THE RELATIONSHIP OF SEVERAL NONVERBAL BEHAVIGRS TO PSYCHOLOGECAL WSTANCE Thesis for the Degree of Ph; D. MICHIGAN STATE UNIVERSITY DIANE BORCHELT 1972 I III; IIIILIIII III I! III II III III II LIBRARY Michigan State; University This is to certify that the thesis entitled THE RELATIONSHIP OF SEVERAL NONUERBAL BEHAVIORS TO PSYCHOLOGICAL DISTANCE presented by DIANE JAY HUISINGA BORCHELT has been accepted towards fulfillment of the requirements for Ph.D. degree in PSXChOIOQY 9626/99; (U ' 77"“ k Major professor 0-7639 ABSTRACT THE RELATIONSHIP OF SEVERAL NONVERBAL BEHAVIORS TO PSYCHOLOGICAL DISTANCE BY Diane Borchelt This study was designed to investigate the relation- ship of several nonverbal behaviors (gesturing, fidgeting, comfort movements, body congruence, posture, and eye con- tact) to the subjective experience of psychological distance. Two hypotheses were made: I. Psychological distance will be inversely related to body congruence, eye contact, and body lean toward each other; that is, the less distance the 85 rate themselves as feeling toward the othe?, the more body congruency and eye contact they will have and the more their bodies will lean toward the other. II. Friends will rate themselves as closer than strangers and also they will exhibit more body congruence, eye contact and body lean in the direction of the other than strangers exhibit. Two groups, each composed of twenty dyads, were used. In one group, dyads were made by using gs who were friends; in the other group, dyads were composed of strangers. Ten dyads of each group had male gs while the other ten dyads were composed of female gs. Subjects in each dyad were asked to discuss ideas Diane Borchelt of abortion and abortion reform. During each discussion the interaction was both videotaped and audiotaped. Follow— ing the discussion, the audiotape was played back for the gs, and they were asked to recapture their feeling as much as possible. At one minute intervals, they were signaled to estimate their momentary, subjective experience of psycho- logical distance from their partner. Estimates were made using a five point scale of psychological distance. Later, raters observed the videotapes and recorded the behaviors being studied as they occurred. To test each hypothesis, a four way repeated measures analysis of variance was performed on the data. Neither hypothesis was confirmed. The results of the analysis for the first hypothesis yielded no significant main effects and only one significant interaction, the measures by sex by relationship inter- action. This interaction suggests that a complex relation- ship exists between psychological distance and nonverbal behavior which varies with the particular nonverbal behavior studied, the sex of, and the relationship between the mem- bers of a dyad. The results of the analysis of the second hypothesis yielded only one significant main effect, the effect of measures. Essentially this confirmed what was presumed, that the nonverbal behaviors used were different from each other. There were no significant interactions. Diane Borchelt Interpretations of the significant findings are suggested. Possible explanations for the lack of support of the hypotheses and evaluations of this study are dis- cussed and suggestions are given for future research. THE RELATIONSHIP OF SEVERAL NONVERBAL BEHAVIORS TO PSYCHOLOGICAL DISTANCE BY Diane Borchelt A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Psychology 1972 “)Qfix I’. 3 CF; ACKNOWLEDGMENTS I wish to acknowledge my committee members Dr. Andrew Porter, Dr. Gary Stollak, and Dr. Barry Wolfe for contributing their time and helpful ideas. I thank Dr. Porter for his patient explanations which enabled me to comprehend the statistical procedures used. I thank Barry for the moral support he often provided during moments of frustration. Special thanks go to my chairman, Dr. Dozier Thornton, who helped me develop and shape my ideas into testable hypotheses. I thank Bob Wilson for the many hours he spent conceptualizing, programming, and reprogramming until we finally finished the analyses. I thank him even more for his friendship. I thank my raters Dave, Elliot, Jerry, Rick, Terry, Tom and Rick. Special thanks go to my good friend Betty Schoepke for the time and energy she spent getting this typed and bound. I thank her most of all for her caring. I thank my beloved husband, Peter, and my many friends for the encouragement and support that they gave me. ii TABLE OF CONTENTS Page INTRODUCTION 0 O O O O O O O O C O O O O l Intimacy and Warmth . . . . . . . . . 6 Eye Contact . . . . . . . . . . . . 8 Physical Distance . . . . . . . . . . 13 Body Lean . . . . . . . . . . . . . 16 Body Congruence . . . Body Harmony with Speech . . . . . . . . 18 STATEMENT OF THE PROBLEM AND HYPOTHESES . . . . . 22 DEFINITIONS 0 O O O O O O O O O O O O O 24 Eye Contact . . . . . . . . . . . . 24 Body Movement . . . . . . . . . . . 24 Postures . . . . . . . . . . . . . 25 Body Congruence . . . . . . . . . . . 26 METHOD 0 O O O O O O I O O O O I O O O 28 Subjects . . . . . . . . . . . . . 28 Apparatus . . . . . . . . . . . . . 29 Data Recording . . . . . . . . . . . 29 Research Room . . . . . . . . . . . 3O PROCEDURE 0 O O O O O I O O I O O O O O 3 1 Data Gathering . . . . . . . . . . . 31 Data Rating . . . . . . . . . . . . 38 RESULTS 0 O O O O O O O O O O O O O O O 40 Subjects . . . . . . . . . . . . f 40 Description of Data . . . . . . . . . 42 Reliability . . . . . . . . . . . . 44 Mean Value . . . . . . . . . . . . 49 Collapsing Data . . . . . . . . . . . 50 Hypothesis I . . . . . . . . . . . . 51 Hypothesis II . . . . . . . . . . . 61 DISCUSSION 0 O O O O O O O O O O O O O C 65 BIBLIOGRAPHY . . . . . . . . . . . . . . 80 APPENDIX 0 O O O O O O O O O O O O C O 86 iii 10. LIST OF TABLES Sampling Information . . . Inter-Rater Reliability Correlations . The Range of Intra-Subject Correlations Percent Agreement for Each Behavior of Those Ss for Whom It Was Impossible to Perform Correlations . Correlation Matrices to Test for Homo- geneity of Body Posture Measures Repeated Measures Analysis of Variance: Hypothesis I . . . . . Univariate F Tests of the Interaction Between Relationship, Gender and Measures . . . . . . The Mean Correlation of Sample I and Sample II for Correlations Between the Mean of a Ss Nonverbal Behavior and His Mean Level of Psychological Distance . . . . . . Repeated Measures Analysis of Variance: Hypothesis II . . . . . Correlations Between N S's Estimate of His Level of Psychological Distance and His Partner's Guess of His Level of Psychological Distance . iv Page 41 45 47 48 52 56 58 62 64 93 LIST OF FIGURES Page List of Measures . . . . . . . . . . 43 Model of the Four Way Repeated Measures Analysis of Variance . . . . . . . . 55 Figure of the Relationship by Gender Interaction for Two Measures: Eye Contact Duration and Gesture Frequency . . 59 INTRODUCTION The field of communications is a complex multi- million dollar business. With a time gap of only seconds, one can speak with someone on the other side of the world; indeed, it only takes a few seconds to communicate with someone on the moon. Television programs are received almost instantly from the other side of the world via satellite. However, for all this SOphistication of our com- munication systems, there are many spouses who cannot con- verse with each other, there are many parents who cannot communicate with their children, and the leaders of the Nations do a dangerously inadequate job of speaking with each other. What is responsible for these communication difficulties? While a great deal of research and knowledge has been gathered about verbal communication, somewhat less has been gathered about what is perhaps our older, more primi- tivecommunication form--nonverbal behavior. Obviously there are many cultural differences in nonverbal communica- tion. Yet for decades, scholars have debated over whether or not nonverbal behavior was completely culture specific or whether there were common elements which cut across cultures. Ekman and Friesen (1971) believe they now have conclusive evidence that there are common elements of non- verbal behavior which are universal among humans. They present convincing data which suggest that all men express emotion through the same set of facial muscles, resulting in similar facial expressions to express a particular emo- tion. Kretch (1969) has data which suggest that there are also transcultural components of vocal intonation. In fact, it is possible that the expression of emotion through facial expressions and vocal intonation may even be an inheritence from our prehuman ancestors. It is interesting to read through articles describing animal communication, especially primate communication, and to notice the similar- ity of vocal intonation patterns and facial expressions between animals and humans (Van Hooff, 1962; Van Lawick- Goodall, 1968). Darwin (1896) suggested a similarity between animals and humans in terms of their manner of expressing emotion; he, too, pointed to the similarity of the facial musculature used for expressing particular emo- tions. While probably man's first mode of communication and still a very important medium, nonverbal behavior has been less well studied than verbal communication. Research on nonverbal behavior is varied and often complex, yet it seems as if we have just made a beginning in exploring it, and we probably are only partially aware of its variety of uses and modes of Operation. Nonverbal behavior functions in many important ways in our daily life: (1) it seems to have some regulative or signal functions for our spoken communication which help keep conversations running smoothly (Kendon, 1967); (2) it helps us express our emotions and communicate about the feelings of others (Darwin, 1896; Knapp, 1963); and (3) it marks us as a member of a particular nationality, social class, and status and helps us to interact smoothly with other members of our society (Hall, 1959). Empirical know- ledge of nonverbal communication patterns can be useful to pe0p1e by possibly improving their communication with others, giving them a better idea of their relationships with others, and aiding them to be more sensitive to and empathic with the moods and feelings of their relatives, friends, and associates and thereby probably bettering their relationships. More knowledge about nonverbal behavior can be especially useful to psychotherapists. Therapists have used and are using nonverbal behavior both consciously and subconsciously as communications of various types of infor- mation about their clients. Judgements about the type and intensity of a client's effect and the pertinence and truth of his verbal communications are often made from nonverbal behavior. Ekman and Friesen (1968) have made the following assumptions: (1) Nonverbal behavior can be considered a relationship 1an ua e, sensitive to, and the primary means of, SignaIing changes in the quality of an ongoing interpersonal relationship. (2) Nonverbal behavior is the primary means of expres— sing or communicating emotion, either because of the physiology of the organism or because of the priority of nonverbal to verbal behavior in the formative years of personality development. While the patient may state his feelings verbally, the presence of nonverbal cues which support or belie the verbalization may sometimes be crucial to the therapist's determination of whether the patient really feels what he says. (3) Nonverbal behavior has special symbolic value, expressing in body language basic, perhaps uncon- scious, attitudes about self or body image. (4) A fourth assumption emphasizes the metacommunicative function of nonverbal behavior to provide qualifiers as to how verbal discourse should be interpreted. (5) Nonverbal behavior is less affected than verbal behavior by attempts to censor communication (pp. 180-181). While the above statements are assumptions and not facts, the existing empirical evidence supports them (Ekman and Friesen, 1969, 1971; Kendon, 1967, 1970; Mehrabian, 1971; Scheflen, 1963, 1964). The investigation of "nonverbal behavior" covers a wide range of research activities. Included in the litera- ture about nonverbal behavior are a wide variety of nonlin— guistic modes of expression, including facial expressions, eye contact, gestures, other body movements, postures, body orientation, voice intonation, speech disturbances such as hesitations and "ah's," dress, physical distance between interactors, and the handling of time. Researchers have f I explored cultural differences and similarities in the ways various groups communicate emotions and behave nonverbally (Hall, 1959; Kretch, 1964; Ekman and Friesen, 1971). Other researchers have looked for relationships between nonverbal behavior and (1) personality characteristics (Exline, 1963; Mobbs, 1968), (2) relationships with and attitude toward others (Scheflen, 1963; 1964; Willis, 1966; Mehrabian, 1968), (3) status of other (Efran, 1968; Mehrabian and Friar, 1969), (4) affect (Darwin, 1896; Schlosberg, 1954; Ekman, 1965), and (5) mood (Dittman, 1962). Some researchers have even studied things such as room size and furniture arrange- ment to see what effect they have on human interaction (Dabbs, 1971; Mehrabian and Diamond, 1971). Some investigators (Birdwhistell, 1967; Condon and Ogsdon, 1966, 1967) have studied very minute segments of movements, whereas other investigators have studied complex patterns of nonverbal interaction (Scheflen, 1964). Infor- mation is gathered by a variety of methods from controlled laboratory experiments to clinically derived hypotheses. Amongst this panorama of data, an area of study which seems both interesting and important is the research which investigates how nonverbal behavior affects, regu- lates, and/or communicates about a relationship between ing individuals.//The research results concerning interact ‘ . W‘ " " ' \w-H—MM/. A‘Mvuu. pthese relationships are quite complex. In general, the literature suggests the following for interacting dyads: (l) the closer or more positive the relationship, the more eye contact there will be, the more alike the participants' body posture and body parts will be positioned, the more they will lean toward each other, and the closer they will be to each other; (2) high status individuals are looked at more than low status individuals; and (3) members of a dyad who are of equal status tend to sit closer together than members of a dyad who are of unequal status. In the follow- ing pages, research investigating nonverbal behavior and the relationship between interacting individuals is reviewed. Particular attention has been paid to studies concerned with the degree of positive feelings between the members of the dyad. Intimacy and Warmth From observations of videotaped interviews of therapists, clients, and sometimes clients' relatives, Scheflen (1963) noted that there appeared to be nonverbal cues between the interactors signaling when to become more or less intimate. In fact, there appeared to be a sequence of nonverbal cues which regulated cycles of intimacy between two or more of the interactors. His observations suggest to him that nonverbal cues may "regulate" periods of intimacy between people. Argyle and Dean (1965) have suggested that eye con- tact, physical proximity between peOple and the level of conversation (how personal) are factors which signify the level of intimacy between people. They have data which support the hypothesis that if one factor is increased, pe0p1e will decrease the level of another to retain a com- fortable level of intimacy. Both frequency and duration of eye contact between their §s decreased as physical proximity was increased. There was considerably less eye contact between mixed sex pairs. Also supporting their hypothesis was their finding that §S would stand closer to a person when his eyes were shut than when they were open. Kendon (1967) compared different points in conversations and discovered less eye contact when mutual smiling occurred. Smiling could also raise the level of intimacy of inter- action. Additional support for Argyle and Dean's hypo- thesis is the data indicating that an interviewer's inquir- ies concerning embarrassing personal topics led to less eye contact from normal gs than his inquiries concerning neutral topics (Exline, Gray, and Schuette, 1965). Aiello and Jones (1971) suggest that their data can be interpreted to support Argyle and Dean's (1965) hypo- thesis on a cross cultural level. They found that when interacting with a friend, Black and Puerto Rican children stood closer than Caucasian children. However, the angle of interaction appeared to be less direct for the Black and Puerto Rican children than for the Caucasian children. That is, Black and Puerto Rican children stood more at a sideways angle while Caucasian children stood more face to face. Reece and Whitman (1962) found that gs produced sig— nificantly more words in a free association experiment when the eXperimenter was "warm" rather than "cold." They defined warmth in terms of nonverbal behavior-~more eye contact with S, more forward body lean toward the g, more frequent smiling, and no finger tapping. Eye Contact Studies investigating eye contact have yielded data which is complex and suggests that eye contact is inter- related with several different factors. In general, however, the data indicate that greater eye contact implies a more positive attitude, a greater desire to interact, more liking, or at least less dislike between participants. Females have been found to engage in more eye contact than males (Exline, 1963; Exline and Winters, 1965; Ashear and Snortum, 1971). While Kendon and Cook's (1969) data support the above findings, they also indicate that this is subject to individual differences and there is no universal tendency for females to engage in more eye contact than males. Ashear and Snortum (1971) have found that for children in kindergarten through eighth grade, eye contact decreased as age increased. Extroverts have been found to engage in more eye contact than introverts (Mobbs, 1968; Kendon and Cook, 1969). Thayer's (1969) data indicate a relationship between greater levels of eye contact and dominance, at least at extreme levels. He had a confederate look into his subject's eyes for either fifty—eight seconds out of every minute or for only two seconds out of every minute. Subjects in the extended look group judged the confederate as more dominant than subjects in the brief look group. These subjects also had a lower estimate of how dominant they had been judged by the confederate than the subjects in the brief looks condition had. Exline and Messick's (1967) data indicate that dominant males reduced their amount of eye contact with listeners who gave few social reinforcements, whereas dependent males engaged in more eye contact with a listener who gave few social reinforcements than with a listener who gave many reinforcements. Competi- tive situations tend to decrease eye contact for high need for affiliation SS and increase eye contact for gs with high dominance needs; when a situation is cooperative and friendly, S3 with high need for affiliation scores engage in more eye contact than S5 with high dominance needs (Exline, 1963). Exline and Winters (1965) found that male gs shared significantly less mutual glances with an interviewer after they were negatively evaluated by him than before the evaluation, whereas gs who were positively evaluated 10 increased mutual glances (nonsignificantly) and gs receiving no evaluation maintained a similar level of eye contact. The half of the S5 in the no evaluation situation who indi- cated greater liking for the interviewer looked at him sig- nificantly more than the half who liked him less. In a second eXperiment, they found that female gs when speaking with two interviewers increased their eye contact with the preferred interviewer and decreased eye contact with the less preferred interviewer. Data for males were in this direction but insignificant. Male gs when listening reduced eye contact with both interviewers during the second of two periods, but they reduced it significantly less with the interviewer they preferred. Mehrabian (1968a) found that standing males engaged in more eye contact with liked per- sons than with disliked persons, but their results were insignificant for females. Sitting male communicators engaged in less eye contact with disliked listeners than with liked listeners regardless of the listener's sex; how- ever female communicators engaged in significantly less eye contact with disliked male listeners than with disliked females, liked females, or liked males (Mehrabian and Friar, 1969). ‘Ellsworth and Carlsmith (1968) found that femalqus engaged in a neutral or positive conversation felt more positively about both the interview and the inter- viewer if she looked at them frequently; however, if the conversation was indirectly yet constantly critical of g, 11 S felt more negatively about both the interview and inter- viewer if the interviewer had frequently looked at her. Exline and Eldridge (1967) found that a verbal message given with more eye contact was considered more favorable than the same message sent with less eye contact. Kendon and Cook (1970) found a positive correlation between the evalua— tion a subject received and the length of his gaze and a negative correlation between the frequency of his gaze and his evaluation, suggesting that people tend to dislike some- one who looks at them with short frequent gazes and tend to like one who looks at them with long infrequent gazes. Data from a study relating eye contact with five degrees of atti- tude toward a listener suggest a parabolic function between eye contact and attitude toward the listener (Mehrabian, 1968b). In other words, eye contact was least for disliked listeners, greatest for neutral listeners, and slightly diminished for well liked listeners. Status of the listener is usually positively cor- related with eye contact. Hearn's (1957) data suggests that eye contact is a parabolic function of the status of the listener. Eye contact is least with a low status listener, most with a moderately high status listener, and moderate with a high status listener. Efran (1968) found that fresh- man §s talking with two confederates had more eye contact with the confederate labeled senior than the confederate labeled freshman. Both male SS and female gs engaged in 12 more mutual glances with high status listeners than with low status listeners (Mehrabian, 1968a). Mehrabian and Friar (1968) found an interaction between sex, status, and liking and their effect on eye contact. Female gs had less eye contact with low status disliked listeners than with either low status-liked, high—status liked, or high-status disliked listeners. The means of the last three do not differ significantly from each other. Male SS engaged in more eye contact with liked high-status listeners than the other three groups. These groups did not differ amongst themselves. Kendon and Cook (1969) explored a variety of looking measurements, including eye contact. Using a total of fifteen subjects, four of those subjects interacted for a half hour with each of the other eleven subjects making a total of forty-four interactions. They discovered that gaze patterns were indeed a consistent aspect of a subject's behavior. In addition, there were high correlations between a subject's amount of eye contact, amount of gaze while listening, and amount of gaze while speaking. This finding suggests that the results of studies with eye contact may be more generally applied to other aspects of looking behavior, especially length of gaze while listening and length of gaze while speaking. While their data reveals that a sub- ject's gaze patterns are consistent, it also indicates that these patterns are somewhat influenced by the individual 13 with whom the subject is interacting. The findings reported above, while complex and not completely clear, strongly suggest that eye contact is related to the content of the interaction, the sex and per- sonalities of the interactors, their status in relationship to each other, and their attitudes and feelings toward each other. The data also suggest that, even though related to all of the above, gaze patterns are consistent for indi- viduals. In summary, research indicates that females, in general, engage in more eye contact than males, individuals with strong social interest will engage in much looking in non-competitive social situations while individuals with high dominance needs look more at others during competitive interactions, high status listeners will be looked at more than low status listeners, and individuals tend to increase eye contact the more they like or regard each other. Some data suggest that status and liking have a parabolic rela- tionship with eye contact, that is, a very high-status or well-liked individual may be looked at slightly less than a moderately high—status or well-liked individual. Physical Distance DevelOpmental studies suggest that children have begun establishing proximic behavior patterns at least by five or six years of age (Bass and Weinstein, 1971; Aiello and Jones, 1971). The physical distance between interacting ‘ , ,, .- . . ._...‘ --_‘_ . ‘ ‘v-- - --.—- ‘n _, 14 individuals is affected by both culture and the relationship mww' 0" '- “FM‘ m’ between individuals. Hall (1959) and Little (1968) found BONE“ . .-‘$~-|‘ differences in the physical distance between interacting individuals of different national groups. Aiello and Jones (1971) found differences in the physical distance in inter- acting dyads of lower class Black and Puerto Rican children and middle class Caucasian children. In terms of relationsh1p, individuals tend to stand MW‘""' “"" " 'V .1..-“ _..- ._. "“ “““'~~H—'——-'——u....._ .1... _. mm"..- W""--~~m closer the more they like each other, the better they know \1111L-~**“m‘" . each other, the less negative the interaction, and the more M‘Irnv-H" ‘ t equalytheir status. Mehrabian (1968b) asked S5 to role play five different attitudes toward a listener. SS decreased their distance linearly as they portrayed more positive attitudes. Mehrabian (1968a) found that gs assumed a person liked them more when he stood nearer to them and that SS stood closer to imagined liked versus imagined disliked individuals. Willis (1966) asked college students to measure the physical distance between themselves and others in their various interactions. His results 1ndicate that y" wax... ...M-¢..--:.5.3‘_.~3‘m° w. friends stand closest then acquaintances, and strangers M‘ “.1.”‘_WM“C£ Mm L “h ”13*” ._ the farthest. Peers stood closer to his 83 than did people v.1...“ _. older than his SS. Parents stood as far away as did strangers. In a cross-cultural study using gs from Sweden, the United States, Scotland, Greece, and Southern Italy, Little (1968) found that while the mean distances which various kinds of interactions were perceived to take 15 place differed between these groups, there was much agree- ment in the ordering of the distance for the relationships. That is, his SS placed dolls who were "friends" closer than dolls who were "acquaintances" and dolls who were "strangers" were placed furthest apart. Pleasant tOpics brought the dolls closer than neutral or unpleasant topics. Martin and Van Dyke (1968) related that women who had interacted with a friendly, accepting experimenter placed human figures closer together on a felt board than women who had inter- acted with a hostile, rejecting experimenter. Little (1965) had SS select distances between line drawings, silhouettes, and actresses for various relationships. gs chose smaller distances between interactors who had closer relationships. Role playing Ss sat closer to an individual when given approval-seeking instructions than when given approval- avoiding instructions (Rosenfelt, 1965). Guardo and Meisels (1971) found that children placed figures of a parent praising a child significantly closer together than figures of a parent reproving the child. Lott and Sommer's (1967) data suggest that several variables influence seating arrangements including the identification of certain posi— tions with level of status and the status of the partici- pants; peOple who are of equal status tend to sit closer together than people of unequal status. It appears that people stand closest to those they know well, have a posi- tive attitude toward, expect a positive attitude from, and 16 with whom they have equal status. However, if the physical distance becomes less than cultural and relationship limits prescribe, a negative atti- tude, confusion and discomfort may occur (Hall, 1959; Gar- finkel, 1964; and Felipe and Sommer, 1966). Yet, this seems to depend upon the age of the invader of one's per- sonal space. Using American Caucasian SS, Fry and Willis (1971) found that five year olds tended to receive a posi- tive reaction from strangers when they stood too close, eight year olds received a neutral or no reaction, whereas ten year olds received negative reactions. Body Lean The lean of the body either forward or backward suggests either a positive attitude or a negative attitude toward the other, respectively. James (1932) asked gs to determine what attitude was being expressed by the postures in photographs of a masked model. His data suggest that forward lean was interpreted as an approach, a posture indicating attention to the other, whereas a backward lean or a turning away was interpreted as a withdrawal, a posture indicating negative feelings. Both male and female listeners felt that a more negative attitude was communi- cated when a communicator leaned backward and away than when he leaned forward, toward them (Mehrabian, 1968a). Mehrabian (1968b) found that Ss attempting to communicate 17 different attitudes had similar postures for neutral or disliked listeners but decreased their angle of reclining for positive attitudes toward a listener. He also found that the sideway lean of a communicator was moderately high with disliked, liked, and intensely liked listeners and lowest for neutral listeners. Females had the most sideway lean with intensely disliked female or male listeners, but male communicators with an intensely disliked male listener had less sideway lean and less body relaxation than with liked male or female listeners or with disliked female listeners. Body Congruence Scheflen (1964) states that a person's posture and postural changes communicate about his relationship to the person with whom he is interacting. His observations sug- gest that body congruence of interacting individuals indi- cates harmony, thought congruence, and equal rank. Charney (1966) supports Scheflen, as his observations indicate that upper body congruence indicates a state of rapport or relat- edness between a client and therapist. In the videotaped interview of one client the congruent periods were correlated with lexical content referring to the other person as well as to the speaker; to positive statements of action, states of being, thought, and feeling; and to specific references to time and place. The noncongruent periods are signi- ficantly (p<.05) different: there was greater use of the first person and much less use of reference to others, there was greater use of negative statements 18 pertaining to action, thought, and feeling; and there was greater use of nonspecific temporal and locative references (p. 311). Dabbs (1969) has data which could be interpreted as sup- porting Scheflen's idea that body congruence is related to positive attitude between interactors. He had an actor interview male gs two at a time and mimic the postures and gestures of one of them. gs who were mimicked had a more positive evaluation of and were more persuaded by the actor. While Dabbs did not report whether or not the actor was blind to his hypothesis, he followed this study with a similar study in which "the actor [was] replaced by a series of gs serving as naive actor confederates." Again, mimicked gs believed that the actor would impress pe0p1e more favor- ably. Other results were insignificant but in the same direction as the original study. Body Harmony with Speech \ Condon and Ogston (1966, 1967) relate an interesting phenomenon. They have done very detailed, frame-by—frame analysis of videotapes of interacting individuals. In thirty frames of "normal" interaction they found that intensive analysis revealed harmonious or synchronous organizations of change between body motion and speech in both intra-individual and interactional behavior. Thus the body of the speaker dances in time with his speech. Further, the body of the listener dances in rhythm with that of the speaker (1966, p. 338). Kendon (1970) using a similar analysis of a video- tape of people conversing in a lounge of a middle class 19 London hotel, supports their finding that both the speaker and the listener's bodies move in synchrony with each other and with the speech. He describes phenomenon somewhat more clearly, "it is found that the points of change in the movement for the separate body parts coincide." He stresses that it is the changes in the flow of motion that coincide and that not all the body parts are necessarily moving in the same way or the same direction. Kendon suggests that the synchronization functions to facilitate the communica— tion process by signaling to the communicants that they are Open to each other. He suggests that the listener synchroni— zation functions to communicate his comprehension of what is being said. The above studies discuss how nonverbal behavior affects and/or is affected by the relationship between interacting individuals. Several aspects of a relationship were discussed-—the status of the interactors in relation- ship to each other, the extent of an individual's liking or positive attitude toward the other, and the level of inti- macy between interactors. An aspect of a relationship which is closely related to those above but which has not been discussed directly is psychological distance. English and English (1958) define psychological distance as "the degree of difficulty a person eXperiences in psycho- logical relationships with another, or the subjective estimate of that difficulty. The difficulty may be due to 20 a sense of difference in manners, ideology, personality or status; or to an awareness of unapproachability or unfriendliness" (p. 159). Psychological distance can be seen as a continuum stretching from feeling very close and "with" another person to feeling very distant and aloof from someone. This author conceptualizes psychological distance as a subjective experience which fluctuates across time. It depends upon the feeling state of the individual toward the other and his experience of their relationshp at that moment. Psychological distance seems to be an impor- tant feature of relationships. The clinical observations of the author and her colleagues suggest the following about the concept of psychological distance: it seems to be especially important in the relationship between therapist and client. It is discussed between some therapists and clients to help them estimate what is happening between them and why. Psycho- logical distance between therapist and client often varies from session to session and, frequently, even within a session. Therapy sessions seem to be facilitated when the therapist and client feel close to each other. Neutral feelings or moderate distance may reflect that little of therapeutic nature is occurring. Great distance or aloof- ness seems to indicate some kind of struggle or crisis between the therapist and client which may or may not be therapeutic, depending on the source of the distance (the 21 therapist getting to the core of things versus the therapist putting the client down) and how the situation is handled. Because of the importance of psychological distance and lack of research about it, it seems as if research in this area would be valuable. An important and interesting problem would be to study the relationship between psycho— logical distance and nonverbal behavior. The following research investigates this relationship; eye contact, body congruence, various postures, and various types of movements are the nonverbal behaviors which were studied. STATEMENT OF THE PROBLEM AND HYPOTHESES The problem is to investigate the correlations between different levels of perceived psychological distance and the nonverbal behavior of college students interacting in same-sex dyads. Psychological distance is defined as a subjective experience of feeling which ranges from feeling very close to another person, caring about him and sensing his caring about you, to feeling very distant and aloof from someone. The following hypotheses are believed to be valid for populations of Western cultures in general. However, they were tested only for same-sex dyads of college students. I. Psychological distance will be inversely related to body congruence, eye contact, and body lean toward each other; that is, the less distance the £5 rate themselves as feeling toward each other, the more body congruence and eye contact they will have, and the more their bodies will lean toward each other. II. Friends will rate themselves as closer than strangers, and also they will exhibit more body congruence, eye contact and body lean in the direction of the other than strangers. Due to the exploratory nature of this research it 22 23 was decided to accept p<.05 for the significance level required over the total experiment. This permits us to accept either of our two hypotheses as significant if the test reached or surpasses the p<.05 level of significance. DEFINITIONS Eye Contact What was measured was those occasions when both subjects (gs) had their eyes oriented in the direction of the eyes of the other; the rater recorded those occasions when it looked as if both gs were looking each other in the eye. While the measure is not precisely eye contact and while it is probably less accurate than measures obtained from direct observation, it has been labeled eye contact because it is believed to be similar or the same phenomena that many of the other researchers measure when they study eye contact. Body Movement Three types of body movements were measured. (1) Gestures--purposeful movements of the arms, hands, or head which facilitated the expression of an idea or attempted to express an idea themselves. (2) Comfort movements-—movements which seemed to be made for the purpose of making g more comfortable. These included changing one's body position scratching, and taking hair out of one's eye. (3) Fidgeting --any movement of the arms, hands, or feet which did not seem purposeful such as wiggling the legs or feet; playing with one's hair, clothes, or hands; or tapping one's finger. 24 25 Postures Five types of postures were measured. These did not include what is considered a "normal" sitting position (back against the chair's back, buttocks occupying most of the chair, and body relaxed). (l) Slouching--these were postures in which the body was obviously relaxed and some- what slovenly. One example would be the position in which one's back was resting against the back of the chair while one's buttocks were resting on the front half of the chair. Another example would be the position in which one was leaning to the side resting his head on his hand and his buttocks were not directly underneath but to the side of his head. (2) Body erect--these were postures in which the body looked stiff or rigid as if the person were tense or was being quite formal. An extreme example would be sitting erect in one's chair without resting one's back against the chair's back. This posture did not occur at all. (3) Side— way lean toward other--these were postures in which the upper half of the body was shifted to the side closest to the other. (4) Sideway lean away from the other--these were postures in which the upper half of the body was shifted to the side away from the other. (5) Forward lean-- these were postures in which the person was leaning forward in his seat; his upper trunk formed an acute angle with his thighs. 26 Body Congruence Body congruence was measured in two ways: (1) four parts of the body were measured separately for congruence and (2) total body congruence was measured. Total body congruence occurred when both gs of a pair were in identical or similar postures and the positions of their limbs were similar. (1) Legs--the legs were considered congruent if the legs of both subjects (g1 and g2) of a dyad were in similar positions such as if both crossed their legs at the knee, both crossed their legs at the ankle, or both had uncrossed legs with knees and feet close together. (2) Arms-—the arms were considered congruent if both arms of El are in identical or similar positions to both arms of §2 or if one arm of g is identical to one arm of g and their other arms 1 are not grossly dissimilar. Examples would be both gs resting their arms on the arms of the chair, both folding their arms across their chest, and both touching their face with one hand while one has his other hand resting on his lap and the other has his arm resting on the arm of the chair. (3) Posture--posture was considered congruent if both g1 and g2 were in a similar position such as if both were in any of the postures defined above or both were in a "normal" sitting position. (4) Head--the heads were con- sidered congruent if both El and g2 had their heads in a similar position; for example, both heads were cocked to 27 the side, erect, or bent down. Both straight replications of body position and mirror images (right side of g1 did what left side of g2 was doing) of body position were con- sidered body congruence. Unfortunately some of the above definitions are not as Operationally specific or precise as would be desired. This required the raters to use some degree of subjective judgment when deciding whether or not a behavior fit into a particular category. However, this researcher feared that making definitions more operationally specific would have led to either definitions that were so long they would have been confusing or definitions so rigid that some valuable observations would have been discounted. The reliability data suggests that the definitions were precise enough that both raters were generally measuring the same thing. '1 METHOD Subjects The SS were eighty undergraduate students attending Michigan State University in Spring Term 1971. The eighty gs were comprised of four groups: (1) ten male dyads com- posed of strangers, (2) ten male dyads composed of friends, (3) ten female dyads composed of strangers, and (4) ten female dyads composed of friends. The Ss were selected in the following manner: A class list of the introductory psychology course was obtained. Each of the 530 students on the list was given a number according to his place on the list. Two mutually exclusive series of random numbers were selected from a table of random numbers. One series was used to select students for the group of friends and the other to select students for the group of strangers. The eXperimentor (E) telephoned those students whose num- bers corresponded with those selected from the table of random numbers. E asked those persons contacted to parti- cipate in her research with either a stranger or a friend. If the student did not agree to participate in the condition for which he had been selected, he was not asked to parti- cipate in the other condition. This was believed to bias the sample less than including a number of pe0p1e in each 28 29 group who had refused to participate in a group to which they had been randomly assigned. gs agreeing to participate with a friend were scheduled at a time convenient for both himself and their friend. Dyads of strangers were created chiefly by the scheduling procedure, i.e., gs who could participate during the same hour made a dyad. Apparatus The apparatus consisted of two audio recorders (a Wollensak 3M 5730 stereo tape recorder and a Revere T3000 monaural tape recorder), a videotape recorder (Ampex VR5100) and a Cohn video camera. Two Esterline Angus event recorders were used to record the frequency and duration of the occur- rence of the behaviors studied. Five boxes containing two buttons each were attached to each recorder (total of ten boxes) so that each button displaced one particular pen when pressed. Data Recording Seven advanced undergraduate social science students observed the videotapes and recorded the data. They were informed that the study concerned nonverbal behavior but they were not told that psychological distance was a vari— able until they had finished rating all of the videotapes. The audio portion of each tape was turned off so that the verbalizations and voice tones could not bias the raters' decisions as to what behaviors were or were not present. 30 Research Room A complex of three adjoining rooms was used. The small (10 feet by 15 feet) outer room contained videotape and audiotape equipment. It was not carpeted, but the other two rooms were. To enter the research room one had to travel through the equipment room. The research room was 10 l/2 feet by 15 feet. It contained the videotape camera, a desk, two straight back arm chairs, a small table and a file cabinet. The chairs were placed facing the camera with their backs to the desk. A wall of windows was to the left of the S5. The windows reached from the ceiling to 2 1/3 feet above the floor. The third room was used only to ask the S proverbs to determine whether they were left or right eye movers (Bakan, 1969).* This room, typically used for group therapy, contained eight to ten chairs and a large round coffee table. Two chairs were positioned so that they were facing each other. The §_was positioned so that his back was to the windows. *The data about left and right eye movers was gathered for use in a future study. It is not directly related to either hypothesis in this study. It is reported here because steps were taken to obtain these data during this research and thus it needs to be mentioned to accurately report the procedure used. PROCEDURE Data Gathering Students from the introductory psychology course 1"" l were randomly selected and telephoned by E_who repeated the t ‘1 following upon contacting a student: Hello. My name is Diane Borchelt. I'm a graduate student in psychology presently working on my disserta— tion. I've randomly selected your name from the psycho- I] . *__4_.‘.‘1_'.;o 1 logy 151 class list because I'd like you to participate ‘ in my research. I'm studying relationships between two v people. Participation would take about an hour to an hour and fifteen minutes of your time. During the first half hour you would be videotaped while talking with a friend/stranger about a selected subject. During the remaining time you would answer questions about your discussion. You would receive two to three half hour credits for your participation. Would you be willing to participate? Students who were willing to participate with a stranger were given appointments and directions to the experimental site. The second person to form a dyad was given the first person's name and asked if he knew the person. If he replied affirmatively, he was given a new appointment with a different partner. Students who agreed to participate with a friend were asked to find a friend of the same sex and of a similar age who would be willing to participate with them. Students who could find a friend were asked to select several times when both could participate. In most cases it was necessary for E to call a second time in order to insure that the potential S could find a friend who would participate and to schedule 31 32 the pair. In a few cases, the student asked his roomate or someone nearby and the arrangements were made during the initial telephone call. When Ss arrived at the experimental site, S intro— duced herself. With pairs of strangers she also introduced them to each other if they had not already introduced them- selves. She asked them whether they knew each other. In every case the reply was negative. In two cases an S said he had "seen the other person around" but had never talked to him. S then led the Ss into the experimental room and asked them to be seated. She asked the Ss to share with each other their views on abortion and abortion reform as openly and honestly as they felt comfortable doing under the eXperimental conditions. She asked them not to move their chairs since they were positioned for the camera. The Ss were told that S would leave the room, set up the equipment, and then knock on the door signaling them to begin. Twenty-five minutes later S would again knock on the door signaling them to stop. S then asked the S5 if they had any questions. The most frequently asked question was "what if we can't talk about the topic for twenty-five minutes?" Ss were told to continue talking even if they could not continue on the selected topic. S suggested that they talk about a related tOpic if possible, but if not they should talk about anything fun or interesting. Other questions were answered if they pertained to clarifying the __ _ ._...r_.__.__. _. A. 33 task. Questions about the nature of the research were not answered. It was explained to the Ss that the question would not be answered to avoid the possibility that the information would affect the experimental results. S then left the room, closed the door, turned on the equipment, and then knocked on the door signaling the S5 to begin. After she knocked, S started a stOp watch. Twenty- five minutes (plus a few seconds) later, S knocked on the door again signaling the S3 to stOp. S then turned off the equipment, unplugged the audio tape recorder, and carried it into the research room. She then asked the S5, "How did it go?" as she set the recorder down, plugged it in and began rewinding it. S then gave the Ss the following instructions: "I'm studying what I consider to be two very important com- ponents of interpersonal relationships, emotion and psycho- logical distance. Psychological distance is defined by English and English (1958) as: The degree of difficulty a person experiences in psycho- logical relationships with another, or the subjective estimate of that difficulty. The difficulty may be due to a sense of differences in manners, ideology, person— ality or status, or to an awareness of inapproachability or unfriendliness (p. 159). This definition seems to describe only the negative side of psychological distance. I see psychological distance as a continuum ranging from feeling very distant and aloof from someone, to feeling neutral, to feeling very close and 'with' another person. One of the things I'd like you to do today 34 is to estimate your subjective experience of psychological distance from each other during your interaction. "I have arbitrarily assigned a five point scale to psychological distance. I would like you to read this now." S then gave each S a copy of the psychological distance scale (see appendix). While they were reading, S finished rewinding the audiotape of their interaction, found the correct place to Start the tape by locating the sound of her knock on the door, and turned on the second tape recorder to warm up. When both Ss had finished, S asked them if the scale "made sense." On a few occasions one or both Ss had a question about the scale. These were answered to their satisfaction. Most Ss agreed that the scale did make sense to them. S then asked them if they had any questions. Again, questions were answered if they concerned clarifica- tion of the task. Questions about the nature of the research were not answered. It was eXplained to the S5 that answering their question could possibly bias the results. S then continued with the instructions. "Through the use of a second tape recorder, I will divide an audiotape of your interaction into one minute seg- ments. The second tape is a recording of my voice saying 'one,' 'two,' 'three,‘ and so on every sixty seconds.* In *This was done to signal Ss when to estimate their level of psychological distance. 35 other words, after sixty seconds I say 'one,’ sixty seconds later I say 'two' and so on. I will replay an audiotape of your interaction. I would like you to try and get back into your interaction as much as possible and remember how you were feeling, particularly how you were feeling toward your friend/partner. "There are three things that I would like you to do at the end of each minute. First, I would like you to circle the number which most closely represents how you were feeling toward your friend/partner. Then I would like you to take a guess about how you think your friend/ partner was feeling about you. You can write that number to the left or beneath or anywhere--just so you're consist— ent. "In the blank along side of each set of numbers, I would like you to write down any emotion or feeling you become aware of experiencing during a minute. This could be anything such as anxiety, uncomfortable, nervous, angry, happy, bored, etc. Don't worry about searching your soul or struggling to find something to put in, just put down feelings and emotions as they come to you. Don't worry if you leave many blank, but do fill the blanks when you can.* "If possible, please don't let stereotypes interfere *This data may be used as a pilot for future research. It was not analyzed for this study as it does not directly relate to either hypothesis. 36 with your ratings of psychological distance. What I mean by this is, various emotions are often associated with cer- tain levels of psychological distance. Anger, for example, is often thought to be distancing-—and often it is-—but sometimes it can bring two people together in close inter- action. Happiness is typically thought of as a warm, close experience, and often it is. However, happiness can be a very personal experience and could cause one to draw inside himself to his own thoughts, and, therefore, become more distant from his companion. So please, try to estimate psychological distance in itself without involving any other feeling which may be present. Are there any ques- tions?" (Again questions concerning clarification of the task were answered while any concerning the nature of the research were not.) After giving the Ss their instructions, S asked one S to turn around and use the desk and the other S to use a little table also in the room. In this way the Ss had their backs to each other. S_made sure both had pencils or pens and then turned on both audiotapes and left the room. After about two minues S returned and checked both S5 to see whether they were following the correct procedure. Ss who were having difficulty had part of the instructions repeated. This occurred infrequently, but on some occasions Ss needed to be reminded to guess what their companion was feeling. 37 S then left the room; she returned briefly once or twice to see that things were going smoothly. When the audiotape of the dyad's interaction was finished, S re-entered the room and turned off both audiotapes. She told the Ss she had two more things for them to do, both of which were quite brief. She gave one S a brief question- naire (see appendix) about the interview as a whole and asked the other to accompany her into another room. There she asked him to interpret five proverbs (see appendix) and watched his eyes, recording the movement after each proverb was given. This was to determine whether they were right or left eye movers (Bakan, 1969). S then had the Ss trade places so that each S performed both tasks. Ss were then asked to sign one of two releases, one permitted the tapes of their interview to be used for this research but then required that they be erased; the second permitted the tapes to be stored and used for other confidential research. For those §§ who needed it, S signed their cards giving them credit for research participation. When there was time, S showed a few minutes of their videotape to those Ss interested in seeing it (about 60 percent of the Ss). S then thanked the S3 for their participation. Any S who asked a question about the nature of the research was given S's telephone number and told to call S during the last week of classes or final exam week when she will expect to have all the data and will at that point be glad to share 38 the eXperimental procedure with them. Data Rating The videotapes were observed and rated twice, once with both Ss shown in order to obtain measures of eye con- tact and body congruence and once utilizing two videotape receivers, each with a different S covered by a sheet of paper, to measure postures and movements.* It is believed that observations can be more accurate if a rater can con- centrate on only the S he is rating, without seeing the other S who may be moving or distracting the rater in some way. The five raters (S) randomly varied the behaviors they measured from tape to tape so that all Ss measured all of the behaviors. For the condition in which both Ss were viewed, one S rated eye contact, a second S rated total body congruence, another S rated head congruence and arm congruence, the fourth S rated posture congruence and leg congruence, and the fifth S measured one of the above to serve as an inter-rater reliability check. For the condi— tion in which one S on each receiver was covered, one R * Generally (about 65 percent of the tapes), body con— congruence and eye contact were rated first, and then postures and movements were rated. Often when the tape was long enough to contain data for two pairs of Ss, both pairs were rated consecutively with congruence and Eye contact measured first from both dyads and then the tape was rewound and then movements and postures were rated for both pairs of 85. In general, the order of the tape rating was done for Eonvenience rather than in an ordered or organized pattern. 39 measured movements for a S while another S measured postures for him, a third S measured m0vements for the other S while a fourth S measured postures for him and the fifth S measured one of the above for the interpersonal reliability check. When an S saw one of his assigned behaviors, he pressed the appropriate key until the behavior terminated. This provided a record of the frequency of a behavior and its duration. Rater's were given three training sessions before they began rating the data. RESULTS Subjects In the manner described above, 80 Ss were selected for the four groups: male friends (mf), female friends (ff), male strangers (ms), and female strangers (fs). Sampling information is presented in Table l. The Ss in the stranger group were indeed strangers to each other, and the S5 in the friends were indeed friends. Only a few pairs of strangers stated that they had "seen each other around." All S5 in the stranger group agreed that they had never spoken to their partner before. All pairs of female friends had known each other for at least eight months. All but one pair were living together, sharing either a room in a dormitory or the same dormitory floor. That one pair had previously lived together on the same dormitory floor. While twelve Ss reported having a friend who was equally close, only two Ss reported having a friend here at Michigan State University who was closer than their partner in this research. All but one pair of male friends reported knowing each other for at least eight months. That one pair had known each other for only six weeks. However, during that six weeks they were rooming together. All but one pair of SS lived together, either in the same house, the same 40 41 .pwumu on ow Moog oou mHmB mommuompfl> HHmQB «« l.mocmuwH© HMOHmOHonommm we mHm>wH Hflmnu mum“ on mamas: mgmB pcmflum no: pom m menu pwnu Hoom 0m mm3 mdmuoflpsm one; .moe ¢.H .moa m.HH .moe m .moe a ..mum ma ..mnm ma ..mum ma ..mum om mm 0 mm m w m c m: Ml Mfi NM WA Umuomucoo mucmpsum mo quESc Hmuoe dl Ml dl .ql Bonm uoc pap umcuumm mQIUHSOB Hflmnu pan ucmE lucflommm uflmzu umwx 0:3 mucmpsum mo Hmflasz o o v H awn» Spas mummflofluumm page: 033 pcmfluw m pcflw uoc pasoo pun mummflo Iflunmm on powwow 0:3 mucmpsum Mo Hogan: one m o o o pmpmmc uo: wum3 mmnu mmswomn pomscs muwa usn powwow 0:3 mucmpsum mo Hmnfisa mze w o H o usmEucHommm Macaw mwox uoc UHU gab mummflofiuumm cu powwow 0:3 pmuomucoo mummUSpm mo Hogans mae o .«e *H o mama manmmscs mm: pan emumafloflu Inna 0:3 pmuomwcoo mucmpspm mo HOQEDG mge a «a ma v mummfloflpumm 0» wmmsmmg 0:3 pmuomucoo mucmpsum Mo Hogans obs om om OH OH mm mm 66m: mumz 0:3 pmpomucoo mucwpspm wo Hones: one muwmcmnum mummcmuum mpcmflum mccmflum mHmEmm mam: meEmm mam: ZOHBfiZmomZH UZHAmZ¢m H mflm<fi 42 dormitory floor, or the same room. This one pair shared the same professional interests. While fourteen of these Ss reported having a friend who was equally close to them as their research partner, only two reported having a friend here at Michigan State University who was closer than their partner. Description of Data Each pair of Ss yielded forty-four bits of data per minute (Figure 1). As data were collected over twenty—five minutes, there are eleven hundred bits of data per pair of Ss. Forty of the forty—four bits of data are measures of thirteen nonverbal behaviors; these are eye contact, five congruence measures (arms, legs, head, posture, and total body congruence), three body movements (gesture, fidget, and comfort) and four types of body lean (toward, away, forward, and slouch). Seven of these nonverbal behaviors (body move- ments and body leans) yield data for each S; and thus for each pair of Ss, fourteen behaviors are measured. However, eye contact and congruence are shared by two people, thus for eye contact and the five congruence variables each S has the same score as his partner. For each of the above twenty behaviors there are two measurements, the frequency of the behavior and its duration. Thus we obtained forty measures. The other four bits of data include a psychological distance self score and a psychological distance guess score for each S. 9. ll. 13. 15. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 43 Psychological Distance 2. Self S Psycholagical Distance 4. Self S2 Eye Contact Frequency 6. Psychological Distance Guess S Psychological Distance Guess S2 Eye Contact Duration Congruence Measures Total Body Congruence 8. Frequency Legs Frequency 10. Arms Frequency 12. Head Frequency 14. Posture Frequency 16. Body Postures Total Body Congruence Duration Legs Duration Arms Duration Head Duration Posture Duration Subject 1 Leaning Toward Frequency 31. Leaning Toward Duration 32. Leaning Away Frequency 33. Leaning Away Duration 34. Leaning Forward Frequency 35. Leaning Forward Duration 36. Body Slouch Frequency 37. Body Slouch Duration 38. Subject 2 Leaning Leaning Leaning Leaning Toward Frequency Toward Duration Away Frequency Away Duration Leaning Forward Frequency Leaning Forward Duration Body Slouch Frequency Body Slouch Duration Body Movements Subject 1 Fidgeting Frequency 39. Fidgeting Duration 40. Gesture Frequency 41. Gesture Duration 42. Comfort Frequency 43. Comfort Duration 44. Fig. Subject 2 Fidgeting Frequency Fidgeting Duration Gesture Frequency Gesture Duration Comfort Frequency Comfort Duration l—-List of Measures 44 Reliability To determine reliability each nonverbal behavior was measured twice for about one-fourth of all Ss. Between rater reliability was calculated for each of the nonverbal behaviors in two manners. The first method was done by correlating the total score (its sum across minutes) of a S's nonverbal behavior as measured by the first rater (R1) with the total of his behavior as measured by the second rater (R2). Thus these correlations were done across Ss; they are inter-subject correlations. A summary of these reliability correlations is presented in Table 2. Basically this method of obtaining reliability fits these data quite well and there were few problems in acquiring the reliabil- ity coefficients in this manner. There is one point worth commenting on, however. There is an additional correlation for both frequency and duration data for the body postures, leaning away and slouching. These adjusted correlations were made because it appears that for two Ss, R1 and R2 interpreted the same behavior differently. Their ratings for the S matched very well but R1 labeled the behavior lean away and R2 labeled the behavior slouch. Thus, an additional correlation for these four measures was performed by switching R2's slouch data with his away data. These are the correlations in parentheses. The second method of obtaining reliability measures intra-subject reliability. The correlations were performed 45 TABLE 2 INTER-RATER RELIABILITY CORRELATIONS Inter-Subject Mean Intra-Subject Correlations Correlations N* Frequency Duration N* Frequency Duration Fidget 20 .4561 .7948 20 .389 .502 Gesture 20 .9113 .9744 20 .762 .747 Comfort 20 .5994 .8940 18 .524 .609 Lean Toward 20 .7376 .7118 8 .508 .834 (.6311) (.9791) (.853) (.830) Lean Away 20 .6160 .2325 9 .909 .978 Lean Forward 20 .7183 .9572 7 .595 .827 (.9956) (.9495) (.880) Body Slouch 20 .6376 .6186 4 .741 .931 Total Body Congruence 11 .1827 .7964 8 .463 .586 Arms 8 .7918 .9139 8 .439 .724 Head 8 .1780 .4377 7 .251 .318 (.739) (.912) Legs 14 .1451 .8734 9 .792 .944 (.780) Posture 14 .1962 .9654 12 .528 .747 Eye Contact 10 .848 .834 10 .444 .534 * For the behaviors Fidget through Slouch, N denotes the number of S rated twice. For the congruence variables, N denotes the number of pairs of S rated twice. 46 across time; for each S, each minute of nonverbal behavior as measured by R1 was paired with the nonverbal behavior of the same minute as measured by R2. All correlations were then changed to Z scores, the Z scores of each behavior were added across Ss and divided by N to obtain the mean, and the mean Z score was retransformed back into a correla- tion. These correlations are also reported in Table 2. Table 3 gives the range of these correlations. They repre- sent the mean correlation for a measure of its intra-S correlations. The data were somewhat less suitable for the intra- subject analysis and several problems arose. One of the problems was that some behaviors did not occur at all for particular Ss. Thus it became impossible to perform the correlation. Table 4 presents percent agreement for each behavior of those S5 for which it was impossible to perform correlations. The percentage was found by using the per- centage of minutes for which R1 and R2 were in agreement that the behavior did not occur at all . Another difficulty is that spuriously low correlation appeared when either (1) the behavior occurred very infre- quently for that S, or (2) the behavior was almost con- stantly there. This is because a correlation with a near constant is always low. For example, the correlation for pair #29 was only +.0852 for posture duration, yet both raters rated the behavior as almost constantly there, R1 47 .‘1‘51‘ Fete-cm H H NNNN. oomo.- N H NNNN. NNNN.- OH Homucoo mNm N N NNNN. NNHN.- m N NONN. NNNH.- HH musumom N N NNNN. NHNo.u N N oooo.H NHNH.- N mNmH N o NNNN. NNNH. m 0 «Hum. omNo.u N mmmm N N ONNN. HNNN.- m H NNNN. NONN.- N NEHN H N NNHN. HNNH. N N NHON. NNNN. N mocmsumcoo Nmom Hmuoe o H NNNN. NNHN. o H ONNN. ONNN. N nosoHN o N NNNN. NNNN. H N NNNN. NNNH. N mumzuom H N oooo.H ONNN.- H N oooo.H NONN.- N mmza N N NNNN. NNNH. H H NNHN. HNNH.- N mumzoe N N NNNN. mnno.- m N NNNN. Nmno.u NH uHoNEOo N NH NNNN. NNNo.- H N NNNN. NNHH. ON musummo N N NNNN. HHoH.- HH N NHNN. NNNH.- oN ummmHN NN. Ne. ENHN 30H NN. mp. gmHm 30H z Bonm m>on< mmcmm Bonm m>on< mmcmm mama coflumuso mumo mucmsvmum. How mcofiumamuuoo How mcoflumamuuou mZOHBm3¢ UHmBOB nonon pum3uom >m3¢ UHmBOB mumm coHumHsQ mpmo wocmswmum mmmDm mo mNm>Hmcd mmusmmwz pmummmmm mmz Hsom may mo Hmpoznl.~ cowumuso uuowEou wocmskum unemfiou coHumuso musummo mocmsqmum musummo coHumuso umOOHN aocmsqmum ummpflm :oNumuso nonon wocmsgmum noson coflumuaa mumBuom wocwswmum pumSuom :oHumuso >m3¢ aocwsgmnh >m3< coNumusn oum3oe mocmsgmum pumzoe coNumuso uomucoo mxm hocmsvmum uomucoo mxm coflumuso Esw mocmsumcoo hocmsvmum Esm mocmsumcou .0: uomhnsm .oc Hflmm Hmpcmo QNmeoHumHmm mummcmnum .mflm IIJ 1234 90 1234 90 11.234 90 1234 90 o o 12 2222 O 34 4444 o. .56 6666 o 78 sO.s.O SHO SOOSHO sea snss.s SON Oosnss .Os 1 2 0 l 2 0 l 2 O l 2 O o . o l l .1. o 2 2 2 o o o 3 3 3 o 4 P P p. P P P P P P P p. P mHmEmm mez mHmEmm mam: mpcmHHm 56 TABLE 6 REPEATED MEASURES ANALYSIS OF VARIANCE: HYPOTHESIS I Sources of Variation SS df MS F Relationship (R) .068 1 .067553 1.525036 Gender (G) .032 1 .032483 .733316 Relationship by Gender (RG) .012 1 .012298 .277632 Error (S:RG) 1.596 36 .044296 Measures (M) .377 17 .022166 .971596 Relationship by Measures (RM) .463 17 .027260 1.144880 Gender by Measures (GM) .601 17 .035356 1.544790 Relationship by Gender by Measures (RGM) 1.002 9* .058939 2.593457** Error (MS:RG) 13.962 316* .022814 Total 18.113 719 .025191 *As determined by Epsilon **Significant at p<.05 The interaction was significant only when using the liberal estimate of the degrees of freedom but was not sig- nificant when using the conservative estimate of the degrees of freedom. Thus the correct degrees of freedom needed to be estimated in order to make a decision about the signifi- cance of the interaction. Using a variance covariance -1Av‘ § v . u 2......__._1 -_ »_ _ 57 matrix, epsilon* was found. Multiplying the liberal degrees of freedom, by epsilon, the estimate of the legiti- mate degrees of freedom were found to be 9 and 316. Using three degrees of freedom, the interaction did prove to be significant. A series of two way unvariate analyses of variance were then performed on these R to Z transformed variables to discover what variables contributed to the significance of the interaction. Because there were eighteen variables, to keep the overall level of the test significant at the .05 level, each individual variable had to be significant at ( least at the .002777 or YLEET'level. None of the variables met this criterion (Table 7). Thus, while our measure, by relationship, by gender interaction proved significant, the individual measures responsible for this significance cannot be determined with any assurance. However, the variables eye contact duration and gesture frequency approach signifi- cance suggesting that further studies of these variables may prove valuable. Graphs of the interactions of these vari- ables are presented in Figure 3. This interaction suggests *In an analysis of variance, an assumption made about the repeated measures is that pairs of observations within 85 are equally correlated. If this assumption is not met, the test of our interaction which uses the liberal degrees of freedom is too liberal. Epsilon is a correction- coefficient which estimates the extent to which the assump- tion was not met. Its use reduces the degrees of freedom for the F of reference making the test more conservative. 58 TABLE 7 UNIVARIATE F TESTS OF THE INTERACTION BETWEEN RELATIONSHIP, GENDER AND MEASURES Variable P Less Than Congruence Sum Frequency .6495 Congruence Sum Duration .2408 Eye Contact Frequency .9198 Eye Contact Duration .0079 * Toward Frequency .1873 Toward Duration .8554 Away Frequency .9942 Away Duration .2693 Forward Frequency .1204 Forward Duration .5695 Slouch Frequency .0960 Slouch Duration .3424 Fidget Frequency .1101 Fidget Duration .0390 Gesture Frequency .0090 * Gesture Duration .0244 Comfort Frequency .6095 Comfort Duration .3739 *Variables which approach significance. To be sig- nificant P needs to be less than .0028. 59 Nocmsqmum musummo mam coNumHso uomucou mmm “mmnsmmmz 039 Hem cofluomumucH Hmpcmo wn mflcmcoflumamm mnu mo musmfimn:.m ; . x." 11 v L , . 1. . a ‘1 I' l‘ , I Vg . #l 0... mm x x mm. . m2. V HNO ONO mocwsgmum musumww omflm N . H N: z \ HNO.0 2O. 0 mm coflumusa uomucou mum 00. No. we. mo. mo. OH. NH. sezoos z s,13qSTJ 60 that for male strangers and female friends the greater the level of psychological distance (the more aloof) the more eye contact interacting dyads exchange, and the more gesturing that occurs. However, the data for male friends and female strangers were in the direction of the hypotheses; the greater the level of psychological distance the less eye contact dyads shared and the less gesturing occurred. It would be interesting to see if further studies of these variables yielded the same type of interaction. The correlation of the grand mean was -.001. It is apparently not different from zero so no significance test was performed on it. Thus, the analysis of alternative one lends no support to the hypothesis. Alternative two of hypothesis I, that there is a cor— relation between a S's overall rating of psychological dis- tance and the kinds of and/or amount of nonverbal behavior he exhibits, was tested in the following manner. First two samples of Ss were formed by randomly placing one member of each pair in each sample. This was done because members of a pair could not meet the criteria of independence necessary for the correlations since it is likely (or at least pos- sible) that one member of a pair influenced the behavior of the other. Five, twenty—by-twenty correlation matrices were computed for each sample. The twenty measures include frequency and duration measures for each of nine nonverbal variables, psychological distance self, and psychological 61 distance guess. The five matrices include one matrix for each of the four groups (MF, FF, MS, FS) and one matrix which was computed across all S5. The correlations of the nine nonverbal behavior with psychological distance self were pulled from each of the matrices. Through the use of a conservative (df=l/2N-l) and liberal (df=N-l) confidence interval procedure, the correlations presented in Table 8 were tested to see if they were significantly different from zero. The test variable was the sum of the two sam- ples (after they had been changed to Fisher's Z scores). The conservative test assured perfect correlation between the Fisher's Z's while the liberal test assured no correla- tion between the Fisher's Z's. None of these correlations were high enough to be considered significant. Thus, the null hypothesis is accepted for alternative two of hypo- thesis I. Sypothesis II Friends will rate themselves as closer than strangers, and also they will exhibit more body congruence, MMHL H eye contact and body lean in the direction of the other, than strangers exhibit. “ I The mean value per minute as described above was the basic unit of this analysis. These scores were averaged for both Ss of a pair as again we could not make the assump- tion that one's behavior was independent from the other. THE MEAN CORRELATIONS OF SAMPLE I AND SAMPLE II FOR TABLE 62 8 CORRELATIONS BETWEEN THE MEAN OF A SS NONVERBAL BEHAVIOR AND HIS MEAN LEVEL—OF PSYCHOLOGICAL DISTANCE n Frequency Duration Eye Contact m friends 10 .400 .215 f friends 10 .065 -.136 m strangers 10 .172 -.090 f strangers 10 .138 -.050 total 40 _.003 -.094 Toward m friends 10 -.095 -.236 f friends 10 -.156 -.013 m strangers 10 .125 -.280 f strangers 10 —.059 -.l98 total 40 .038 -.106 Away m friends 10 _.110 .112 f friends 10 .267 .324 m strangers 10 .331 .171 f strangers 10 .300 —.034 total 40 .270 .189 Forward m friends 10 -.128 -.388 f friends 10 .130 .125 m strangers 10 —.271 -.231 f strangers 10 -.063 —.132 total 40 .022 -.125 Slouch m friends 10 -.098 -.212 f friends 10 .090 .090 m strangers 10 .201 .186 f strangers 10 .164 .164 total 40 .170 .139 Fidgit m friends 10 _ 051 -.187 f friends 10 -.052 -.184 m strangers 10 .189 -.277 f strangers 10 .104 .222 total 40 -.067 -.036 Gesture m friends 10 .280 .294 f friends 10 —.235 .322 m strangers 10 .377 .340 f Strangers 10 .050 .338 total 40 .105 .248 Comfort m friends 10 -.068 .112 f friends 10 -.267 —.178 m strangers 10 .106 .181 f strangers 10 .022 —.111 total 40 -.016 .029 Congruence Sum m friends 10 -.070 .093 f friends 10 - 023 —.110 m strangers 10 .011 .009 f strangers 10 —.031 .042 total 40 -.091 —.l35 63 In order to facilitate the variables being in a common metric each measure was divided by its standard deviation. A four way repeated measures analysis of variance was performed on the data as described above. ‘Figure 2 is also a model of this analysis as it is essentially the same analysis as the one used in alternative one of hypothesis I. The differences are (l) in this analysis the mean value per minute of each measure was used, whereas in hypothesis I the correlation of each measure with the level of psycho- logical distance was used; and (2) in this analysis the variables psychological distance self and psychological distance guess were added. The sources of this analysis are presented in Table 9. The results of this analysis yielded only one significant main effect. This was the effect of measures. Essentially, this confirms what was already pre- sumed: the nonverbal measures used are different from each other, e.g. eye contact duration differs from gesture frequency. This analysis yielded no significant inter- actions. 64 TABLE 9 REPEATED MEASURES ANALYSIS OF VARIANCE: HYPOTHESIS II Source of Variance SS df MS F Relationship (R) 774.919 1 774.918915 .443286 Gender (G) 4497.177 1 4497.177032 2.572578 Relationship by Gender (RG) 1416.522 1 1416.522415 .810311 Error (S:RG) 62932.330 36 1748.120282 Measures (M) Relationship by Measures (RM) Gender by Measures (GM) Relationship by Gender by Measures (RGM) Error (MS:RG) Total 2030652.299 21 23795.035 21 40313.339 21 15559.047 21 1388067.380 756 3568008.048 879 96697.728539 52.665658* 1133.096896 .617132 1919.682796 1.045540 740.907020 .403529 1836.067962 4059.167291 *Significant at p <.05 _-..._ ._ ,“‘;:3-. ‘ u‘ ‘1"! n' . DISCUSSION Neither of the hypotheses of this research were supported. However, there was one significant interaction in the analysis of variance performed to test hypothesis I. This was the most complex interaction of the study; it was the relationship by gender by measures interaction for the correlations of the nonverbal behaviors with psychological distance. This relationship would suggest that the sex of an interacting dyad, the relationship of its members to each other (either friends or strangers), and their level of psychological distance at the moment, are all related to their nonverbal behavior. In addition, this interaction suggests that these variables interact in a complex way and that they are related to different nonverbal behaviors in different ways. Because this was an exploratory study, results were accepted as significant at the p .05 level. Since there were two hypotheses tested, each with several main effects and several interactions in their analysis of variance, it is possible that the one significant interaction found in this research is a function of chance rather than a true relationship. This possibility seems even more likely when one considers that it was not possible to determine which 65 _..._‘u.s-_b-‘ o. 1.. 3L! 66 variables contributed to the significance of this inter- action. Therefore, these data must be interpreted cautiously. Since one cannot determine which behaviors are affected when the levels of psychological distance change, it is difficult to make any interpretations of the data which have a practical application in the "real world." Basically all that can be said is similar to what has been stated above; the data suggest that in same sex dyads of college undergraduates, the sex of the participants, their relationship, and their momentary level of psychological distance will interact and influence their nonverbal behavior. However, it will be related to various nonverbal behaviors in various ways. At this time there is insuffi- cient data to say exactly how or what nonverbal behaviors are affected by the above variables. While this finding has few meaningful implications or applications for either society or the individuals in it, it does have some important implications for further research. The fact that there was a positive finding sug— gests that further research on the relationship between psychological distance and nonverbal behavior would be valu- able. Several methods of improving future research will be discussed below. The data also suggest that the measures frequency of gesture and duration of eye contact would be particularly important measures to study in future research. F—h H ‘ .. I An. n‘ d)» F ._ ‘1‘. 1...ka 67 Other possibly fruitful measures to study include duration of gesture and duration of fidget.* There are several explanations as to why the first hypothesis was not supported. The simplest and perhaps the most parsimonious is that it is incorrect. The data indi- cates that psychological distance is not inversely related to eye contact, body lean toward the other, and body con- gruence. Our analyses would suggest that psychological distance has no simple linear relationship to the nonverbal behavior studied. Perhaps psychological distance is not A related to nonverbal behavior at all. However, there are several other explanations as to why this research did not support the hypothesis and why only one significant inter- action was found. It was possible that the lack of significant findings is related to the manner of obtaining levels of psycho- logical distance. This could have created problems in *These measures are also suggested because for them p<.05. When eighteen variables are tested using p<.05 to make a decision to accept a result as significant, it becomes very likely that at least one variable will pass this criterion merely by chance. If several of these vari- ables reach this criterion, it is impossible to determine which (if any or all) have been judged significant due to chance and which reached it due to a true difference. In this research, to avoid making a Type I error and to main- tain the overall level of the test significant at the p<.05 level, each variable had to reach p<.05/18 or .0027 to be judged significant. However, this increased the possibility of making a Type II error. To this author it appears that any measure for which p<.05 may be valuable to study in future research. 68 several ways. Our Ss may have been naive in the sense that they may have been relatively unfamiliar with the concept of psychological distances at least in terms of thinking about it in relationship to themselves and others. Being unac- customed to thinking in terms of psychological distance might have made the task somewhat difficult and ratings less accurate, especially at first. Some Ss may have even been unaccustomed to introspection of any sort and thus r._k._.__~‘--..—a ‘1‘:- :» Lane L this may have been a very novel task for them. This could have led to less accurate results. It is recommended that future studies utilize Ss who may be more SOphisticated in terms of psychological distance or at least in terms of introspection. If possible to obtain, clients and thera- pists would probably make a good population to study as they typically are quite trained in introspection and they .usually are trained in evaluating their relationships with others. Training a sample of college students in estimat— ing their level of psychological distance before performing the research would be another alternative to consider. A second problem in estimating the level of psycho- logical distance in this study was the lack of variability along the psychological distance scale. Although a five point scale was used,only 11.25% of the S5 used any fives at all and almost half (46.25%) of the Ss used only a range of two or three units along the scale. (11.25% of the Ss 69 used the full range, 42.50% of the Ss used a range of four, 27.50% of the Ss used a range of three, and 18.75% of the Ss had a range of only two.) Correlations across a small range tend to be low. If there had been a wider range across the psychological distance scores it is possible that the correlations of psychological distance with the nonverbal behaviors would have been larger. It is suggested that in future research, attempts are made to obtain a wider range in their psychological distance scores. Since they would allow for more variation, a seven point scale or a five point scale permitting half point scores may be more valuable for correlational research on psychological dis- tance than a mere five point scale. There are a variety of eXperimental manipulations that one could attempt which may facilitate the Ss using the full range available or at least a wider part of it. One could pretest Ss on attitude scales or personality variables and then pair them for agreement and disagreement or for compatibility and incompatability. SS could be requested to talk on several selected topics rather than just one, perhaps selecting one topic upon which they agree and one upon which they disagree. Married couples discussing both topics about which they agree and topics about which they disagree would most likely produce a wide range of levels of psychological distance. A confederate could be used who would agree with 1/2 the Ss and disagree with the other. A confederate could be warm and accepting 70 with l/2 the Ss and cool and aloof toward the other 1/2 of the S5. S5 could be given one tOpic about which they were told to disagree and another topic about which they were to agree, regardless of their true beliefs. Another possibility is that the lack of significant findings may be related to the conceptualization of the problem. It is possible that the way in which psychological distance was operationally defined in this study was inade— quate. The definitions of the different levels of psycho- logical distance may have been interpreted differently by different S5 or they may not have understood them as well as this researcher had been led to believe. Reports of sub- jective states may be inconsistent or unreliable or at least an inappropriate measure of psychological distance. On the other hand, it may be that the concept, itself, is so elusive to measurement that it is a meaningless concept for research purposes. While psychological distance is a meaningful con- cept to this author, and at least to a few other psycho- logists, it is possible that it has little or no meaning for the general pOpulace. An attempt was made to provide some support or validity for the concept of psychological distance as it was Operationally defined in this study. This effort met with limited success. A positive correlation between a S's estimate of his level of psychological distance and his partner's guess of 71 his level of psychological distance would suggest that the concept has some validity, as it would indicate that S2 could partially perceive how distant S was feeling from him. If 1 S2 could perceive Sl's level of psychological distance, then S1 was not responding merely to chance but to something that S2 could at least partially perceive. These correlations were performed and are listed in Table 10 in the Appendix. Again there were two methods of testing this question, within a dyad across minutes and across dyads. The correlation for the intra-dyad test across the whole sample was positive but not significant. However, there was high consistency amongst the correlations for each of the four groups. All were low, insignificant but posi- tive. The inter-dyad correlation across the whole sample was also positive but not significant. However, the correla- tion for female friends was positive and significant using the liberal test but not significant using the conservative test.* The correlation for male strangers was negative; it also was significant using the liberal test but not signifi- cant using the conservative test. A correlation made using friends alone was significant using either the liberal or *The same statistical method was used for these cor- relations that was used to test alternative two of hypo- thesis I. 72 the conservative test. Also, a correlation using all Ss but male strangers was significant using either the liberal or the conservative test. However, these tests were per- formed post hoc, after seeing that the correlation for male strangers was negative. The above statistics appear to give mild support to the validity of the concept of psychological distance as it was operationally defined in this research. The data were analyzed in the most appropriate way to test the hypothesis. The hypothesis that a linear relation- ship exists between nonverbal behavior and psychological distance was not supported. However, it is possible that a relationship exists between psychological distance which is other than linear. Mehrabian (1968) found a parabolic rela- tionship between eye contact and five degrees of liking of a listener. Hearn (1957) found a parabolic relationship between eye contact and status of a listener. It would be interesting for future studies to be designed so that they can test for other than linear relationships between psycho- logical distance and nonverbal behavior. Testing for a parabolic relationship would be especially valuable. How- ever, researchers interested in exploring this problem need to be aware that such a study would be extremely complex and require great sophistication of design. Of particular dif— ficulty would be obtaining all five levels of psychological distance. A possible explanation for the lack of 73 significant findings in this research is that the relation— ship between psychological distance and nonverbal behavior is parabolic rather than linear. The lack of support for the second hypothesis is sur— prising and seems inconsistent with other research on non- verbal behavior. Willis (1966) found that friends stood closer to each other than strangers. The Ss in Little's (1968) study placed dolls who were "friends" closer together . _?,.__. ......_’..-.4-.-r¢--:.b j ' -.o:d than dolls who were "strangers." Mehrabian (1968a, 1968b) has found that distance decreased linearly the more posi- L tively the Ss felt toward the other. Several researchers have found more eye contact with liked individuals than with disliked individuals (Exline and Winters, 1965; Mehrabian, 1968a; Mehrabian, 1968b; Mehrabian and Friar, 1969; Ellsworth and Carlsmith, 1968). Mehrabian (1968b) found that listeners felt that a more negative attitude was com— municated when the communicator leaned backward and away than when he leaned forward and toward. While it is difficult to eXplain the results of this study in light of the above research, the brief question— naires given at the end of each research session (see Appendix) may shed some light on this problem. Ss in the stranger group indicated that they did like their partner. (To question six, "Do you like him/her?" thirty-one answered "yes," four answered "a lot," three answered "a little," and two answered "I don't know.") Thus it appears that partners 74 in the stranger group did like each other. While it seems unlikely that they liked or cared about each other as much as members of the friends group, it does appear that in terms of the level of positive feelings toward one's partner there was not as much difference between the two groups as was eXpected when this study was designed. Another consideration is that strangers really may not have been strange to each other. Twenty—eight Ss saw their partner as being about as much like them as the average Michigan State University student, eight saw him as more alike themself, and only four saw him as more different. All were of the same sex, all were college students, and all were placed in the same situation. In fact, going through the research process together may have led to a feeling of comradeship. Also, Ss did have twenty-five minutes in which to get to know each other. Since they did indicate liking toward each other, after several minutes they may have been responding more as friends than as strangers. Thus the lack of significant differences between the groups of friends and groups of strangers could be related to the fact that there was less difference in the amount of positive feelings and perhaps in degree in familiarity (or strangeness) than was originally anticipated. If this is indeed the case, it may suggest that the level of positive feelings toward each other is more important in influencing nonverbal behavior than the length of time two people have 75 known each other.* Future researchers investigating differences between friends and strangers in their nonverbal behavior may wish to utilize different populations so that a strangers group would be more "strange." Using pe0p1e living in a city as Ss rather than college students may yield quite different results than those obtained in this study. One could use both college students and individuals of similar age who were not attending college. However, one would need to be cautious not to confuse differences due to cultural influ- ences with differences due to the difference in the rela- tionship. The most surprising result of our analysis and the most difficult to explain was that neither analysis of variance yielded any main effect for sex differences. This is particularly surprising for the variable eye contact where sex differences have frequently been found in other research (Exline, 1963; Exline and Winters, 1965; Kendon and Cook, 1969; Ashear and Snortum, 1971). As noted in the review of the literature, nonverbal behavior is a complex phenomenon that is related to many variables (sex, culture, personality, relationship, status, *The means of the two groups were not consistently in the predicted direction. Generally, they were in the pre- dicted direction for eye contact and body congruence, but in the Opposite direction for body positions. 76 attitudes). It is possible that in this particular research situation the effects of variables other than the ones studied were predominant and that these effects camouflaged or blocked out any effect of the variables examined in this research. In particular various aspects of the experimental design may have had an effect. In several ways it was an “unnatural" situation. Ss were provided with a topic of conversation and directed to talk rather than being allowed to develOp their conversation naturally or spontaneously. Ss both knew they were being videotaped and were able to see the camera plainly. It is possible that under an "unnatural" situation a person's nonverbal behavior and/or his level of psychological distance are different from his nonverbal behavior and/or psychological distance level under more "natural" circumstance. A more logical consideration than "unnaturalness" is anxiety. A number of Ss appeared uncomfortable. Several made comments reflecting uneasiness about being videotaped. This research design included a variety of factors which alone or in combination could have contributed to anxiety in the Ss; these include knowing they were being videotaped, seeing the camera, knowing they were being studied, meeting a stranger (for those Ss in the stranger group), talking about a controversial topic such as abortion reform, being aware of at least a little ambiguity about what they would do after their discussion (even though they knew they would 77 answer questions they did not know what kind of questions) and perhaps experiencing a good deal of ambiguity about what the researcher would do with their data. It is possi- ble that anxiety influenced the Ss nonverbal behavior much more than psychological distance did and in the process camouflaged any results due to psychological distance or g a any of the other variables. This does not seem to be the most parsimonious eXplanation of the results as there is no evidence that anxiety influenced the experimental results. Also many of the same anxiety factors have existed in other research which has obtained significant results. It is mentioned as a mere possibility. The experience gained through this research enables the proposal of several suggestions which should prove valuable for future research. As suggested above, for future research involving psychological distance, it would probably be valuable to use Ss trained in introspection and Ss who are accustomed to thinking about their interpersonal relationships. It also would be valuable in correlational Itype research to attempt to obtain a wider range of levels of psychological distance than was obtained in this study. The raters in this research received three training sessions. While the reliability quotients obtained in this research seem adequate, further training may lead to even greater reliability which in turn would make the data analy— sis more sensitive to significant findings. For the reason 78 explained below, this may be particularly valuable when the research requires analyses using S's data over minutes as was necessary in this research. Although somewhat difficult to ascertain with assur- ance because of the difficulties in obtaining intra S reliability, it does appear that inter S reliability is generally higher than intra S reliability indicating that a SS total score for a nonverbal behavior is more reliable measure than his data across minute scores. It is suggested that when possible,studies utilize a Ss total score rather than utilizing a data across minute approach. When not posSible it is suggested that the raters be particularly well trained and the variables be as well defined as possi- ble. Researchers using body postures as a variable may wish to pay particular attention in differentiating leaning away from slouching to their raters, as raters in this study had some trouble differentiating between them. Scanning the reliability table suggests that there may be a positive relationship between the clear definitions of a nonverbal behavior and high reliability correlations. Thus, it is suggested that future researchers define their variables as clearly as practical for their research. It is also suggested that when research facilities permit, videotaping with a hidden camera would be prefer- able to using one in a position that is visible to the S5. A particularly frustrating aSpect of the results of 79 this research is that the one significant finding is too open to question to be sure further research of psycho- logical distance will lead to greater understanding of non- verbal behavior, and yet it is too enticing (at least to this researcher) to suggest that psychological distance is totally unrelated to nonverbal behavior. Thus, this researcher feels that these data suggest that psychological distance, gender and relationship are related to nonverbal behavior, but not in the simple linear way that was prOposed in the hypotheses. Hopefully, future research will clarify what these relationships are. It is recommended that further research be undertaken to examine the relationship between the above variables. An interesting problem would be investigating dyads composed of married or dating (or living together) couples. The rela- tionship between nonverbal behavior and psychological dis- tance in therapists and clients could make a valuable study. A study designed to investigate the relationship between psychological distance and nonverbal behavior that was sensitive to a parabolic relationship as well as a linear relationship would be particularly valuable. BIBLIOGRAPHY 80 BIBLIOGRAPHY Aiello, John R. and Jones, S. E. Field study of the proxemic behavior of young school children in three subcultural groups. Journal of Personality and Social Psychology. 1971, 19, 351-356. Argyle, M. and Dean, J. Eye contact, distance, and affilia- tion. Sociometgy, 1965, 28, 289-304. Ashear, Victor and Snortum, J. R. Eye contact in children as a function of age, sex, social and intellective variables. Developmental Psychology, 1971, 4, 479. Bakan, P. Hypnotizability, laterality of eye movements and functional brain asymmetry. Perceptual and Motor Skills, 1969, 28, 927-932. Bass, M. H. and Weinstein, M. 8. Early development of interpersonal distance in children. Canadian Journal of Behavioral Science, 1971, 3, 368-376. Birdwhistell, R. L. Some body motion elements accompanying spoken American English. In Lee Thayer (ed.) Communication: Concepts and Perspectives. Wash- ington, D. C.: Spartan Books, 1967, 53-76. Charny, J. E. Psychosomatic manifestations of rapport in psychotherapy. Psychosomatic Medicine, 1966, 28, 305—315. Condon, W. S. and Ogston, W. D. Sound film analysis of normal and pathological behavior patterns. Journal of Nervous and Mental Disease, 1966, 143, 338— 7. Condon, W. S. and Ogston, W. D. A segmentation of behavior. Journal of Psychiatric Research, 1967, 5, 221-235. Dabbs, J. M. Similarity of gestures and interpersonal influence. Proceedings of 77th Annual Convention of American Psychological Association, 1969, 47’337-338. Dabbs, J. M. Physical closeness and negative feelings. 'PSychonomic Science, 1971, 23, 141-143. 81 82 Darwin, C. R. The EXpression of Emotions in Man and Animals. New York: D. Appleton and Company, 1896. Dittman, A. T. The relationship between body movements and moods in interviews. Journal of Consulting‘ Psycholggy, 1962, 26, 480. ~ Efran, J. S. Looking for approval: Effects on visual behavior of approbation from persons differing in importance. Journal of Personality and Social Psychology, 1968, 10, 21—25. Ekman, P. Differential communication of affect by head and body cues. Journal of Personality and Social Psychology, 1965, 2, 725-735. Ekman, P. and Friesen, W. V. Nonverbal Behavior in Psycho- therapy Research. In John Shlien (ed.), Research in Psychotherapy. Washington, D. C.: American Psycho— logiEal Association, Inc., 1968, 179—216. Ekman, P. and Friesen, W. V. Nonverbal Leakage and Clues to Deception. Psychiatry, 1969, 32, 88-105. Ekman, P. and Friesen, W. V. Constants across cultures in the face and emotion. Journal of Personality and Social Psychology, 1971, 124-129. Ellsworth, P. C. and Carlsmith, J. M. Effects of eye con— tact and verbal content on affective response to a dyadic interaction. Journal of Personality and Social Psychology, 1968, 10, 15-20. English, H. B. and English, A. C. A Comprehensive Diction— ary of Psychological and Psychoanalytical Terms. New York: David McKay Company, Inc., 1958. Exline, R. V. Explorations in the process of person per- ception: visual interaction in relation to com- petition, sex, and need for affiliation. Journal of Personality, 1963, 31, 1-20. Exline, R. V. and Eldridge, C. Effects of two patterns of of a speaker's visual behavior upon the perception of the authenticity of his verbal message. Paper presented at the meeting of the Eastern Psychological Association, Boston, April, 1967. Exline, R. V., Gray, D. and Schuette, D. Visual behavior in a dyad as affected by interview content and sex of respondent. Journal of Personality and Social Psycho- logy, 1965, 201-209. 83 Exline, R. V. and Messick, D. The effects of dependency and social reinforcement upon visual behavior during an interview. British Journal of Social and Clinical Psychology, 1967) 6, 256-266. Exline, R. V. and Winters, L. C. Affective relations and mutual glances in dyads. S. Tomkins and C. Izard (eds.) Affect ngnition and Personality. New York: Springer, 1965, 319-350. Felipe, N. J. and Sommer,R. Invasions of personal space. Social Problems, 1966, 14, 206-214. Fry, A. M. and Willis, F. N. Invasion of personal space as a function of the age of other invader. Psycho- logical Record, 1971, 21, 385-9. Garfinkel, H. Studies of the routine grounds of everyday activities. Social Problems, 1964, 11, 225-250. Guardo, C. J. and Meiselo, M. Child-parent Spatial patterns under praise and reproof. Developmental Psychology, 1971, 365. Hall, E. T. The Silent Language. New York: Faucett, 1959. Hays, W. L. Statistics for Psychologists. New York: Holt, Rinehart and Winston, 1963. Hearn, G. Leadership and the spatial factor in small groups. Journal of Abnormal and Social Psychology, 1957, 54, 269-272. James, W. T. A study of the eXpression of bodily posture. Journal of General Psychology, 1932, 7, 405-437. Kendon, A. Some functions of gaze direction in social inter- action. Acta Psychologica, 1967, 26, 22-63. Kendon, A. Movement coordination in social interaction: some examples described. Acta Psychologica, 1970, 32, 100-125. Kendon, A. and Cook, M. The consistency of gaze patterns in social interaction. British Journal of Psychology. 1969, 60, 481-494. Knapp, P. H. EXpression of the Emotions in Man. New York: InternaEion Universities Press, 1963. 84 Kretch, R. Cross cultural communication. In J. R. Davitz (ed.) 'The Communication of Emotional Meaning. New York: McGraw Hill, 1964, 148—150. Little, K. B. Personal space. ‘Journal'of Experimental Social Psychology, 1965, 1, 237-247. Little, K. B. Cultural variation in social schemata. Journal of Personality and Social Psychology, 1968, 10, 1-7. Lott, D. F. and Sommer, R. Seating arrangements and status. Journal of Personality and Social Psychology, 1967, 7' 90—940 Martin, D. C. and Van Dyke, W. K. The eXpression of social experience in a figure placement task. Psychonomic Science, 1968, 12, 355-356. Mehrabian, A. Inference of attitude from the posture, orientation, and distance of a communicator. Journal of Consulting and Clinical Psychology, 1968, 32, 296-308 (a). Mehrabian, A. Relationship of attitude to seated posture, orientation, and distance. Journal of Personality and Social Psychology, 1968, 10, 26-30 (b). Mehrabian, A. Nonverbal betrayal of feeling. Journal of Experimental Research in Personality, 1971, 5, 64-73. Mehrabian, A. and Diamond, S. G. Effects of furniture arrangement, props, and personality on social inter- action. Journal of Personality and Social Psycho- logy, 1971, 20, 18—30. Mehrabian, A. and Friar, J. T. Encoding of attitude by a stated communicator via posture and position cues. Journal of Consulting and Clinical Psychology, 1969, 33, 330-336. Mobbs, N. A. Eye-contact in relation to social introver- sion/extroversion. British Journal of Social and Clinical Psychology, 1968, 7, 305—306. Reece, M. M. and Whitman, R. N. Expressive movements, warmth and verbal reinforcement. 'Journal'of Abnormal and Social Psychology, 1962, 64, 234—236. 85 Rosenfeld, H. M. Effect of an approval-seeking induction on interpersonal proximity, Psychological Reports, 1965, 17, 120—122. Scheflen, A. E. Communication and regulation in therapy. Psychiatgy, 1963, 26, 126-136. Scheflen, A. E. The significance of posture in communica- tion systems. Psychiatry, 1964, 27, 316-331. Schlosberg, H. Three dimensions of emotion. The Psygho— logical Review, 1954, 61, 81-88. Thayer, S. The effect of interperson looking duration on dominance judgments. Journal of Social Psychology, 1969, 79, 285-86. Van Lawick-Goodall, J. The behaviour of free living chim- panzees in the Gombe Stream Reserve. Animal Behaviour Monographs, Vol. 1, 1968, 161—311. Van Hooff, J. A. R. A. M. Facial expressions in higher primates. Sumposia of the Zoological Society of London, 1962, 8, 77-125. Willis, F. N. Initial speaking distance as a function of the speaker's relationship. Psychonomic Science, 1966, 5, 221-222. APPENDIX 86 \DmflmUluh-WNH FJH #40 12. P4P bk» 0 15. PSYCHOLOGICAL DISTANCE SCALE VERY CLOSE--You are very involved with your companion. You are probably sharing something important to one of you and probably caring about him and/or sensing his caring about you. You are feeling accepted and under- stood or feeling as if you accept and understand your companion. SOMEWHAT CLOSE--You feel somewhat close to and "with" your companion but the feeling is not intense. You are communicating reasonably well; you understand each other fairly well. You may feel similar to him in some important way. NEUTRAL-—You feel neither distant nor close to your companion. He's there and so are you. This is the way you may feel when chatting with a classmate before a class. SOMEWHAT DISTANT--You feel slightly remote. You may feel as if the other is different from you and that you don't quite understand each other. You may feel as if you or the other is placing some limitation on your interpersonal involvement. Something may seem to be missing from your communication. VERY DISTANT--You feel very distant and aloof. It may feel almost as if the other were not there or were not real or as if he were pushing you away. You may feel alienated from the other. You feel either that he does not accept you or that you cannot accept him at this moment. 1 2 3 4 5 16. 1 2 3 4 5 1 2 3 4 5 l7. 1 2 3 4 5 l 2 3 4 5 18. l 2 3 4 5 l 2 3 4 5 19. 1 2 3 4 5 1 2 3 4 5 20. l 2 3 4 5 1 2 3 4 5 21. 1 2 3 4 5 1 2 3 4 5 22. 1 2 3 4 5 1 2 3 4 5 23. 1 2 3 4 5 1 2 3 4 5 24. 1 2 3 4 5 1 2 3 4 5 25. 1 2 3 4 5 1 2 3 4 5 26. 1 2 3 4 5 1 2 3 4 5 27. 1 2 3 4 5 l 2 3 4 5 28. 1 2 3 4 5 1 2 3 4 5 29. l 2 3 4 5 1 2 3 4 5 ’ 30. l 2 3 4 5 87 88 BRIEF QUESTIONNAIRE FOR STRANGERS What is the date of your birth (include year)? How would you rate your discussion as a whole in terms of psychological distance? 1 2 3 4 5 In regard to the topic(s) discussed do you and he/she strongly agree have similar views agree on some points but differ on others differ strongly differ Do you think he/she understood what you were trying to say to him/her? 1. yes, completely 2. mostly 3. a little 4. not at all Did you understand what he/she was saying to you? 1. yes, completely 2. mostly 3. a little 4. not at all Do you like him/her? no a little yes a lot I don't know Does he/she like you? no a little yes' a lot I don't know Considering this person as a whole do you think he/she is very much like you more like you than the average MSU student like you about as much as other MSU students more different from you than the average MSU student very different from you Do you plan to see him/her again? 89 BRIEF QUESTIONNAIRE FOR FRIENDS What is the date of your birth (include year)? How would you rate your discussion as a whole in terms of psychological distance? 1 2 3 4 5 In regard to the topic(s) discussed do you and your friend strongly agree generally agree agree on some points but differ on others differ strongly differ How long have you known your friend? Circle all of the following which you do or share with this friend: a room a dormitory floor a dorm meals sports activities mixed social activities (i.e., dates) philosophical or ethical talks talks about either or both of your personal problems a home town playing cards studying gossiping listening to records What other if any activities do you share? Has this friend ever been to your home? Here at MSU would you say this friend is (l) the closest like-sex friend you have (2) as close as but not closer than several other friends (3) not as close as another friend you have here at MSU 90 RELEASE FORM FOR AUDIO AND VIDEO TAPES I, agree that information obtained during the two videotaped counseling and testing sessions may be stored and periodically used for scientific purposes. This permission covers the use of test results, videotapes, and sound recordings. This permission is given with the understanding that all informa- tion will be treated in a professional manner and that adequate measures will be taken to safeguard confidentiality. Signed Date 91 RELEASE FORM FOR AUDIO AND VIDEO TAPES I, agree that the information obtained during the videotaped counseling and testing session on may be used for Diane Borchelt's dissertation research. This per— .mission covers the use of test results, videotapes, and sound recordings. This permission is given with the under- standing that all information will be treated in a profes- sional manner and that adequate measures will be taken to safeguard confidentiality. It is also given with the understanding that the audiotape and videotape will be erased after her data is collected unless I sign an addi- tional release form permitting these tapes to be stored for future research. Signed Date 92 P ROVE RBS If you throw dirt at a man, some of it will stick. Every slip is not a fall. Silence gives consent. The child is father to the man. Must is a hard nut to crack. The outsider sees the most of the game. A handful of knowledge is worth a bushelful of learning. The blow that liberates the slave, sets the master free. Little strokes fell great oaks. When one door shuts another opens. 93 TABLE 10 CORRELATIONS BETWEEN N S's ESTIMATE OF HIS LEVEL OF PSYCHOLOGICAL DISTANCE AND HIS PARTNER'S GUESS OF HIS LEVEL OF PSYCHOLOGICAL DISTANCE Mean Intra Dyad Inter Dyad Correlations Correlations Male Friends .243 .285 Female Friends .220 .545 * Male Strangers .204 -.532 * Female Strangers .161 .078 Total .208 .207 Friends .444** Friends plus Female Strangers .372** *Significant using the liberal test (df=N—l) but not significant using the conservative (df=l/2N-l) test. **Significant using either the liberal or the con- servative test. in - "11111111111111“