Nonverbal communication of couples in conflict

Nonverbal communication of couples in conflict

JOURNAL OF RESEARCH Nonverbal IN PERSONALITY 253-269 (1981) 15, Communication of Couples in Conflict JOHN E. LOCHMAN Duke University Medical...

1001KB Sizes 25 Downloads 129 Views

JOURNAL

OF RESEARCH

Nonverbal

IN PERSONALITY

253-269 (1981)

15,

Communication

of Couples in Conflict

JOHN E. LOCHMAN Duke

University

Medical

Center

AND GEORGE University

ALLEN

of Connecticut

This article examined the rates, contextual meanings, and attributional meanings of nonverbal behavior occurring during role-played conflict between dating couples. Eighty couples reported perceptions of their own behaviors and their partners’ behaviors and were observed in a laboratory setting by trained raters. The nonverbal channel was used significantly more by females than by males and to express approval rather than disapproval. However, the nonverbal-disapproval summary category demonstrated weak validity. Eye contact had the most complex, unexpected contextual meanings. Reduced eye contact appeared to variously convey disapproval, cues of lower interpersonal power, and a reduction of intimacy to compensate for accompanying increases in other approval behaviors. Despite significant correlations between nonverbal and verbal behaviors, dating partners rarely based their perceptions of their behavior or their partners’ behavior upon their rates of nonverbal behaviors.

Assessment and treatment strategies are needed to help couples cope with interpersonal conflicts (e.g., Lochman & Allen, 1979). However, little relevant literature currently exists which describes how disapproval and approval are communicated within established couples in conflict. While the most obvious way in which people’s intentions, feelings, and evaluations are communicated is through their verbal statements, considerable information about their feelings and reactions presumably becomes transmitted through the nonverbal channel of behavior (Speer, 1972). It is likely that the choice of which channel to emphasize depends both on some individual difference variables (e.g., sex of partner) and the purpose of the communication (e.g., disapproval, approval). Requests for reprints should be sent to John E. Lochman, Ph.D., Division of Medical Psychology, Department of Psychiatry, Duke University Medical Center, Durham, NC 27710. 253 0092-6566/81/020253-17/$02.00/O Copyright @ 1981 by Academic Press, Inc. All rights of reproduction in any form reserved.

254

LOCHMAN

AND

ALLEN

Previous research has found females to be more adept than males at encoding variations of interpersonal liking and disliking through nonverbal cues in nondyadic laboratory settings (Buck, Miller, & Caul, 1974; Zaidei & Mehrabian, 1969). This superiority of females in nonverbal encoding was apparent regardless of the sex of the decoder. Zaidel and Mehrabian (1969) concluded that females were better encoders because they probably used nonverbal cues more often due to cultural training that encourages females to inhibit overt verbal expression of negative feedback. While this assumption of differential frequency rates of nonverbal cues for male and females has not been examined within cross-sex dyads, evidence does exist for this assumption within same-sex dyads of strangers. Thus, pairs of females have been found to engage in more eye contact in discussion groups (Exline, 1972) and more affiliative behavior (e.g., eye contact, head nods, pleasant facial expressions, verbal reinforcers) in a waiting-room situation (Mehrabian, 1971) than did pairs of males. Zaidel and Mehrabian (1969) also found that negative attitudes were more accurately communicated through the nonverbal channel than were positive attitudes. While Zaidel and Mehrabian (1969) found females to be significantly better than males in communicating variations in negative attitude, other researchers have found no sex differences in the ability to send pleasant versus unpleasant cues (Buck et al., 1974). Previous research has not indicated whether this evident superiority in transmitting disapproval nonverbally is associated with actual higher rates of disapproving behaviors in the nonverbal channel. A more basic concern than the specific rates of behavior is whether nonverbal cues have the same meaning to the actual participants in interactions. In reality, the nonverbal channel is “noisy” @peer, 1972). Nonverbal behaviors may be subject to different interpretations more often than verbal constructs and messages, which are themselves often imprecise. While a nonverbal behavior such as eye contact may be related to concepts of warmth (Bayes, 1972), affiliation (Mehrabian & Ksionsky, 1972), and social reinforcement (Patterson & Reid, 1970), it can also convey quite different messages in certain situations. Thus, Ellsworth and Carlsmith (1968) found that frequent eye contact, in conjunction with negative verbal messages, was negatively evaluated by a receiver; Frommes and Beam (1974) found that low-dominant males had negative reactions to a confederate’s frequent eye gaze, indicating that they interpreted gaze as a dominance cue; Ellsworth and Ross (1975) found direct eye gaze to promote intimacy between female dyads but reticence between males; and Ellsworth, Friedman, Perlich, and Hoyt (1978) found subjects to be satisfied with high rates of eye contact when in situations that elicited fear, but not in situations that induced embarrassment. Similarly, in some

NONVERBAL

COMMUNICATION

OF

COUPLES

255

research, female smiling has been found to be unrelated to the positiveness or negativeness of verbal messages (Bugental, Love, & Gianetto, 1971). In the present investigation, the contextual and attributional meanings of the frequency of couples’ nonverbal behaviors will be assessed. The contextual meaning examines how various nonverbal behavior codes relate to each other and to verbal evaluative behaviors. The behaviors from the Marital Interaction Coding System (Hops, Wills, Patterson, & Weiss, 1971), which is the rating system used in the present study, have been classified as Verbal Positive Social Reinforcenent, Nonverbal Positive Social Reinforcement, Verbal Negative Social Reinforcement, and Nonverbal Negative Social Reinforcement, on an a priori basis (Birchler, Weiss, & Vincent, 1975). The accuracy of these summary categories for couples in conflict has yet to be determined and will be investigated here. The attributional meanings of the nonverbal cues will be examined by correlating individuals’ stated perceptions of their own evaluative behavior and their partners’ behavior with the actual behavior rates that had occurred. In summary, the present study hypothesizes that higher rates of nonverbal behavior will occur (1) for females and (2) for disapproval. The contextual and attributional meaning of the couples’ nonverbal behaviors will also be examined. Couples will be observed interacting during roleplayed conflicts in a laboratory setting. METHOD Subjects A sample of 80 dating couples was recruited through announcements in upper-level university classes and newspaper advertisements. To participate, couples were required to (a) have known one another for at least 6 months, (b) interact at least twice a week, and (c) mutually agree that their dating relationship was the most important one they had at the time. All participants received $5.00 after completing the experiment. Subjects ranged from 16 to 24 years of age (Mdn = 19.0 years). The average dating dyad had known one another for 2.4 years and had been actively dating for the past 1.5 years. Partners usually interacted between five and six times weekly, with 81% of the couples reporting that they were exclusively dating their partners.

Procedure Every couple was tested separately during a 90-min session. After arriving, the couple was informed of the existence of observers located behind a one-way mirror, of the confidentiality of the data they provided, and of their right to refuse or terminate their participation. One couple decided to drop out at this point. The Inventory of Marital Conflicts Task (IMC) (Olson & Ryder, 1970) was described, using instructions that were slight modifications of those used by Birchler et al. (1975). For each of the IMC vignettes used in the present study, the causes of the conflicts in the vignettes were slanted in opposite directions on each partner’s form, although the basic details of the conflicts remained the same. Differences in the description of the situation

256

LOCHMAN

AND

ALLEN

would thus lead the male partner to infer that the woman was at fault for the conflict and vice versa. Each of the IMC vignettes used in the present study involved unresolved interpersonal conflicts. The vignettes did not indicate who the dominant partner was or who won the conflicts. All of the couples role-played the same vignettes. The vignettes included disagreements between individuals about whether they should seek couples’ counseling and about whether leisure time should be spent in joint activities or in watching a favorite television show by oneself. The IMC was used only as a basis for creating standardized interactions. The IMC scoring system for determining who won the conflicts was not used. To create more intense task involvement, couples were asked to role-play participants in the vignettes in a manner similar to the way Raush, Barry, Hertel, and Swain (1974) used issue-oriented improvisation scenes. The decision of whether to discuss or to role-play the conflicts was made after a pilot study with five couples was conducted. By role-playing, partners were found to engage more actively in direct criticism and approval of each other, rather than just discuss in a detached manner the conflicts of the characters in the vignettes. Couples were encouraged to identify how the themes of the conflicts were similar to issues in their own relationships. Throughout the role-played conflicts, the partners remained seated in two swivel chairs which were approximately 2 ft apart. The first vignette served as a warm-up for the couple, and observational data for this episode were excluded from analysis. The next two 4-min vignettes provided the behavioral data for the present study. At the end of this role-play session, partners were taken to separate rooms and asked to provide ratings of their own behavior and their partners’ behavior during the interaction. In a second study following a manipulation condition, the couples role-played several more vignettes. Details and findings from the second study can be found in Lochman and Allen (1979) and will not be discussed here.

Assessment Procedure and Instruments Subjecfive ratings. Subjects rated their behavior and their partners’ behavior on seven semantic differential items after the role-playing session. The questionnaire instructed the subjects to “Describe the way your partner reacted to you,” and “Describe the way you reacted to your partner.” These ratings were made once and were based only on the previous 8-min interaction. The same seven items were used for the self-attribution and the partners’-attribution ratings. Each item consisted of a scale ranging from 1 to 7 between a pair of bipolar adjectives. The items reflected the dimensions of activity (silent-talkative, passive-active), potency (helpless-powerful), and evaluation (cold-warm, acceptingcritical, attentive-inattentive, approving-disapproving) as described by Osgood, Suci, and Tannenbaum (1957). The first adjective in each of these pairs (silent, passive, etc.) was at the higher end of the 1 to 7 scales. A factor analysis of the participants’ ratings yielded the two expected factors for evaluation and activity, but found that the “powerful” item loaded on both factors, rather than representing a third separate factor. Behavioral observations. Four observers rated participant behavior by means of the Marital Interaction Coding System (Hops et al., 1971). using four summary categories derived from Birchler et al. (1975). Each summary category consisted of three specific behavioral codes. Verbal Approval consisted of behaviors indicating simple verbal agreement (AG, Agreement), brief praise (PA, Passive Approval), and extended positive feedback (AA, Active Approval). Nonverbal Approval included head nods of assent (AS, Assent), positive physical contact (PP, Positive Physical), and smiling (SM, Smiling). Verbal Disapproval consisted of Disagreements (DG), Complaints (CP), and Criticism (CR) directed at the partner, while Nonverbal Disapproval was coded when one partner ignored the other’s statement or question (NR, No Response), failed to keep eye contact (NT, Not Tracking), or changed facial expression or body position so as to “turn off’ the other (TO, Turn Off).

NONVERBAL

COMMUNICATION

OF

COUPLES

257

Four observer pairs were trained until a 70% rate of interrater agreement had been consistently achieved. The raters were not informed of the experimental hypotheses. The frequency of the behaviors of each couple were simultaneously recorded by two observer pairs in continuous 30-set intervals. Each behavior was counted only once between its initiation and its ending, no matter how long this behavioral cycle lasted. Thus, a lack of eye contact (NT) which lasted for 1 min was recorded as 1 NT. Since frequency was only one measure of these nonverbal cues, other methods of coding the cues, such as assessing their duration, could yield different patterns of results than found in this study. One pair of raters coded verbal behaviors by listening through earphones without being able to visually observe the couple. The second pair provided independent ratings of nonverbal behaviors by observing the couple without being able to hear the verbal exchange between partners. Throughout the course of the experiment, reliability for verbal codings averaged 74% with agreement averaging 86% for the nonverbal categories. For purposes of data analysis, the frequencies of the codings made by both raters in a pair over the last two vignettes were averaged.

RESULTS

Relationships

among the Behavioral

Categories

As shown in Table 1, the relationships between the 12 individual behavior codings and the 4 summary categories were examined by computing Pearson product-moment correlations between codes. In three out of the four instances (VA, VD, and NVA), the summary categories were significantly and usually more highly correlated with the three appropriate individual behavior codings within each summary category than with the codes that had been assumed to be within the other summary categories. These three categories appear to be validly constructed. The Nonverbal Disapproval category (NVD) demonstrated less divergent validity than the other summary categories for both males and females. Results for this category need to be interpreted with the most caution. While males failed to produce significant intercorrelations between VA, VD, and NVA, their NVD rates actually had positive unexpected correlations with VA and NVA. Females displayed anticipated significant intercorrelations between VA, VD, and NVA, but not with NVD. Of the 66 intercorrelations among the dozen individual behavior codes, females had 13 significant correlations, and males had 9. Both of these rates were above what would have been expected by chance, indicating that the verbal and nonverbal behaviors emitted by the couples were interrelated. However, 7 of the 9 significant correlations for males, and 3 of the 13 correlations for females, were in a direction that was opposite to the assumption of the a priori MICS categories. Most of the unexpected correlations occurred with the NT (lack of eye contact) and NR (No response) codes. However, the meaning of the correlations with NR was difficult to determine because of the very low frequencies of NR and the resultant potential unreliability of these correlations. Of all the codes, NT had the most and the largest significant correla-

PP

AS

SM

CR

CP

DG

AA

PA

AG

.I7 -.07

.02 .25*

.02 .03

AA

PA

-.13 -.13

.05 -.I5

.16 .03

DG

PRODUCT-M•

PEARSON

.06 .04

-.05 .Ol

-.07 -.I2

-.13 -.26*

CP

.24* .39***

.24* .29**

-.19 -.05

-.16 -.I6

.04 -.I4

CR

MENTCORRELATION~

-.19 -.06

-.14 -.24*

.12 .04

-.07 -.03

.06 -.02

.12 .27*

SM

MISC

BETWEEN

.09 .14

-.05 -.25*

-.I6 -.12

-.04 -.15

-.lO -.05

-.03 -.04

.18 .17

AS

codes

-.03

.oo

-.07 .20

-.04 .14

.18 -.Ol

-.06 -.16

.02 .20

-.05 -.07

- .03 .30**

PP

and summary

-.17 -.18

.I9 .23*

.49*** .19

-.08 .14

-.26* -.31**

.30** .02

-.I5 - .02

.06 -.09

.27* .13

NT

.25* -.04

-.04 .26*

-.04 -.12

-.lO -.09

-.lO .Ol

-.Ol .05

.ll -.07

.27* -.05

-.06 .02

NR

categories

.09 -.06

.41*** .05

-.02 -.Ol

.02 .23*

.oo -.14

.12 .25*

-.I8 .oo

-.lO -.04

.19 -.07

TO

-.03 .32**

.14 .13

.11 .23*

-.05 -.18

-.15 -.26*

.13 -.04

.27* .39***

.35*** .25*

.95*** .91***

VA

TABLE 1 BEHAVIORALCODESANDSUMMARYCATEGORIES,FORMALESANDFEMALES

-.02 .Ol

- .09 -.26*

-.06 -.ll

.74*** .86***

.40*** .64***

.79x** .60***

-.20 -.08

-.07 -.20

.lO -.17

VD

.19 .64***

.33** .33**

.94+** .84***

-.20 -.04

-.12 -.20

.09 -.09

-.OS .06

.04 - .06

.15 .38***

NVA

-.14 -.18

.23* .24*

.47*** .16

- .08 -.06

-.25* -.32***

.30** .10

-.I7 -.02

.05 -.09

.29** .14

NVD

*** p < .oOl.

* p -c .05. ** p < .Ol.

a Upper correlation in pair indicates males’ scores, N = 80. Lower correlation

NVA

VD

VA

TO

NR

NT .07 .07

.16 .I6

.11 .07

-.03 -.Ol

.22* .I4

.03 -.22*

.09 .18

-.08 -.03

.09 -.I9

in pair indicates females’ scores, N = 80.

.08 .I3

-.08 -.14

.12 .36***

.ll -.03

.Ol -.04

.46*** .lO

.46*** .08

.lO -.12

.23* .lO

.26* .45***

.lO .19

.99*** .95***

260

LOCHMAN

AND ALLEN

tions with other behaviors. For males, less eye contact was associated with more frequent disagreements (DG) as expected. However, NT was also found to have an unanticipated negative correlation with CP (Complaints) for both sexes, and positive correlations with verbal agreements (AG, males), smiles (SM, males), and head nods (AS, females). Rates of Nonverbal Behaviors

The mean frequencies of the 12 MICS specific behavioral codes and the four summary categories are presented in Table 2. The rates of Verbal Approval were substantially lower than the other summary categories. Three ratio scores were derived to test this study’s hypotheses. The Total Nonverbal Ratio was computed by dividing the sum of the two MICS nonverbal summary categories (NVA, NVD) for each partner by the sum of all four MICS summary categories (NVA, NVD, VA, VD). The Nonverbal Approval Ratio was calculated by dividing the Nonverbal Approval category (NVA) by the sum of both approval categories (VA, NVA). Similarly, the Nonverbal Disapproval Ratio was computed by dividing NVD by VD plus NVD. The means and t tests for these ratios are presented in Table 3. In general, over 60% of the behaviors used by partners to express approval and disapproval were nonverbal behaviors. Compared to males, females were found to emit significantly higher proportions of nonverbal beTABLE MEAN

FREQUENCIES

OF 12 MICS

2

BEHAVIORAL CODES AND FOUR FOR MALES AND FEMALES

Males Verbal approval (VA) Agreement (AG) Passive approval (PA) Active approval (AA) Verbal disapproval (VD) Disagreement (DG) Complaint (CP) Criticism (CR) Nonverbal approval (NVA) Smile (SM) Assent (AS) Positive physical (PP) Nonverbal disapproval (NVD) Not tracking (NT) No response (NR) Turn off (TO)

10

2 2 6

SUMMARY

CATEGORIES

Females

NONVERBAL

COMMUNICATION

MEANS AND t VALUES FOR

TABLE 3 FOUR RATIOS

OF

OF NONVERBAL

Mean Nonverbal ratio Males Females

.60 .68

Nonverbal ratio Approval Disapproval

261

COUPLES

BEHAVIORS

t value l.%*

.87 .43

12.29***

Nonverbal approval ratio Males Females

.85 .88

.84

Nonverbal disapproval ratio Males Females

.40 .46

.85

a df =78. * p < .05. *** p < .ool.

haviors, t(78) = 1.96, p < .05. When the Nonverbal Disapproval Ratio was tested against the Nonverbal Approval Ratio, it was found that, in comparison to disapproval behaviors, significantly greater proportions of approval behaviors were emitted nonverbally, t(78) = 12.29, p < .OOl. Tests for sex differences on the Nonverbal Approval Ratio and on the Nonverbal Disapproval Ratio were not significant. Self-Attribution

of Behavior

The correlations between the 12 MICS behavior codes and the 7 semantic-differential self-report items are presented in Table 4. Of the 84 comparisons between items, 22 reached statistical significance for females, and 15 for males, again well above a chance level. The four self-ratings for accepting, approval, cold, and attentive all correlate in expected directions with the frequencies of approval and disapproval behaviors. This indicates that couples are accurately aware of their rates of disapproval and approval behavior in conflict situations, at least in the verbal communication channel. Fewer correlations existed with the

nonevaluative items, particularly with the helpless-powerful self-rating. While females’ self-ratings were generally more related to their behavior than were males’ self-ratings, both sexes utilized nonverbal cues much less than verbalizations when making attributions of their behavior. Only 3 out of 15 significant males’ correlations and 4 out of 22 females’ correlations involved nonverbal behaviors. Of all the nonverbal cues,

262

LOCHMAN

ALLEN

TABLE 4 PRODUCT-MOMENT CORRELATION+ BETWEEN BEHAVIORAL SELF-PERCEPTIONS FOR MALES AND FEMALES

PEARSON

MISC codes

AND

Silent

Accepting

Helpless

Cold

Approving

CODES AND

Passive

Attentive

AC

.15 -.07

-.Ol .20

.22 -.05

.04 -.20

.l2 .29**

-.02 -.04

-.Ol .09

PA

- .06 .oo

.06 .25*

.03 -.07

-.I3 -.21

.26* .I7

.03 -.I3

-.04 .l6

AA

-.21 -.12

.I0 .32**

-.09 -.03

-.19 -.33**

.06 .26*

- .26* -.24*

DG

-.03 -.23*

-.24* -.26*

- .02 .02

.22* .25*

-.26* -.19

-.Ol .05

-.32** -.ll

CP

-.15 -.03

-.l3 -.27*

-.06 -.10

.17 .23*

- .08 -.31**

-.06 .05

-.09 -.l2

CR

-.35** -.23*

-.32** -.46***

-.30** -.13

.30** .32**

- .03 - .45***

-.24* -.05

.02 -.23*

SM

.37*** -.20

.08 .I4

.17 -.09

-.06 -.23*

-.05 .17

.26* -.21

-.05 .oo

AS

.04 .I2

-.04 .08

-.I1 .I5

-.05 -.I8

-.02 .I5

-.07 .20

-.I3 .I2

.oo .19

-.31** -.I1

.23* .25*

PP

-.18 -.ll

-.l8 .09

-.09 -.23*

-.I7 - .30**

NT

.18 -.lO

- .05 .05

.I0 -.Ol

.07 -.06

-.03 .09

.06 .04

NR

-.05 -.I1

-.14

.ll -.06

.06 .lO

.15 .02

-.08 .oo

-.Ol .12

-.08 -.04

-.09 -.14

- .03 -.07

.ll .02

-.lO .07

-.I8 .oo

TO

.oo

0 Upper number in pair indicates males’ indicates females’ correlations, N = 80. **

***

* p < .05. p < .Ol.

p < .OOl.

correlations,

-.04 -.06

N = 80. Lower

number

.07 .22* -.04 -.ll

in pair

NONVERBAL

COMMUNICATION

OF

COUPLES

263

partners appeared to rely most on their rates of smiles (SM) and positive physical contacts (PP) when reporting self-attribution of their behavior. Males and females appeared to interpret these cues differently, however, when they made their self-ratings. Males who were more silent and more passive had higher rates of smiles (SM) and fewer positive physical contacts (PP). Females who had more PP perceived they were more powerful, warm, and attentive. Their perception of their warmth was also positively associated with their rates of smiles. In direct contrast to males, females’ rates of smiles had nearly significant negative correlations with their self-perceptions of silence and passivity. Attribution

of Others’ Behavior

The correlations between the MICS codes and the seven partner-rating items are presented in Table 5. The total number of significant correlations is similar to that reported for the self-attributions in Table 4, with 18 significant male perceptions of female behavior and 17 significant female perceptions of male behavior. The partners’ ratings of each other generally follow the same patterns as the self-ratings. This congruence between self- and others’ ratings is particularly evident for both sexes for the Accepting and Cold items and for the correlations involving verbal behaviors. While agreement was found on 20 out of 25 (80%) of the significant correlations involving verbal behavior, only 1 out of 12 (8%) of the significant correlations involving nonverbal behaviors demonstrated similar agreement. As with self-perceptions, nonverbal cues were less related to the ratings of others than were verbal cues. Of the 17 significant correlations of females’ perceptions of male behavior, 5 involved nonverbal cues, while of 18 significant correlations of male perceptions of female behavior, only 1 included a nonverbal cue. Males attributed more warmth to partners who smiled more. Females perceived that partners who had less eye contact were more silent, more passive, and more inattentive. Lower rates of male PP were associated with females’ perceptions of more silence, and higher rates of TO (Turn Offs) were correlated with greater perceived inattentiveness. Females appeared to utilize somewhat different nonverbal cues when making attributions of their partners’ behavior (NT, TO, PP) than when making self-attributions (SM, PP). DISCUSSION

The results of the intercorrelations between MICS behavior codes indicate the ambiguity in meaning of several nonverbal behaviors and of the a priori MICS summary categories for nonverbal disapproval. The behaviors clustered in expected patterns more for females than males.

264

LOCHMAN

AND

TABLE PEARSON

MICS Codes

ALLEN

5

PRODUCT-MOMENT CORRELATION!? BETWEEN BEHAVIORAL OTHERS’ PERCEPTIONS FOR MALES AND FEMALES

Silent

Accepting

Helpless

Cold

Approving

CODES

AND

Passive

Attentive

AG

.09 -.13

.20 .10

-.07 -.02

-.27* -.08

.19 .24*

.Ol -.09

-.02 .06

PA

-.lO .oo

.17 .26*

-.12 .lO

-.17 -.17

.I5 .23*

-.12 .I2

.08 .17

AA

-.ll -.Ol

.09 .2t?*

.02 .lO

-.16 -.32**

.27* .17

-.14 -.04

.24* .22* -.38*** .21

DG

.09 .05

-.29** -.26*

-.17 -.12

.22 .39***

-.30** -.34**

-.Ol

.03

CP

-.07 .ll

-.21 -.24*

-.14 -.I7

.22 .37***

-.22 - .39***

-.Ol -.I8

-.ll -.13

CR

-.14 -.06

-.28* -.34**

-.28* -.12

.28* .28**

-.25* -.45***

-.I3 -.23*

-.26* -.31**

.15 .05

-.Ol .18

.19 .04

-.03 -.14

-.05

AS

-.Ol .16

-.Ol .05

-.08 .04

-.17 -.15

-.03 .12

-.07 -.04

-.Ol .I0

PP

-.23* -.13

-.12 .07

-.02 -.03

-.07 -.31**

-.08 .09

-.20 -.17

.15 .18

NT

.33** -.06

.07 .08

.ll

.04

.02 -.I3

.oo .09

.29** .oo

.oo .08

-.lO -.04

.03 .oo

.I1 .05

SM

NR TO

.06 .OS -.09 -.lO

.03 -.04 -.18 -.17

-.I2 -.13

-.13 -.18

.19

-.21 -.18

.lO .oo

-.13 -.03

-.16 -.07

-.22* .14 .13 .Ol -.26* .oo

cI Upper number in pair indicates correlations between males’ behaviors and females’ perceptions of males, N = 80. Lower number in pair indicates correlations between females’ behaviors and males’ perceptions of females, N = 80. * p < .05. ** p < .Ol. *** p < .OOl.

NONVERBAL

COMMUNICATION

OF

COUPLES

265

The unanticipated positive correlations between nonverbal disapproval and the approval categories were markedly stronger for the male partners. Males’ behaviors were generally more unrelated to each other than females’ behaviors were, and when males’ behaviors were related, they were more likely to indicate unexpected correlations. Therefore, males’ communication in conflict situations appears to be more difficult to interpret and may lead to more mixed behavioral messages. These weaknesses in the a priori categories reduce the strength of some of this study’s subsequent findings. The nonverbal behavioral category which demonstrated the greatest complexity and ambiguity in its apparent contextual meaning was lack of eye contact (NT). While the frequency of NT appeared to act as a nonverbal cue for disapproval on some occasions, the presence or lack of eye contact also conveyed other messages as well. Thus, when an individual gave brief statements of disagreement, he or she emitted little eye contact, as expected, but when a person gave more prolonged, intense verbal complaints to the partner, he or she actually increased the amount of eye contact. These higher rates of eye contact occurring in conjunction with high rates of verbal complaints appeared to serve either as an ingratiation tactic or as an attempt to assert greater power or dominance in the interaction, as suggested in previous research on eye contact (Ellsworth & Carlsmith, 1968; Frommes & Beam, 1974). When the rates of eye contact were analyzed in conjunction with the frequencies of the verbal and nonverbal approval codes, eye contact within the couples’ interaction appeared to be utilized for an even different purpose and to take on a different meaning. Lower rates of eye contact were found to be positively associated with more frequent smiles and brief verbal agreements for males, and head nods for females. These mixed findings suggest that lack of eye contact not only conveys disapproval, but also may provide some support for the compensation hypothesis of nonverbal cues advanced by Argyle and Dean (1965) and Patterson (1973). The compensation hypothesis predicted that when the threshold of one nonverbal reinforcement cue was surpassed (e.g., interpersonal distance became too close), the frequency of other nonverbal reinforcement cues (e.g., eye contact) needed to be decreased in order to avoid discomfort between dyad members. The discomfort was assumed to be generated by reinforcement that was too intense and intimate for the couple’s expectations within a specific situation. In the present study, eye contact did appear to be reduced to compensate for increases in some nonverbal behaviors (e.g., head nods, smiles) that indicated attention and approval. Rates of Nonverbal

Behaviors

Although there were many individual differences in the ways that dating couples demonstrated approval or disapproval toward each other, certain

266

LOCHMAN

AND

ALLEN

typical patterns emerged in these role-played conflicts. Females were found to communicate evaluative feedback proportionately more often in the nonverbal channel than did males, as predicted in the first hypothesis. This finding is consistent with previous research that has found females to be more skillful in encoding nonverbal cues. However, when nonverbal cues for approval and disapproval were examined separately, no significant sex differences emerged. This finding supports the Buck et al. (1974) conclusion that females were not significantly more nonverbal than males in their natural expressions of approval or disapproval separately. The comparative proportions of nonverbal behaviors used presumably to express approval versus disapproval were also examined. Contrary to the implications of previous findings that negative attitudes were more accurately and effectively communicated in the nonverbal channel than positive attitudes (Zaidel & Mehrabian, 1969), the dating couples’ proportions of nonverbal approval behaviors appeared to be much higher than their proportions of nonverbal disapproval. However, these results must be interpreted with caution, since weak validity was evident for the NVD category for both sexes and, to a lesser extent, for all the summary categories for males. The NVD behaviors may not be accurate indicators of disapproval. In addition, the generality of these high rates of NVA may be limited by the effects of either (I) the experimental setting and task, which required the couples to role-play conflict to hidden observers, or (2) the particular types of couples studied. Either of these factors may have led the dating couples to handle interpersonal conflict with unusual tact and sensitivity. Established marital couples may have more intense and frequent disapproval, and less approval, in their naturally occurring conflicts. In the context of this role-played conflict task, dating couples were relatively verbal and direct in their expressions of disapproval. Their high rates of nonverbal approval may have served to modulate the intensity of the conflicts, allowing the couples to disagree openly and criticize each other without escalating their anger too sharply. This use of nonverbal approval corresponds closely to descriptions of ingratiation tactics in dyadic interactions (Jones & Gerard, 1967). A person’s nonverbal approval behaviors can express the subject’s general attraction to his or her partner despite their specific disagreement. Since it is known that when dyad members are more interpersonally dependent on each other they use more ingratiation tactics, it would be expected that further research may discover higher rates of nonverbal approval behaviors with close, established couples than with dyads of strangers, or with couples whose relationships are troubled and distant. Attributional

Meaning

of Nonverbal

Behaviors

These results have indicated that, based upon the observed context of behaviors, the rates of certain nonverbal behaviors, such as smiles, head

NONVERBAL

COMMUNICATION

OF

COUPLES

267

nods, positive physical contacts, and frowns, appear to communicate approval and disapproval as expected, especially for females, and that lack of eye contact may communicate other messages. However, neither males nor females relied primarily on their rates of nonverbal cues to ascribe meaning to their own behavior. The dating partners appeared to be aware of their own rates of smiles and positive physical contacts only when they made certain self-attributions. Sex differences emerged in the meaning of these cues. While males perceived that higher rates of smiles were associated with greater passivity and silence, females tended to perceive that their own rates of smiles were positively correlated with higher rates of activity and talkativeness. While females perceived that they had greater power when they had more positive physical contacts, males perceived their own power was only associated with verbal criticism. This latter finding suggests that females utilized positive physical contact (PP) as an ingratiation tactic to augment their power, as previously discussed, while males relied primarily on overt verbalizations to exercise power. Other significant positive associations were found between males’ rates of positive physical contacts and their self-perceptions of greater passivity, between females’ frequency of smiles and their perceived warmth, and between females’ frequencies of positive physical contacts and their self ratings of warmth and attentiveness. In general, nonverbal cues appeared to have most relevance to males as indicators of their activity level, while females perceived SM and PP as related to their warmth, attentiveness, and power. Two possibilities exist to account for the couples’ relatively minor reliance on their other nonverbal cues of head nods, eye contact, and frowns when making self-attributions. Either (1) the partners lacked awareness of the frequencies of these nonverbal behaviors or (2) the frequencies of these nonverbal cues were generally irrelevant to the partners when they assessed how approving or disapproving they had been. The couples may have used other cues to develop these selfattitudes. The partners’ attributions of each others’ behavior does not resolve this uncertainty. Individuals were seemingly quite aware of how their partners interpreted their own verbal behavior. For example, while males did not appear to significantly utilize their rates of approval (PA, AA) or complaints (CP) in making self-attributions of how accepting they had been, they did use these same cues in their perceptions of how accepting their females partners had been, in a pattern identical to how females perceived their own behavior. Thus, while males did not significantly utilize certain verbal behaviors in their own self-attributions, they were accurately aware of these behaviors in others. This suggests that similar verbal behaviors have different relevance for males and for females as they make self-attributions.

268

LOCHMAN

AND

ALLEN

However, partners, especially males, utilized relatively few nonverbal cues in their attributions of each others’ behaviors. When frequencies of nonverbal cues were significantly associated with attributions, partners did not accurately perceive how nonverbal cues may have affected their partners’ self-attributions. For example, while females felt more powerful when they made frequent physical contact (PP) with their partners, males did not perceive that females with more PP had more power. Females perceived that positive associations existed between males’ lack of eye contact (NT) and males’ silence, passivity and inattentiveness, between males’ positive physical contacts (PP) and their talkativeness, and between males’ negative facial expressions (Turn Off, TO) and their inattentiveness. None of these correlations had been significant for males’ selfperceptions. The only similar meaning which individuals and their partners significantly ascribed to a nonverbal cue was the perceived warmth that was positively correlated with the females’ rate of physical positives. The issue of whether frequencies of nonverbal cues have less importance in attributional meanings of behavior is still unsettled. Since rates of nonverbal cues are utilized infrequently for both self- and others’ attributions, these cues may be relatively irrelevant for attributions of disapproval. However, since the contextual meaning of the nonverbal cues had been stronger, and since male and female partners had difficulty interpreting their cues in a congruent manner, the dating couples may be unaware of the potential meaning of their own and their partners’ nonverbal behavior rates. Further research will be required to clarify this important distinction. Additional research could also examine whether couples in conflict who have similar self-perceptions have different patterns of nonverbal behaviors than couples who have differing, complementary selfperceptions. Other fruitful areas of investigation include whether partners who “win” conflicts have different behavioral patterns from partners who “lose” and whether different patterns are evident during resolved versus unresolved conflicts. REFERENCES Argyle, M., & Dean, J. Eye contact, distance,

and

affiliation. So&merry,

1%5, 28, 289-

304.

Bayes, M. A. Behavioral cues of interpersonal warmth. Journal of Consulting and Clinical Psychology, 1972, 39, 333-339. Birchler, G. R., Weiss, R. L., & Vincent, J. P. Multimethod analysis of social reinforcement exchange between maritally distressed and nondistressed spouse and stranger dyads. Journal of Personality and Social Psychology, 1975, 31, 349-360. Buck, R., Miller, R. E., & Caul, W. F. Sex, personality, and physiological variables in the communication of affect via facial expression. Journal of Personality and Social Psychology, 1974, 30, 587-5%.

NONVERBAL

COMMUNICATION

OF COUPLES

269

Bugental, D. E., Love, L. R., & Gianetto, R. M. Perifidious feminine faces. Journal of Personality and Social Psychology, 1971, 17, 314-318. Ellsworth, P. C., & Carlsmith, J. M. Effects of eye contact and verbal content on affective response to a dyadic interaction, Journal of Personality and Social Psychology, 1968, 10, 15-20. Ellsworth, P. C., Friedman, H. S., Perlich, D., & Hoyt, M. E. Some effects of gaze on subjects motivated to seek or to avoid social comparison. Journal of Experimental Social Psychology, 1978, 14, 69-87. Ellsworth, P. C., & Ross, L. Intimacy in response to direct gaze. Journal of Experimentaf Social Psychology, 1975, 11, 592-613. Exline, R. V. Visual interaction: The glances of power and preference. In J. Cole (Ed.), Nebraska symposium on motivation, 1971. Lincoln: Univ. of Nebraska Press, 1972. Frommes, D. K., & Beam, D. C. Dominance and sex differences in nonverbal responses to differential eye contact. Journal of Research in Personality, 1974, 8, 76-87. Hops, H., Wills, T., Patterson, G. R., & Weiss, R. L. Marital interaction coding system. Unpublished manuscript, University of Oregon, 1971. Jones, E. E., & Gerard, H. B. Foundations of social psychology. New York: Wiley, 1967. Lochman, J. E., & Allen, G. Elicited effects of approval and disapproval: An examination of parameters having implications for counseling couples in conflict. Journal of Consulting and Clinical

Mehrabian,

Psychology,

1979, 47, 634-636.

A., & Ksionzky, S. Categories of social behavior. Comparative

Group

Studies,

1972, 3, 425436.

Olson, D. H., & Ryder, R. G. Inventory of Marital Conflicts: An experimental interaction procedure. Journal of Marriage and the Family, 1970, 32, 443-448. Osgood, C. E., Suci, G. J., & Tannenbaum, P. A. The measurement of meaning. Urbana: Univ. of Illinois Press, 1957. Patterson, G. R., &Reid, J. B. Reciprocity and coercion: Two facets of social systems. In C. Neuringer & J. L. Michaels (Eds.), Behnvior modification in clinical psychology. New York: Appleton-Century-Crofts, 1970. Patterson, M. L. Compensation in nonverbal immediacy behaviors: A review. Sociometry. 1973, 36, 237-252.

Raush, H. L., Barry, W. A., Hertel, R. K., & Swain, M. A. Communication, conflict and marriage. San Francisco: Jossey-Bass, 1974. Speer, D. C. Nonverbal communication of affective information: Some laboratory findings pertaining to an interactional process. Comparative Group Studies, 1972, 3,409-423. Zaidel, S. F., & Mehrabian, A. The ability to communicate and infer positive and negative attitudes facially and vocally. Journal of Research in Personality, 1%9, 3, 233-241.