussion with all its verbal and nonverbal cues and watching a light board that represented purely the amount of vocal activity was extremely high. The fourth and final study compared rankings made from the light-board flashes and descriptions of the relative amount of talk in which each group member engaged. This latter condition was the same as the method used in the present study. The correlations between these two types of stimuli, in terms of the interpersonal judgements made, were again very high. Allgeier9 in a recently completed dissertation replicated much of the Hayes and Meltzer research, extending their efforts to interpersonal attraction ratings and measures of perceived adjustment. In addition, while Hayes and Meltzer used male subjects, Allgeier chose to use females. His conclusions were similar to those of Hayes and Meltzer. Hayes and Sievers10 utilized the Hayes and Meltzer findings in a way very similar to our approach. They sought to determine the traits associated with varying levels of vocal activity. Just as in our study, descriptions consisted of the percentages of time individuals characteristically spend talking. These talk-time percentages ranged from 5% to 75%. Recently Daly, McCroskey, and Richmond11 utilized the description approach in determining the relationship between vocal activity and perceptions of credibility, attraction, homophily, and power. Results were generally as hypothesized, given previous research with actual interactants. Other earlier research by Mehrabian12 and Ward13 also provides support for our experimental approach.

Finally, a check was made of our method via the examination of previously well-evidenced relationships between vocal activity and perceived leadership and influence. If our approach is justified, we would expect significant positive correlations between vocal activity and these two perceptions. Scales designed to measure perceived leadership and influence were included among the items on the research instrument. The obtained correlations between vocal activity and the perceptions reported on these scales were virtually identical to those reported in previous research.


A total of 520 evaluations were collected. Two evaluations were incomplete and were discarded prior to analysis. There were approximately an equal number of males and females responding. Analysis of the data indicated no significant effects for either sex of the respondent or sex of the described individual. As a check on our use of descriptions of hypothetical individuals, we examined the relationship between described vocal activity and perceived leadership and influence. As expected, significant positive correlations were found--correlations that closely resembled those obtained in previous research with live stimuli material. Perceived leadership (r=.68, p<.05) and perceived influence (r=.54, p<.05) were both significantly and positively related to described vocal activity.

The data for each question were analyzed separately.14 While initial examination of the data indicated significant linear relationships, higher non-linear relationships were indicated by eta correlations.15 Quality and vocal activity were found to be positively correlated (r=.31, p<.05). A significant correlation ratio (eta=.50, p<.05) indicated that the relationship was nonlinear (eta>r, F=26.24, p<.01). Listening and vocal activity, as suggested by the hypothesis, were negatively and significantly related (r=-.31, p<.05), with a significant correlation ratio (eta=.51,p<.05) that was greater than r (F=28.36, p<.01), suggesting therefore a nonlinear relationship. Finally, understanding and vocal activity were found to be negatively correlated (r=-.25, p<.05) with a significant correlation ratio (eta=.36, p<.05) that again was greater than the r (F=9.87, p<.01), indicating non-linearity.

Table 1 indicates the relationships between the dependent variables and vocal activity grouped in twenty-point intervals.


The present research provides a further indication of the major role that vocal activity may play in interpersonal interactions. Individuals are able to make numerous judgments based entirely on vocal activity. Past research has noted the relationships between vocal activity and variables such as leadership, influence and status rankings. For two of these variable (leadership and influence) the relationship has been essentially linear, while for liking, the relationship seems to be more curvilinear.16



Level Quality Listening Understanding

0-15 3.74 6.49 5.40

20-35 5.42 6.83 6.01

40-55 6.00 5.46 5.50

60-75 6.27 4.77 4.90

80-95 5.65 3.49 4.06

The quality-activity relationship in this investigation appears to be basically linear, provided one accepts the assumption that an individual will not lack more than 70 to 75 percent of the time in a normal group interaction of four or five persons. Since an expected frequency would be about 20 to 25 percent, any level as high as 70 to 75 percent would seem extremely unrealistic. The results suggest that the more one talks, up to a very high level, the more others will attribute quality to that person's contributions in a discussion. This finding offers replication of earlier findings to a large extent.17 Two possible interpretations may be offered in explaining these results. One might suggest that there is some sort of transfer from the perceived leadership and influence relationships with vocal activity to the attribution of quality to ideas offered. Another explanation may lie in the past experiences of individuals. An individual who talks a great deal will have to provide some sort of significant content in that time spent talking. The number of quality ideas contributed ought to increase, simply as a function of time spent communicating. If this is true in most situations, then individuals may generalize their observation to all settings. Such an interpretation would explain the findings of Riecken, who noted that even when an individual did not have the correct solution, but talked a great deal of the time, he was ranked highly for the "correctness" of his contributions. In Riecken's work, individuals who had low vocal activity, yet who offered the "correct" solution, found that others in the group did not attribute the ideas to them.

The listening-vocal activity relationship seems essentially linear. While one may observe some upward trend in the lower levels of activity, after the level reaches 20 to 35 percent, the trend becomes noticeably downward. These results suggest that, in the eyes of the others, one who talks more than average will not have enough time to listen well. Individuals who talk less than average are seen as better listeners than those high in vocal activity. This relates strongly to the work by Pope and Nudler,18 among others, which suggests that vocal activity is made up of silence patterns and articulation patterns. Apparently, individuals believe that if one is talking a great deal, he or she can't be listening very well.

The nonlinearity of the understandingness-vocal activity relationship is obvious from Table 1. The optimal level of vocal activity for perceived understandingness seems to fall within the range of 25 to 30 percent. If one talks about average, or a little more than average, but no excessively, he or she tends to be perceived as attempting to understand the others' point of view. But if a person talks more than that, the perceived understandingness drops off sharply.

From this research and other studies cited above, it is clear that the vocal activity of individuals serves as a major mediating variable in group and interpersonal interactions. Further research is needed to explain the causes of these relationships, as well as to identify predictors and other consequences of that activity.


1H. W. Riecken, "The Effects of Talkativeness on Ability to Influence Group Solutions to Problems," Sociometry, 21 (1958), 309-21.

2D. J. Stang, "Effect of Interaction Rate on Ratings of Leadership and Liking," Journal of Personality and Social Psychology, 27 (1973), 405-08.

3B. Pope and S. Nudler, "Some Clinical and Sociometric Correlations of Interviewee Verbal Behavior," Proceedings of the 31st Annual Convention of the American Psychological Association, 8 (1973), 561-62.

4D. Willard and F. L. Strodtbeck, "Latency of Verbal Responses and Participation in Small Groups," Sociometry, 35 (1972), 161.

5E. Aiken, "Changes in Interpersonal Descriptions Accompanying the Operant Conditioning of Verbal Frequencies in Groups," Journal of Verbal Leaning and Verbal Behavior, 4 (1965), 243-247; M. Argyle, The Psychology of Interpersonal Behavior (Baltimore: Penguin Books, 1967); R. F. Bales, "The Equilibrium Problem in Small Groups," Working Papers in the Theory of Action, R. F. Bales and E. Shils, eds. (Glenocoe: Free Press, 1953); R. F. Bales, Personality and Interpersonal Behavior (New York: Holt, Rinehart and Winston, 1970); B. Bass, "An Analysis of Leaderless Group Discussion," Journal of Applied Psychology, 33 (1949), 527-33; B. Bass, "Situational Tests II: Variables in Leaderless Groups Discussion," Educational Psychology and Measurement, 11 ) 1951, 196-207; A Bavelas, A. Hastorf, A. Gross and R. Kite, "Experiments on the Alteration of Group Structure," Journal of Experimental Social Psychology, 1 (1965, 55-70; E. F. Borgatta and R. F. Bales, "Sociometry Status Patterns and Characteristics of Interaction," Journal of Abnormal and Social Psychology, 43 (1956), 289-97; R. Butler and E. Cureton, "Factor Analysis of Small Group Leadership Behavior," Journal of Social Psychology, 89 (1973), 85-90; L. F. Carter, W. Haythorn, B. Shriver and J. Lanzetta, "The Behavior of Leaders and Other Groups Members," Journal of Abnormal and Social Psychology, 46 (1951), 589-95; D. Hayes and L. Meltzer, "Interpersonal Judgements Based on Talkativeness: Fact or Artifact?" Sociometry, 35 (1972), 531-61; C. L. Jaffee and R. L. Lucas, "Effcts of Rates of Talking and Correctness of Decisions on Leader Choice in Small Groups," Journal of Social Psychology, 79 (1969), 247-54; J. B. Kirscht, T. M. Lodahl and M. Haire, "Some Factors in the Selection of Leaders by Members of Small Groups," Journal of Abnormal and Social Psychology, 58 (1959), 406-08; A. Knutson, "Quiet and Vocal Groups," Sociometry, 23 (1960), 36-49; R. Mann, "Dimensions of Interpersonal Performance in Small Groups under Task and Social Emotional Conditions," Journal of Abnormal and Social Psychology, 62 (1961), 674-82; Riecken; M. E. Shaw and J. C. Gilchrist, "Intragroup Communication and Leader Choice," Journal of Social Psychology, 43 (1956), 133-38; P. E. Slater, "Role Differentiation in Small Groups," American Sociological Review, 20 (1955), 300-10; Stang; F. L. Strodtbeck, "Husband-Wife Interaction Over Revealed Differences," American Sociological Review, 16 (1951), 468-73; Willard and Strodtbeck; S. Zdep, "Intra-group Reinforcement and Its Effects on Leadership Behavior," Organizational Behavior and Human Performance, 4 (1969), 284-98.

6Argyle; Bass (1949); Bavelas, Hastoroff, Gross and Kite; Borgatta and Bales; Hayes and Meltzer; B. Norfleet, "Interpersonal Relations and Group Productivity," Journal of Social Issues, 4 (1948), 66-69; Pope and Nudler; L. Simkins and J. West, "Reinforcement of Durations of Talking in Triad Groups," Psychological Reports, 18 (1966), 231-36; Slater; Stang; F. L. Strodtbeck and L. Hook, "The Social Dimensions of A Twelve Man Jury Table," Sociometry, 24 (1961), 397-415.

7Riecken; Jaffee and Lucas

8Hayes and Meltzer

9A. R. Allgeier, "The Effects of Differential Amounts of Talkativeness on Interpersonal Judgements," Diss. Purdue University 1974.

10D. Hayes and S. Sievers, "A Sociolinguistic Investigation of the Dimensions of Interpersonal Behavior," Journal of Personality and social Psychology, 24 (1972), 254-61.

11J. A. Daly, J. C. McCroskey, and V. P. Richmond, "The Relationship Between Vocal Activity and Perceptions of Communication in Small Group Interaction," Paper presented at the Annual Convention of the Speech Communication Association, Chicago, December, 1974.

12A. Mehrabian, "Communication Length as an Index of Communicator Attitude, Psychological Reports, 17 (1965), 519-22.

13C. D. Ward, "Length of Attitude Statements as an Indicator of Attitudes," Psychological Reports, 27 (1970), 398.

14Both Pearson r and eta correlations were obtained. Since each subject was asked to respond to four individuals, there was some error present that could be attributed to the repeated measurement. Consequently, the preferred analysis would have removed this error through intra-class correlation procedures. However, because the subjects did not respond to all the hypothetical sources, and because the ones to which they responded and the order of response were determined randomly, no known intra-class correlational procedure appropriate to the data was available. Therefore, the error attributable to the repeated measure was treated as random error, and may have resulted in lower estimates of relationship than actually existed in the data. Fortunately, since all of the tested relationships were statistically significant, this potential source of Type II error may be discounted in this study. The N used for estimating statistical significance in this study was 130, the number of subjects, rather than 518, the number of observations.

15N. N. Downie and R. W. Health, Basic Statistical Methods, 3rd edn. (New York: Harper and Row, 1970).


17Riecken; Jaffee and Lucas.

18Pope and Nudler.

Click Here To Go Back To PERIODICALS