Citation
Interpersonal orientation correlates of nonverbal behavior in conversational interaction

Material Information

Title:
Interpersonal orientation correlates of nonverbal behavior in conversational interaction
Alternate title:
Nonverbal behavior in conversational interaction
Creator:
Campbell, William Wallace, 1939-
Publication Date:
Language:
English
Physical Description:
viii, 80 leaves : ill. ; 28 cm.

Subjects

Subjects / Keywords:
Conversation ( jstor )
Correlations ( jstor )
Hands ( jstor )
Mathematical variables ( jstor )
Personality psychology ( jstor )
Psychology ( jstor )
Self esteem ( jstor )
Social interaction ( jstor )
Soot ( jstor )
Variable coefficients ( jstor )
Conversation ( lcsh )
Dissertations, Academic -- Psychology -- UF ( lcsh )
Human behavior ( lcsh )
Psychology thesis Ph. D ( lcsh )
Social interaction ( lcsh )
City of Jacksonville ( local )
Genre:
bibliography ( marcgt )
theses ( marcgt )
non-fiction ( marcgt )

Notes

Thesis:
Thesis--University of Florida.
Bibliography:
Includes bibliographical references (leaves 74-78).
Additional Physical Form:
Also available online.
General Note:
Typescript.
General Note:
Vita.
Statement of Responsibility:
by William Wallace Campbell.

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
Copyright [name of dissertation author]. Permission granted to the University of Florida to digitize, archive and distribute this item for non-profit research and educational purposes. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder.
Resource Identifier:
025628279 ( ALEPH )
03114647 ( OCLC )

Downloads

This item has the following downloads:


Full Text









INTERPERSONAL ORIENTATION CORRELATES OF NONVERBAL BEHAVIOR IN CONVERSATIONAL INTERACTION







By

WILLIAM WALLACE CAMPBELL














A DISSERTATION PRESENTED TO T11E GRADUATE COUNCIL OF
TlE UNIVERSITY OF FLORIDA
IN PARTIAL FULFILLMENT OF Ti-E- REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY
















UNIVERSITY OF FLORIDA 1976













AC K NOWL ED GE1MENT S



The list of friends, fellow students, and professors whose encouragement, criticism, and support merits their being listed here is too long to include. Two friends, however, cannot be omitted, they are Jeannie Clay and Dr. Cynthia Gallois.

My supervisory committee members were truly important in many ways. Dr. Wiley Rasbury, Dr. Robert Ilornberger and Dr. Richard McGee played major roles in my clinical development. Dr. Robert Ziller, with his broad and varied interests, rejuvenated my research interests in each interaction we had. Many thanks are extended to Dr. Thomas Samne for his help and his suggestions through this lengthy research experience. Dr. Norman Markel with his concern for learning, and his concern for people, has had tremendous impact on me as a person, as a student and as a researcher. Without his insights, his suggestions, his many hours of support (given seemingly without limit) this research and this dissertation would have had a much lower probability of successful completion.














TABLE OF CONTENTS



Page

ACKNOWLEDGEMENTS . . . . . . . . . ii

LIST OF TABLES . . . . . . . . . . v

ABSTRACT . . . . . . . . . . . vi

CHAPTER

I INTRODUCTION . . . . . . . . 1

Purpose of the Study . . . . . 4
Coverbal Behavior : * * * S Interpersonal Orientation . . . . 8 Review of the Research Literature . . 9 Research Rationale . . . . . . 12
Coverbal Behaviors . . . . 12 Turn length . . . . . 13 Facing . . . . . . 13
Hand movement . . . . 14 Body touching . . . . IS Head tilts. . . . . is Derived Coverbal Measures . . . is Interpersonal Orientation . . . 16 Fundamental Interpersonal. Relations Orientation-Behavior. 16 Self-Other Orientation Tasks. 19
The two interpersonal orientation instruments . . . 23 Statistical Analysis . . . . 24
Hypotheses . . . . . . . . 27

11 METHOD . . . . . . . . . 31

Subjects . . . . . . . . 31
Equipment . . . . . . . 31
Discussion Problems . . . . . 32
Interpersonal Orientation Instruments . 33 Procedures . . . . . . . 3S
Data Reduction . . . . . . . 36
Derived Coverbal Measures . . . . 38 Data Analysis . . . . . . . 39



iii








TABLE OF CONTENTS
(Continued)



Page

CHAPTER

111 RESULTS. ....................41

Monitors? Reliability. ..........41
Data Analysis ................43

IV DISCUSSION ....................48

Personality Type and Coverbal Behavior. 48
Methodological Considerations. .......52 Future Research and Application. .......5 APPENDICIES ......................57

REFERENCES ........................74

BIOGRAPHICAL SKETCH ...................79































iv














LIST OF TABLES



Table Page

1 Coverbal Behavior Definitions. ...........37

2 Pearson Correlation Coefficients for
Reliability Data..................42

3 Normalized Canonical Variable Coefficients 45







































V














Abstract of Dissertation Presented to the Graduate Council of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy



INTERPERSONAL ORIENTATION CORRELATES OF NONVERBAL
BEHAVIOR IN CONVERSATIONAL INTERACTION

By

William Wallace Campbell

December, 1976

Chairperson: Norman N. Markel, Ph.D. Major Department: Psychology

This study investigated the relationships between

interpersonal orientation and nonverbal behavior. Twelve male and 12 female students enrolled in undergraduate speech classes were the subjects. Each subject engaged in two conversations which were videotaped so that monitors could record the occurrence of five nonverbal 1 behaviors. The five nonverbal behaviors studied were: turn length

(TL), facing (F), hand movement (M), body touching (T), and head tilts (HT). Each subject completed Schutz' Fundamental Interpersonal Relations Orientation-Behavior (FIRO-B) scale and the Self-Other Orientation Tasks (SOOT) developed by Killer. The seven interpersonal orientation measures studied were: expressed control (EQ, wanted control (WC), expressed inclusion (EI), and wanted inclusion (WI) from FIRO-B, and self-esteem (SE), self-centrality (Cent), and



vi








self-complexity (Comp) from SOOT.

The following hypotheses were proposed:

1. There is an overall relationship between interpersonal orientation and nonverbal behavior.

2. F and HT are positively related to Wl and inversely related to El.

3. TL and M are positively related to WC and Cent.

4. High HT and low TL are associated with high WC and low EC.

5. High T and low TL are associated with low SE and low EC.

6. Comp is not correlated with nonverbal behavior.

The canonical correlation analysis was the statistical procedure employed in this study. Hypothesis 1 was statistically supported (p < .004). The other five hypotheses were evaluated through examination of the canonical coefficients. Hypotheses 4 and 6 appeared strongly supported. Hypotheses 3 and S seemed partially supported. Hypothesis

2 was not supported.

Statistical support for hypothesis I was viewed as support for further study of the relationships between interpersonal orientation and nonverbal behavior. The overall personality types which emerged were the abdicrat, the autocrat, and the dominant personality. The main behavioral manifestations of the abdicratic (or submissive) type were a large amount of head tilting, little body touching, and short speaking turns. The behavioral



vii








manifestations of the autocratic (or dominant) type were the mirror image: little head tilting, a large amount of body touching, and long speaking turns.
















































viii













CHAPTER I

INTRODUCT10N



There is a great deal of speech related behavior

other than the systematic production of sounds which determines the denotation of face-to-face communication. Of particular interest in this study are behaviors which are frequently referred to as kinesic, or body motion. Man's interest in these 'non-content' aspects of communication is not new. Almost fifty years ago Sapir (1927) expressed the belief that if we study a person's communication style we will learn much about him as a psychophysical unit when he stated that, if we analyze the speech of an individual, in its social perspective, we obtain valuable information about his psychiatric characteristics. Allport (1961) more recently posited that the expressive components, i.e., onels manner or style of behaving, reflect personality structure and serve as a potential guide to personality assessment.

The communication setting of particular interest in this study is the dyadic conversation, that isthe twoperson face-to-face conversational interaction. This particular setting was selected for several reasons, practical as well as heuristic. This mode of communication is



-I-





-2


after all the most commonly employed means by which sentiments and information are exchanged. In spite of its ubiquitous usage and in spite of a growing and impressive amount of research, much is left to be understood about the dyadic conversation. Cherry (1961) observed,

A conversation is one of the commonest
phenomena we encounter, yet it is one which raises very great scientific problems, many
still unsolved. It is often our commonest
experiences, which we take for granted, that
are elusive of exploration and description.
(P. 10)

Most of us have had the experience of being disturbed as a result of something that has been said to us and then realized that it was not what was said, but how it was said and what behaviors accompanied the spoken word. That mystical 'third ear' ascribed to some people is largely their ability to attend, at some level, to some of the behaviors accompanying the spoken message. Psychotherapists are utilizing impressions derived from observations with their 'third ear' both during psychotherapy and as a means of evaluating past therapeutic efforts. It is anticipated that the findings of this study, and future studies, will make a contribution in the effort, in the current vernacular, 'to raise the level of consciousness' of therapists in such situations as they respond to these behavioral cues.

Many terms have been used to describe what a person does while he is also speaking, some of those terms are: Inon-content communication', kinesiscs, 'non-verbal





-3


communication', and 'body language'. The term that most adequately describes the phenomena, and the term to be used throughout this study is Icoverbal', one proposed by Markel (1975). He defines Icoverbal' as, "behavior of interlocutors which occurs in association with or accompanying words, but which is not essential for the articulation or grammatical functioning of those words" (p. 189). Further, if we consult the dictionary (Woolf, 1973) we find that 'col is a prefix form of 'complement' which is defined in the same source as, "to make complete." So this coverbal behavior combines with the verbal behavior to make the complete communication.

The remainder of this chapter will include a section on the purpose of this study, a discussion of coverbal behavior and interpersonal orientation, and a general review of the literature. Following the review of the literature, a section is presented which discusses the research rationale with specific references to each coverbal behavior and each measure of interpersonal orientation. Next a discussion of the canonical correlation analysis is presented. The somewhat extensive nature of this discussion of the statistical analysis is based on the fact that the canonical correlation analysis is a somewhat complex, relatively infrequently utilized statistical analysis. The chapter is concluded with the presentation of the hypotheses of the study.





-4


Purpose of the Study


The purpose of this study is to investigate the relationship between coverbal behavior and interpersonal orientation. The point will be made that research has shown coverbal behavior to be rather stable over time for an individual, while each person has a unique style of coverbal behavior. Thus coverbal behavior offers a reliable measure with which to discriminate between individuals. Interpersonal orientation is also relatively stable and unique for the individual (Maslow, 1968; Ziller, 1973). As Ziller (1973) points out, interpersonal orientation can be, and is, modified over time, else it would be nonadaptive. However, this change is gradual and this characteristic of being gradually modified gives stability to the person.

The major question addressed by this study is, "is

there a relationship between coverbal behavior and interpersonal orientation?" Following that question, an attempt will be made to identify the nature of specific relationships within that overall relationship. The coverbal behaviors under investigation are: Turn Length (Tb), Facing (F) Hand Movement (M) Body Touching (TF), and Head Tilts (HT). The interpersonal orientation characteristics to be employed are: Wanted Inclusion (WC), Expressed Inclusion (El) Wanted Control (WC) and Expressed Control

(EC) as measured by the Fundamental Interpersonal Relations Orientation-Behavior (FJRO-B) scale developed by





-5


Schutz (1966, 1967), and Self-esteem (SE), Self-centrality (Cent), and Self-complexity (Comp) as measured by the SelfOther Orientation Tasks (SOOT) developed by Ziller (1973).



Coverbal Behavior


By definition coverbal behavior always accompanies

verbal behavior. Whatever and wherever we are verbalizing, we are sim ultaneously engaged in some behavior other than the articulation of words according to culturally prescribed rules. Under certain circumstances the coverbal behavior may not be a perceived component of the communication. For example, most coverbal behavior of interest in this research is transmitted via the visual channel (Markel, 1969). If that channel is unavailable to the person to whom the communication is directed, obviously these coverbal behaviors cannot be a part of the total communication as perceived by that listener. This would be the case when a sightless person is the listener, when the interaction takes place in total darkness, etc. These types of special conditions certainly have impact on the listener's sensitivity to other communication channels (e.g., auditory, olfactory, etc.), but that is not the focus of this research. A basic premise of this research is that the coverbal behaviors under investigation are available for processing by the listener or listeners.

Coverbal behavior can communicate information which is complementary to, contradictory to, or independent of,





-6


the verbal message. When a father speaks comforting words to his frightened child while holding the child in a gentle, caring manner the father is transmitting complementary verbal and coverbal messages. The man at the beach who tells his lover he would never look at another woman, while he is watching every bikini-clad lovely in sight, is sending verbal and coverbal messages which are contradictory. The college professor, expounding on a certain theory to his class, while he still has shaving cream under his ear, is probably sending a more prominent message about himself (from the listeners' points of view he may be sending messages about his grooming, the fact that he overslept that morning, or whatever) with his coverbal message (the shaving cream) than with his verbal message concerning the theory at hand. In this example, the coverbal message probably does not complement or contradict the verbal message, most likely it is independent of the verbal message. Nonetheless, the coverbal message is a prominent, possible dominant, component of the communication.

Bateson, Jackson, Haley and Weakland (1956) presented a now well-known case for the need for agreement between the verbal and coverbal message, proposing their theory that through consistent conflict between the verbal and coverbal messages the speaker creates what they termed the 'double-bind'. Their theory went on to suggest the double-bind as a prime factor in the creation of the






-7


schizophrenic personality. Argyle (1969) pointed out the importance of coverbal cues as they combine with the verbal message. Markel, Bein and Phillis (1973) found a normative relationship between voice and content. Their terms 'synchrony' and 'dysynchrony' are analagous to the terms 'complementary' and 'contradictory' as used in the examples in the previous paragraph.

Whatever the type coverbal communication (i.e.,

complementary, contradictory, or independent) it tends to consistently communicate something about the speaker himself, sometimes in addition to, sometimes instead of what the verbal message has to say about the speaker. For several reasons, the coverbal message may tell more about the speaker than does the verbal message. First, the speaker himself may be unaware of his coverbal message and therefore makes no attempt to 'filter' it so that he will 'look good' or 'look right'. Second, the rules governing coverbal behavior are less stringent (sometimes less understood) than the rules governing verbal behavior. Finally, another reason that coverbal behavior can tell us so much about the speaker is that it can be utilized in such a way that communication of sensitive messages is not so threatening. That is because the coverbal message does not place the same kind of demands on the person to whom it is directed. The coverbal message is easier for that person to simply ignore. That act of ignoring, or nonresponse, may be less traumatic to the sender of the message





-8


than would an overt, verbal rebuke. An increased amount of eye contact (within acceptable bounds) can communicate to a woman that a man would like to initiate a relationship with her. At the same time, the demands for an overt response from her are not nearly so great as they would have been had he used verbal messages to communicate that same desire. The man can feel less threat of rejection, and possibly embarrassment, knowing that he has not required a verbal response of the, woman. Thus, more sensitive, more personal data may more freely be communicated via coverbal behavior. She is free to simply ignore his overture, in that case his message has gotten a subtle response, and his pride remains more intact for his next adventure. Hopefully, he correctly decoded her message. For a discussion of the different rules governing verbal and coverbal behavior, and some of the implications of those differences, see Markel (1975).

The above examples illustrate two points: 1) the

coverbal message can be as important as, or more important than, the verbal message, and 2) the receiver of the coverbal message, especially if untrained in decoding coverbal communication, can misinterpret the message. Both of these points are addressed by Fast (1972).



Interpersonal Orientation


Interpersonal orientation is the way an individual views himself in the context of his social milieu. For









some major personality theorists, understanding the importance of other people to the individual, and the individual's need to relate to other people, is basic to understanding the individual. Fromm (1941, 1947) spoke of man's feelings of isolation and threat as he becomes increasingly separated from other people. Adler (1939) felt that man's social interest and need for other people wereinstinctive. That is, man does not merely learn to need contact and affiliation with others, rather he is born with that social need as a biological given. Sullivan (1953, 1964), probably the most significant single figure in advancing the social perspective as a means for understanding the individual, felt that it was.meaningless to attempt viewing the individual in any other context than the social context. In his interpersonal approach to psychiatry lie pointed out that we all enter and leave this world as social beings and that the most fruitful way of conceptualizing and treating humans is as social beings. If we look at man's basic needs as postulated by Maslow (1967, 1968) we find among them social needs such as the need for affiliation and affection. An accurate and adequately high self-esteem, shaped through social interaction, is also a basic need according to Maslow (1968).



Review of the Research Literature


For some time researchers have been interested in the psychological importance of coverbal behavior. As with any









behavior under investigation, certain basic issues arose relating to quantification and analysis of the coverbal behavior data. One practical, and apparently meaningful, scheme for quantifying coverbal behavior is based on its temporal characteristics (e.g., Chapple, 1939; GoldmanEisler, 1968; Mahl, 1956; Norwine and Murphy, 1938), with the quantity often being simply the duration of time spent engaging in the behavior. Simple duration, along with proportion of available time spent engaging in a behavior, will be the approach to quantifying coverbal behavior in this study.

As interest in the research field grew, along with an increase in man's electronic sophistication, new techniques for measuring, recording, and analyzing coverbal behavior emerged. Almost thirty years ago, Norwine and Murphy (1938) developed a sound-activated device for recording speech behavior. Chapple (1939) was developing his device, later to be known as the Interaction Chronograph (Chapple, 1949), which utilized both the human observer and mechanical devices for recording verbal and coverbal behavior. In his earlier work (Chapple, 1940; Chapple and Arensberg, 1940), Chapple had demonstrated that personality characteristics were associated with certain stable properties of individuals' interaction style.

Some ten years after Chapple (1939) began his research which led to the Interaction Chronography, Verzeano and Finesinger (1949) developed the Automatic Speech Analyzer








which eliminated the necessity of the human observer for obtaining data during interaction. Lorenz and Cobb (1952), using the Automatic Speech Analyzer, were able to differentiate normals and psychiatrically impaired on the basis of temporal aspects of their speech patterns. Chapple (1953) went on to develop his standardized interview, with which, using his Interaction Chronograph., he investigated subjects' interview behavior and its relationship to certain personality dimensions (Chapple, Chapple, and Repp, 1954).

Man's imagination and increasingly sophisticated

electronic equipment available to him, continue to create innovative methods for recording and analyzing coverbal behavior. As can be noted from previously cited research, as different techniques were developing for collecting coverbal data, the data were being examined in numerous ways and contexts. In the area of psychological interests alone, researchers were relating coverbal behavior to social orientation, transient emotional states, degree of psychological health, and interpersonal needs, to name a few. With this wealth of techniques and the potential applications of the obtained data, a primary task of the researcher becomes one of choosing which techniques to utilize and to what end. That brings us to this study. A discussion of how those decisions were made in this study follows.





-12


Research Rationale


Coverbal Behaviors


The point was made earlier that there is always coverbal behavior present during conversation. There is, in fact, so much behavior occurring that it becomes an important task of the researcher to choose just what behavior will be selected out for analyzing. Four questions were asked in this research so that the final behavior selection could be accomplished. Is the behavior detectably recorded when a dyadic conversation is videotaped? Can the relatively untrained observer reliably record the onset and termination of the behavior as the videotape is monitored? Are the behaviors easily observed during normal conversational interaction? Are the behaviors related to the interpersonal orientation of the speaker?

Previous research and pilot studies prior to this

research answered the first three questions affirmatively. The outcome of this research will have to answer the fourth question. Since the results of this study were obviously not available when the behaviors were selected, selection was based on previous research which asked similar questions (e.g., Chapple, 1940; Chapple and Arensberg, 1940; Chapple et al., 1954; Lorenz and Cobb, 1952; Norwine and Murphy, 1938). The coverbal behaviors selected for investigation were: Turn length, Facing, Hand movement, Body





-13


touching, and Head tilts. Brief research references for each behavior follow. Operational definitions of each behavior are presented in Table I in the Method chapter.

Turn length. Following early interest in temporal

aspects of communication behavior, this behavior, or some rough equivalent, has been studied frequently. GoldmanEisler (19S4), looking at action time (percentage of total time which was spent in talking), found a relationship with the content variable of self-reference. That is, people who referred more to themselves, and presumably talked more about themselves, spoke during a greater percentage of the total time. Cervin (19S7) found that individuals scoring high on a scale of emotional responsiveness also spoke during a greater portion of the available time. Markel, Bein., Campbell, and Shaw (1976) demonstrated a greater use of available time by speakers who scored high on a measure of needed inclusion. This often investigated behavior of amount of speaking time (referred to by several roughly equivalent terms such as 'action time' (Goldman-Eisler, 19S4), 'participation quotient' (Prebor, 1972), 'mean percentage', (Matarazzo, Weins, Matarazzo and Saslow, 1968), etc.), seems to be related to various aspects of personality.

Facing. Facing behavior was selected because of its high correlation with eye contact. Eye contact itself is virtually impossible to detect by observers as they view videotapes of conversations. Facing, on the other hand,






14


is rather reliably observed and recorded by the relatively untrained observer. Clinically, eye contact is a frequently utilized diagnostic tool. Socially, this behavior seems to be thought of as an indicator of the speaker's candor. This is a good example of coverbal behavior that can be easily mis-interpreted if the listener fails to take into account cultural factors such as the norms of the culture from which the speaker comes. Argyle and Dean (1965) and Exline (1971) found facing behavior positively correlated with personal attraction. Exline (1971) and LaFrance and Mayo (1976) found facing associated with dominance. Duncan (1972) conducted research which concluded that facing is an effective tool for controlling speaking turn during conversations. When the speaking person looks away from the listener, our culture tells us that that is not a polite time for the listener to attempt taking the role of speaker. A review of the literature on facing can be found in Ellsworth and Ludwig (1972) .

Hand movement. Hand movement seems to be a method of attracting attention and controlling the interaction. Dittmann(1962) found it related to approval seeking in subjects. Mehrabian (1970, 1971) found hand movement to be associated with attraction and the desire for affiliation. While some investigators have devised elaborate systems of quantitatively and qualitatively classifying hand movement (e.g. Ekman and Friesen, 19 72 ; Frey, 1975) this study looks at hand movement simply in terms





-is


of duration of occurrence.

Body touching. Research on body touching (touching one's own body with his hand) has not been extensive. Freedman and Hoffman (1967) found body touching to be related to anxiety level. Scheflen (1965) found body touching associated with the need for self-protection. Fast (1970), summarizing research in the field, asserts that body touching can be a coverbal method of sending a 'hands off' message to others. In other words, the person who does not want to be included can say so with his hands. The findings of these three authors are seen as being consistent in that the person who does not want to be included may experience anxiety and feel the need for self-protection.

Head tilts. Head tilt behavior appears to be an

indication of attentiveness and possibly cognitive activity during conversation. In animal research it appears to be associated with submissiveness. This relationship may also hold for humans. Mehrabian (1971) and Mehrabian and Ksionzky (1970) found head tilt related to relaxation of the person.



Derived Coverbal Measures


The derived measure for Turn Length in this study was the duration of the speaking turn as defined in Table 1 in the Method chapter. The derived measures for Facing, Hand Movement, Body Touching and [lead Tilts were computed





-16


proportions of the individual's speaking turns during which he was engaged in the given behavior. The proportion measure was used rather than straight duration measure in order to control for turn length. The proportion tells us not merely how much time was spent in each behavior, rather it tells how much of the available time was spent engaging in each behavior.



Interpersonal Orientation


The interpersonal orientation measures used in this study are Wanted Control (WC), Expressed Control (EC), Wanted Inclusion (WI), and Expressed Inclusion (EI) as measured with the Funadmental Interpersonal Relations Orientations-Behavior (FIRO-B), and Self-esteem (SE), Selfcentrality (Cent), and Self-complexity (Comp) as measured with the Self-Other Orientation Tasks (SOOT).

Fundamental Interpersonal Relations OrientationBehavior. A steadily increasing amount of research is reporting various approaches to testing Schutz's (1966) theory of interpersonal orientation. For a summary of that research see Schutz (1966). Schutz bases his theory on the existence of three interpersonal needs: Inclusion, Control, and Affection. Inclusion and affection have been found to be highly correlated in research (Argyle, 1969). Therefore, it is meaningful to think of the individual in terms of the two dimensions inclusion and control. For that





-17


reason, only the inclusion and control scales were used from FIRO-B. Schutz describes interpersonal needs for inclusion and control as follows (Schutz, 1966):

1) Interpersonal need for Inclusion (1): the need

to establish and maintain a satisfactory relation with other people with respect to interaction and

association, and

2) Interpersonal need for Control (C): the need to

establish and maintain a satisfactory relation with other people with respect to control and

power.

Schutz (1966) refers to each person's behavior in each of these dimensions in terms of the behavior expressed (E) towards others and the behavior wanted (W) from others. Emerging then are four dimensions of the individual's interpersonal orientation: Wanted Inclusion (11M, Expressed Inclusion (EI), Wanted Control (WC), and Expressed Control

(EC). Schutz (1966, 1967) developed the FIRO-B as a selfreport means of tapping each of these dimension. The FIRO-B is presented in Appendix A.

Schutz (1966) describes individuals whose behavior is consistently involved in satisfying his needs in the different dimensions. He describes the high I person with terms such as, "interact, communicate, and attend" (p. 21). The person with high C needs is described as displaying behavior such as, "dominance and control" (p. 22), while the low C person is seen as,








"submissive and a follower" (p. 22). How these behaviors are manifested will be influenced by the individual's needs in the I and C dimensions as to whether he wants to be included (or controlled) or whether he wants to include (or control) others. Schutz (1966) further makes the point that there is not necessarily a complementary relationship between wanted and expressed needs in a given area. The person with a high need to control others may also have a high need to be controlled by others. The domineering sergeant may need and gratefully accept control from his lieutenant.

According to his classification the loversociall

person is the one who exhibits excessive inclusion behavior and the lundersociall is the person who makes little or no attempt to include or be included. The 'autocrat' tends towards the domineering personality and is characterized by the excessive desire for power and control. Much of the autocrat's behavior is'directed towards controlling others. The labdicrat' is the person who makes no attempt to exert control over others. Schutz (1966) points out the extreme case of the abdicrat who wants no control over others and at the same time has a high need for others to control him. This person does not say, 'I am going to leave you alone and I would like for you to leave me alone', rather he is saying, 'I do not want to control you, but please tell me what to do,.

The above personality types with their accompanying






-19


behavioral manifestations represent deviations from an ideal state of need satisfaction according to FIRO theory (Schutz, 1966). In the inclusion and control areas talking is seen as an available means by which the person can strive towards the ideal need satisfaction state. The person with the high need to control can attempt to gain control by talking a lot and minimizing the opportunity for the other person to talk. The person with little need to control may talk less and look more to the other person for control cues. The person with high need to be included can

seek attention with hand movements and by talking a great deal. The person with little desire to be included may talk less and send, 'I am not interested in being included' messages with his high rate of body touching.

Self-Other Orientation Tasks. Ziller (1970) is one among many (e.g., Maslow, 1967, 1968; Schutz, 1966; Sullivan, 1953, 1964) who proposes that man should be studied as a social unit. In his social psychological theory of personality, social adaptation is viewed in terms of selfother concepts. For a review of his theory see Ziller (1973). Three of Ziller's self-other dimensions are employed in this study.

His Self-esteem (SE) and Self-centrality (Cent) scales utilize what Ziller (1973) refers to as cognitive mapping. Through the use of ostensibly value-free symbols SE and Cent are meansured without the bias of mediating 'social oughts', so often found to be a problem when verbal





-20


self-reports are used for tapping these aspects of personality. Subjects are asked to place symbols, representing themselves and other people, somewhere in a prescribed area. The SE scale, for instance, utilizes the proclivity in our culture for people to place items of higher value toward the left in a horizontal ordering. On the Cent scale the subject is presented with a large circle and asked to place two smaller circles, representing self and a friend, somewhere in the large circle. Scoring of the Cent scale is on the basis of whether the self or the friend is placed nearer the center of the large circle. The more often the self is placed nearer the center, the higher his Cent score. The SE and the Cent scales each contain six items as described above. All six items on the SE scale are identical except with respect to which five significant other people are represented by the symbols accompanying the ever present sixth symbol representing self. All six items on the Cent scale are identical in every respect. Because of the similarity of the items within each scale, only one sample item from each scale is presented. The sample items are found in Appendix B. The Comp scale is an adjective checklist scale containing 109 adjectives. The Comp scale, in its entirety, is presented in Appendix B. Subjects were instructed to check as many adjectives as they felt applied to them. The Comp score is computed by totaling the number of adjectives checked by the subject. The higher the number of adjectives checked, the higher





-21


the Comp score, and the more complex is the self construct.

Ziller (1973) referred to the self-esteem as that component of the individual which mediates modification of the self in response to new information which is received about the self in social interaction. This view is consistent with the notion that self-esteem is the individuals perception of his own self-worth (Osgood, Suci, and Tannenbaum, 1957). Following these positions, as Ziller states, the individual with high self-esteem is better able to evaluate and assimilate new information. He is not the helpless, vulnerable pawn, subject to restructuring the self-concept immediately upon receipt of information which is in conflict with the then existing self-concept. The high self-esteem person uses new information as it fits for him, the low self-esteem is ever seeking new information so that he may modify the self to fit what the information tells him others think he should be.

The self-centrality scale is concerned with whether the person uses the self or others as his reference point (Ziller, 1973). The higher a person's self-centrality score, the more he uses the self as the reference point. Ziller compares high and low self-centrality with Ausubel's (1952) terms egocentrism and sociocentrism. The high selfcentrality person appears incapable of perceiving his environment from others' viewpoints. He withdraws and does not want new information from others. By implication,





-22


the person with low self -centrality can only view his environment from others' perspectives, lie will constantly seek cues from others as to how he should process his environment and in turn, how he should behave. This is clearly a scale on which the healthy, well-balanced person would score in the mid-range.

As Ziller (1973) defines his measure of self-complexity it could be seen as a measure of personal adjustment. Viewing it in this sense, it is different from his other scales. In the case of his self-esteem and self-centrality scales, and in Schutz' scales, there appears a curvilinear relationship between personal adjustment and scores on the scales. That is, the well-adjusted, socially adept person will fall in the mid-range on these scales. A possible exception to this curvilinear relationship with the'other scales is the self-esteem scale. Ziller (1973) presented his self-esteem-complexity matrix wherein the person falling in the high SE and high Comp cell has a differentiated and integrated theory of social behavior. This person may be the super-well-adjusted person. It appears, ,however, that it is in the self-complexity scale that a direct, positive correlation between scale score and adjustment is found. The high Comp person is the person capable of self evaluation, the one capable of viewing his environment from different perspectives. This is the person who is well balanced, well adjusted, and less likely to have exceptional needs or unrealistic perceptions of self or others.






-23


The two interpersonal orientation instruments. FIRO-B, utilizing the verbal self-report technique, attempts to measure directly the individual's interpersonal orientation. FIRO-B solicits information from the person as to his behavior with other people. Through divulging these behavioral preferences, the level of the person's needs in different interpersonal areas is assessed. Since these ratings are based on the person's expressed behavioral preferences, the ratings should prove to be good predictors of how the person will behave in an interpersonal situation. A source of error in such predictions lies in the extent to which the person honestly responded to the instrument.

SOOT, while tapping aspects of the personality which

are similar to those tapped by FIRO-B, is in some important ways looking at the person with a different approach. SOOT, except in the case of the complexity scale, avoids the use of words in the assigned tasks. Even with the complexity scale, words are not used in the same sense as in the FlRO-B. A verbal self-report is not involved with the complexity scale, rather it is an adjective checklist. Throughout SOOT there is the attempt to avoid 'good-bad', fright-wrong', 'acceptable-unacceptable' type choices on the part of the responder. This is accomplished through the extensive use of symbols rather than words. Another important difference between SOOT and FIRO-B lies in the interpersonal characteristics being measured. SOOT measures characteristics which are shaped through interpersonal






-24


experience. They are also characteristics which have considerable impact on interpersonal behavior. SOOT is not, however, attempting to directly measure behavioral preferences, as does FIRO-B. Therefore, SOOT should be a useful predictor of behavior, but in a less direct manner than is FIRO-B.

Both FIRO-B and SOOT measure interpersonal orientation characteristics which were of interest in this study. They measure similar, yet different, characteristics, with different, verbal as opposed to nonverbal, approaches. It was for the above reasons that the two instruments were included in this study.



Statistical Analysis


Canonical correlation analysis was the statistical procedure employed in this study. This is a relatively recent statistical technique which social scientists are beginning to realize can be a very useful method of looking at two sets of data. The canonical correlation (Kelley,

Beggs and McNeil, 1961; Morrison, 1967) is similar to the more familiar Pearson product moment correlation where two variables are analyzed to ascertain the amount of common information contained in the two. The multiple regression equation goes a step further, providing a method for combining the variables in one group of variables and correlating the newly created variable with a single





-25


variable of interest. With the multiple regression analysis equation not only is there a single indicator of the relationship between the single variable and the linearly combined group of variables, but we also are supplied with the coefficients of the variables themselves as they appear in the regression equation. These coefficients are the weights assigned to each of the original variables in the group. The coefficients are indications of the magnitude (numerical value of the coefficient) and the direction (algebraic sign with the coefficient) of the contribution made by each of the original variables as they were combined. Thus, we have an idea of to what extent and in what direction each variable makes its contribution.

The canonical correlation goes still one step beyond the multiple regression analysis. The canonical correlation linearly combines variables from two groups of data so that a pair of canonical variables emerge, one from each group of data. The correlation between these newly created canonical variables is computed and then it is possible to address the question of whether the two groups of data are related. The level of the correlation indicates the strength of the relationship, that is, it is an indication of how much common data is contained in the two groups. Similarly to the regression equation, we are supplied with coefficients which indicate the magnitude and direction of the influence of each original variable on the canonical variable.





-26


The canonical correlation analysis is a very useful, somewhat complex, method of attempting to identify complex relationships. Although there is no statistical test beyond the overall correlation, specific relationships are inferred by examining the canonical coefficients. In this study the relationships being searched for are complex. They are complex in the sense that it would be too simplistic, possibly meaningless, to single out one coverbal behavior and attempt to establish its relationship with a single interpersonal orientation variable. Humans, their behavior, and their personalities are not that simple. To state that the person who needs to control others will talk longer may be true. However, there are factors other than his need to control which contribute to his behavior, and, there are behaviors other than turn length through which his interpersonal orientation will manifest itself.

After one pair of canonical variables has been identified and tested, others may also exist with the same data sets. In fact, there can be as many pairs of canonical variables as there are variables in the smaller group of data. The procedure is continued until the correlations are not significant or until the number of correlations reaches the number of variables in the smaller data group. In each new relationship, if more than one exists, there is a new set of coefficients and the statistical tests are carried out in the same manner. A more detailed discussion of these procedures is presented in the Method chapter.





-27


Hypotheses


Utilizing five coverbal behavior measures obtained

from 24 videotaped conversations, and seven interpersonal orientation scores, four from FIRO-B and three from SOOT, this study attempts to support the existence of a relationship between coverbal behavior and interpersonal orientation.* Naturalness of the conversations was maximized as subjects were placed in a comfortable environment and asked to discuss relevant social discussion problems with virtually no time constraints placed on the length of their conversations.

Canonical correlation analysis is applied to the data. The canonical correlation is conceptually more complex than a t-test or a pairwise correlation such as the Pearson product moment correlation. Similarly, its interpretations are more complex. To predict an overall relationship between the two groups of data is straightforward enough, and that prediction is directly testable with the canonical correlation analysis. However, to look inside the data groups and make specific, directional relationship predictions involving several variables involves risks beyond the obvious risk of being wrong by predicting a relationship that does not exist. There exists the risk of failing to support an hypothesis, not necessarily because the hypothesized relationship does not hold true, but because of the way the variables were combined in the computation





-28


of the canonical variable. As pointed out earlier, several canonical relationships are possible. If, in the construction of the canonical variable(s) for a given group of the data, the relative directions of two variables within one group, both of which are in an hypothesized relationship, are not the same as they appear in the hypothesis, full support of the hypothesis is not possible. An example makes this point more concrete. Noting the second hypothesis in this study, it predicts that head tilts and facing will be positively related to WI scores and inversely related to SE scores. In addition to the main, overt prediction that two coverbal behaviors will vary directly with one interpersonal orientation score and inversely with another interpersonal orientation score, another prediction is implied. The other prediction is that facing and head tilts will vary together and that WI and SE scores will vary inversely. If the variables are not combined in a way consistent with these latter, implied predictions, full support of the hypothesis is not possible. Nonetheless,

to resort to pairwise predictions involves a level of simplicity inappropriate for this study and the relationships at interest.

Research has shown that interpersonal aspects of

personality are related to coverbal behavior (Allport, 1961; Chapple et al., 1954; Goldman-Eisler, 1954).

Hypothesis 1: There is a relationship between coverbal behavior and interpersonal orientation.





-29


Facing has been shown to be associated with personal attraction (Argyle and Dean, 1965; Exline, 1971). Head tilting behavior appears related to submissiveness (Argyle, 1969).

Hypothesis 2: F and HT are positively related to WI and inversely related to EI.

Goldman-Eisler (1954) found that subjects talked more if their self reference was greater. Dittmann(1962) found hand movement associated with the desire to get approval from others and to control the other person in conversation. The person who wants to control his conversational mate will talk a great deal (Schutz, 19.66).

Hypothesis 3: TL and M are positively related to WC and Cent.

Head tilting is associated with submissiveness (Argyle, 1969). The person desiring to control will talk more than the person not wishing to control (Schutz, 1966).

Hypothesis 4: High HT and low TL are associated with high WC and low EC.

Scheflen (1965) found body touching associated with the need for self-protection. The person wanting control over others will talk a great deal (Schutz, 1966).

Hypothesis 5: High T and low TL are associated with low SE and low EC.

Killer's (1973) high self-complexity individual is a well-balanced person. This person is less likely to have outstanding need areas which are unsatisfied. As the





-30


self-complexity drops, idiosynchratic needs will emerge. It seems that these needs will emerge in an unsystematic fashion, not predictable from the self-complexity level of the person.

Hypothesis 6: Comp is not correlated with coverbal behavior.














CHAPTER II

METHOD



Subjects


Twelve male and twelve female students enrolled in introductory speech class at the University of Florida volunteered to participate in this study. All subjects were caucasion; ages ranged from 18 to 27, with a mean age of 20.3. Two males and two females were selected from the volunteers from each speech section and those four subjects comprised a group for this study. Other than the specified gender requirement, the only condition imposed on the composition of the group was that subjects' familiarity with other group members be limited to the interaction in the speech course in which they were currently enrolled. The resulting sample then was made up thusly: six groups of two males and two females; each group drawn from a different section of the speech course; each subject's prior interaction with his group members was limited to interaction in speech class.



Equipment


Conversations between pairs of subjects were recorded



-31-





-32


in a carpeted room with dimensions of approximately 10 x 15 feet. Subjects were seated at adjacent sides of a padded card table in comfortable, padded arm chairs in one corner of the room. Each subject was seated at approximately a 4SI angle to a Sony 3210 video camera equipped with zoom lens. The video camera was connected to a Sony AV 3600 Videocorder which was the recorder used for recording the audio, as well as the video, signals during conversations. A Sony microphone was mounted on a stand between and slightly behind the subjects. Subsequent to the exeprimental sessions, continuous lapsed time in minutes, seconds, and tenths of seconds from beginning to end of each conversation was superimposed on the video tape by means of a Model VTG33 Odetic Video Timer.



Discussion Problems


Each of the discussion problems was of assumed and apparent contemporary social interest. One problem, the "abortion" problem, involved a young married couple groping with the decision of whether or not the pregnant wife should have an abortion. The other problem, the "living together" problem, concerned a young, heterosexual, unmarried couple, living together, trying to decide how they should present their living situation to his parents, who were unaware of their cohabitation, while the parents were in town for a brief visit. Judging from observed level of





-33


apparent interest on subjects part and based on verbal response of subjects after the experimental sessions were completed, subjects actually became quite involved in the discussion problems. The texts of these problems are

presented in Appendix C.



Interpersonal Orientation Instruments


The two instruments utilized for measuring interpersonal orientation of subjects were the Fundamental Interpersonal Relations Orientation-Behavior (FIRO-B) developed by William Schutz and Self-Other Orientation (SOOT) scale developed by Robert Ziller.

FIRO-B is composed of 54 items and has been used with subjects ranging from eighth grade to adults. The results of this instrument yield six scales: Expressed Inclusion, Wanted Inclusion, Expressed Control, Wanted Control, Expressed Affection, and Wanted Affection. Only the first four of these scales were used in this study. The possible range of scores on these six scales is from 1 to 9, with all scores being integers. This instrument was designed by Schutz (1966, 1967) to measure the relatively stable attitudes of an individual which determine his interactive behavior. Schutz has published considerable research demonstrating validity for his instrument.

SOOT was designed to measure individuals' self-other orientation in terms of their perceived relationships. The method of assessing self-esteem and self-centrality is





-34


the use of ostensibly value-free symbols which result in subjects' supplying more valid information about themselves than would be expected in more traditional, verbal, selfreport inventories. In the self-esteem scale symbols represent the self, significant other people such as family members, teachers, friends, etc. The subjects' tasks were to arrange the symbols according to certain instructions. Scoring is accomplished by obtaining the sum of the assigned numerical values according to the position in which the symbol representing the self is placed. The self-centrality score is based on whether the 'self' is placed nearer to or farther from the center of a

large circle than is the symbol representing a friend. Each scale has six items. The self-centrality items are identical throughout the scale. The only difference between the six items of the self-esteem scale is in the significant other people which the symbols represent. The complexity scale consists of a 109-word adjective checklist. Even though the SOOT is a relatively new instrument, a considerable amount of research has demonstrated the reliability and validity of the scales (Ziller, 1973).

The SOOT scales used in this study are Self-Esteem,

Self-Centrality, and Self-Complexity. Sample items from the self-esteem and self-c'entrality scales are presented in Appendix B. Because of the similarity of the items, only

one item per scale is presented here. The entire self complexity scale and the general instructions for SOOT are also





-35


presented in Appendix B.

FIRO-B is presented in its entirety in Appendix A.



Procedures


Subjects were assigned experimental session times by group so that one group participated in an evening. After all group members had arrived for a session one experimenter explained that the purpose of the study was to study the behavior of persons in conversation in a dyadic setting. They were told that each of them would participate in two conversations, one each with two of the other members of their group, and that the conversations would be recorded on video tape. They were also informed that they would complete several paper and pencil instruments which would call for information about themselves and information about how they felt about their experience of participating in the study. Subjects then signed consent forms for participation in the experiment.

A second experimenter then took two subjects to the taping room where they would have their conversations. The camera and recorder was started by the experimenter. Subjects were seated at the table with copies of one of the discussion problems affixed to the table in front of each of them. They were instructed to read the discussion problem and discuss as long as they needed to in order to come to some agreement on a solution. The experimenter read the subjects' identification numbers into the





-36


microphone, told the subjects to call him when they were

finished, and left the room.

During the first conversation the first experimenter had the subjects who were not participating in the conversation complete half of the paper and pencil tasks that they were to complete. Following the first conversation, the subjects just completing their conversation were taken to other rooms where they completed half of their paper and pencil tasks while the other two subjects had their

first conversation.

The process was continued until each subject had:

participated in two conversations (one each with a samesex and an other-sex partner), completed the background information form, the FIRO-B, the SOOT, and semantic differentials relating to how they felt about the experiment, the experimenters, and their partners. Of the paper and pencil instruments, only the FIRO-B and the SOOT were used in this study.

Subjects were randomly assigned to: either complete FIRO-B or SOOT first., discuss the 'abortion' problem or the 'living together' problem first, and converse with a same-sex or opposite-sex partner first.



Data Reduction


Monitors were trained to observe the tapes and record the occurrence of the coverbal behaviors listed in Table 1. Each conversation was monitored once at normal tape speed





-37





Table I

Coverbal Behavior Definitions



Behavior Definition

1. Turn Begins when one speaker starts
talking and ends when the other speaker starts solo talking

2. Facing Subject's nose is pointed
roughly at the center of
partner's face

3. Right hand movement Any movement of subject's
right hand and wrist, unless it results from a movement of the whole torso

4. Left hand movement Any movement of subject's
left hand and wrist, unless it results from a movement of the whole torso

5. Right hand body touching Any part of subject's right
hand touching any part of his body

6. Left hand body touching Any part of subject's left hand touching any part of his body

7. Head tilt Subject's eyes are not in a
horizontal plane






-38


for each behavior for each subject. During that first monitoring the approximate beginning and ending times of that behavior were recorded. Using those approximate times as guides, the tape was then played at slow-motion speed (approximately 1/15 of normal speed) to record more precisely (usually within 0.1 second) the beginning and ending times. These times were obtained by using lapsed time which had been superimposed on the tape with an Odetic Video Timer after conversations were completed.

Monitors were assisted in determining beginnings and endings of turns by having a typed verbatim transcript of the conversations. Periodic checks indicated that tracking was reliable when monitors had been given several hours training.



Derived Coverbal Measures


A computer program was developed which used the

beginning and ending times for each coverbal behavior and for the turns and computed durations and frequencies of all behaviors per turn. For the purposes of this study, total hand movement and total body touching were of interest, rather than hand movement and body touching broken down by left and right hand. Therefore, right hand movement and left hand movement were combined into one measure: Hand Movement (M). Also, right hand body touching and left hand body touching were combined into one measure: Touching (T). The coverbal measures used in the statistical





-39


analysis were the grand mean per behavior for subjects across both their conversations, after summing across turns. In the case of the measure Turn Length (TL), it was the grand mean of the turn lengths across both conversations. The behaviors: Facing (F), Hand Movement (M), Touching (T), and Head Tilt (HT) were computed as proportions of the respective turn which was spent engaging in those behaviors. The grand mean for each subject that was used for these behaviors then was the mean obtained across all speaking turns of both conversations using those proportions.



Data Analysis


The statistical method for analysis of the relationship between the two groups of data (interpersonal orientation and coverbal) was the canonical correlation. There were five variables in the coverbal data: Turn Length (TL), Facing (F), Movement (M), Touching (T), and Head Tilt (11T). Seven Personality variables made up the other group: Expressed Control (EC), Wanted Control (WC), Expressed Inclusion (EI), Wanted Inclusion (WI), Self-Esteem (SE), Self-Centrality (Cent), and Complexity (Comp). The first four of the personality variables are scales from the Fundamental Interpersonal Relations Orientation-Behavior (FIRO-B) and the last three are scales from the Self-Other Orientation Tasks (SOOT).

Canonical correlation analysis is a method whereby the question, "Is there a significant relationship between the





-40


two groups of data?", can be addressed. Further, an indication of the strength of the relationship(s) is indicated by the level of statistical significance. More specific ideas about to what extent each original variable makes its input into the relationship is inferred from the canonical variable coefficients. Further elaboration on this statistical procedure and subsequent inferences is presented in the Result and Discussion chapters.













CHAPTER III

RESULTS



Monitors' Reliability


In order to assess the reliability with which the three monitors recorded the coverbal behavior, a monitor other than the original monitor viewed 30-second segments of each conversation tape, recording the beginning and ending times of each behavior. Although five coverbal behaviors were used in the data analysis, seven behaviors were observed in the raw data. This is because two pairs of behaviors (left and right hand movement and left and right hand body touching) were combined into two behaviors (Hand Movement and Touching respectively) prior to data analysis. As pointed out in the Method chapter, the behaviors of interest were hand movement and body touching, with no concern for which hand was involved.

For continuous data, such as the measures in this

study, a satisfactory indication of inter-monitor reliability is provided by the Pearson product moment correlation coefficient. Table 2 contains the Pearson correlation coefficients for the seven coverbal measures. The coefficients range from a low of .79 for Left Hand Movement to a high of .99 for Left Hand Body Touching. Four of the


-41-





-42





Table 2

Pearson Correlation Coefficients for Reliability Data a



Behavior R

Turn Length .98

Facing .96

Left Hand Movement .92

Right Hand Movement 79

Left Hand Body Touching .99

Right Hand Body Touching .98

Head Tilt .89



a Correlations were computed on duration of behavior as originally recorded and as recorded by a different monitor during 30-second intervals for each behavior of each conversation.






-43


seven measures were greater than .9S. The inter-monitor reliability was found to be quite satisfactory for all coverbal measures. Clearly, the behaviors under investigation are observable to the extent that an individual with minimal training can reliably record their occurrence.



Data Analysis


The statistical method used for analyzing the data was canonical correlation. Since the coverbal behaviors Facing (F), Hand Movement (M), Touching (T), and Head Tilt

(HT) were expressed as proportions, these data were transformed before computing the canonical correlations so that a better approximation of a normal distribution would be obtained. The angle arcsin of the square root was the transformation employed. The mean values of the coverbal behaviors and the interpersonal orientation measures appear in Appendix D.

As pointed out in the Method chapter, the canonical correlation is analogous to the more familiar Pearson product moment correlations. However, the test for significance is not carried out in the same manner. In the case of the canonical correlation it is first converted to a chi square statistic which is in turn tested for significance. Similarly to the multiple regression analysis, the canonical correlation yields coefficient values for the original variables. These are the values which, when used as coefficients for the original variables, combine the





-44


original variables to produce the respective canonical variables. That is, using the canonical coefficients to determine the weight (magnitude and direction) of the input of each original variable a new pair of variables (the canonical variables) is created. The resulting two variables (one for each of the two groups of data) have the highest possible correlation of any pair that can be created by linearly combining the variables in the respective groups of data. In order to remove (arbitrary) influence of scale and variability, the coefficients are normalized. Normalizing is accomplished by first multiplying each coefficient by the observed standard deviation of the corresponding variable. The resulting values are then recalled so that the coefficient is equal to 1.0 (absolute value) with each coefficient retaining the appropriate algebraic sign. It is through examination of these normalized values that we get an idea of the relative contribution of each variable to the common information shared by the two groups of data. The canonical variables (C r = coverbal, 10 r = interpersonal orientation) and their normalized coefficients are presented in Table 3.

The first canonical correlation (R I ) was significant

(RI = .89, X2 = 61.56, df = p < .004). The hypothesis that there is a relationship between the coverbal behaviors observed and the interpersonal orientation measures derived from FIRO-B and SOOT was supported. While no further statistical test is appropriate, further information as to








-45












0

4--)
r- 4-J
0


ITI
4--) 0 4-J
4- C:
CL
E cli
0
u
V) tn 11
4J C) C) it 4J



u cli
.1-4 +j
4-4 :?: r- o r
Ln 0 -4 ::3 0
CD 4J -I
0 r- u
0
+ 4-J
r.
0) C rl::l
r-4 C) 4-j 0) 0
A V) tn Ln
LO 0 m 0) M
Ln H 1 0)
> -4 4-4 4
m 0

tn (1) V)
V)
U 4
-4 (1) u
r- r). 4-J r0 CD ;-4 4-) -H
4-J .,.A Q
LO >
0
7 u C: 4-3
A ct 4 V
4 0 a) Z3
.-q C) 0 4-4
4-4
0
LO
4 -1 4-)
0 c cri
4-)
4

Cfi
> 0
.14 0
>
cli 1-4
u r-4
H r- C.)
CD C 0 0 ) 0
CD CD 0 V)
r- 1-4


C)
4 > 4-J
4 0 4 CD 0
-4 P--4 u






-46


the relative importance of each variable is inferred from the individual coefficients.

Looking-at the coefficients of the coverbal behaviors, Head Tilt (HT) with its coefficient of +1.00 provides the greatest input into C r Touching (T) and Movement (M) with coefficients of -.61 and -.54 respectively are of considerable importance. Turn Length (TL) with a coefficient of -.45 is of some importance. The coefficient of -.23 assigned to Facing (F) indicates that facing behavior was not meaningful in the context of this particular canonical relationship.

The strongest single variable from the interpersonal orientation data was Expressed Control (EC) with its coefficient of -1.00.Wanted Control (WC) with a rather large coefficient of +.76 is the next strongest variable entering into 10 r Self-Esteem (SE), Self-Centrality (Cent), and Expressed Inclusion (EI) are of considerable importance with their respective coefficients of -.60, -.60, and -.SS. Wanted Inclusion (WI) and Complexity (Comp) with their respective coefficients of +.05 and -.03 are of virtually no importance in the canonical variable 10 r in this relationship.

The second canonical correlation was not significant
2
(R 2 78 X 34.41, df = 24, p > AS). Thus it was concluded that there were no other linear combinations of the two groups of data which would yield canonical variables which were significantly correlated and which were





-47


uncorrelated with the previously obtained canonical variables. No further statistical analyses were carried out.














CHAPTER IV

DISCUSSION



Personality Type and Coverbal Behavior


The first canonical correlation was statistically

significant. This indicates that there is a significant relationship between the coverbal behaviors observed during the conversations and the interpersonal orientation characteristics measured by FIRO-B and SOOT.

Specifically, Hypothesis 1 stated that there is, overall, a relationship between the coverbal behaviors observed and the interpersonal orientation measures in this study. This hypothesis followed from the findings of previous research (e.g., Allport, 1961; Chapple et al.,

1954; Goldman-Eisler, 1954) and was significantly supported. The principal personality dimension to emerge from this relationship was that of need for control. Expressed control and wanted control, two of the scales from FIRO-B, were the two strongest interpersonal orientation factors in the study. As pointed out in an earlier chapter, it is possible for there to be significant canonical relationships beyond the first relationship. None, however, was found beyond the first relationship in this study. If others had been identified, the coverbal behavior variables and the


-48-






-49


interpersonal orientation measures would have been combined in ways different from the combinations in the first relationship. That would have permitted further discussion of other personality types.

Hypothesis 6 predicted that self-complexity would not

be related to coverbal behavior. This hypothesis was viewed as strongly supported as the self-complexity coefficient was extremely low, the lowest, in fact, of all the coefficients. This hypothesis was consistent with Ziller's (1973) description of self-complexity. Killer's high selfcomplexity individual is seen as a well-balanced person without extreme interpersonal needs, whereas, this study was looking at the behavioral correlates of elevated interpersonal needs. Therefore, support of this hypothesis is seen as further validation of Ziller's concept of selfcomplexity.

Hypothesis 4 predicted that a large amount of head tilting and short turns would be associated with high wanted control and low expressed control. This hypothesis was strongly supported as the predicted relationships-were consistent with the relationships indicated by the canonical coefficients. The person with a high need to be controlled and a low need to control others (the abdicrat) does, in fact, exhibit submissive, non-controlling behavior in the form of a great amount of head tilting and short turns. The relationship equally supports Schutz' concept of the autocrat, that is the person with a high need to control





_50


others and a low need to be controlled by others, as this person exhibits minimal head tilting behavior and long turns. The abdicrat and the autocrat are examples of excessive needs in the interpersonal orientation dimension that Schutz identifies as the interpersonal need for control. It appears then, that two behavioral manifestations of the interpersonal need for control are head tilts and turn length.

Hypothesis 3 predicted that turn length and hand

movement would be directly related to wanted control and self-centrality. This hypothesis received rather strong, although partial, support. Turn length and hand movement are directly related to self-centrality. However, they are inversely related to wanted control. All variables varied in the predicted direction with the exception of wanted control. It should be noted that, of all the hypotheses predicting directional relationships, this is the only hypothesis in which a control variable varied differently from the predicted direction. It is also the only one of the directional hypotheses which makes a prediction about self-centrality. It is possible that self-centrality is more significantly related to turn length and hand movement than is-wanted control. That is, given that a person exhibits lengthy speaking turns and a great deal of hand movement, there may be reason to believe that he is highly self-centered, but insufficient data for predicting his need to control others. His high self-centeredness may






_51


produce lengthy speaking turns and a great deal of hand movement regardless of his need to control others.

Hypothesis 5 predicted that a large amount of body

touching behavior and short turns would be associated with low self-esteem and low expressed control. This hypothesis received partial support. The only measure which did not vary in the predicted direction was body touching. Body touching was directly related to self-esteem and expressed control, rather than inversely related as predicted. In fact, the extent to which body touching contradicted the hypothesized relationship (judging from the relatively large magnitude of its coefficient) was surprising. The strength of this relationship indicates that body touching may not be well understood in terms of its psychological significance. Little previous research has been conducted with this coverbal behavior. Body touching was seen at the formulation of this study as a behavioral manifestation of feelings of anxiety, of the need for self-protection, and the need to not become involved with others (Fast, 1970;

Freedman and Hoffman, 1967; Scheflen, 1965). In fact, however, the canonical relationship indicated that body touching is related to the need for control. This result is more in keeping with the suggestion by Scheflen and Scheflen (1972) that body touching is related to dominance.

Hypothesis 2 predicted that facing and head tilts were directly related to wanted inclusion and inversely related to expressed inclusion. In view of the weak input of three






-52


of the variables (facing was the weakest of all coverbal behaviors, and expressed inclusion and wanted inclusion fell fifth and sixth in order of magnitude of the seven interpersonal orientation measures) there was no support for this hypothesis.

In sum, this study provided statistical support for

the relationship between coverbal behavior and interpersonal orientation. The most salient interpersonal orientation characteristic of this relationship was the interpersonal need for control as measured by the Fundamental Interpersonal Relations Orientation-Behavior scale (Schutz, 1966, 1967). The most salient of the coverbal behaviors of this relationship were head tilts and body touching.

The overall personality types that emerged are clearly the abdicrat and the autocrat as described by Schutz (1966) and the dominance dimension as described by Scheflen and Scheflen (1972). The main coverbal behavioral manifestations of the abdicratic (or submissive) type is a large amount of head tilting, little body touching, and short speaking turns. The coverbal manifestations of the autocratic (or dominant) type is the mirror image: little head tilting, a large amount of body touching, and long speaking turns.



Methodological Considerations


Preserving the naturalness of the experimental

environment was emphasized throughout the study. Subjects






-53


were placed in a comfortable, familiar-type of setting with much care given to making the environment as distraction-free as possible. Data recording during the .interactions was accomplished as unobtrusively as possible, with all recording equipment (except the microphone) at least ten feet from the subjects. Subjects discussed relevant social issues with their peers. No time constraints were placed on the conversations, and no one was present during the conversations except the two interactants. It was felt that, to the extent that naturalness could be preserved, individuals' authentic, unique coverbal styles would be displayed during the conversation. The interactive aspect of the behavior was viewed as critical for this research. In the past, some researchers have studied coverbal behavior as it related to personality by analyzing

data which were obtained during subjects I speaking a monologue (e.g., Gottschalk, Winget, and Gleser, 1969; Vargas, 1968). The monologue seems to be an inappropriate method

for gathering data which are hoped to be shown related to interpersonal orientation. By definition, the interpersonal component of communication is removed from the monologue.

A point should be stated about the sample size in

this study. The sample size was rather small. The principal reason for this is the extremely lengthy work involved in reducing the data from video tapes of conversations to analyzable data. Additionally, this was an exploratory study. As Hays (1963) pointed out, small






-54


sample sizes are often utilized in exploratory studies because the exploratory studies are trying to identify relationships for further study and refinement. The exploratory study serves as a guide for further research, and that is one of the purposes of this study. Further justification for the use of a rather small sample size lies in the test for statistical significance. As is the case with all statistical tests, sample size was taken into consideration in the test for significance in this study. While one could virtually always make the statement, 'the larger the sample size, the better', in this case, there seemed to be adequate support for the use of the rather small sample.

A final point is made in reference to the methodology. Modern videotaping equipment, elaborate computer reduction of the data, and complex statistical analysis were utilized. However, the human observer was an integral part*of the research. No behavioral data were used in this study which could not be reliably observed and recorded by the relatively untrained observer. In order to demonstrate, in a hard-data sense, the relationship between coverbal behavior and interpersonal orientation, the modern electronic equipment was employed. Future application of the findings of the study can, however, be accomplished, without any special equipment and without disruption of normal face-toface conversational interaction. The only equipment required is the ever-present, best of all machines, the human.






-55


Future Research and Application


Adequate evidence was presented to encourage and

warrant further investigation of the relationships between coverbal behavior and interpersonal orientation. Of the interpersonal orientation measures studied, Schutz' control dimensions appeared especially salient. Ziller's selfesteem and self-centrality measures also appeared promising in terms of further research. Following the identification of these personality characteristics as related to coverbal behavior, future research can focus more specifically on these behaviors and their relationships with coverbal behavior. Self-complexity was supported as a measure of personal adjustment and should be explored further in that context.

For the researcher interested in coverbal behavior, there was no evidence pointing to the elimination of any of the five coverbal behaviors for consideration in future research. Turn length and facing, two often studied and apparently important behaviors, appeared as the least powerful of the coverbal behaviors. It may be that the correct combination of behaviors and personality characteristics has not yet been identified in the case of these two variables. This study hoped to identify basic relationships for further study. Which of those relationships are selected is partly a function of the interest of the researcher. Whichever relationships are investigated, the





-56


naturalistic conversational interaction seems to be an excellent setting within which to study them.

Taking a more. long-range viewpoint, the application

of the more refined and more clearly defined relationships has important implications for humans. Scheflen and Scheflen (1972) pointed out in quite some detail the pervasive extent to which coverbal behavior is a vital part of our communication. They presented impressive evidence of the fact that this subtle means of communication can be, and in fact is, extensively utilized to control members of society. Not only is coverbal behavior a powerful means of control, but it is also so subtle that it can be used outside the awareness of the controlled person. Further research will lead to a better understanding of the 'coverbal language'. The subliminal message is a difficult one to which to respond. As heightened awareness is accomplished, more rational, adaptive responses should follow. This may lead to a new freedom for the individual from undesirable control imposed by society. In a more positive way, this increased awareness can result in increased effectiveness on the part of those individuals who are attempting to bring more freedom and individuality for the person, for example in the work of the better trained psychotherapist.
































APPENDICES

































APPENDIX A

FIRO-B









SUBJECT NUMBER DATE

DIRECTIONS: This questionnaire is designed to explore the typical ways you interact with people. There are, of course, no right or wrong answers; each person has his own ways of behaving.

Sometimes people are tempted to answer questions like these in terms of what they think a person should do. This is not what is wanted here. We would like to know how you actually behave.

Some items seem similar to others. However, each item is different so please answer each one without regard to the others. There is no time limit, but do not debate long over any item.

For each statement below, decide which of the following answers best applies to you. Place the number of the answer on the left of the statement. Please be as honest as you can.

1. usually 2. often 3. sometimes 4. occasionally S. rarely 6. never

1. 1 try to be with people.

2. 1 let other people decide what to do.

3. 1 join social groups.

4. 1 try to have close relationships with people.

S. I tend to join social organizations when I have the
opportunity.

6. 1 let other people strongly influence my actions.

7. 1 try to be included in informal social activities.

8. 1 try to have close, personal relationships with
people.

9. 1 try to include other people in my plans.

10. 1 leg other people control my actions.

11. 1 try to have people around me.

12. 1 try to get close and personal with people.

13. When people are doing things together I tend to
join them.



-59-






-60


14. 1 am easily led by people.

15. 1 try to avoid being alone.

16. 1 try to participate in group activities.



For each of the next group of statements, choose one of the following answers:

1. most 2. many 3. some 4. a few
people people people people

5. one or two 6. nobody
people

17. 1 try to be friendly to people.

18. 1 let other people decide what to do.

19. My personal relations with people are cool and
distant.

20. 1 let other people take charge of things.

21. 1 try to have close relationships with people.

22. 1 let other people strongly influence my actions.

23. 1 try to get close and personal with people.

24. 1 let other people control my actions.

2S. I act cool and distant with people.

26. 1 am easily led by people.

27. 1 try to have close, personal relationships with
people.

28. 1 like people to invite me to things.

29. 1 like people to act close and personal with me.

30. 1 try to influence strongly other people's actions.

31. 1 like people to invite me to join in their
activities.

32. 1 like people to act close toward me.

33. 1 try to take charge of things when I am with
people.





-61


34. 1 like people to include me in their activities.

35. 1 like people to act cool and distant toward me.

36. 1 try to have other people do things the way I
want them done.

37. 1 like people to ask me to participate in their
discussions.

38. 1 like people to act friendly toward me.

39. 1 like people to invite me to participate in their
activities.

40. 1 like people to act distant toward me.



For each of the next group of statements, choose one of the following answers:

1. usually 2. often 3. sometimes 4. occasionally S. rarely 6. never

41. 1 try to be the dominant person when I am with
people.

42. 1 like people to invite me to things.

43. 1 like people to act close toward me.

44. 1 try to have other people do things I want done.

45. 1 like people to invite me to join their activities.

46. 1 like people to act cool and distant toward me.

47. 1 try to influence strongly other people's actions.

48. 1 like people to include me in their activities.

49. 1 like people to act close and personal with me.

50. 1 try to take charge of things when I am with
people.

51. 1 like people to invite me to participate in their
activities.

52. 1 like people to act distant toward me.






-62


53. 1 try to have others do things the way I want them
done.

54. 1 take charge of things when I am with people.





























APPENDIX B

GENERAL INSTRUCTIONS FOR SOOT

SAMPLE ITEM FROM SLEF ESTEEM SCALE

SAMPLE ITEM FROM SELF CENTRALITY SCALE

COMPLEXITY SCALE









Social Orientation Tasks


The questions which follow are designed to provide an indication of the way you look at yourself and significant other people. In this description of yourself and others, words are avoided. This is a social psychological instrument designed for research purposes only. Hopefully, it will tell us something about differences among people in

their perceptions of self and others.

This instrument has been approved by the Department of Health, Education and Welfare, Office of Education.

Please work as quickly as possible. It should require little more than ten minutes.































-64-






-65





The circles below stand for people. Mark each circle with the letter standing for one of the people in the list. Do this in any way you like, but use each person only once and do not omit anyone.

F someone who is flunking S yourself 14 the happiest person you Su someone you know who is
know successful
K someone you know who is St the strongest person you
kind know










OOOC)OO





-66


Self Centrality Item


In the large circle below, draw two circles -- one to stand for yourself and a second to stand for a friend. Place an S in the circle for self and an F in the circle for your friend.





-67


Complexity Scale


Instructions: Here is a list of words. You are to read the words quickly and check each one that you think describes YOU. You may check as many or as few words as you like -- but be HONEST. Don't check words that tell what kind of person you should be. Check words that tell what kind of a person you really are.

1. able 20. careless 39. false

2. active 21. charming 40. fine

3. afraid 22. cheerful 41. fierce

4. alone 23. clean 42. foolish

S. angry 24. clever 43. friendly

6. anxious 25. comfortable 44. funny

7. ashamed 26. content 45. generous

8. attractive 27. cruel 46. gentle

9. bad 28. curious 47. glad

10. beautiful 29. delicate 48. good

11. big 30. delightful 49. great

12. bitter 31. different so. happy

13. bold 32. difficult 51. humble

14. brave 33. dirty 52. idle

is. bright 34. dull 53. important

16. busy 35. dumb 54. independent

17. calm 36. eager SS. jealous

18. ca-Dable 37. fair 56. kind

19. careful 38. faithful S7. large






-68


58. lazy 85. serious

59. little 86. sharp

60. lively 87. silly

61. lonely 88. slow

62. loud 89. small

63. lucky 90. smart

64. mild 91. soft

6s. miserable 92. special

66. modest 93. strong

67. neat 94. stupid

68. old 95. strong

69. patient 96. sweet

70. peaceful 97. terrible

71. perfect 98. ugly

72. pleasant 99. unhappy

73. polite 100. unusual

74. poor 101. useful

75. popular 102. valuable

76. proud 103. warm

77. quiet 104. weak

78. quick 105. wild

79. responsible 106. wise

80. rough 107. wonderful

81. rude 108. wrong

82. sad 109. young

83. selfish

84. sensible































APPENDIX C

DISCUSSION PROBLEM I
(ABORTION PROBLEM)

DISCUSSION PROBLEM 2
(LIVING-TOGETHER PROBLEM)









Discussion Problem 1
TKUortion Problem)


Mary is four weeks pregnant. Both she and her husband Bob are somewhat ambivalent about having a baby at this point in their lives. Bob tells Mary that she should do whatever she would be most happy with. Mary can go ahead and have the baby, or she can have an abortion. The couple has a secure income and a baby would present no great financial burden, but Mary is still not sure about having the baby. She has other interests in her life and would like to pursue them; on the other hand, she is not completely adverse to being a mother. What do you think Mary should do? How do you feel about Bob's position on the matter?





























-70-






-71


Discussion Problem 2
jLiving- Together Problem)


Sandra has just gotten home from work and finds her

boyfriend John just ending a telephone conversation. John tells her that he was speaking with his parents and that they are arriving the next morning for his birthday. He then tries to convince Sandra that she will have to make it look like she is not living there, but is merely spending the day. While John feels very committed to Sandra now after three months of living together, he is afraid his parents would not react very well to finding out about them at this time. Sandra, however, is of a different opinion, and states that she does not need to pretend for anyone. What do you think Sandra should do? How do you feel about John's position in the matter?
































APPENDIX D

COVERBAL BEHAVIOR MEAN VALUES

INTERPERSONAL ORIENTATION MEAN SCORES























'cl V)
Cd 4-J rq 4
0 (1)

4-) 00
Cli
Ic:
U




g_
4-)
0 +j C:
0
r-4 0
CD cn u
u

0
Cd u
V) 4J U0

Z3
r 4 u Cli
03 4J Cd
>
--1 4-4
cli cd 0 0
4-J .14
9 L 4J
Q) 0 0 Cd
C) H 4J
0 (A 4-J
,H -71 ;- Q)
> 0 r- 0 H
0 fa4
u 0 0

tn in.


;-4 0
4-) V)


0 U Sa4 00
4-J 0 00 0
r : U) 4-J
Ct -4 CO

>
u
0

4J (1) 0
r=
V)



9:- 4-J 4.J LO
tb 00 4-)
Z
Lo (1) 4-)
V) U


Lo

F 4











-73-












REFERENCES



Adler, A. Social interest. New York: Putnam, 1939.

Allport, G.W. Pattern and growth in personality. New York:
Holt, Rinehart and Winston, 1961.

Argyle, M., and Dean, J. Eye contact, distance, and
affiliation. Sociometry. 1965, 28, 289-304.

Argyle, M. Social Interaction. New York: Atherton Press,
1969.

Ausubel, D. Ego development and the personality disorders.
New Yrok: Grune and Stratton, 1952.

Bateson, G., Jackson, D., Haley, J. and Weakland, J. Toward
a theory of schizophrenia. Behavioral Science, 1956,
1, 251-264.

Cervin, V. Relationship of ascendant-submissive behavior
in dyadic groups of human subjects to their emotional
responsiveness. Journal of Abnormal and Social
Psychology, 1957, 54, 241-249.

Chapple, E.D. Quantitative analysis of the interaction of
individuals. Proceedings of the National Academy of
Science, 1939, 25, 58-67.

Chapple, E.D. Personality differences as described by
invariant properties of individuals in interaction.
Proceedings of the National Academy of Science, 1940,
26, 10-16.

Chapple, E.D. The interaction chronograph: Its evolution
and present application. Personnel, 1949, 25, 285-307.

Chapple, E.D. The standard experimental (stress) interviews as used in interaction chronograph investigations.
Human Organization, 1953, 12, 23-32.

Chapple, E.D., and Arensberg, C.M. Measuring human relations: An introduction to the study of the interaction
of individuals. Genetic Psychological Monographs,
1940, 22, 3-147.



-74-





-75


Chapple, E.D., Chapple, M.F., and Repp, J.A. Behavioral
definitions of personality and temperament characteristics. Human Organization, 1954, 13, 34-39.

Cherry, C. On Human Communication: A Review, a Survey,
and a Criticism. New York: Sciences Editions, 1961.

Dittmann, A.T. The relationship between body movements and
moods in interviews. Journal of Consulting Psychology,
1962, 26, 480.

Duncan, S.D., Jr. Some signals and rules for taking speaking turns in conversation. Journal of Personality and
Social Psychology, 1972, 23, 283-292.

Ekman, P., and Friesen, W.V. Hand movements. Journal of
Communication, 1972, 22, 353-374.

Ellsworth, P.C., and Ludwig, L.M. Visual behavior in
social interaction. Journal of Communication, 1972,
22, 375-403.

Exline, R.V. Visual interaction: The glances of power and
preference. In J.K. Cole (Ed.), Nebraska Symposium on
Motivation (Vol. 19). Lincoln: University of
Nebraska Press, 1971.

Fast, J. Body language. New York: M. Evans and Company,
1970.

Freedman, N., and Hoffman, S. Kinetic behavior in altered
clinical states: Approach to objective analysis of
motor behavior during clinical interviews. Perceptual
and Motor Skills, 1967, 24, 527-539.

Frey, S. Tonic aspects of behavior in interaction. In
A. Kendon, R.M. Harris, and M.R. Key (Eds.), Organization of behavior in face-to-face interaction. The
Hague: Mouton, 1975.

Fromm, E. Escape from freedom. New York: Rinehart, 1941.

Fromm, E. Man for himself. New York: Rinehart, 1947.

Goldman-Eisler, F. A study of individual differences and
of interaction in the behavior of some aspects of
language in interviews. Journal of Mental Science,
1954, 100, 177-197.

Goldman-Eisler, F. Psycholinguistics. London: Academic
Press, 1968.








Gottschalk, L., Winget, C., and Gleser, G. Manual of
Instructions for using the Gottschalk-Gleser Content
Analysis Scales. Berkeley: University of California
Press, 1969.

Hays, W.L. Statistics. New York: Holt, Rinehart and
Winston, 1963.

Kelly, F., Beggs, D. and McNeil, A. Multiple regression
approach. Carbondale, Illinois: Southern Illinois
University Press, 1969.

LaFrance, M., and Mayo, C. Racial differences in gaze
behavior during conversations: Two systematic
observational studies. Journal of Personality and
Social Psychology, 1976, 33, 547-552.

Lorenz, M., and Cobb, S. Language behavior in manic
patients. Archives of Neurology and Psychiatry, 1952,
67, 763-770.

Mahl, G.F. Disturbances and silences in patients' speech
in psychotherapy. Journal of Abnormal and Social
Psychology, 1956, 53, 1-15.

Markel, N.N. (Ed.). Psycholinguistics: An introduction
to the study of speech and personality. Homewood,
Illinois: Dorsey Press, 1969.

Markel, N.N. Coverbal behavior associated with conversation turns. In A. Kendon, R.M. Harris, and M.R. Key
(Eds.), Organization of behavior in face-to-face
interaction. The Hague: Mouton, 1975.

Markel, N.N., Bein, M.F., Campbell, W.W., and Shaw, M.E.
The relationship of self-rating of expressed inclusion
and speaking time. Language and Speech, in press.

Markel, N., Bein, M., Phillis, J. The relationship between
words and tone-of-voice. Language and Speech, 1973,
16, 15-21.

Maslow, A.H. A theory of metamotivation: the biological
rootery of the value of life. Journal of Humanistic
Psychology, 1967, 7, 93-127.

Maslow, A.H. Toward a psychology of being. (2nd ed.).
Princeton: Van Nostrand, 1968.

Matarazzo, J.D., Wiens, A.N., Matarazzo, R.G., and Saslow,
G. Speech and silence behavior in clinical psychotherapy and its laboratory correlates. In J.M. Shlien (Ed.), Research in psychotherapy, Vol. 3. Washington:
American Psychological Association, 1968.






-77


Mehrabian, A. A semantic space for nonverbal behavior.
Journal of Consulting and Clinical Psychology, 1970,
35, 248-257.

Mehrabian, A. Nonverbal communication. In J.K. Cole (Ed.),
Nebraska Symposium on Motivation (Vol. 19). Lincoln:
University of Nebraska Press, 1971.

Mehrabian, A., and Ksionzky, S. Models for affiliative
and conformity behavior. Psychological Bulletin, 1970,
74, 110-126.

Morrison, D.F. Multivariate Statistical Methods. New
York: McGraw-Hill, 1967.

Norwine, A., and Murphy, 0. Characteristic time intervals
in telephonic conversation. Bell System Technical
Journal, 1938, 17, 281-291.

Osgood, C.E., Suci, G.J., and Tannenbaum, P.H. The
measurement of meaning. Urbana, Illinois: University
of Illinois Press, 1957.

Prebor, L.D. The natural history of a conversation.
Doctoral dissertation, University of
Florida, 1972.

Sapir, E. Speech as a personality trait. American Journal
of Sociology, 1927, 82, 895.

Scheflen, A.E. Quasi-courtship behavior in psychotherapy.
Psychiatry, 1965, 28, 245-257.

Scheflen, A.E., and Scheflen, A. Body language and social
order communication or behavioral control. Englewood
Cliffs, N.J.: Prentice-Hall, Inc., 1972.

Schutz, W.C. The interpersonal underworld. Palo Alto,
California: Science and Behavior Books, 1966.

Schutz, W.C. The FIRO scales manual. Los Angeles:
Consulting Psychologists Press, Inc., 1967.

Sullivan, H.S. The interpersonal theory of psychiatry.
New York: Norton, 1953.

Sullivan, H.S. The fusion of psychiatry and social science.
New York: Norton, 1964.

Vargas, R.A. A study of certain personality characteristics
of male college students who report frequent positive
experiencing and behaving. Doctoral
dissertation, University of Florida, 1968.






-78


Verzeano, M., and Finesinger, J.E. An automatic analyzer
for the study of speech in interaction and in free
association. Science, 1949, 110, 45-46.

Woolf, H.B. (Editor in Chief). Webster's New Collegiate
Dictionary. Springfield,. Massachusetts: G.C. Merriam
Co., 1973.

Ziller, R.C. The social self. New York: Pergamon Press
Inc., 1973.














BIOGRAPHICAL SKETCH



William Wallace Campbell was born on September 28,

1939. After graduating from Murrah High School in Jackson, Mississippi, he attended Millsaps College for two years. After working in Washington, D.C. for two years and owning a business in Atlanta for four years, he returned to college in January, 1968. He received his B.A. degree from Georgia State University in Atlanta, Georgia. After working briefly for the Fulton County Family and Children Services in Atlanta he entered graduate school at the University of Florida in September, 1970 to begin his studies in clinical psychology. He was awarded the M.A. degree in December, 1971. He completed his course work in clinical psychology in the summer of 1973. Mr. Campbell completed a one-year internship in clinical psychology at the Department of Clinical Psychology in September, 1974. Shortly after completing his internship he began working for the Jacksonville Drug Abuse Program in Jacksonville, Florida while continuing to work on his doctoral dissertation. He is presently still employed at the Drug Program in Jacksonville. During his first year of graduate study he was a Graduate School Fellow. From the second year of graduate study through his internship he was a United States Public Health


-79-






-80


Service trainee.

Mr. Campbell married the former Jeannie Clay in

December, 1968. He is the father of one daughter, Brooke Elaine, and one son, Ross Bartley.











I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy.



I J Q
Norman N. Markel, Ph.D.,
Chairperson
Professor of Speech, Anthropology, and Psychology



I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy.




Robert C. Killer, Ph.D.
Professor of Psychology



I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor*of Philosophy.




Wiley Rasbury, Ph.\D.
Assistant Professor of
Clinical Psychology









I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy.




Robert H. Hornberger, Ph.D. Assistant Professor of
Clinical Psychology



I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy.




Thomag J. Same, Ph.D.
Assistant Professor of
Speech





This dissertation was submitted to the Graduate Faculty of the Deaprtment of Psychology in the College of Arts and Sciences and to the Graduate Council, and was accepted as partial fulfillment of the requirements for the degree of Doctor of Philosophy.

December, 1976




Dean, Graduate School




Full Text
-3-
communication', and 'body language'. The term that most
adequately describes the phenomena, and the term to be
used throughout this study is 'coverbal', one proposed by
Markel (1975). He defines 'coverbal' as, "behavior of
interlocutors which occurs in association with or
accompanying words, but which is not essential for the
articulation or grammatical functioning of those words"
(p. 189). Further, if we consult the dictionary (Woolf,
1973) we find that 'co' is a prefix form of 'complement'
which is defined in the same source as, "to make complete."
So this coverbal behavior combines with the verbal behavior
to make the complete communication.
The remainder of this chapter will include a section
on the purpose of this study, a discussion of coverbal
behavior and interpersonal orientation, and a general
review of the literature. Following the review of the
literature, a section is presented which discusses the
research rationale with specific references to each coverbal
behavior and each measure of interpersonal orientation.
Next a discussion of the canonical correlation analysis is
presented. The somewhat extensive nature of this discussion
of the statistical analysis is based on the fact that the
canonical correlation analysis is a somewhat complex,
relatively infrequently utilized statistical analysis. The
chapter is concluded with the presentation of the hypotheses
of the study.


xml version 1.0 encoding UTF-8
REPORT xmlns http:www.fcla.edudlsmddaitss xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.fcla.edudlsmddaitssdaitssReport.xsd
INGEST IEID EVMCGIXE2_V4E9FH INGEST_TIME 2014-10-07T00:07:42Z PACKAGE AA00025755_00001
AGREEMENT_INFO ACCOUNT UF PROJECT UFDC
FILES


CHAPTER III
RESULTS
Monitors' Reliability
In order to assess the reliability with which the three
monitors recorded the coverbal behavior, a monitor other
than the original monitor viewed 30-second segments of each
conversation tape, recording the beginning and ending times
of each behavior. Although five coverbal behaviors were
used in the data analysis, seven behaviors were observed
in the raw data. This is because two pairs of behaviors
(left and right hand movement and left and right hand
body touching) were combined into two behaviors (Hand
Movement and Touching respectively) prior to data analysis.
As pointed out in the Method chapter, the behaviors of
interest were hand movement and body touching, with no
concern for which hand was involved.
For continuous data, such as the measures in this
study, a satisfactory indication of inter-monitor reliabil
ity is provided by the Pearson product moment correlation
coefficient. Table 2 contains the Pearson correlation
coefficients for the seven coverbal measures. The coef
ficients range from a low of .79 for Left Hand Movement to
a high of .99 for Left Hand Body Touching. Four of the
-41-


-24-
experience. They are also characteristics which have con
siderable impact on interpersonal behavior. SOOT is not,
however, attempting to directly measure behavioral prefer
ences, as does FIRO-B. Therefore, SOOT should be a useful
predictor of behavior, but in a less direct manner than is
FIRO-B.
Both FIRO-B and SOOT measure interpersonal orientation
characteristics which were of interest in this study. They
measure similar, yet different, characteristics, with
different, verbal as opposed to nonverbal, approaches. It
was for the above reasons that the two instruments were
included in this study.
Statistical Analysis
Canonical correlation analysis was the statistical
procedure employed in this study. This is a relatively
recent statistical technique which social scientists are
beginning to realize can be a very useful method of looking
at two sets of data. The canonical correlation (Kelley,
Beggs and McNeil, 1961; Morrison, 1967) is similar to the more
familiar Pearson product moment correlation where two
variables are analyzed to ascertain the amount of common
information contained in the two. The multiple regression
equation goes a step further, providing a method for
combining the variables in one group of variables and
correlating the newly created variable with a single


-36-
microphone, told the subjects to call him when they were
finished, and left the room.
During the first conversation the first experimenter
had the subjects who were not participating in the con
versation complete half of the paper and pencil tasks that
they were to complete. Following the first conversation,
the subjects just completing their conversation were taken
to other rooms where they completed half of their paper
and pencil tasks while the other two subjects had their
first conversation.
The process was continued until each subject had:
participated in two conversations (one each with a same-
sex and an other-sex partner), completed the background
information form, the FIRO-B, the SOOT, and semantic
differentials relating to how they felt about the experi
ment, the experimenters, and their partners. Of the paper
and pencil instruments, only the FIRO-B and the SOOT were
used in this study.
Subjects were randomly assigned to: either complete
FIRO-B or SOOT first, discuss the 'abortion' problem or
the 'living together' problem first, and converse with a
same-sex or opposite-sex partner first.
Data Reduction
Monitors were trained to observe the tapes and record
the occurrence of the coverbal behaviors listed in Table 1.
Each conversation was monitored once at normal tape speed


REFERENCES
Adler, A. Social interest. New York: Putnam, 1939.
Allport, G.W. Pattern and growth in personality. New York:
Holt, Rinehart and Winston, 1961.
Argyle, M., and Dean, J. Eye contact, distance, and
affiliation. Sociometry. 1965 2_8, 289- 304.
Argyle, M. Social Interaction. New York: Atherton Press,
1969.
Ausubel, D. Ego development and the personality disorders.
New YrokT Grue and Stratton, 1952.
Bateson, G., Jackson, D., Haley, J. and Weakland, J. Toward
a theory of schizophrenia. Behavioral Science, 1956,
1, 251-264.
Cervin, V. Relationship of ascendant-submissive behavior
in dyadic groups of human subjects to their emotional
responsiveness. Journal of Abnormal and Social
Psychology, 1957, 5j4, 241-249.
Chappie, E.D. Quantitative analysis of the interaction of
individuals. Proceedings of the National Academy of
Science, 1939, 2j>_, 58-67.
Chappie, E.D. Personality differences as described by
invariant properties of individuals in interaction.
Proceedings of the National Academy of Science, 1940,
26, 10-16.
Chappie, E.D. The interaction chronograph: Its evolution
and present application. Personnel, 1949 2_5 2 85- 307.
Chappie, E.D. The standard experimental (stress) inter
views as used in interaction chronograph investigations.
Human Organization, 1953, 1_2, 23- 32.
Chappie, E.D., and Arensberg, C.M. Measuring human rela
tions: An introduction to the study of the interaction
of individuals. Genetic Psychological Monographs,
1940, 22, 3-147.
- 74-


-43-
seven measures were greater than .95. The inter-monitor
reliability was found to be quite satisfactory for all
coverbal measures. Clearly, the behaviors under investi
gation are observable to the extent that an individual with
minimal training can reliably record their occurrence.
Data Analysis
The statistical method used for analyzing the data was
canonical correlation. Since the coverbal behaviors
Facing (F), Hand Movement (M), Touching (T), and Head Tilt
(HT) were expressed as proportions, these data were trans
formed before computing the canonical correlations so that
a better approximation of a normal distribution would be
obtained. The angle arcsin of the square root was the
transformation employed. The mean values of the coverbal
behaviors and the interpersonal orientation measures appear
in Appendix D.
As pointed out in the Method chapter, the canonical
correlation is analagous to the more familiar Pearson
product moment correlations. However, the test for sig
nificance is not carried out in the same manner. In the
case of the canonical correlation it is first converted to
a chi square statistic which is in turn tested for signi
ficance. Similarly to the multiple regression analysis,
the canonical correlation yields coefficient values for the
original variables. These are the values which, when used
as coefficients for the original variables, combine the


-60-
14. I am easily led by people.
15. I try to avoid being alone.
16. I try to participate in group activities.
For each of the next group of statements, choose one of the
following answers:
1. most 2. many 3. some 4. a few
people people people people
5. one or two 6. nobody
people
17.
I
try
to be
friendly to people

18.
I
let
other
people
decide what
to do.
19.
My
r personal
relations with peo
pie are cool and
di
stan
t.
20.
I
let
other
people
take charge
of things.
21.
I
try
to have close
relationsh
ips with people.
22.
I
let
other
people
strongly in
fluence my actions.
23.
I
t ry
to get
close
and persona
1 with people.
24.
I
let
other
people
control my
actions.
25.
I
act
cool and dist
ant with pe
ople.
26.
I
am e
asi ly
led by
people.
27.
I
try
to have close
, personal
relationships with
pe
ople

28.
I
like
peop 1
e to in
vite me to
things.
29.
I
like
peopl
e to ac
t close and
personal with me.
30.
I
try
to influence
strongly other people's action
31.
I
like
peopl
e to in
vite me to
join in their
ac
tivi
ties.
32.
I
like
peopl
e to ac
t close toward me.
33.
I
try
to take charge of things
when I am with
people.


-35-
presented in Appendix B.
FIRO-B is presented in its entirity in Appendix A.
Procedures
Subjects were assigned experimental session times by
group so that one group participated in an evening. After
all group members had arrived for a session one experi
menter explained that the purpose of the study was to study
the behavior of persons in conversation in a dyadic
setting. They were told that each of them would partici
pate in two conversations, one each with two of the other
members of their group, and that the conversations would
be recorded on video tape. They were also informed that
they would complete several paper and pencil instruments
which would call for information about themselves and in
formation about how they felt about their experience of
participating in the study. Subjects then signed consent
forms for participation in the experiment.
A second experimenter then took two subjects to the
taping room where they would have their conversations.
The camera and recorder was started by the experimenter.
Subjects were seated at the table with copies of one of
the discussion problems affixed to the table in front of
each of them. They were instructed to read the discussion
problem and discuss as long as they needed to in order to
come to some agreement on a solution. The experimenter
read the subjects' identification numbers into the


APPENDICES


CHAPTER II
METHOD
Subj ects
Twelve male and twelve female students enrolled in
introductory speech class at the University of Florida
volunteered to participate in this study. All subjects
were caucasion; ages ranged from 18 to 27, with a mean age
of 20.3. Two males and two females were selected from the
volunteers from each speech section and those four sub
jects comprised a group for this study. Other than the
specified gender requirement, the only condition imposed
on the composition of the group was that subjects' famil
iarity witli other group members be limited to the inter
action in the speech course in which they were currently
enrolled. The resulting sample then was made up thusly:
six groups of two males and two females; each group drawn
from a different section of the speech course; each sub
ject's prior interaction with his group members was limited
to interaction in speech class.
Equipment
Conversations between pairs of subjects were recorded
- 31-


-19-
behavioral manifestations represent deviations from an ideal
state of need satisfaction according to FIRO theory
(Schtz, 1966). In the inclusion and control areas talking
is seen as an available means by which the person can
strive towards the ideal need satisfaction state. The
person with the high need to control can attempt to gain
control by talking a lot and minimizing the opportunity for
the other person to talk. The person with little need to
control may talk less and look more to the other person for
control cues. The person with high need to be included can
seek attention with hand movements and by talking a great
deal. The person with little desire to be included may
talk less and send, 'I am not interested in being included'
messages with his high rate of body touching.
Self-Other Orientation Tasks. Ziller (1970) is one
among many (e.g., Maslow, 1967, 1968; Schtz, 1966;
Sullivan, 1953, 1964) who proposes that man should be
studied as a social unit. In his social psychological theory
of personality, social adaptation is viewed in terms of self-
other concepts. For a review of his theory see Ziller
(1973). Three of Ziller's self-other dimensions are
employed in this study.
His Self-esteem (SE) and Self-centrality (Cent) scales
utilize what Ziller (1973) refers to as cognitive mapping.
Through the use of ostensibly value-free symbols SE and
Cent are meansured without the bias of mediating 'social
oughts', so often found to be a problem when verbal


-28-
of the canonical variable. As pointed out earlier, several
canonical relationships are possible. If, in the con
struction of the canonical variable(s) for a given group
of the data, the relative directions of two variables within
one group, both of which are in an hypothesized relation
ship, are not the same as they appear in the hypothesis,
full support of the hypothesis is not possible. An example
makes this point more concrete. Noting the second hypoth
esis in this study, it predicts that head tilts and facing
will be positively related to WI scores and inversely
related to SE scores. In addition to the main, overt
prediction that two coverbal behaviors will vary directly
with one interpersonal orientation score and inversely with
another interpersonal orientation score, another prediction
is implied. The other prediction is that facing and head
tilts will vary together and that WI and SE scores will
vary inversely. If the variables are not combined in a way
consistent with these latter, implied predictions, full
support of the hypothesis is not possible. Nonetheless,
to resort to pairwise predictions involves a level of
simplicity inappropriate for this study and the relationships
at interest.
Research has shown that interpersonal aspects of
personality are related to coverbal behavior (Allport, 1961;
Chappie et al., 1954; Goldman-Eisler, 1954).
Hypothesis 1: There is a relationship between coverbal
behavior and interpersonal orientation.


-16-
proportions of the individual's speaking turns during which
he was engaged in the given behavior. The proportion
measure was used rather than straight duration measure in
order to control for turn length. The proportion tells us
not merely how much time was spent in each behavior, rather
it tells how much of the available time was spent engaging
in each behavior.
Interpersonal Orientation
The interpersonal orientation measures used in this
study are Wanted Control (WC), Expressed Control (EC),
Wanted Inclusion (WI), and Expressed Inclusion (El) as
measured with the Funadmental Interpersonal Relations
Orientations Behavior (FIRO-B), and Self-esteem (SE), Self
centrality (Cent), and Se 1f-complexity (Comp) as measured
with the Self-Other Orientation Tasks (SOOT).
Fundamental Interpersonal Relations Orientation-
Behavior. A steadily increasing amount of research is
reporting various approaches to testing Schtz's (1966)
theory of interpersonal orientation. For a summary of
that research see Schtz (1966). Schtz bases his theory
on the existence of three interpersonal needs: Inclusion,
Control, and Affection. Inclusion and affection have been
found to be highly correlated in research (Argyle, 1969).
Therefore, it is meaningful to think of the individual in
terms of the two dimensions inclusion and control. For that


-29-
Facing has been shown to be associated with personal
attraction (Argyle and Dean, 1965; Exline, 1971). Head
tilting behavior appears related to submissiveness (Argyle,
1969) .
Hypothesis 2: F and HT are positively related to WI
and inversely related to El.
Goldman-Eisler (1954) found that subjects talked more
if their self reference was greater. Dittmann (1962) found
hand movement associated with the desire to get approval
from others and to control the other person in conversation.
The person who wants to control his conversational mate will
talk a great deal (Schtz, 1966).
Hypothesis 3: TL and M are positively related to WC
and Cent.
Head tilting is associated with submissiveness (Argyle,
1969). The person desiring to control will talk more than
the person not wishing to control (Schtz, 1966).
Hypothesis 4: High HT and low TL are associated with
high WC and low EC.
Scheflen (1965) found body touching associated with
the need for self-protection. The person wanting control
over others will talk a great deal (Schtz, 1966).
Hypothesis 5: High T and low TL are associated with
low SE and low EC.
Ziller's (19 73) high se 1f-complexity individual is a
well-balanced person. This person is less likely to
outstanding need areas which are unsatisfied. As the
have


-38-
for each behavior for each subject. During that first
monitoring the approximate beginning and ending times of
that behavior were recorded. Using those approximate times
as guides, the tape was then played at slow-motion speed
(approximately 1/15 of normal speed) to record more pre
cisely (usually within 0.1 second) the beginning and ending
times. These times were obtained by using lapsed time
which had been superimposed on the tape with an Odetic
Video Timer after conversations were completed.
Monitors were assisted in determining beginnings and
endings of turns by having a typed verbatim transcript of
the conversations. Periodic checks indicated that tracking
was reliable when monitors had been given several hours
training.
Derived Coverbal Measures
A computer program was developed which used the
beginning and ending times for each coverbal behavior and
for the turns and computed durations and frequencies of
all behaviors per turn. For the purposes of this study,
total hand movement and total body touching were of in
terest, rather than hand movement and body touching broken
down by left and right hand. Therefore, right hand move
ment and left hand movement were combined into one measure:
Hand Movement (M). Also, right hand body touching and
left hand body touching were combined into one measure:
Touching (T). The coverbal measures used in the statistical


-26-
The canonical correlation analysis is a very useful,
somewhat complex, method of attempting to identify complex
relationships. Although there is no statistical test
beyond the overall correlation, specific relationships
are inferred by examining the canonical coefficients. In
this study the relationships being searched for are com
plex. They are complex in the sense that it would be too
simplistic, possibly meaningless, to single out one coverbal
behavior and attempt to establish its relationship with a
single interpersonal orientation variable. Humans, their
behavior, and their personalities are not that simple. To
state that the person who needs to control others will talk
longer may be true. However, there are factors other than
his need to control which contribute to his behavior, and,
there are behaviors other than turn length through which
his interpersonal orientation will manifest itself.
After one pair of canonical variables has been identi
fied and tested, others may also exist with the same data
sets. In fact, there can be as many pairs of canonical
variables as there are variables in the smaller group of
data. The procedure is continued until the correlations
are not significant or until the number of correlations
reaches the number of variables in the smaller data group.
In each new relationship, if more than one exists, there is
a new set of coefficients and the statistical tests are
carried out in the same manner. A more detailed discussion
of these procedures is presented in the Method chapter.


LIST OF TABLES
Table
1 Coverbal Behavior Definitions
2 Pearson Correlation Coefficients for
Reliability Data
*
3 Normalized Canonical Variable Coefficients .
Page
37
42
45
v


-49-
interpersonal orientation measures would have been combined
in ways different from the combinations in the first
relationship. That would have permitted further discussion
of other personality types.
Hypothesis 6 predicted that self-complexity would not
be related to coverbal behavior. This hypothesis was viewed
as strongly supported as the self-complexity coefficient
was extremely low, the lowest, in fact, of all the coef
ficients. This hypothesis was consistent with Ziller's
(1973) description of self-complexity. Ziller's high self
complexity individual is seen as a well-balanced person
without extreme interpersonal needs, whereas, this study
was looking at the behavioral correlates of elevated inter
personal needs. Therefore, support of this hypothesis is
seen as further validation of Ziller's concept of self-
complexity.
Hypothesis 4 predicted that a large amount of head
tilting and short turns would be associated with high
wanted control and low expressed control. This hypothesis
was strongly supported as the predicted relationships were
consistent with the relationships indicated by the canonical
coefficients. The person with a high need to be controlled
and a low need to control others (the abdicrat) does, in
fact, exhibit submissive, non-controlling behavior in the
form of a great amount of head tilting and short turns.
The relationship equally supports Schtz' concept of the
autocrat, that is the person with a high need to control


-73-
Coverbal Behavior Mean Values
Turn
Length
Facing
Movement
Touching
Head
Tilts
5.8
.38
.20
. 70
.22
cl
Turn Length is the mean time in seconds of all turns. All other behavior means
represent the mean value of the proportion of each turn during which the
respective behavior was occurring.
Interpersonal Orientation Mean Scores
FIRO-B
SOOT
WI
El WC
EC
SE
Cent
Comp
5. 7
5.1 3.4
3.8
26.9
2.0
51. 8


Abstract of Dissertation Presented to the Graduate Council
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy
INTERPERSONAL ORIENTATION CORRELATES OF NONVERBAL
BEHAVIOR IN CONVERSATIONAL INTERACTION
By
William Wallace Campbell
December, 1976
Chairperson: Norman N. Markel, Ph.D.
Major Department: Psychology
This study investigated the relationships between
interpersonal orientation and nonverbal behavior. Twelve
male and 12 female students enrolled in undergraduate
speech classes were the subjects. Each subject engaged in
two conversations which were videotaped so that monitors
could record the occurrence of five nonverbal behaviors.
The five nonverbal behaviors studied were: turn length
(TL), facing (F), hand movement (M), body touching (T), and
head tilts (HT). Each subject completed Schtz' Funda
mental Interpersonal Relations Orientation-Behavior (FIRO-B)
scale and the Self-Other Orientation Tasks (SOOT) developed
by Ziller. The seven interpersonal orientation measures
studied were: expressed control (EC), wanted control (WC),
expressed inclusion (El), and wanted inclusion (WI) from
FIRO-B, and self-esteem (SE), self-centrality (Cent), and
vi


-61-
34 .
I
like
people
to
inc
lude me i:
n their acti
viti
Les .
35.
I
like
people
to
act
cool
and
distant tow
ard
me.
36.
I
try
to have
other
people
do
things the
way
I
want them done
37.
I
like
people
to
ask
me to
pa
rticipate in
the
?ir
dj
LSCUS
sions.
38.
I
like
people
to
act
f rien
dly
toward me.
39.
I
like
people
to
invite me
to
participate
in
their
activi
ties.
40.
I
like
people
to
act
distant
toward me.
For each of the next group of statements, choose one of the
following answers:
1. usually 2. often 3. sometimes 4. occasionally
5. rarely 6. never
41.
I
t ry
to be the
dominant pe
rson when I am with
pe
:ople

42.
I
like
people
to
invite me
to
things.
43.
I
like
people
to
act close
toward me.
44.
I
try
to have
other people
do
things I want done.
45.
I
like
people
to
invite me
to
join their activiti
46.
I
like
people
to
act cool and
distant toward me.
47.
I
try
to infli
jen
ce strongly
other people's actions
48.
I
like
people
to
include me
in
their activities.
49.
I
like
people
to
act close
and
personal with me.
50.
I
t ry
to take
ch,
arge of thi
ngs
when I am with
pe
ople

51.
I
like
people
to
invite me
to
participate in their
ac
ti vi
ties.
52.
I
like
people
to
act distan
t t
oward me.


I certify that
opinion it conforms
presentation and is
as a dissertation fo
I have read this study and that in my
to acceptable standards of scholarly
fully adequate, in scope and quality,
r the degree of Doctor of Philosophy.
Robert H. Hornberger, Ph.D.
Assistant Professor of
Clinical Psychology
I certify that
opinion it conforms
presentation and is
as a dissertation fo
I have read this study and that in my
to acceptable standards of scholarly
fully adequate, in scope and quality,
r the degree of Doctor of Philosophy.
1 11 Ulll CL O U U U 111 C j ill LJ
Assistant Professor of
Speech
This dissertation was submitted to the Graduate Faculty of
the Deaprtment of Psychology in the College of Arts and
Sciences and to the Graduate Council, and was accepted as
partial fulfillment of the requirements for the degree of
Doctor of Philosophy.
December, 1976
Dean, Graduate School


-78-
Verzeano, M., and Finesinger, J.E. An automatic analyzer
for the study of speech in interaction and in free
association. Science, 1949, 110, 45-46.
Woolf, H.B. (Editor in Chief). Webster's New Collegiate
Dictionary. Springfield, Massachusetts: G.C. Merriam
Co., 1973.
Ziller, R.C. The social self.
Inc., 1973.
New York: Pergamon Press


-10-
behavior under investigation, certain basic issues arose
relating to quantification and analysis of the coverbal
behavior data. One practical, and apparently meaningful,
scheme for quantifying coverbal behavior is based on its
temporal characteristics (e.g., Chappie, 1939; Goldman-
Eisler, 1968; Mahl, 1956; Norwine and Murphy, 1938),
with the quantity often being simply the duration of time
spent engaging in the behavior. Simple duration, along
with proportion of available time spent engaging in a
behavior, will be the approach to quantifying coverbal
behavior in this study.
As interest in the research field grew, along with an
increase in man's electronic sophistication, new techniques
for measuring, recording, and analyzing coverbal behavior
emerged. Almost thirty years ago, Norwine and Murphy
(1938) developed a sound-activated device for recording
speech behavior. Chappie (1939) was developing his device,
later to be known as the Interaction Chronograph (Chappie,
1949), which utilized both the human observer and mechanical
devices for recording verbal and coverbal behavior. In
his earlier work (Chappie, 1940; Chappie and Arensberg,
1940) Chappie had demonstrated that personality character
istics were associated with certain stable properties of
individuals' interaction style.
Some ten years after Chappie (1939) began his research
which led to the Interaction Chronography, Verzeano and
Finesinger (1949) developed the Automatic Speech Analyzer


APPENDIX C
DISCUSSION PROBLEM 1
(ABORTION PROBLEM)
DISCUSSION PROBLEM 2
(LIVING-TOGETHER PROBLEM)


-8-
than would an overt, verbal rebuke. An increased amount
of eye contact (within acceptable bounds) can communicate
to a woman that a man would like to initiate a relation
ship with her. At the same time, the demands for an overt
response from her are not nearly so great as they would
have been had he used verbal messages to communicate that
same desire. The man can feel less threat of rejection,
and possibly embarrassment, knowing that he has not re
quired a verbal response of the woman. Thus, more sensi
tive, more personal data may more freely be communicated
via coverbal behavior. She is free to simply ignore his
overture, in that case his message has gotten a subtle
response, and his pride remains more intact for his next
adventure. Hopefully, he correctly decoded her message.
For a discussion of the different rules governing verbal
and coverbal behavior, and some of the implications of
those differences, see Markel (1975).
The above examples illustrate two points: 1) the
coverbal message can be as important as, or more important
than, the verbal message, and 2) the receiver of the
coverbal message, especially if untrained in decoding
coverbal communication, can misinterpret the message. Both
of these points are addressed by Fast (1972).
Interpersonal Orientation
Interpersonal orientation is the way an individual
views himself in the context of his social milieu. For


-13-
touching, and Head tilts. Brief research references for
each behavior follow. Operational definitions of each
behavior are presented in Table 1 in the Method chapter.
Turn length. Following early interest in temporal
aspects of communication behavior, this behavior, or some
rough equivalent, has been studied frequently. Goldman-
Eisler (1954), looking at action time (percentage of total
time which was spent in talking), found a relationship
with the content variable of self-reference. That is,
people who referred more to themselves, and presumably
talked more about themselves, spoke during a greater per
centage of the total time. Cervin (1957) found that
individuals scoring high on a scale of emotional respon
siveness also spoke during a greater portion of the
available time. Markel, Bein, Campbell, and Shaw (1976)
demonstrated a greater use of available time by speakers
who scored high on a measure of needed inclusion. This
often investigated behavior of amount of speaking time
(referred to by several roughly equivalent terms such as
'action time' (Goldman-Eisler, 1954), 'participation
quotient' (Prebor, 1972), 'mean percentage', (Matarazzo,
Weins, Matarazzo and Saslow, 1968), etc.), seems to be
related to various aspects of personality.
Facing. Facing behavior was selected because of its
high correlation with eye contact. Eye contact itself is
virtually impossible to detect by observers as they view
videotapes of conversations. Facing, on the other hand,


BIOGRAPHICAL SKETCH
William Wallace Campbell was born on September 28,
1939. After graduating from Murrah High School in Jackson,
Mississippi, he attended Millsaps College for two years.
After working in Washington, D.C. for two years and owning
a business in Atlanta for four years, he returned to college
in January, 1968. He received his B.A. degree from Georgia
State University in Atlanta, Georgia. After working
briefly for the Fulton County Family and Children Services
in Atlanta he entered graduate school at the University of
Florida in September, 1970 to begin his studies in clinical
psychology. He was awarded the M.A. degree in December,
1971. He completed his course work in clinical psychology
in the summer of 1973. Mr. Campbell completed a one-year
internship in clinical psychology at the Department of
Clinical Psychology in September, 1974. Shortly after
completing his internship he began working for the Jackson
ville Drug Abuse Program in Jacksonville, Florida while
continuing to work on his doctoral dissertation. He is
presently still employed at the Drug Program in Jacksonville.
During his first year of graduate study he was a Graduate
School Fellow. From the second year of graduate study
through his internship he was a United States Public Health
- 79-


-54-
sample sizes are often utilized in exploratory studies
because the exploratory studies are trying to identify
relationships for further study and refinement. The ex
ploratory study serves as a guide for further research, and
that is one of the purposes of this study. Further justi
fication for the use of a rather small sample size lies in
the test for statistical significance. As is the case with
all statistical tests, sample size was taken into considera
tion in the test for significance in this study. While one
could virtually always make the statement, 'the larger the
sample size, the better', in this case, there seemed to be
adequate support for the use of the rather small sample.
A final point is made in reference to the methodology.
Modern videotaping equipment, elaborate computer reduction
of the data, and complex statistical analysis were utilized
However, the human observer was an integral part of the
research. No behavioral data were used in this study which
could not be reliably observed and recorded by the rela
tively untrained observer. In order to demonstrate, in a
hard-data sense, the relationship between coverbal behavior
and interpersonal orientation, the modern electronic
equipment was employed. Future application of the findings
of the study can, however, be accomplished, without any
special equipment and without disruption of normal face-to-
face conversational interaction. The only equipment
required is the ever-present, best of all machines, the
human.


-2-
after all the most commonly employed means by which senti
ments and information are exchanged. In spite of its
ubiquitous usage and in spite of a growing and impressive
amount of research, much is left to be understood about
the dyadic conversation. Cherry (1961) observed,
A conversation is one of the commonest
phenomena we encounter, yet it is one which
raises very great scientific problems, many
still unsolved. It is often our commonest
experiences, which we take for granted, that
are elusive of exploration and description.
(p. 10)
Most of us have had the experience of being disturbed
as a result of something that has been said to us and then
realized that it was not what was said, but how it was
said and what behaviors accompanied the spoken word. That
mystical 'third ear' ascribed to some people is largely
their ability to attend, at some level, to some of the
behaviors accompanying the spoken message. Psychotherapists
are utilizing impressions derived from observations with
their 'third ear' both during psychotherapy and as a means
of evaluating past therapeutic efforts. It is anticipated
that the findings of this study, and future studies, will
make a contribution in the effort, in the current vernac
ular, 'to raise the level of consciousness' of therapists
in such situations as they respond to these behavioral
cues .
Many terms have been used to describe what a person
does while he is also speaking, some of those terms are:
'non-content communication', 'kinesics', 'non-verbal


APPENDIX A
FIRO-B


-40-
two groups of data?", can be addressed. Further, an
indication of the strength of the relationship(s) is in
dicated by the level of statistical significance. More
specific ideas about to what extent each original variable
makes its input into the relationship is inferred from the
canonical variable coefficients. Further elaboration on
this statistical procedure and subsequent inferences is
presented in the Result and Discussion chapters.


-33-
apparent interest on subjects' part and based on verbal
response of subjects after the experimental sessions were
completed, subjects actually became quite involved in the
discussion problems. The texts of these problems are
presented in Appendix C.
Interpersonal Orientation Instruments
The two instruments utilized for measuring inter
personal orientation of subjects were the Fundamental
Interpersonal Relations Orientation-Behavior (FIRO-B)
developed by William Schtz and Self-Other Orientation
(SOOT) scale developed by Robert Ziller.
FIRO-B is composed of 54 items and has been used with
subjects ranging from eighth grade to adults. The results
of this instrument yield six scales: Expressed Inclusion,
Wanted Inclusion, Expressed Control, Wanted Control,
Expressed Affection, and Wanted Affection. Only the first
four of these scales were used in this study. The possible
range of scores on these six scales is from 1 to 9, with
all scores being integers. This instrument was designed
by Schtz (1966, 1967) to measure the relatively stable
attitudes of an individual which determine his interactive
behavior. Schtz has published considerable research
demonstrating validity for his instrument.
SOOT was designed to measure individuals' self-other
orientation in terms of their perceived relationships.
The method of assessing self-esteem and self-centrality is


-68-
58.
lazy
59.
little
60.
lively
61.
lonely
62.
loud
63.
lucky
64.
mild
65.
miserable
66.
modest
67.
neat
68.
old
69.
patient
70.
peace ful
71.
perfect
72.
pleasant
73.
polite
74 .
poor
75.
popular
76.
proud
77.
quiet
78.
quick
79.
responsible
80.
rough
81.
rude
82.
sad
83.
se 1fish
84.
sensible
85.
serious
86.
sharp
87.
silly
88.
slow
89.
small
90.
smart
91.
soft
92 .
special
93.
stronge
94.
stupid
95.
strong
96.
sweet
97.
terrible
98.
ugly
99.
unhappy
100.
unusual
101.
useful
102.
valuable
103.
warm
104.
weak
105.
wild
106 .
wise
107.
wonderful
108.
wrong
109.
young


Discussion Problem 1
(Abortion Problem)
Mary is four weeks pregnant. Both she and her husband
Bob are somewhat ambivalent about having a baby at this
point in their lives. Bob tells Mary that she should do
whatever she would be most happy with. Mary can go ahead
and have the baby, or she can have an abortion. The couple
has a secure income and a baby would present no great
financial burden, but Mary is still not sure about having
the baby. She has other interests in her life and would
like to pursue them; on the other hand, she is not complete
ly adverse to being a mother. What do you think Mary should
do? How do you feel about Bob's position on the matter?
- 70-


-21-
the Comp score, and the more complex is the self construct.
Ziller (1973) referred to the self-esteem as that
component of the individual which mediates modification
of the self in response to new information which is re
ceived about the self in social interaction. This view
is consistent with the notion that self-esteem is the
individuals perception of his own self-worth (Osgood, Suci,
and Tannenbaum, 1957). Following these positions, as
Ziller states, the individual with high self-esteem is
better able to evaluate and assimilate new information.
He is not the helpless, vulnerable pawn, subject to re
structuring the self-concept immediately upon receipt of
information which is in conflict with the then existing
self-concept. The high self-esteem person uses new infor
mation as it fits for him, the low self-esteem is ever
seeking new information so that he may modify the self
to fit what the information tells him others think he
should be.
The self-centrality scale is concerned with whether
the person uses the self or others as his reference point
(Ziller, 1973). The higher a person's self-centrality
score, the more he uses the self as the reference point.
Ziller compares high and low self-centrality with Ausubel's
(1952) terms egocentrism and sociocentrism. The high self
centrality person appears incapable of perceiving his
environment from others' viewpoints. He withdraws and
does not want new information from others. By implication,


-50-
others and a low need to be controlled by others, as this
person exhibits minimal head tilting behavior and long
turns. The abdicrat and the autocrat are examples of
excessive needs in the interpersonal orientation dimension
that Schtz identifies as the interpersonal need for con
trol. It appears then, that two behavioral manifestations
of the interpersonal need for control are head tilts and
turn length.
Hypothesis 3 predicted that turn length and hand
movement would be directly related to wanted control and
self-centrality. This hypothesis received rather strong,
although partial, support. Turn length and hand movement
are directly related to self-centrality. However, they are
inversely related to wanted control. All variables varied
in the predicted direction with the exception of wanted
control. It should be noted that, of all the hypotheses
predicting directional relationships, this is the only
hypothesis in which a control variable varied differently
from the predicted direction. It is also the only one of
the directional hypotheses which makes a prediction about
self-centrality. It is possible that self-centrality is
more significantly related to turn length and hand move-
9
ment than is wanted control. That is, given that a person
exhibits lengthy speaking turns and a great deal of hand
movement, there may be reason to believe that he is highly
self-centered, but insufficient data for predicting his
need to control others. His high self-centeredness may


-5-
Schutz (1966, 1967), and Self-esteem (SE) Self-centrality
(Cent), and Self-complexity (Comp) as measured by the Self-
Other Orientation Tasks (SOOT) developed by Ziller (1973).
Coverbal Behavior
By definition coverbal behavior always accompanies
verbal behavior. Whatever and wherever we are verbalizing,
we are simultaneously engaged in some behavior other than
the articulation of words according to culturally pre
scribed rules. Under certain circumstances the coverbal
behavior may not be a perceived component of the communi
cation. For example, most coverbal behavior of interest
in this research is transmitted via the visual channel
(Markel, 1969). If that channel is unavailable to the
person to whom the communication is directed, obviously
these coverbal behaviors cannot be a part of the total
communication as perceived by that listener. This would
be the case when a sightless person is the listener, when
the interaction takes place in total darkness, etc. These
types of special conditions certainly have impact on the
listener's sensitivity to other communication channels
(e.g., auditory, olfactory, etc.), but that is not the focus
of this research. A basic premise of this research is that
the coverbal behaviors under investigation are available
for processing by the listener or listeners.
Coverbal behavior can communicate information which
is complementary to, contradictory to, or independent of,


-15-
of duration of occurrence.
Body touching. Research on body touching (touching
one's own body with his hand) has not been extensive.
Freedman and Hoffman (1967) found body touching to be
related to anxiety level. Scheflen (1965) found body
touching associated with the need for self-protection.
Fast (1970), summarizing research in the field, asserts
that body touching can be a coverbal method of sending a
'hands off' message to others. In other words, the person
who does not want to be included can say so with his hands.
The findings of these three authors are seen as being
consistent in that the person who does not want to be
included may experience anxiety and feel the need for
self-protection.
Head tilts. Head tilt behavior appears to be an
indication of attentiveness and possibly cognitive activity
during conversation. In animal research it appears to be
associated with submissiveness. This relationship may
also hold for humans. Mehrabian (1971) and Mehrabian and
Ksionzky (1970) found head tilt related to relaxation of
the person.
Derived Coverbal Measures
The derived measure for Turn Length in this study was
the duration of the speaking turn as defined in Table 1 in
the Method chapter. The derived measures for Facing, Hand
Movement, Body Touching and Head Tilts were computed


-30-
self-complexity drops, idiosynchratic needs will emerge.
It seems that these needs will emerge in an unsystematic
fashion, not predictable from the self-complexity level of
the person.
Hypothesis 6: Comp is not correlated with coverbal
behavior.


-65-
The circles below stand for people. Mark each circle with
the letter standing for one o
this in any way you like, but
and do not omit anyone.
F someone who is flunking
H the happiest person you
know
K someone you know who is
kind
the people in the list. Do
use each person only once
S yourself
Su someone you know who is
successful
St the strongest person you
know


-11-
which eliminated the necessity of the human observer for
obtaining data during interaction. Lorenz and Cobb (1952),
using the Automatic Speech Analyzer, were able to differ
entiate normals and psychiatrically impaired on the basis
of temporal aspects of their speech patterns. Chappie
(1953) went on to develop his standardized interview, with
which, using his Interaction Chronograph., he investigated
subjects' interview behavior and its relationship to
certain personality dimensions (Chappie, Chappie, and Repp,
1954) .
Man's imagination and increasingly sophisticated
electronic equipment available to him, continue to create
innovative methods for recording and analyzing coverbal
behavior. As can be noted from previously cited research,
as different techniques were developing for collecting
coverbal data, the data were being examined in
numerous ways and contexts. In the area of psychological
interests alone, researchers were relating coverbal behavior
to social orientation, transient emotional states, degree
of psychological health, and interpersonal needs, to name
a few. With this wealth of techniques and the potential
applications of the obtained data, a primary task of the
researcher becomes one of choosing which techniques to
utilize and to what end. That brings us to this study.
A discussion of how those decisions were made in this study
follows.


- 37-
Table 1
Coverbal Behavior Definitions
Behavior
Definition
1. Turn
2. Facing
3. Right hand movement
4. Left hand movement
5. Right hand body touching
6. Left hand body touching
7. Head tilt
Begins when one speaker starts
talking and ends when the
other speaker starts solo
talking
Subject's nose is pointed
roughly at the center of
partner's face
Any movement of subject's
right hand and wrist, unless
it results from a movement of
the whole torso
Any movement of subject's
left hand and wrist, unless
it results from a movement of
the whole torso
Any part of subject's right
hand touching any part of
his body
Any part of subject's left
hand touching any part of
his body
Subject's eyes are not in a
horizontal plane


-80-
Service trainee.
Mr. Campbell married the former Jeannie Clay in
December, 1968. He is the father of one daughter, Brooke
Elaine, and one son, Ross Bartley.


CHAPTER IV
DISCUSSION
Personality Type and Coverbal Behavior
The first canonical correlation was statistically
significant. This indicates that there is a significant
relationship between the coverbal behaviors observed
during the conversations and the interpersonal orientation
characteristics measured by FIRO-B and SOOT.
Specifically, Hypothesis 1 stated that there is,
overall, a relationship between the coverbal behaviors
observed and the interpersonal orientation measures in
this study. This hypothesis followed from the findings of
previous research (e.g., Allport, 1961; Chappie et al.,
1954; Goldman-Eisler, 1954), and was significantly supported.
The principal personality dimension to emerge from this
relationship was that of need for control. Expressed con
trol and wanted control, two of the scales from FIRO-B,
were the two strongest interpersonal orientation factors
in the study. As pointed out in an earlier chapter, it is
possible for there to be significant canonical relationships
beyond the first relationship. None, however, was found
beyond the first relationship in this study. If others had
been identified, the coverbal behavior variables and the
-48-


-51-
produce lengthy speaking turns and a great deal of hand
movement regardless of his need to control others.
Hypothesis 5 predicted that a large amount of body
touching behavior and short turns would be associated with
low self-esteem and low expressed control. This hypothesis
received partial support. The only measure which did not
vary in the predicted direction was body touching. Body
touching was directly related to self-esteem and expressed
control, rather than inversely related as predicted. In
fact, the extent to which body touching contradicted the
hypothesized relationship (judging from the relatively
large magnitude of its coefficient) was surprising. The
strength of this relationship indicates that body touching
may not be well understood in terms of its psychological
significance. Little previous research has been conducted
with this coverbal behavior. Body touching was seen at the
formulation of this study as a behavioral manifestation of
feelings of anxiety, of the need for self-protection, and
the need to not become involved with others (Fast, 1970;
Freedman and Hoffman, 1967; Scheflen, 1965). In fact, however,
the canonical relationship indicated that body touching is
related to the need for control. This result is more in
keeping with the suggestion by Scheflen and Scheflen (1972)
that body touching is related to dominance.
Hypothesis 2 predicted that facing and head tilts were
directly related to wanted inclusion and inversely related
to expressed inclusion. In view of the weak input of three


Table 3
Normalized Canonical Variable Coefficients
C = 1.00HT .6IT .54M .45TL .23F
r
10 = -l.OOEC + .76WC .60SE .60Cent .55 El + .05WI .03Comp
C = Canonical variable for coverbal behavior
r
10 = Canonical variable for interpersonal orientation
Coverbal Behavior: HT = Head Tilt, T = Touching, M = Hand Movement,
TL = Turn Length, F = Facing
Interpersonal Orientation Data: EC = Expressed control, WC = Wanted control,
SE = Self-esteem, Cent = Self-centrality,
El = Expressed inclusion, WI = Wanted
inclusion, Comp = Complexity
-45-


-14-
is rather reliably observed and recorded by the relatively
untrained observer. Clinically, eye contact is a frequently
utilized diagnostic tool. Socially, this behavior seems to
be thought of as an indicator of the speaker's candor. This
is a good example of coverbal behavior that can be easily
mis-interpreted if the listener fails to take into account
cultural factors such as the norms of the culture from
which the speaker comes. Argyle and Dean (1965) and Exline
(1971) found facing behavior positively correlated with
personal attraction. Exline (1971) and LaFrance and Mayo
(1976) found facing associated with dominance. Duncan
(1972) conducted research which concluded that facing is
an effective tool for controlling speaking turn during
conversations. When the speaking person looks away from
the listener, our culture tells us that that is not a
polite time for the listener to attempt taking the role
of speaker. A review of the literature on facing can be
found in Ellsworth and Ludwig (1972).
Hand movement. Hand movement seems to be a method
of attracting attention and controlling the interaction.
Dittmann(1962) found it related to approval seeking in
subjects. Mehrabian (1970, 1971) found hand movement to
be associated with attraction and the desire for affilia
tion. While some investigators have devised elaborate
systems of quantitatively and qualitatively classifying
hand movement (e.g., Ekman and Friesen, 1972; Frey,
1975), this study looks at hand movement simply in terms


-44-
original variables to produce the respective canonical
variables. That is, using the canonical coefficients to
determine the weight (magnitude and direction) of the input
of each original variable a new pair of variables (the
canonical variables) is created. The resulting two vari
ables (one for each of the two groups of data) have the
highest possible correlation of any pair that can be
created by linearly combining the variables in the respec
tive groups of data. In order to remove (arbitrary) in
fluence of scale and variability, the coefficients are
normalized. Normalizing is accomplished by first multi
plying each coefficient by the observed standard deviation
of the corresponding variable. The resulting values are
then rescaled so that the coefficient is equal to 1.0
(absolute value) with each coefficient retaining the
appropriate algebraic sign. It is through examination of
these normalized values that we get an idea of the relative
contribution of each variable to the common information
shared by the two groups of data. The canonical variables
(Cr = coverbal, 10 = interpersonal orientation) and their
normalized coefficients are presented in Table 3.
The first canonical correlation (R^) was significant
(R^ = .89, y2 = 61.56, df = p < .004). The hypothesis that
there is a relationship between the coverbal behaviors
observed and the interpersonal orientation measures derived
from FIRO-B and SOOT was supported. While no further
statistical test is appropriate, further information as to


Social Orientation Tasks
The quest
indication of
other people,
words are avoi
ment designed
will tell us s
their percepti
This inst
Health, Educat
Please wo
little more th
ions which follow are designed to provide an
the way you look at yourself and significant
In this description of yourself and others,
ded. This is a social psychological instru-
for research purposes only. Hopefully, it
omething about differences among people in
ons of self and others.
rument has been
approved by
the Depar
tment of
ion
and Welfare,
Office of
Education.
rk
as quickly as
possible.
It should
require
an
ten minutes.
-64-


-47-
uncorrelated with the previously obtained canonical vari
ables. No further statistical analyses were carried out


-4-
Purpose of the Study
The purpose of this study is to investigate the rela
tionship between coverbal behavior and interpersonal
orientation. The point will be made that research has
shown coverbal behavior to be rather stable over time for
an individual, while each person has a unique style of
coverbal behavior. Thus coverbal behavior offers a reliable
measure with which to discriminate between individuals.
Interpersonal orientation is also relatively stable and
unique for the individual (Maslow, 1968; Ziller, 1973). As
Ziller (1973) points out, interpersonal orientation can be,
and is, modified over time, else it would be nonadaptive.
However, this change is gradual and this characteristic of
being gradually modified gives stability to the person.
The major question addressed by this study is, "Is
there a relationship between coverbal behavior and inter
personal orientation?" Following that question, an attempt
will be made to identify the nature of specific relation
ships within that overall relationship. The coverbal
behaviors under investigation are: Turn Length (TL) ,
Facing (F), Hand Movement (M), Body Touching (T), and Head
Tilts (HT). The interpersonal orientation characteristics
to be employed are: Wanted Inclusion (WC), Expressed
Inclusion (El), Wanted Control (WC), and Expressed Control
(EC), as measured by the Fundamental Interpersonal
Relations Orientation-Behavior (FIRO-B) scale developed by


-62-
53. I try to have others do things the way I want them
done.
54. I take charge of things when I am with people.


I certify that I have read this study and that in my
opinion it conforms to acceptable standards of scholarly
presentation and is fully adequate, in scope and quality,
as a dissertation for the degree of Doctor of Philosophy.
y )'] }'?1 CUt
Norman N. Markel, Ph.D.,
Chairperson
Professor of Speech, Anthro
pology, and Psychology
I certify that
opinion it conforms
presentation and is
I have read this study and that in my
to acceptable standards of scholarly
fully adequate, in scope and quality,
as a dissertation for the degree of Doctor of Philosophy.
Robert C. Ziller, Ph.D.
Professor of Psychology
I certify that
opinion it conforms
presentation and is
as a dissertation fo
I have read this study and that in my
to acceptable standards of scholarly
fully adequate, in scope and quality,
r the degree of Doctor of Philosophy.
Wiley Q. Rasbury, Ph.\
Assistant Professor of
Clinical Psychology


-32-
in a carpeted room with dimensions of approximately 10 x 15
feet. Subjects were seated at adjacent sides of a padded
card table in comfortable, padded arm chairs in one corner
of the room. Each subject was seated at approximately a
45 angle to a Sony 3210 video camera equipped with zoom
lens. The video camera was connected to a Sony AV 3600
Videocorder which was the recorder used for recording the
audio, as well as the video, signals during conversations.
A Sony microphone was mounted on a stand between and
slightly behind the subjects. Subsequent to the exepri-
mental sessions, continuous lapsed time in minutes, seconds,
and tenths of seconds from beginning to end of each conver
sation was superimposed on the video tape by means of a
Model VTG33 Odetic Video Timer.
Discussion Problems
Each of the discussion problems was of assumed and
apparent contemporary social interest. One problem, the
"abortion" problem, involved a young married couple groping
with the decision of whether or not the pregnant wife
should have an abortion. The other problem, the "living
together" problem, concerned a young, heterosexual, un
married couple, living together, trying to decide how they
should present their living situation to his parents, who
were unaware of their cohabitation, while the parents were
in town for a brief visit. Judging from observed level of


ACKNOWLEDGEMENTS
The list of friends, fellow students, and professors
whose encouragement, criticism, and support merits their
being listed here is too long to include. Two friends,
however, cannot be omitted, they are Jeannie Clay and
Dr. Cynthia Gallois.
My supervisory committee members were truly important
in many ways. Dr. Wiley Rasbury, Dr. Robert Hornberger
and Dr. Richard McGee played major roles in my clinical
development. Dr. Robert Ziller, with his broad and varied
interests, rejuvenated my research interests in each
interaction we had. Many thanks are extended to Dr. Thomas
Sain for his help and his suggestions through this lengthy
research experience. Dr. Norman Markel with his concern
for learning, and his concern for people, has had tremen
dous impact on me as a person, as a student and as a
researcher. Without his insights, his suggestions, his
many hours of support (given seemingly without limit), this
research and this dissertation would have had a much lower
probability of successful completion.


manifestations of the autocratic (or dominant) type were
the mirror image:
body touching, and
little head tilting, a large amount of
long speaking turns.
vi ii


-75-
Gottschalk, L., Winget, C., and Gleser, G. Manual of
Instructions for using the Gottschalk-Gleser Content
Analysis Scales^ Berkeley: University of California
Press, 1969.
Hays, W.L. Statistics. New York: Holt, Rinehart and
Winston, 1963.
Kelly, F., Beggs, D. and McNeil, A. Multiple regression
approach. Carbondale, Illinois: Southern Illinois
University Press, 1969.
LaFrance, M., and Mayo, C. Racial differences in gaze
behavior during conversations: Two systematic
observational studies. Journal of Personality and
Social Psychology, 1976, 3_3, 547-552.
Lorenz, M., and Cobb, S. Language behavior in manic
patients. Archives of Neurology and Psychiatry, 1952,
67, 763-770.
Mahl, G.F. Disturbances and silences in patients' speech
in psychotherapy. Journal of Abnormal and Social
Psychology, 1956 5J5, 1-15.
Markel, N.N. (Ed.). Psycholinguistics: An introduction
to the study of speech and personality. Homewood,
Illinois: Dorsey Press, 1969.
Markel, N.N. Coverbal behavior associated with conversa
tion turns. In A. Kendon, R.M. Harris, and M.R. Key
(Eds.), Organization of behavior in face-to-face
interaction. The Hague: Mouton, 1975.
Markel, N.N., Bein, M.F., Campbell, W.W., and Shaw, M.E.
The relationship of self-rating of expressed inclusion
and speaking time. Language and Speech, in press.
Markel, N., Bein, M., Phillis, J. The relationship between
words and tone-of-voice. Language and Speech, 1973,
IS, 15-21.
Maslow, A.H. A theory of metamotivation: the biological
rootery of the value of life. Journal of Humanistic
Psychology, 1967 7_, 93- 127.
Maslow, A.H. Toward a psychology of being. (2nd ed.).
Princeton^ Van Nost rand, 1968.
Matarazzo, J.D., Wiens, A.N., Matarazzo, R.G., and Saslow,
G. Speech and silence behavior in clinical psycho
therapy and its laboratory correlates. In J.M. Shlien
(Ed.), Research in psychotherapy, Vol. 3. Washington:
American Psychological Association, 1968.


-53-
were placed in a comfortable, familiar-type of setting
with much care given to making the environment as dis
traction-free as possible. Data recording during the
interactions was accomplished as unobtrusively as possible,
with all recording equipment (except the microphone) at
least ten feet from the subjects. Subjects discussed
relevant social issues with their peers. No time con
straints were placed on the conversations, and no one was
present during the conversations except the two interactants.
It was felt that, to the extent that naturalness could be
preserved, individuals' authentic, unique coverbal styles
would be displayed during the conversation. The interactive
aspect of the behavior was viewed as critical for this
research. In the past, some researchers have studied
coverbal behavior as it related to personality by analyzing
data which were obtained during subjects' speaking a mono
logue (e.g., Gottschalk, Winget, and Gleser, 1969; Vargas,
1968). The monologue seems to be an inappropriate method
for gathering data which are hoped to be shown related to
interpersonal orientation. By definition, the interpersonal
component of communication is removed from the monologue.
A point should be stated about the sample size in
this study. The sample size was rather small. The prin
cipal reason for this is the extremely lengthy work
involved in reducing the data from video tapes of conver
sations to analyzable data. Additionally, this was an
exploratory study. As Hays (1963) pointed out, small


-34-
the use of ostensibly value-free symbols which result in
subjects' supplying more valid information about themselves
than would be expected in more traditional, verbal, self-
report inventories. In the self-esteem scale symbols
represent the self, significant other people such as
family members, teachers, friends, etc. The subjects'
tasks were to arrange the symbols according to certain
instructions. Scoring is accomplished by obtaining the
sum of the assigned numerical values according to the
position in which the symbol representing the self is
placed. The self-centrality score is based on whether the
'self' is placed nearer to or farther from the center of a
large circle than is the symbol representing a friend.
Each scale has six items. The self-centrality items are
identical throughout the scale. The only difference between
the six items of the self-esteem scale is in the significant
other people which the symbols represent. The complexity
scale consists of a 109-word adjective checklist. Even
though the SOOT is a relatively new instrument, a con
siderable amount of research has demonstrated the relia
bility and validity of the scales (Ziller, 1973).
The SOOT scales used in this study are Self-Esteem,
Self-Centrality, and Self-Complexity. Sample items from the
self-esteem and self-centrality scales are presented in
Appendix B. Because of the similarity of the items, only
one item per scale is presented here. The entire self-complex
ity scale and the general instructions for SOOT are also


APPENDIX B
GENERAL INSTRUCTIONS FOR SOOT
SAMPLE ITEM FROM SLEF ESTEEM SCALE
SAMPLE ITEM FROM SELF CENTRALITY SCALE
COMPLEXITY SCALE


-39-
analysis were the grand mean per behavior for subjects
across both their conversations, after summing across turns
In the case of the measure Turn Length (TL), it was the
grand mean of the turn lengths across both conversations.
The behaviors: Facing (F), Hand Movement (M), Touching (T)
and Head Tilt (HT) were computed as proportions of the
respective turn which was spent engaging in those behaviors
The grand mean for each subject that was used for these
behaviors then was the mean obtained across all speaking
turns of both conversations using those proportions.
Data Analysis
The statistical method for analysis of the relation
ship between the two groups of data (interpersonal orienta
tion and coverbal) was the canonical correlation. There
were five variables in the coverbal data: Turn Length (TL)
Facing (F) Movement (M), Touching (T), and Head Tilt (HT).
Seven Personality variables made up the other group:
Expressed Control (EC) Wanted Control (WC), Expressed
Inclusion (El), Wanted Inclusion (WI), Self-Esteem (SE),
Self-Centrality (Cent), and Complexity (Comp). The first
four of the personality variables are scales from the
Fundamental Interpersonal Relations Orientation-Behavior
(FIRO-B) and the last three are scales from the Self-Other
Orientation Tasks (SOOT).
Canonical correlation analysis is a method whereby the
question, "Is there a significant relationship between the


-18-
"submissive and a follower" (p. 22). How these
behaviors are manifested will be influenced by the
individual's needs in the I and C dimensions as to whether
he wants to be included (or controlled) or whether he wants
to include (or control) others. Schtz (1966) further
makes the point that there is not necessarily a comple
mentary relationship between wanted and expressed needs
in a given area. The person with a high need to control
others may also have a high need to be controlled by others.
The domineering sergeant may need and gratefully accept
control from his lieutenant.
According to his classification the 'oversocial'
person is the one who exhibits excessive inclusion behavior
and the 'undersocial' is the person who makes little or no
attempt to include or be included. The 'autocrat' tends
towards the domineering personality and is characterized
by the excessive desire for power and control. Much of
the autocrat's behavior is directed towards controlling
others. The 'abdicrat' is the person who makes no attempt
to exert control over others. Schtz (1966) points out the
extreme case of the abdicrat who wants no control over
others and at the same time has a high need for others to
control him. This person does not say, 'I am going to
leave you alone and I would like for you to leave me alone',
rather he is saying, 'I do not want to control you, but
please tell me what to do'.
The above personality types with their accompanying


-9-
some major personality theorists, understanding the
importance of other people to the individual, and the
individual's need to relate to other people, is basic to
understanding the individual. Fromm (1941, 1947) spoke
of man's feelings of isolation and threat as he becomes
increasingly separated from other people. Adler (1939)
felt that man's social interest and need for other people
were instinctive. That is, man does not merely learn to
need contact and affiliation with others, rather he is born
with that social need as a biological given. Sullivan
(1953, 1964), probably the most significant single figure
in advancing the social perspective as a means for under
standing the individual, felt that it was meaningless to
attempt viewing the individual in any other context than
the social context. In his interpersonal approach to
psychiatry he pointed out that we all enter and leave this
world as social beings and that the most fruitful way of
conceptualizing and treating humans is as social beings.
If we look at man's basic needs as postulated by Maslow
(1967, 1968) we find among them social needs such as the
need for affiliation and affection. An accurate and
adequately high self-esteem, shaped through social inter
action, is also a basic need according to Maslow (1968).
Review of the Research Literature
For some time researchers have been interested in the
psychological importance of coverbal behavior. As with any


-20-
self-reports are used for tapping these aspects of per
sonality. Subjects are asked to place symbols, representin
themselves and other people, somewhere in a prescribed area
The SE scale, for instance, utilizes the proclivity in our
culture for people to place items of higher value toward
the left in a horizontal ordering. On the Cent scale the
subject is presented with a large circle and asked to place
two smaller circles, representing self and a friend, some
where in the large circle. Scoring of the Cent scale is
on the basis of whether the self or the frient is placed
nearer the center of the large circle. The more often the
self is placed nearer the center, the higher his Cent score
The SE and the Cent scales each contain six items as
described above. All six items on the SE scale are iden
tical except with respect to which five significant other
people are represented by the symbols accompanying the ever
present sixth symbol representing self. All six items on
the Cent scale are identical in every respect. Because of
the similarity of the items within each scale, only one
sample item from each scale is presented. The sample
items are found in Appendix B. The Comp scale is an
adjective checklist scale containing 109 adjectives. The
Comp scale, in its entirity, is presented in Appendix B.
Subjects were instructed to check as many adjectives as
they felt applied to them. The Comp score is computed by
totaling the number of adjectives checked by the subject.
The higher the number of adjectives checked, the higher


-77-
Mehrabian, A. A semantic space for nonverbal behavior.
Journal of Consulting and Clinical Psychology, 1970,
15, 248-257.
Mehrabian, A. Nonverbal communication. In J.K. Cole (Ed.),
Nebraska Symposium on Motivation (Vol. 19). Lincoln:
University of Nebraska Press, 1971.
Mehrabian, A., and Ksionzky, S. Models for affiliative
and conformity behavior. Psychological Bulletin, 1970,
_74 110- 126.
Morrison, D.F. Multivariate Statistical Methods. New
York: McGraw-Hill, 1967.
Norwine, A., and Murphy, 0. Characteristic time intervals
in telephonic conversation. Bell System Technical
Journal, 1938 _17 281- 291.
Osgood, C.E., Suci, G.J., and Tannenbaum, P.H. The
measurement of meaning. Urbana, Illinois: University
of Illinois Press, 1957.
Prebor, L.D. The natural history of a conversation.
Doctoral dissertation, University of
Florida, 1972.
Sapir, E. Speech as a personality trait. American Journal
of Sociology, 1927, 82, 895.
Scheflen, A.E. Quasi-courtship behavior in psychotherapy.
Psychiatry, 1965, 2_8, 245-257.
Scheflen, A.E., and Scheflen, A. Body language and social
order communication or behavioral control. Englewood
Cliffs, N.J.: Prentice-Hall, Inc., 1972.
Schtz, W.C. The interpersonal underworld. Palo Alto,
California: Science and Behavior Books, 1966.
Schtz, W.C. The FIRO scales manual. Los Angeles:
Consulting Psychologists Press, Inc., 1967.
Sullivan, H.S.
New York:
The interpersonal theory of psychiatry.
Norton, 1953.
Sullivan, H.S.
New York:
The fusion of psychiatry and social science.
Norton, 1964.
Vargas, R.A. A study of certain personality characteristics
of male college students who report frequent positive
experiencing and behaving. Doctoral
dissertation, University of Florida, 1968.


TABLE OF CONTENTS
(Continued)
Page
CHAPTER
III RESULTS 41
Monitors' Reliability 41
Data Analysis 43
IV DISCUSSION 48
Personality Type and Coverbal Behavior. 48
Methodological Considerations 52
Future Research and Application 55
APPENDICIES 57
REFERENCES 74
BIOGRAPHICAL SKETCH 79
i v


-67-
Complexity Scale
Instructions: Here is a list of words. You are to
read the words quickly and check each one that you think
describes YOU. You may check as many or as few words as
you
like -- but be
HONEST.
Don't check words
that tell
what
kind of person
you should be. Check
words that tell
what
kind of a person you
really are.
1.
ab le
20.
careless
39.
false
2.
active
21.
charming
40.
fine
3.
afraid
22.
cheerful
41.
fierce
4.
alone
23.
clean
42.
foolish
5 .
angry
24.
clever
43.
friendly
6.
anxious
25.
comfortable
44.
funny
7.
ashamed
26 .
content
45.
generous
8.
attractive
27.
cruel
46.
gentle
9.
bad
28.
curious
47.
glad
10.
beautiful
29.
delicate
48.
good
11.
big
30.
delightful
49.
great
12.
bitter
31.
different
50.
happy
13.
bold
32.
difficult
51.
humble
14.
brave
33.
dirty
52.
idle
15.
bright
34.
dull
53.
important
16.
busy
35.
dumb
54.
independent
17.
calm
36.
eager
55.
jealous
18.
capable
37.
fair
56.
kind
19.
careful
38.
faithful
57.
large


-22-
the person with low self-centrality can only view his
environment from others' perspectives. He will constantly
seek cues from others as to how he should process his
environment and in turn, how he should behave. This is
clearly a scale on which the healthy, well-balanced person
would score in the mid-range.
As Ziller (1973) defines his measure of self-complexity
it could be seen as a measure of personal adjustment.
Viewing it in this sense, it is different from his other
scales. In the case of his self-esteem and se 1f-centrality
scales, and in Schtz' scales, there appears a curvilinear
relationship between personal adjustment and scores on the
scales. That is, the well-adjusted, socially adept person
will fall in the mid-range on these scales. A possible
exception to this curvilinear relationship with the other
scales is the self-esteem scale. Ziller (1973) presented
his se 1f-esteem-complexity matrix wherein the person
falling in the high SE and high Comp cell has a differen
tiated and integrated theory of social behavior. This
person may be the super-well-adjusted person. It appears,
however, that it is in the self-complexity scale that a
direct, positive correlation between scale score and
adjustment is found. The high Comp person is the person
capable of self evaluation, the one capable of viewing his
environment from different perspectives. This is the
person who is well balanced, well adjusted, and less likely
to have exceptional needs or unrealistic perceptions of
self or others.


SUBJECT NUMBER DATE
DIRECTIONS: This questionaire is designed to explore the
typical ways you interact with people. There are, of
course, no right or wrong answers; each person has his own
ways of behaving.
Sometimes people are tempted to answer questions like
these in terms of what they think a person should do. This
is not what is wanted here. We would like to know how you
actually behave.
Some items seem similar to others. However, each item
is different so please answer each one without regard to
the others. There is no time limit, but do not debate long
over any item.
For each statement below, decide which of the following
answers best applies to you. Place the number of the
answer on the left of the statement. Please be as honest
as you can.
1.usually 2. often 3. sometimes 4. occasionally
5. rarely 6. never
1. I try to be with people.
2.I let other people decide what to do.
3.I join social groups.
4.I try to have close relationships with people.
5. I tend to join social organizations when I have the
opportunity.
6. I let other people strongly influence my actions.
7.I try to be included in informal social activities.
8. I try to have close, personal relationships with
people.
9. I try to include other people in my plans.
10.I leg other people control my actions.
11.I try to have people around me.
12.I try to get close and personal with people.
13.When people are doing things together I tend to
join them.
-59-


-66-
Self Centrality Item
In the large circle below, draw two circles -- one to stand
for yourself and a second to stand for a friend. Place an
S in the circle for self and an F in the circle for your
friend.


-7-
schizophrenic personality. Argyle (1969) pointed out the
importance of coverbal cues as they combine with the
verbal message. Markel, Bein and Phillis (1973) found a
normative relationship between voice and content. Their
terms 'synchrony' and 'dysynchrony' are analagous to the
terms 'complementary' and 'contradictory' as used in the
examples in the previous paragraph.
Whatever the type coverbal communication (i.e.,
complementary, contradictory, or independent), it tends
to consistently communicate something about the speaker
himself, sometimes in addition to, sometimes instead of
what the verbal message has to say about the speaker. For
several reasons, the coverbal message may tell more about
the speaker than does the verbal message. First, the
speaker himself may be unaware of his coverbal message
and therefore makes no attempt to 'filter' it so that he
will 'look good' or 'look right'. Second, the rules
governing coverbal behavior are less stringent (sometimes
less understood) than the rules governing verbal behavior.
Finally, another reason that coverbal behavior can tell us
so much about the speaker is that it can be utilized in
such a way that communication of sensitive messages is not
so threatening. That is because the coverbal message does
not place the same kind of demands on the person to whom
it is directed. The coverbal message is easier for that
person to simply ignore. That act of ignoring, or non
response, may be less traumatic to the sender of the message


CHAPTER I
INTRODUCTION
There is a great deal of speech related behavior
other than the systematic production of sounds which deter
mines the denotation of face-to-face communication. Of
particular interest in this study are behaviors which are
frequently referred to as kinesic, or body motion. Man's
interest in these 'non-content' aspects of communication
is not new. Almost fifty years ago Sapir (1927) expressed
the belief that if we study a person's communication style
we will learn much about him as a psychophysical unit when
he stated that, if we analyze the speech of an individual,
in its social perspective, we obtain valuable information
about his psychiatric characteristics. Allport (1961) more
recently posited that the expressive components, i.e.,
one's manner or style of behaving, reflect personality
structure and serve as a potential guide to personality
assessment.
The communication setting of particular interest in
this study is the dyadic conversation, that is, the two-
person face-to-face conversational interaction. This
particular setting was selected for several reasons, prac
tical as well as heuristic. This mode of communication is
-1-


INTERPERSONAL ORIENTATION CORRELATES OF NONVERBAL
BEHAVIOR IN CONVERSATIONAL INTERACTION
By
WILLIAM WALLACE CAMPBELL
A DISSERTATION PRESENTED TO THE GRADUATE COUNCIL OF
THE UNIVERSITY OF FLORIDA
IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE
DEGREE OF DOCTOR OF PHILOSOPHY
UNIVERSITY OF FLORIDA
1976

ACKNOWLEDGEMENTS
The list of friends, fellow students, and professors
whose encouragement, criticism, and support merits their
being listed here is too long to include. Two friends,
however, cannot be omitted, they are Jeannie Clay and
Dr. Cynthia Gallois.
My supervisory committee members were truly important
in many ways. Dr. Wiley Rasbury, Dr. Robert Hornberger
and Dr. Richard McGee played major roles in my clinical
development. Dr. Robert Ziller, with his broad and varied
interests, rejuvenated my research interests in each
interaction we had. Many thanks are extended to Dr. Thomas
Sain for his help and his suggestions through this lengthy
research experience. Dr. Norman Markel with his concern
for learning, and his concern for people, has had tremen
dous impact on me as a person, as a student and as a
researcher. Without his insights, his suggestions, his
many hours of support (given seemingly without limit), this
research and this dissertation would have had a much lower
probability of successful completion.

TABLE OF CONTENTS
Page
ACKNOWLEDGEMENTS ii
LIST OF TABLES v
ABSTRACT vi
CHAPTER
I INTRODUCTION 1
Purpose of the Study 4
Coverbal Behavior 5
Interpersonal Orientation 8
Review of the Research Literature .... 9
Research Rationale 12
Coverbal Behaviors 12
Turn length 13
Facing 13
Hand movement 14
Body touching 15
Head tilts 15
Derived Coverbal Measures 15
Interpersonal Orientation 16
Fundamental Interpersonal
Relations Orientat ion-Behavior. 16
Self-Other Orientation Tasks. 19
The two interpersonal orien
tation instruments 2 3
Statistical Analysis 24
Hypotheses 27
II METHOD 31
Subjects 31
Equipment 31
Discussion Problems 32
Interpersonal Orientation Instruments . 33
Procedures 35
Data Reduction 36
Derived Coverbal Measures 38
Data Analysis 39
i i i

TABLE OF CONTENTS
(Continued)
Page
CHAPTER
III RESULTS 41
Monitors' Reliability 41
Data Analysis 43
IV DISCUSSION 48
Personality Type and Coverbal Behavior. 48
Methodological Considerations 52
Future Research and Application 55
APPENDICIES 57
REFERENCES 74
BIOGRAPHICAL SKETCH 79
i v

LIST OF TABLES
Table
1 Coverbal Behavior Definitions
2 Pearson Correlation Coefficients for
Reliability Data
*
3 Normalized Canonical Variable Coefficients .
Page
37
42
45
v

Abstract of Dissertation Presented to the Graduate Council
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy
INTERPERSONAL ORIENTATION CORRELATES OF NONVERBAL
BEHAVIOR IN CONVERSATIONAL INTERACTION
By
William Wallace Campbell
December, 1976
Chairperson: Norman N. Markel, Ph.D.
Major Department: Psychology
This study investigated the relationships between
interpersonal orientation and nonverbal behavior. Twelve
male and 12 female students enrolled in undergraduate
speech classes were the subjects. Each subject engaged in
two conversations which were videotaped so that monitors
could record the occurrence of five nonverbal behaviors.
The five nonverbal behaviors studied were: turn length
(TL), facing (F), hand movement (M), body touching (T), and
head tilts (HT). Each subject completed Schtz' Funda
mental Interpersonal Relations Orientation-Behavior (FIRO-B)
scale and the Self-Other Orientation Tasks (SOOT) developed
by Ziller. The seven interpersonal orientation measures
studied were: expressed control (EC), wanted control (WC),
expressed inclusion (El), and wanted inclusion (WI) from
FIRO-B, and self-esteem (SE), self-centrality (Cent), and
vi

self-complexity (Comp) from SOOT.
The following hypotheses were proposed:
1. There is an overall relationship between inter
personal orientation and nonverbal behavior.
2. F and HT are positively related to WI and in
versely related to El.
3. TL and M are positively related to WC and Cent.
4. High HT and low TL are associated with high WC
and low EC.
5. High T and low TL are associated with low SE and
low EC.
6. Comp is not correlated with nonverbal behavior.
The canonical correlation analysis was the statistical
procedure employed in this study. Hypothesis 1 was statis
tically supported (p < .004). The other five hypotheses
were evaluated through examination of the canonical coef
ficients. Hypotheses 4 and 6 appeared strongly supported.
Hypotheses 3 and 5 seemed partially supported. Hypothesis
2 was not supported.
Statistical support for hypothesis 1 was viewed as
support for further study of the relationships between
interpersonal orientation and nonverbal behavior. The
overall personality types which emerged were the abdicrat,
the autocrat, and the dominant personality. The main be
havioral manifestations of the abdicratic (or submissive)
type were a large amount of head tilting, little body
touching, and short speaking turns. The behavioral
vi 1

manifestations of the autocratic (or dominant) type were
the mirror image:
body touching, and
little head tilting, a large amount of
long speaking turns.
vi ii

CHAPTER I
INTRODUCTION
There is a great deal of speech related behavior
other than the systematic production of sounds which deter
mines the denotation of face-to-face communication. Of
particular interest in this study are behaviors which are
frequently referred to as kinesic, or body motion. Man's
interest in these 'non-content' aspects of communication
is not new. Almost fifty years ago Sapir (1927) expressed
the belief that if we study a person's communication style
we will learn much about him as a psychophysical unit when
he stated that, if we analyze the speech of an individual,
in its social perspective, we obtain valuable information
about his psychiatric characteristics. Allport (1961) more
recently posited that the expressive components, i.e.,
one's manner or style of behaving, reflect personality
structure and serve as a potential guide to personality
assessment.
The communication setting of particular interest in
this study is the dyadic conversation, that is, the two-
person face-to-face conversational interaction. This
particular setting was selected for several reasons, prac
tical as well as heuristic. This mode of communication is
-1-

-2-
after all the most commonly employed means by which senti
ments and information are exchanged. In spite of its
ubiquitous usage and in spite of a growing and impressive
amount of research, much is left to be understood about
the dyadic conversation. Cherry (1961) observed,
A conversation is one of the commonest
phenomena we encounter, yet it is one which
raises very great scientific problems, many
still unsolved. It is often our commonest
experiences, which we take for granted, that
are elusive of exploration and description.
(p. 10)
Most of us have had the experience of being disturbed
as a result of something that has been said to us and then
realized that it was not what was said, but how it was
said and what behaviors accompanied the spoken word. That
mystical 'third ear' ascribed to some people is largely
their ability to attend, at some level, to some of the
behaviors accompanying the spoken message. Psychotherapists
are utilizing impressions derived from observations with
their 'third ear' both during psychotherapy and as a means
of evaluating past therapeutic efforts. It is anticipated
that the findings of this study, and future studies, will
make a contribution in the effort, in the current vernac
ular, 'to raise the level of consciousness' of therapists
in such situations as they respond to these behavioral
cues .
Many terms have been used to describe what a person
does while he is also speaking, some of those terms are:
'non-content communication', 'kinesics', 'non-verbal

-3-
communication', and 'body language'. The term that most
adequately describes the phenomena, and the term to be
used throughout this study is 'coverbal', one proposed by
Markel (1975). He defines 'coverbal' as, "behavior of
interlocutors which occurs in association with or
accompanying words, but which is not essential for the
articulation or grammatical functioning of those words"
(p. 189). Further, if we consult the dictionary (Woolf,
1973) we find that 'co' is a prefix form of 'complement'
which is defined in the same source as, "to make complete."
So this coverbal behavior combines with the verbal behavior
to make the complete communication.
The remainder of this chapter will include a section
on the purpose of this study, a discussion of coverbal
behavior and interpersonal orientation, and a general
review of the literature. Following the review of the
literature, a section is presented which discusses the
research rationale with specific references to each coverbal
behavior and each measure of interpersonal orientation.
Next a discussion of the canonical correlation analysis is
presented. The somewhat extensive nature of this discussion
of the statistical analysis is based on the fact that the
canonical correlation analysis is a somewhat complex,
relatively infrequently utilized statistical analysis. The
chapter is concluded with the presentation of the hypotheses
of the study.

-4-
Purpose of the Study
The purpose of this study is to investigate the rela
tionship between coverbal behavior and interpersonal
orientation. The point will be made that research has
shown coverbal behavior to be rather stable over time for
an individual, while each person has a unique style of
coverbal behavior. Thus coverbal behavior offers a reliable
measure with which to discriminate between individuals.
Interpersonal orientation is also relatively stable and
unique for the individual (Maslow, 1968; Ziller, 1973). As
Ziller (1973) points out, interpersonal orientation can be,
and is, modified over time, else it would be nonadaptive.
However, this change is gradual and this characteristic of
being gradually modified gives stability to the person.
The major question addressed by this study is, "Is
there a relationship between coverbal behavior and inter
personal orientation?" Following that question, an attempt
will be made to identify the nature of specific relation
ships within that overall relationship. The coverbal
behaviors under investigation are: Turn Length (TL) ,
Facing (F), Hand Movement (M), Body Touching (T), and Head
Tilts (HT). The interpersonal orientation characteristics
to be employed are: Wanted Inclusion (WC), Expressed
Inclusion (El), Wanted Control (WC), and Expressed Control
(EC), as measured by the Fundamental Interpersonal
Relations Orientation-Behavior (FIRO-B) scale developed by

-5-
Schutz (1966, 1967), and Self-esteem (SE) Self-centrality
(Cent), and Self-complexity (Comp) as measured by the Self-
Other Orientation Tasks (SOOT) developed by Ziller (1973).
Coverbal Behavior
By definition coverbal behavior always accompanies
verbal behavior. Whatever and wherever we are verbalizing,
we are simultaneously engaged in some behavior other than
the articulation of words according to culturally pre
scribed rules. Under certain circumstances the coverbal
behavior may not be a perceived component of the communi
cation. For example, most coverbal behavior of interest
in this research is transmitted via the visual channel
(Markel, 1969). If that channel is unavailable to the
person to whom the communication is directed, obviously
these coverbal behaviors cannot be a part of the total
communication as perceived by that listener. This would
be the case when a sightless person is the listener, when
the interaction takes place in total darkness, etc. These
types of special conditions certainly have impact on the
listener's sensitivity to other communication channels
(e.g., auditory, olfactory, etc.), but that is not the focus
of this research. A basic premise of this research is that
the coverbal behaviors under investigation are available
for processing by the listener or listeners.
Coverbal behavior can communicate information which
is complementary to, contradictory to, or independent of,

-6-
the verbal message. When a father speaks comforting words
to his frightened child while holding the child in a
gentle, caring manner the father is transmitting comple
mentary verbal and coverbal messages. The man at the beach
who tells his lover he would never look at another woman,
while he is watching every bikini-clad lovely in sight,
is sending verbal and coverbal messages which are contra
dictory. The college professor, expounding on a certain
theory to his class, while he still has shaving cream under
his ear, is probably sending a more prominent message about
himself (from the listeners' points of view he may be
sending messages about his grooming, the fact that he
overslept that morning, or whatever) with his coverbal
message (the shaving cream) than with his verbal message
concerning the theory at hand. In this example, the co
verbal message probably does not complement or contradict
the verbal message, most likely it is independent of the
verbal message. Nonetheless, the coverbal message is a
prominent, possible dominant, component of the communica
tion.
Bateson, Jackson, Haley and Weakland (1956) presented
a now well-known case for the need for agreement between
the verbal and coverbal message, proposing their theory
that through consistent conflict between the verbal and
coverbal messages the speaker creates what they termed the
'double-bind'. Their theory went on to suggest the
double-bind as a prime factor in the creation of the

-7-
schizophrenic personality. Argyle (1969) pointed out the
importance of coverbal cues as they combine with the
verbal message. Markel, Bein and Phillis (1973) found a
normative relationship between voice and content. Their
terms 'synchrony' and 'dysynchrony' are analagous to the
terms 'complementary' and 'contradictory' as used in the
examples in the previous paragraph.
Whatever the type coverbal communication (i.e.,
complementary, contradictory, or independent), it tends
to consistently communicate something about the speaker
himself, sometimes in addition to, sometimes instead of
what the verbal message has to say about the speaker. For
several reasons, the coverbal message may tell more about
the speaker than does the verbal message. First, the
speaker himself may be unaware of his coverbal message
and therefore makes no attempt to 'filter' it so that he
will 'look good' or 'look right'. Second, the rules
governing coverbal behavior are less stringent (sometimes
less understood) than the rules governing verbal behavior.
Finally, another reason that coverbal behavior can tell us
so much about the speaker is that it can be utilized in
such a way that communication of sensitive messages is not
so threatening. That is because the coverbal message does
not place the same kind of demands on the person to whom
it is directed. The coverbal message is easier for that
person to simply ignore. That act of ignoring, or non
response, may be less traumatic to the sender of the message

-8-
than would an overt, verbal rebuke. An increased amount
of eye contact (within acceptable bounds) can communicate
to a woman that a man would like to initiate a relation
ship with her. At the same time, the demands for an overt
response from her are not nearly so great as they would
have been had he used verbal messages to communicate that
same desire. The man can feel less threat of rejection,
and possibly embarrassment, knowing that he has not re
quired a verbal response of the woman. Thus, more sensi
tive, more personal data may more freely be communicated
via coverbal behavior. She is free to simply ignore his
overture, in that case his message has gotten a subtle
response, and his pride remains more intact for his next
adventure. Hopefully, he correctly decoded her message.
For a discussion of the different rules governing verbal
and coverbal behavior, and some of the implications of
those differences, see Markel (1975).
The above examples illustrate two points: 1) the
coverbal message can be as important as, or more important
than, the verbal message, and 2) the receiver of the
coverbal message, especially if untrained in decoding
coverbal communication, can misinterpret the message. Both
of these points are addressed by Fast (1972).
Interpersonal Orientation
Interpersonal orientation is the way an individual
views himself in the context of his social milieu. For

-9-
some major personality theorists, understanding the
importance of other people to the individual, and the
individual's need to relate to other people, is basic to
understanding the individual. Fromm (1941, 1947) spoke
of man's feelings of isolation and threat as he becomes
increasingly separated from other people. Adler (1939)
felt that man's social interest and need for other people
were instinctive. That is, man does not merely learn to
need contact and affiliation with others, rather he is born
with that social need as a biological given. Sullivan
(1953, 1964), probably the most significant single figure
in advancing the social perspective as a means for under
standing the individual, felt that it was meaningless to
attempt viewing the individual in any other context than
the social context. In his interpersonal approach to
psychiatry he pointed out that we all enter and leave this
world as social beings and that the most fruitful way of
conceptualizing and treating humans is as social beings.
If we look at man's basic needs as postulated by Maslow
(1967, 1968) we find among them social needs such as the
need for affiliation and affection. An accurate and
adequately high self-esteem, shaped through social inter
action, is also a basic need according to Maslow (1968).
Review of the Research Literature
For some time researchers have been interested in the
psychological importance of coverbal behavior. As with any

-10-
behavior under investigation, certain basic issues arose
relating to quantification and analysis of the coverbal
behavior data. One practical, and apparently meaningful,
scheme for quantifying coverbal behavior is based on its
temporal characteristics (e.g., Chappie, 1939; Goldman-
Eisler, 1968; Mahl, 1956; Norwine and Murphy, 1938),
with the quantity often being simply the duration of time
spent engaging in the behavior. Simple duration, along
with proportion of available time spent engaging in a
behavior, will be the approach to quantifying coverbal
behavior in this study.
As interest in the research field grew, along with an
increase in man's electronic sophistication, new techniques
for measuring, recording, and analyzing coverbal behavior
emerged. Almost thirty years ago, Norwine and Murphy
(1938) developed a sound-activated device for recording
speech behavior. Chappie (1939) was developing his device,
later to be known as the Interaction Chronograph (Chappie,
1949), which utilized both the human observer and mechanical
devices for recording verbal and coverbal behavior. In
his earlier work (Chappie, 1940; Chappie and Arensberg,
1940) Chappie had demonstrated that personality character
istics were associated with certain stable properties of
individuals' interaction style.
Some ten years after Chappie (1939) began his research
which led to the Interaction Chronography, Verzeano and
Finesinger (1949) developed the Automatic Speech Analyzer

-11-
which eliminated the necessity of the human observer for
obtaining data during interaction. Lorenz and Cobb (1952),
using the Automatic Speech Analyzer, were able to differ
entiate normals and psychiatrically impaired on the basis
of temporal aspects of their speech patterns. Chappie
(1953) went on to develop his standardized interview, with
which, using his Interaction Chronograph., he investigated
subjects' interview behavior and its relationship to
certain personality dimensions (Chappie, Chappie, and Repp,
1954) .
Man's imagination and increasingly sophisticated
electronic equipment available to him, continue to create
innovative methods for recording and analyzing coverbal
behavior. As can be noted from previously cited research,
as different techniques were developing for collecting
coverbal data, the data were being examined in
numerous ways and contexts. In the area of psychological
interests alone, researchers were relating coverbal behavior
to social orientation, transient emotional states, degree
of psychological health, and interpersonal needs, to name
a few. With this wealth of techniques and the potential
applications of the obtained data, a primary task of the
researcher becomes one of choosing which techniques to
utilize and to what end. That brings us to this study.
A discussion of how those decisions were made in this study
follows.

-12-
Research Rationale
Coverbal Behaviors
The point was made earlier that there is always co
verbal behavior present during conversation. There is, in
fact, so much behavior occurring that it becomes an
important task of the researcher to choose just what
behavior will be selected out for analyzing. Four ques
tions were asked in this research so that the final
behavior selection could be accomplished. Is the behavior
detectably recorded when a dyadic conversation is video
taped? Can the relatively untrained observer reliably
record the onset and termination of the behavior as the
videotape is monitored? Are the behaviors easily observed
during normal conversational interaction? Are the
behaviors related to the interpersonal orientation of the
speaker?
Previous research and pilot studies prior to this
research answered the first three questions affirmatively.
The outcome of this research will have to answer the fourth
question. Since the results of this study were obviously
not available when the behaviors were selected, selection
was based on previous research which asked similar questions
(e.g., Chappie, 1940; Chappie and Arensberg, 1940; Chappie
et al., 1954; Lorenz and Cobb, 1952; Norwine and Murphy,
1938). The coverbal behaviors selected for investigation
Turn length, Facing, Hand movement, Body
were:

-13-
touching, and Head tilts. Brief research references for
each behavior follow. Operational definitions of each
behavior are presented in Table 1 in the Method chapter.
Turn length. Following early interest in temporal
aspects of communication behavior, this behavior, or some
rough equivalent, has been studied frequently. Goldman-
Eisler (1954), looking at action time (percentage of total
time which was spent in talking), found a relationship
with the content variable of self-reference. That is,
people who referred more to themselves, and presumably
talked more about themselves, spoke during a greater per
centage of the total time. Cervin (1957) found that
individuals scoring high on a scale of emotional respon
siveness also spoke during a greater portion of the
available time. Markel, Bein, Campbell, and Shaw (1976)
demonstrated a greater use of available time by speakers
who scored high on a measure of needed inclusion. This
often investigated behavior of amount of speaking time
(referred to by several roughly equivalent terms such as
'action time' (Goldman-Eisler, 1954), 'participation
quotient' (Prebor, 1972), 'mean percentage', (Matarazzo,
Weins, Matarazzo and Saslow, 1968), etc.), seems to be
related to various aspects of personality.
Facing. Facing behavior was selected because of its
high correlation with eye contact. Eye contact itself is
virtually impossible to detect by observers as they view
videotapes of conversations. Facing, on the other hand,

-14-
is rather reliably observed and recorded by the relatively
untrained observer. Clinically, eye contact is a frequently
utilized diagnostic tool. Socially, this behavior seems to
be thought of as an indicator of the speaker's candor. This
is a good example of coverbal behavior that can be easily
mis-interpreted if the listener fails to take into account
cultural factors such as the norms of the culture from
which the speaker comes. Argyle and Dean (1965) and Exline
(1971) found facing behavior positively correlated with
personal attraction. Exline (1971) and LaFrance and Mayo
(1976) found facing associated with dominance. Duncan
(1972) conducted research which concluded that facing is
an effective tool for controlling speaking turn during
conversations. When the speaking person looks away from
the listener, our culture tells us that that is not a
polite time for the listener to attempt taking the role
of speaker. A review of the literature on facing can be
found in Ellsworth and Ludwig (1972).
Hand movement. Hand movement seems to be a method
of attracting attention and controlling the interaction.
Dittmann(1962) found it related to approval seeking in
subjects. Mehrabian (1970, 1971) found hand movement to
be associated with attraction and the desire for affilia
tion. While some investigators have devised elaborate
systems of quantitatively and qualitatively classifying
hand movement (e.g., Ekman and Friesen, 1972; Frey,
1975), this study looks at hand movement simply in terms

-15-
of duration of occurrence.
Body touching. Research on body touching (touching
one's own body with his hand) has not been extensive.
Freedman and Hoffman (1967) found body touching to be
related to anxiety level. Scheflen (1965) found body
touching associated with the need for self-protection.
Fast (1970), summarizing research in the field, asserts
that body touching can be a coverbal method of sending a
'hands off' message to others. In other words, the person
who does not want to be included can say so with his hands.
The findings of these three authors are seen as being
consistent in that the person who does not want to be
included may experience anxiety and feel the need for
self-protection.
Head tilts. Head tilt behavior appears to be an
indication of attentiveness and possibly cognitive activity
during conversation. In animal research it appears to be
associated with submissiveness. This relationship may
also hold for humans. Mehrabian (1971) and Mehrabian and
Ksionzky (1970) found head tilt related to relaxation of
the person.
Derived Coverbal Measures
The derived measure for Turn Length in this study was
the duration of the speaking turn as defined in Table 1 in
the Method chapter. The derived measures for Facing, Hand
Movement, Body Touching and Head Tilts were computed

-16-
proportions of the individual's speaking turns during which
he was engaged in the given behavior. The proportion
measure was used rather than straight duration measure in
order to control for turn length. The proportion tells us
not merely how much time was spent in each behavior, rather
it tells how much of the available time was spent engaging
in each behavior.
Interpersonal Orientation
The interpersonal orientation measures used in this
study are Wanted Control (WC), Expressed Control (EC),
Wanted Inclusion (WI), and Expressed Inclusion (El) as
measured with the Funadmental Interpersonal Relations
Orientations Behavior (FIRO-B), and Self-esteem (SE), Self
centrality (Cent), and Se 1f-complexity (Comp) as measured
with the Self-Other Orientation Tasks (SOOT).
Fundamental Interpersonal Relations Orientation-
Behavior. A steadily increasing amount of research is
reporting various approaches to testing Schtz's (1966)
theory of interpersonal orientation. For a summary of
that research see Schtz (1966). Schtz bases his theory
on the existence of three interpersonal needs: Inclusion,
Control, and Affection. Inclusion and affection have been
found to be highly correlated in research (Argyle, 1969).
Therefore, it is meaningful to think of the individual in
terms of the two dimensions inclusion and control. For that

-17-
reason, only the inclusion and control scales were used
from FIRO-B. Schtz describes interpersonal needs for
inclusion and control as follows (Schtz, 1966):
1) Interpersonal need for Inclusion (I): the need
to establish and maintain a satisfactory relation
with other people with respect to interaction and
association, and
2) Interpersonal need for Control (C): the need to
establish and maintain a satisfactory relation
with other people with respect to control and
power.
Schtz (1966) refers to each person's behavior in each
of these dimensions in terms of the behavior expressed (E)
towards others and the behavior wanted (W) from others.
Emerging then are four dimensions of the individual's
interpersonal orientation: Wanted Inclusion (WI), Expressed
Inclusion (El), Wanted Control (WC), and Expressed Control
(EC). Schtz (1966, 1967) developed the FIRO-B as a self-
report means of tapping each of these dimension. The
FIRO-B is presented in Appendix A.
Schtz (1966) describes individuals whose behavior
is consistently involved in satisfying his needs in the
different dimensions. He describes the high I person with
terms such as, "interact, communicate, and attend" (p. 21).
The person with high C needs is described as displaying
behavior such as, "dominance and control" (p. 22),
while the low C person is seen as,

-18-
"submissive and a follower" (p. 22). How these
behaviors are manifested will be influenced by the
individual's needs in the I and C dimensions as to whether
he wants to be included (or controlled) or whether he wants
to include (or control) others. Schtz (1966) further
makes the point that there is not necessarily a comple
mentary relationship between wanted and expressed needs
in a given area. The person with a high need to control
others may also have a high need to be controlled by others.
The domineering sergeant may need and gratefully accept
control from his lieutenant.
According to his classification the 'oversocial'
person is the one who exhibits excessive inclusion behavior
and the 'undersocial' is the person who makes little or no
attempt to include or be included. The 'autocrat' tends
towards the domineering personality and is characterized
by the excessive desire for power and control. Much of
the autocrat's behavior is directed towards controlling
others. The 'abdicrat' is the person who makes no attempt
to exert control over others. Schtz (1966) points out the
extreme case of the abdicrat who wants no control over
others and at the same time has a high need for others to
control him. This person does not say, 'I am going to
leave you alone and I would like for you to leave me alone',
rather he is saying, 'I do not want to control you, but
please tell me what to do'.
The above personality types with their accompanying

-19-
behavioral manifestations represent deviations from an ideal
state of need satisfaction according to FIRO theory
(Schtz, 1966). In the inclusion and control areas talking
is seen as an available means by which the person can
strive towards the ideal need satisfaction state. The
person with the high need to control can attempt to gain
control by talking a lot and minimizing the opportunity for
the other person to talk. The person with little need to
control may talk less and look more to the other person for
control cues. The person with high need to be included can
seek attention with hand movements and by talking a great
deal. The person with little desire to be included may
talk less and send, 'I am not interested in being included'
messages with his high rate of body touching.
Self-Other Orientation Tasks. Ziller (1970) is one
among many (e.g., Maslow, 1967, 1968; Schtz, 1966;
Sullivan, 1953, 1964) who proposes that man should be
studied as a social unit. In his social psychological theory
of personality, social adaptation is viewed in terms of self-
other concepts. For a review of his theory see Ziller
(1973). Three of Ziller's self-other dimensions are
employed in this study.
His Self-esteem (SE) and Self-centrality (Cent) scales
utilize what Ziller (1973) refers to as cognitive mapping.
Through the use of ostensibly value-free symbols SE and
Cent are meansured without the bias of mediating 'social
oughts', so often found to be a problem when verbal

-20-
self-reports are used for tapping these aspects of per
sonality. Subjects are asked to place symbols, representin
themselves and other people, somewhere in a prescribed area
The SE scale, for instance, utilizes the proclivity in our
culture for people to place items of higher value toward
the left in a horizontal ordering. On the Cent scale the
subject is presented with a large circle and asked to place
two smaller circles, representing self and a friend, some
where in the large circle. Scoring of the Cent scale is
on the basis of whether the self or the frient is placed
nearer the center of the large circle. The more often the
self is placed nearer the center, the higher his Cent score
The SE and the Cent scales each contain six items as
described above. All six items on the SE scale are iden
tical except with respect to which five significant other
people are represented by the symbols accompanying the ever
present sixth symbol representing self. All six items on
the Cent scale are identical in every respect. Because of
the similarity of the items within each scale, only one
sample item from each scale is presented. The sample
items are found in Appendix B. The Comp scale is an
adjective checklist scale containing 109 adjectives. The
Comp scale, in its entirity, is presented in Appendix B.
Subjects were instructed to check as many adjectives as
they felt applied to them. The Comp score is computed by
totaling the number of adjectives checked by the subject.
The higher the number of adjectives checked, the higher

-21-
the Comp score, and the more complex is the self construct.
Ziller (1973) referred to the self-esteem as that
component of the individual which mediates modification
of the self in response to new information which is re
ceived about the self in social interaction. This view
is consistent with the notion that self-esteem is the
individuals perception of his own self-worth (Osgood, Suci,
and Tannenbaum, 1957). Following these positions, as
Ziller states, the individual with high self-esteem is
better able to evaluate and assimilate new information.
He is not the helpless, vulnerable pawn, subject to re
structuring the self-concept immediately upon receipt of
information which is in conflict with the then existing
self-concept. The high self-esteem person uses new infor
mation as it fits for him, the low self-esteem is ever
seeking new information so that he may modify the self
to fit what the information tells him others think he
should be.
The self-centrality scale is concerned with whether
the person uses the self or others as his reference point
(Ziller, 1973). The higher a person's self-centrality
score, the more he uses the self as the reference point.
Ziller compares high and low self-centrality with Ausubel's
(1952) terms egocentrism and sociocentrism. The high self
centrality person appears incapable of perceiving his
environment from others' viewpoints. He withdraws and
does not want new information from others. By implication,

-22-
the person with low self-centrality can only view his
environment from others' perspectives. He will constantly
seek cues from others as to how he should process his
environment and in turn, how he should behave. This is
clearly a scale on which the healthy, well-balanced person
would score in the mid-range.
As Ziller (1973) defines his measure of self-complexity
it could be seen as a measure of personal adjustment.
Viewing it in this sense, it is different from his other
scales. In the case of his self-esteem and se 1f-centrality
scales, and in Schtz' scales, there appears a curvilinear
relationship between personal adjustment and scores on the
scales. That is, the well-adjusted, socially adept person
will fall in the mid-range on these scales. A possible
exception to this curvilinear relationship with the other
scales is the self-esteem scale. Ziller (1973) presented
his se 1f-esteem-complexity matrix wherein the person
falling in the high SE and high Comp cell has a differen
tiated and integrated theory of social behavior. This
person may be the super-well-adjusted person. It appears,
however, that it is in the self-complexity scale that a
direct, positive correlation between scale score and
adjustment is found. The high Comp person is the person
capable of self evaluation, the one capable of viewing his
environment from different perspectives. This is the
person who is well balanced, well adjusted, and less likely
to have exceptional needs or unrealistic perceptions of
self or others.

-23-
The two interpersonal orientation instruments. FIRO-B,
utilizing the verbal self-report technique, attempts to
measure directly the individual's interpersonal orientation.
FIRO-B solicits information from the person as to his
behavior with other people. Through divulging these be
havioral preferences, the level of the person's needs in
different interpersonal areas is assessed. Since these
ratings are based on the person's expressed behavioral
preferences, the ratings should prove to be good predictors
of how the person will behave in an interpersonal situation.
A source of error in such predictions lies in the extent
to which the person honestly responded to the instrument.
SOOT, while tapping aspects of the personality which
are similar to those tapped by FIRO-B, is in some important
ways looking at the person with a different approach. SOOT,
except in the case of the complexity scale, avoids the use
of words in the assigned tasks. Even with the complexity
scale, words are not used in the same sense as in the
FIRO-B. A verbal self-report is not involved with the
complexity scale, rather it is an adjective checklist.
Throughout SOOT there is the attempt to avoid 'good-bad',
'right-wrong', 'acceptable-unacceptable' type choices on
the part of the responder. This is accomplished through
the extensive use of symbols rather than words. Another
important difference between SOOT and FIRO-B lies in the
interpersonal characteristics being measured. SOOT measures
characteristics which are shaped through interpersonal

-24-
experience. They are also characteristics which have con
siderable impact on interpersonal behavior. SOOT is not,
however, attempting to directly measure behavioral prefer
ences, as does FIRO-B. Therefore, SOOT should be a useful
predictor of behavior, but in a less direct manner than is
FIRO-B.
Both FIRO-B and SOOT measure interpersonal orientation
characteristics which were of interest in this study. They
measure similar, yet different, characteristics, with
different, verbal as opposed to nonverbal, approaches. It
was for the above reasons that the two instruments were
included in this study.
Statistical Analysis
Canonical correlation analysis was the statistical
procedure employed in this study. This is a relatively
recent statistical technique which social scientists are
beginning to realize can be a very useful method of looking
at two sets of data. The canonical correlation (Kelley,
Beggs and McNeil, 1961; Morrison, 1967) is similar to the more
familiar Pearson product moment correlation where two
variables are analyzed to ascertain the amount of common
information contained in the two. The multiple regression
equation goes a step further, providing a method for
combining the variables in one group of variables and
correlating the newly created variable with a single

-25-
variable of interest. With the multiple regression analysis
equation not only is there a single indicator of the
relationship between the single variable and the linearly
combined group of variables, but we also are supplied with
the coefficients of the variables themselves as they appear
in the regression equation. These coefficients are the
weights assigned to each of the original variables in the
group. The coefficients are indications of the magnitude
(numerical value of the coefficient) and the direction
(algebraic sign with the coefficient) of the contribution
made by each of the original variables as they were com
bined. Thus, we have an idea of to what extent and in
what direction each variable makes its contribution.
The canonical correlation goes still one step beyond
the multiple regression analysis. The canonical correla
tion linearly combines variables from two groups of data
so that a pair of canonical variables emerge, one from each
group of data. The correlation between these newly created
canonical variables is computed and then it is possible to
address the question of whether the two groups of data are
related. The level of the correlation indicates the
strength of the relationship, that is, it is an indication
of how much common data is contained in the two groups.
Similarly to the regression equation, we are supplied with
coefficients which indicate the magnitude and direction of
the influence of each original variable on the canonical
variable.

-26-
The canonical correlation analysis is a very useful,
somewhat complex, method of attempting to identify complex
relationships. Although there is no statistical test
beyond the overall correlation, specific relationships
are inferred by examining the canonical coefficients. In
this study the relationships being searched for are com
plex. They are complex in the sense that it would be too
simplistic, possibly meaningless, to single out one coverbal
behavior and attempt to establish its relationship with a
single interpersonal orientation variable. Humans, their
behavior, and their personalities are not that simple. To
state that the person who needs to control others will talk
longer may be true. However, there are factors other than
his need to control which contribute to his behavior, and,
there are behaviors other than turn length through which
his interpersonal orientation will manifest itself.
After one pair of canonical variables has been identi
fied and tested, others may also exist with the same data
sets. In fact, there can be as many pairs of canonical
variables as there are variables in the smaller group of
data. The procedure is continued until the correlations
are not significant or until the number of correlations
reaches the number of variables in the smaller data group.
In each new relationship, if more than one exists, there is
a new set of coefficients and the statistical tests are
carried out in the same manner. A more detailed discussion
of these procedures is presented in the Method chapter.

-27-
Hypotheses
Utilizing five coverbal behavior measures obtained
from 24 videotaped conversations, and seven interpersonal
orientation scores, four from FIRO-B and three from SOOT,
this study attempts to support the existence of a relation
ship between coverbal behavior and interpersonal orientation.
Naturalness of the conversations was maximized as subjects
were placed in a comfortable environment and asked to
discuss relevant social discussion problems with virtually
no time constraints placed on the length of their conver
sations .
Canonical correlation analysis is applied to the data.
The canonical correlation is conceptually more complex than
a t-test or a pairwise correlation such as the Pearson
product moment correlation. Similarly, its interpretations
are more complex. To predict an overall relationship
between the two groups of data is straightforward enough,
and that prediction is directly testable with the canonical
correlation analysis. However, to look inside the data
groups and make specific, directional relationship pre
dictions involving several variables involves risks beyond
the obvious risk of being wrong by predicting a relation
ship that does not exist. There exists the risk of failing
to support an hypothesis, not necessarily because the
hypothesized relationship does not hold true, but because
of the way the variables were combined in the computation

-28-
of the canonical variable. As pointed out earlier, several
canonical relationships are possible. If, in the con
struction of the canonical variable(s) for a given group
of the data, the relative directions of two variables within
one group, both of which are in an hypothesized relation
ship, are not the same as they appear in the hypothesis,
full support of the hypothesis is not possible. An example
makes this point more concrete. Noting the second hypoth
esis in this study, it predicts that head tilts and facing
will be positively related to WI scores and inversely
related to SE scores. In addition to the main, overt
prediction that two coverbal behaviors will vary directly
with one interpersonal orientation score and inversely with
another interpersonal orientation score, another prediction
is implied. The other prediction is that facing and head
tilts will vary together and that WI and SE scores will
vary inversely. If the variables are not combined in a way
consistent with these latter, implied predictions, full
support of the hypothesis is not possible. Nonetheless,
to resort to pairwise predictions involves a level of
simplicity inappropriate for this study and the relationships
at interest.
Research has shown that interpersonal aspects of
personality are related to coverbal behavior (Allport, 1961;
Chappie et al., 1954; Goldman-Eisler, 1954).
Hypothesis 1: There is a relationship between coverbal
behavior and interpersonal orientation.

-29-
Facing has been shown to be associated with personal
attraction (Argyle and Dean, 1965; Exline, 1971). Head
tilting behavior appears related to submissiveness (Argyle,
1969) .
Hypothesis 2: F and HT are positively related to WI
and inversely related to El.
Goldman-Eisler (1954) found that subjects talked more
if their self reference was greater. Dittmann (1962) found
hand movement associated with the desire to get approval
from others and to control the other person in conversation.
The person who wants to control his conversational mate will
talk a great deal (Schtz, 1966).
Hypothesis 3: TL and M are positively related to WC
and Cent.
Head tilting is associated with submissiveness (Argyle,
1969). The person desiring to control will talk more than
the person not wishing to control (Schtz, 1966).
Hypothesis 4: High HT and low TL are associated with
high WC and low EC.
Scheflen (1965) found body touching associated with
the need for self-protection. The person wanting control
over others will talk a great deal (Schtz, 1966).
Hypothesis 5: High T and low TL are associated with
low SE and low EC.
Ziller's (19 73) high se 1f-complexity individual is a
well-balanced person. This person is less likely to
outstanding need areas which are unsatisfied. As the
have

-30-
self-complexity drops, idiosynchratic needs will emerge.
It seems that these needs will emerge in an unsystematic
fashion, not predictable from the self-complexity level of
the person.
Hypothesis 6: Comp is not correlated with coverbal
behavior.

CHAPTER II
METHOD
Subj ects
Twelve male and twelve female students enrolled in
introductory speech class at the University of Florida
volunteered to participate in this study. All subjects
were caucasion; ages ranged from 18 to 27, with a mean age
of 20.3. Two males and two females were selected from the
volunteers from each speech section and those four sub
jects comprised a group for this study. Other than the
specified gender requirement, the only condition imposed
on the composition of the group was that subjects' famil
iarity witli other group members be limited to the inter
action in the speech course in which they were currently
enrolled. The resulting sample then was made up thusly:
six groups of two males and two females; each group drawn
from a different section of the speech course; each sub
ject's prior interaction with his group members was limited
to interaction in speech class.
Equipment
Conversations between pairs of subjects were recorded
- 31-

-32-
in a carpeted room with dimensions of approximately 10 x 15
feet. Subjects were seated at adjacent sides of a padded
card table in comfortable, padded arm chairs in one corner
of the room. Each subject was seated at approximately a
45 angle to a Sony 3210 video camera equipped with zoom
lens. The video camera was connected to a Sony AV 3600
Videocorder which was the recorder used for recording the
audio, as well as the video, signals during conversations.
A Sony microphone was mounted on a stand between and
slightly behind the subjects. Subsequent to the exepri-
mental sessions, continuous lapsed time in minutes, seconds,
and tenths of seconds from beginning to end of each conver
sation was superimposed on the video tape by means of a
Model VTG33 Odetic Video Timer.
Discussion Problems
Each of the discussion problems was of assumed and
apparent contemporary social interest. One problem, the
"abortion" problem, involved a young married couple groping
with the decision of whether or not the pregnant wife
should have an abortion. The other problem, the "living
together" problem, concerned a young, heterosexual, un
married couple, living together, trying to decide how they
should present their living situation to his parents, who
were unaware of their cohabitation, while the parents were
in town for a brief visit. Judging from observed level of

-33-
apparent interest on subjects' part and based on verbal
response of subjects after the experimental sessions were
completed, subjects actually became quite involved in the
discussion problems. The texts of these problems are
presented in Appendix C.
Interpersonal Orientation Instruments
The two instruments utilized for measuring inter
personal orientation of subjects were the Fundamental
Interpersonal Relations Orientation-Behavior (FIRO-B)
developed by William Schtz and Self-Other Orientation
(SOOT) scale developed by Robert Ziller.
FIRO-B is composed of 54 items and has been used with
subjects ranging from eighth grade to adults. The results
of this instrument yield six scales: Expressed Inclusion,
Wanted Inclusion, Expressed Control, Wanted Control,
Expressed Affection, and Wanted Affection. Only the first
four of these scales were used in this study. The possible
range of scores on these six scales is from 1 to 9, with
all scores being integers. This instrument was designed
by Schtz (1966, 1967) to measure the relatively stable
attitudes of an individual which determine his interactive
behavior. Schtz has published considerable research
demonstrating validity for his instrument.
SOOT was designed to measure individuals' self-other
orientation in terms of their perceived relationships.
The method of assessing self-esteem and self-centrality is

-34-
the use of ostensibly value-free symbols which result in
subjects' supplying more valid information about themselves
than would be expected in more traditional, verbal, self-
report inventories. In the self-esteem scale symbols
represent the self, significant other people such as
family members, teachers, friends, etc. The subjects'
tasks were to arrange the symbols according to certain
instructions. Scoring is accomplished by obtaining the
sum of the assigned numerical values according to the
position in which the symbol representing the self is
placed. The self-centrality score is based on whether the
'self' is placed nearer to or farther from the center of a
large circle than is the symbol representing a friend.
Each scale has six items. The self-centrality items are
identical throughout the scale. The only difference between
the six items of the self-esteem scale is in the significant
other people which the symbols represent. The complexity
scale consists of a 109-word adjective checklist. Even
though the SOOT is a relatively new instrument, a con
siderable amount of research has demonstrated the relia
bility and validity of the scales (Ziller, 1973).
The SOOT scales used in this study are Self-Esteem,
Self-Centrality, and Self-Complexity. Sample items from the
self-esteem and self-centrality scales are presented in
Appendix B. Because of the similarity of the items, only
one item per scale is presented here. The entire self-complex
ity scale and the general instructions for SOOT are also

-35-
presented in Appendix B.
FIRO-B is presented in its entirity in Appendix A.
Procedures
Subjects were assigned experimental session times by
group so that one group participated in an evening. After
all group members had arrived for a session one experi
menter explained that the purpose of the study was to study
the behavior of persons in conversation in a dyadic
setting. They were told that each of them would partici
pate in two conversations, one each with two of the other
members of their group, and that the conversations would
be recorded on video tape. They were also informed that
they would complete several paper and pencil instruments
which would call for information about themselves and in
formation about how they felt about their experience of
participating in the study. Subjects then signed consent
forms for participation in the experiment.
A second experimenter then took two subjects to the
taping room where they would have their conversations.
The camera and recorder was started by the experimenter.
Subjects were seated at the table with copies of one of
the discussion problems affixed to the table in front of
each of them. They were instructed to read the discussion
problem and discuss as long as they needed to in order to
come to some agreement on a solution. The experimenter
read the subjects' identification numbers into the

-36-
microphone, told the subjects to call him when they were
finished, and left the room.
During the first conversation the first experimenter
had the subjects who were not participating in the con
versation complete half of the paper and pencil tasks that
they were to complete. Following the first conversation,
the subjects just completing their conversation were taken
to other rooms where they completed half of their paper
and pencil tasks while the other two subjects had their
first conversation.
The process was continued until each subject had:
participated in two conversations (one each with a same-
sex and an other-sex partner), completed the background
information form, the FIRO-B, the SOOT, and semantic
differentials relating to how they felt about the experi
ment, the experimenters, and their partners. Of the paper
and pencil instruments, only the FIRO-B and the SOOT were
used in this study.
Subjects were randomly assigned to: either complete
FIRO-B or SOOT first, discuss the 'abortion' problem or
the 'living together' problem first, and converse with a
same-sex or opposite-sex partner first.
Data Reduction
Monitors were trained to observe the tapes and record
the occurrence of the coverbal behaviors listed in Table 1.
Each conversation was monitored once at normal tape speed

- 37-
Table 1
Coverbal Behavior Definitions
Behavior
Definition
1. Turn
2. Facing
3. Right hand movement
4. Left hand movement
5. Right hand body touching
6. Left hand body touching
7. Head tilt
Begins when one speaker starts
talking and ends when the
other speaker starts solo
talking
Subject's nose is pointed
roughly at the center of
partner's face
Any movement of subject's
right hand and wrist, unless
it results from a movement of
the whole torso
Any movement of subject's
left hand and wrist, unless
it results from a movement of
the whole torso
Any part of subject's right
hand touching any part of
his body
Any part of subject's left
hand touching any part of
his body
Subject's eyes are not in a
horizontal plane

-38-
for each behavior for each subject. During that first
monitoring the approximate beginning and ending times of
that behavior were recorded. Using those approximate times
as guides, the tape was then played at slow-motion speed
(approximately 1/15 of normal speed) to record more pre
cisely (usually within 0.1 second) the beginning and ending
times. These times were obtained by using lapsed time
which had been superimposed on the tape with an Odetic
Video Timer after conversations were completed.
Monitors were assisted in determining beginnings and
endings of turns by having a typed verbatim transcript of
the conversations. Periodic checks indicated that tracking
was reliable when monitors had been given several hours
training.
Derived Coverbal Measures
A computer program was developed which used the
beginning and ending times for each coverbal behavior and
for the turns and computed durations and frequencies of
all behaviors per turn. For the purposes of this study,
total hand movement and total body touching were of in
terest, rather than hand movement and body touching broken
down by left and right hand. Therefore, right hand move
ment and left hand movement were combined into one measure:
Hand Movement (M). Also, right hand body touching and
left hand body touching were combined into one measure:
Touching (T). The coverbal measures used in the statistical

-39-
analysis were the grand mean per behavior for subjects
across both their conversations, after summing across turns
In the case of the measure Turn Length (TL), it was the
grand mean of the turn lengths across both conversations.
The behaviors: Facing (F), Hand Movement (M), Touching (T)
and Head Tilt (HT) were computed as proportions of the
respective turn which was spent engaging in those behaviors
The grand mean for each subject that was used for these
behaviors then was the mean obtained across all speaking
turns of both conversations using those proportions.
Data Analysis
The statistical method for analysis of the relation
ship between the two groups of data (interpersonal orienta
tion and coverbal) was the canonical correlation. There
were five variables in the coverbal data: Turn Length (TL)
Facing (F) Movement (M), Touching (T), and Head Tilt (HT).
Seven Personality variables made up the other group:
Expressed Control (EC) Wanted Control (WC), Expressed
Inclusion (El), Wanted Inclusion (WI), Self-Esteem (SE),
Self-Centrality (Cent), and Complexity (Comp). The first
four of the personality variables are scales from the
Fundamental Interpersonal Relations Orientation-Behavior
(FIRO-B) and the last three are scales from the Self-Other
Orientation Tasks (SOOT).
Canonical correlation analysis is a method whereby the
question, "Is there a significant relationship between the

-40-
two groups of data?", can be addressed. Further, an
indication of the strength of the relationship(s) is in
dicated by the level of statistical significance. More
specific ideas about to what extent each original variable
makes its input into the relationship is inferred from the
canonical variable coefficients. Further elaboration on
this statistical procedure and subsequent inferences is
presented in the Result and Discussion chapters.

CHAPTER III
RESULTS
Monitors' Reliability
In order to assess the reliability with which the three
monitors recorded the coverbal behavior, a monitor other
than the original monitor viewed 30-second segments of each
conversation tape, recording the beginning and ending times
of each behavior. Although five coverbal behaviors were
used in the data analysis, seven behaviors were observed
in the raw data. This is because two pairs of behaviors
(left and right hand movement and left and right hand
body touching) were combined into two behaviors (Hand
Movement and Touching respectively) prior to data analysis.
As pointed out in the Method chapter, the behaviors of
interest were hand movement and body touching, with no
concern for which hand was involved.
For continuous data, such as the measures in this
study, a satisfactory indication of inter-monitor reliabil
ity is provided by the Pearson product moment correlation
coefficient. Table 2 contains the Pearson correlation
coefficients for the seven coverbal measures. The coef
ficients range from a low of .79 for Left Hand Movement to
a high of .99 for Left Hand Body Touching. Four of the
-41-

-42-
Table 2
Pearson Correlation Coefficients
for Reliability Data
Behavior
R
Turn Length
.98
Facing
.96
Left Hand Movement
.92
Right Hand Movement
. 79
Left Hand Body Touching
.99
Right Hand Body Touching
.98
Head Tilt
. 89
Correlations were computed on duration of behavior as
originally recorded and as recorded by a different
monitor during 30-second intervals for each behavior of
each conversation.

-43-
seven measures were greater than .95. The inter-monitor
reliability was found to be quite satisfactory for all
coverbal measures. Clearly, the behaviors under investi
gation are observable to the extent that an individual with
minimal training can reliably record their occurrence.
Data Analysis
The statistical method used for analyzing the data was
canonical correlation. Since the coverbal behaviors
Facing (F), Hand Movement (M), Touching (T), and Head Tilt
(HT) were expressed as proportions, these data were trans
formed before computing the canonical correlations so that
a better approximation of a normal distribution would be
obtained. The angle arcsin of the square root was the
transformation employed. The mean values of the coverbal
behaviors and the interpersonal orientation measures appear
in Appendix D.
As pointed out in the Method chapter, the canonical
correlation is analagous to the more familiar Pearson
product moment correlations. However, the test for sig
nificance is not carried out in the same manner. In the
case of the canonical correlation it is first converted to
a chi square statistic which is in turn tested for signi
ficance. Similarly to the multiple regression analysis,
the canonical correlation yields coefficient values for the
original variables. These are the values which, when used
as coefficients for the original variables, combine the

-44-
original variables to produce the respective canonical
variables. That is, using the canonical coefficients to
determine the weight (magnitude and direction) of the input
of each original variable a new pair of variables (the
canonical variables) is created. The resulting two vari
ables (one for each of the two groups of data) have the
highest possible correlation of any pair that can be
created by linearly combining the variables in the respec
tive groups of data. In order to remove (arbitrary) in
fluence of scale and variability, the coefficients are
normalized. Normalizing is accomplished by first multi
plying each coefficient by the observed standard deviation
of the corresponding variable. The resulting values are
then rescaled so that the coefficient is equal to 1.0
(absolute value) with each coefficient retaining the
appropriate algebraic sign. It is through examination of
these normalized values that we get an idea of the relative
contribution of each variable to the common information
shared by the two groups of data. The canonical variables
(Cr = coverbal, 10 = interpersonal orientation) and their
normalized coefficients are presented in Table 3.
The first canonical correlation (R^) was significant
(R^ = .89, y2 = 61.56, df = p < .004). The hypothesis that
there is a relationship between the coverbal behaviors
observed and the interpersonal orientation measures derived
from FIRO-B and SOOT was supported. While no further
statistical test is appropriate, further information as to

Table 3
Normalized Canonical Variable Coefficients
C = 1.00HT .6IT .54M .45TL .23F
r
10 = -l.OOEC + .76WC .60SE .60Cent .55 El + .05WI .03Comp
C = Canonical variable for coverbal behavior
r
10 = Canonical variable for interpersonal orientation
Coverbal Behavior: HT = Head Tilt, T = Touching, M = Hand Movement,
TL = Turn Length, F = Facing
Interpersonal Orientation Data: EC = Expressed control, WC = Wanted control,
SE = Self-esteem, Cent = Self-centrality,
El = Expressed inclusion, WI = Wanted
inclusion, Comp = Complexity
-45-

-46-
the relative importance of each variable is inferred from
the individual coefficients.
Looking at the coefficients of the coverbal behaviors,
Head Tilt (HT) with its coefficient of +1.00 provides the
greatest input into C^,. Touching (T) and Movement (M) with
coefficients of -.61 and -.54 respectively are of con
siderable importance. Turn Length (TL) with a coefficient
of -.45 is of some importance. The coefficient of -.23
assigned to Facing (F) indicates that facing behavior was
not meaningful in the context of this particular canonical
relationship.
The strongest single variable from the interpersonal
orientation data was Expressed Control (EC) with its
coefficient of -1.00.Wanted Control (WC) with a rather
large coefficient of +.76 is the next strongest variable
entering into 10 Self-Esteem (SE), Self-Centrality
(Cent), and Expressed Inclusion (El) are of considerable
importance with their respective coefficients of -.60, -.60,
and -.55. Wanted Inclusion (WI) and Complexity (Comp) with
their respective coefficients of +.05 and -.03 are of
virtually no importance in the canonical variable 10 in
this relationship.
The second canonical correlation was not significant
(R^ = .78, = 34.41, df = 24, p > .05). Thus it was
concluded that there were no other linear combinations of
the two groups of data which would yield canonical variables
which were significantly correlated and which were

-47-
uncorrelated with the previously obtained canonical vari
ables. No further statistical analyses were carried out

CHAPTER IV
DISCUSSION
Personality Type and Coverbal Behavior
The first canonical correlation was statistically
significant. This indicates that there is a significant
relationship between the coverbal behaviors observed
during the conversations and the interpersonal orientation
characteristics measured by FIRO-B and SOOT.
Specifically, Hypothesis 1 stated that there is,
overall, a relationship between the coverbal behaviors
observed and the interpersonal orientation measures in
this study. This hypothesis followed from the findings of
previous research (e.g., Allport, 1961; Chappie et al.,
1954; Goldman-Eisler, 1954), and was significantly supported.
The principal personality dimension to emerge from this
relationship was that of need for control. Expressed con
trol and wanted control, two of the scales from FIRO-B,
were the two strongest interpersonal orientation factors
in the study. As pointed out in an earlier chapter, it is
possible for there to be significant canonical relationships
beyond the first relationship. None, however, was found
beyond the first relationship in this study. If others had
been identified, the coverbal behavior variables and the
-48-

-49-
interpersonal orientation measures would have been combined
in ways different from the combinations in the first
relationship. That would have permitted further discussion
of other personality types.
Hypothesis 6 predicted that self-complexity would not
be related to coverbal behavior. This hypothesis was viewed
as strongly supported as the self-complexity coefficient
was extremely low, the lowest, in fact, of all the coef
ficients. This hypothesis was consistent with Ziller's
(1973) description of self-complexity. Ziller's high self
complexity individual is seen as a well-balanced person
without extreme interpersonal needs, whereas, this study
was looking at the behavioral correlates of elevated inter
personal needs. Therefore, support of this hypothesis is
seen as further validation of Ziller's concept of self-
complexity.
Hypothesis 4 predicted that a large amount of head
tilting and short turns would be associated with high
wanted control and low expressed control. This hypothesis
was strongly supported as the predicted relationships were
consistent with the relationships indicated by the canonical
coefficients. The person with a high need to be controlled
and a low need to control others (the abdicrat) does, in
fact, exhibit submissive, non-controlling behavior in the
form of a great amount of head tilting and short turns.
The relationship equally supports Schtz' concept of the
autocrat, that is the person with a high need to control

-50-
others and a low need to be controlled by others, as this
person exhibits minimal head tilting behavior and long
turns. The abdicrat and the autocrat are examples of
excessive needs in the interpersonal orientation dimension
that Schtz identifies as the interpersonal need for con
trol. It appears then, that two behavioral manifestations
of the interpersonal need for control are head tilts and
turn length.
Hypothesis 3 predicted that turn length and hand
movement would be directly related to wanted control and
self-centrality. This hypothesis received rather strong,
although partial, support. Turn length and hand movement
are directly related to self-centrality. However, they are
inversely related to wanted control. All variables varied
in the predicted direction with the exception of wanted
control. It should be noted that, of all the hypotheses
predicting directional relationships, this is the only
hypothesis in which a control variable varied differently
from the predicted direction. It is also the only one of
the directional hypotheses which makes a prediction about
self-centrality. It is possible that self-centrality is
more significantly related to turn length and hand move-
9
ment than is wanted control. That is, given that a person
exhibits lengthy speaking turns and a great deal of hand
movement, there may be reason to believe that he is highly
self-centered, but insufficient data for predicting his
need to control others. His high self-centeredness may

-51-
produce lengthy speaking turns and a great deal of hand
movement regardless of his need to control others.
Hypothesis 5 predicted that a large amount of body
touching behavior and short turns would be associated with
low self-esteem and low expressed control. This hypothesis
received partial support. The only measure which did not
vary in the predicted direction was body touching. Body
touching was directly related to self-esteem and expressed
control, rather than inversely related as predicted. In
fact, the extent to which body touching contradicted the
hypothesized relationship (judging from the relatively
large magnitude of its coefficient) was surprising. The
strength of this relationship indicates that body touching
may not be well understood in terms of its psychological
significance. Little previous research has been conducted
with this coverbal behavior. Body touching was seen at the
formulation of this study as a behavioral manifestation of
feelings of anxiety, of the need for self-protection, and
the need to not become involved with others (Fast, 1970;
Freedman and Hoffman, 1967; Scheflen, 1965). In fact, however,
the canonical relationship indicated that body touching is
related to the need for control. This result is more in
keeping with the suggestion by Scheflen and Scheflen (1972)
that body touching is related to dominance.
Hypothesis 2 predicted that facing and head tilts were
directly related to wanted inclusion and inversely related
to expressed inclusion. In view of the weak input of three

-52-
of the variables (facing was the weakest of all coverbal
behaviors, and expressed inclusion and wanted inclusion fell
fifth and sixth in order of magnitude of the seven inter
personal orientation measures), there was no support for
this hypothesis.
In sum, this study provided statistical support for
the relationship between coverbal behavior and interpersonal
orientation. The most salient interpersonal orientation
characteristic of this relationship was the interpersonal
need for control as measured by the Fundamental Inter
personal Relations Orientation-Behavior scale (Schtz,
1966, 1967). The most salient of the coverbal behaviors
of this relationship were head tilts and body touching.
The overall personality types that emerged are clearly
the abdicrat and the autocrat as described by Schtz (1966)
and the dominance dimension as described by Scheflen and
Scheflen (1972). The main coverbal behavioral manifesta
tions of the abdicratic (or submissive) type is a large
amount of head tilting, little body touching, and short
speaking turns. The coverbal manifestations of the auto
cratic (or dominant) type is the mirror image: little head
tilting, a large amount of body touching, and long speaking
turns.
Methodological Considerations
Preserving the naturalness of the experimental
environment was emphasized throughout the study. Subjects

-53-
were placed in a comfortable, familiar-type of setting
with much care given to making the environment as dis
traction-free as possible. Data recording during the
interactions was accomplished as unobtrusively as possible,
with all recording equipment (except the microphone) at
least ten feet from the subjects. Subjects discussed
relevant social issues with their peers. No time con
straints were placed on the conversations, and no one was
present during the conversations except the two interactants.
It was felt that, to the extent that naturalness could be
preserved, individuals' authentic, unique coverbal styles
would be displayed during the conversation. The interactive
aspect of the behavior was viewed as critical for this
research. In the past, some researchers have studied
coverbal behavior as it related to personality by analyzing
data which were obtained during subjects' speaking a mono
logue (e.g., Gottschalk, Winget, and Gleser, 1969; Vargas,
1968). The monologue seems to be an inappropriate method
for gathering data which are hoped to be shown related to
interpersonal orientation. By definition, the interpersonal
component of communication is removed from the monologue.
A point should be stated about the sample size in
this study. The sample size was rather small. The prin
cipal reason for this is the extremely lengthy work
involved in reducing the data from video tapes of conver
sations to analyzable data. Additionally, this was an
exploratory study. As Hays (1963) pointed out, small

-54-
sample sizes are often utilized in exploratory studies
because the exploratory studies are trying to identify
relationships for further study and refinement. The ex
ploratory study serves as a guide for further research, and
that is one of the purposes of this study. Further justi
fication for the use of a rather small sample size lies in
the test for statistical significance. As is the case with
all statistical tests, sample size was taken into considera
tion in the test for significance in this study. While one
could virtually always make the statement, 'the larger the
sample size, the better', in this case, there seemed to be
adequate support for the use of the rather small sample.
A final point is made in reference to the methodology.
Modern videotaping equipment, elaborate computer reduction
of the data, and complex statistical analysis were utilized
However, the human observer was an integral part of the
research. No behavioral data were used in this study which
could not be reliably observed and recorded by the rela
tively untrained observer. In order to demonstrate, in a
hard-data sense, the relationship between coverbal behavior
and interpersonal orientation, the modern electronic
equipment was employed. Future application of the findings
of the study can, however, be accomplished, without any
special equipment and without disruption of normal face-to-
face conversational interaction. The only equipment
required is the ever-present, best of all machines, the
human.

-55-
Future Research and Application
Adequate evidence was presented to encourage and
warrant further investigation of the relationships between
coverbal behavior and interpersonal orientation. Of the
interpersonal orientation measures studied, Schtz' control
dimensions appeared especially salient. Ziller's self
esteem and self-centrality measures also appeared promising
in terms of further research. Following the identification
of these personality characteristics as related to coverbal
behavior, future research can focus more specifically on
these behaviors and their relationships with coverbal
behavior. Self-complexity was supported as a measure of
personal adjustment and should be explored further in that
context.
For the researcher interested in coverbal behavior,
there was no evidence pointing to the elimination of any
of the five coverbal behaviors for consideration in future
research. Turn length and facing, two often studied and
apparently important behaviors, appeared as the least
powerful of the coverbal behaviors. It may be that the
correct combination of behaviors and personality characteris
tics has not yet been identified in the case of these two
variables. This study hoped to identify basic relation
ships for further study. Which of those relationships are
selected is partly a function of the interest of the
researcher. Whichever relationships are investigated, the

-56-
naturalistic conversational interaction seems to be an
excellent setting within which to study them.
Taking a more long-range viewpoint, the application
of the more refined and more clearly defined relationships
has important implications for humans. Scheflen and
Scheflen (1972) pointed out in quite some detail the per
vasive extent to which coverbal behavior is a vital part
of our communication. They presented impressive evidence
of the fact that this subtle means of communication can be,
and in fact is, extensively utilized to control members of
society. Not only is coverbal behavior a powerful means of
control, but it is also so subtle that it can be used out
side the awareness of the controlled person. Further
research will lead to a better understanding of the
'coverbal language'. The subliminal message is a difficult
one to which to respond. As heightened awareness is
accomplished, more rational, adaptive responses should
follow. This may lead to a new freedom for the individual
from undesirable control imposed by society. In a more
positive way, this increased awareness can result in in
creased effectiveness on the part of those individuals who
are attempting to bring more freedom and individuality for
the person, for example in the work of the better trained
psychotherapist.

APPENDICES

APPENDIX A
FIRO-B

SUBJECT NUMBER DATE
DIRECTIONS: This questionaire is designed to explore the
typical ways you interact with people. There are, of
course, no right or wrong answers; each person has his own
ways of behaving.
Sometimes people are tempted to answer questions like
these in terms of what they think a person should do. This
is not what is wanted here. We would like to know how you
actually behave.
Some items seem similar to others. However, each item
is different so please answer each one without regard to
the others. There is no time limit, but do not debate long
over any item.
For each statement below, decide which of the following
answers best applies to you. Place the number of the
answer on the left of the statement. Please be as honest
as you can.
1.usually 2. often 3. sometimes 4. occasionally
5. rarely 6. never
1. I try to be with people.
2.I let other people decide what to do.
3.I join social groups.
4.I try to have close relationships with people.
5. I tend to join social organizations when I have the
opportunity.
6. I let other people strongly influence my actions.
7.I try to be included in informal social activities.
8. I try to have close, personal relationships with
people.
9. I try to include other people in my plans.
10.I leg other people control my actions.
11.I try to have people around me.
12.I try to get close and personal with people.
13.When people are doing things together I tend to
join them.
-59-

-60-
14. I am easily led by people.
15. I try to avoid being alone.
16. I try to participate in group activities.
For each of the next group of statements, choose one of the
following answers:
1. most 2. many 3. some 4. a few
people people people people
5. one or two 6. nobody
people
17.
I
try
to be
friendly to people

18.
I
let
other
people
decide what
to do.
19.
My
r personal
relations with peo
pie are cool and
di
stan
t.
20.
I
let
other
people
take charge
of things.
21.
I
try
to have close
relationsh
ips with people.
22.
I
let
other
people
strongly in
fluence my actions.
23.
I
t ry
to get
close
and persona
1 with people.
24.
I
let
other
people
control my
actions.
25.
I
act
cool and dist
ant with pe
ople.
26.
I
am e
asi ly
led by
people.
27.
I
try
to have close
, personal
relationships with
pe
ople

28.
I
like
peop 1
e to in
vite me to
things.
29.
I
like
peopl
e to ac
t close and
personal with me.
30.
I
try
to influence
strongly other people's action
31.
I
like
peopl
e to in
vite me to
join in their
ac
tivi
ties.
32.
I
like
peopl
e to ac
t close toward me.
33.
I
try
to take charge of things
when I am with
people.

-61-
34 .
I
like
people
to
inc
lude me i:
n their acti
viti
Les .
35.
I
like
people
to
act
cool
and
distant tow
ard
me.
36.
I
try
to have
other
people
do
things the
way
I
want them done
37.
I
like
people
to
ask
me to
pa
rticipate in
the
?ir
dj
LSCUS
sions.
38.
I
like
people
to
act
f rien
dly
toward me.
39.
I
like
people
to
invite me
to
participate
in
their
activi
ties.
40.
I
like
people
to
act
distant
toward me.
For each of the next group of statements, choose one of the
following answers:
1. usually 2. often 3. sometimes 4. occasionally
5. rarely 6. never
41.
I
t ry
to be the
dominant pe
rson when I am with
pe
:ople

42.
I
like
people
to
invite me
to
things.
43.
I
like
people
to
act close
toward me.
44.
I
try
to have
other people
do
things I want done.
45.
I
like
people
to
invite me
to
join their activiti
46.
I
like
people
to
act cool and
distant toward me.
47.
I
try
to infli
jen
ce strongly
other people's actions
48.
I
like
people
to
include me
in
their activities.
49.
I
like
people
to
act close
and
personal with me.
50.
I
t ry
to take
ch,
arge of thi
ngs
when I am with
pe
ople

51.
I
like
people
to
invite me
to
participate in their
ac
ti vi
ties.
52.
I
like
people
to
act distan
t t
oward me.

-62-
53. I try to have others do things the way I want them
done.
54. I take charge of things when I am with people.

APPENDIX B
GENERAL INSTRUCTIONS FOR SOOT
SAMPLE ITEM FROM SLEF ESTEEM SCALE
SAMPLE ITEM FROM SELF CENTRALITY SCALE
COMPLEXITY SCALE

Social Orientation Tasks
The quest
indication of
other people,
words are avoi
ment designed
will tell us s
their percepti
This inst
Health, Educat
Please wo
little more th
ions which follow are designed to provide an
the way you look at yourself and significant
In this description of yourself and others,
ded. This is a social psychological instru-
for research purposes only. Hopefully, it
omething about differences among people in
ons of self and others.
rument has been
approved by
the Depar
tment of
ion
and Welfare,
Office of
Education.
rk
as quickly as
possible.
It should
require
an
ten minutes.
-64-

-65-
The circles below stand for people. Mark each circle with
the letter standing for one o
this in any way you like, but
and do not omit anyone.
F someone who is flunking
H the happiest person you
know
K someone you know who is
kind
the people in the list. Do
use each person only once
S yourself
Su someone you know who is
successful
St the strongest person you
know

-66-
Self Centrality Item
In the large circle below, draw two circles -- one to stand
for yourself and a second to stand for a friend. Place an
S in the circle for self and an F in the circle for your
friend.

-67-
Complexity Scale
Instructions: Here is a list of words. You are to
read the words quickly and check each one that you think
describes YOU. You may check as many or as few words as
you
like -- but be
HONEST.
Don't check words
that tell
what
kind of person
you should be. Check
words that tell
what
kind of a person you
really are.
1.
ab le
20.
careless
39.
false
2.
active
21.
charming
40.
fine
3.
afraid
22.
cheerful
41.
fierce
4.
alone
23.
clean
42.
foolish
5 .
angry
24.
clever
43.
friendly
6.
anxious
25.
comfortable
44.
funny
7.
ashamed
26 .
content
45.
generous
8.
attractive
27.
cruel
46.
gentle
9.
bad
28.
curious
47.
glad
10.
beautiful
29.
delicate
48.
good
11.
big
30.
delightful
49.
great
12.
bitter
31.
different
50.
happy
13.
bold
32.
difficult
51.
humble
14.
brave
33.
dirty
52.
idle
15.
bright
34.
dull
53.
important
16.
busy
35.
dumb
54.
independent
17.
calm
36.
eager
55.
jealous
18.
capable
37.
fair
56.
kind
19.
careful
38.
faithful
57.
large

-68-
58.
lazy
59.
little
60.
lively
61.
lonely
62.
loud
63.
lucky
64.
mild
65.
miserable
66.
modest
67.
neat
68.
old
69.
patient
70.
peace ful
71.
perfect
72.
pleasant
73.
polite
74 .
poor
75.
popular
76.
proud
77.
quiet
78.
quick
79.
responsible
80.
rough
81.
rude
82.
sad
83.
se 1fish
84.
sensible
85.
serious
86.
sharp
87.
silly
88.
slow
89.
small
90.
smart
91.
soft
92 .
special
93.
stronge
94.
stupid
95.
strong
96.
sweet
97.
terrible
98.
ugly
99.
unhappy
100.
unusual
101.
useful
102.
valuable
103.
warm
104.
weak
105.
wild
106 .
wise
107.
wonderful
108.
wrong
109.
young

APPENDIX C
DISCUSSION PROBLEM 1
(ABORTION PROBLEM)
DISCUSSION PROBLEM 2
(LIVING-TOGETHER PROBLEM)

Discussion Problem 1
(Abortion Problem)
Mary is four weeks pregnant. Both she and her husband
Bob are somewhat ambivalent about having a baby at this
point in their lives. Bob tells Mary that she should do
whatever she would be most happy with. Mary can go ahead
and have the baby, or she can have an abortion. The couple
has a secure income and a baby would present no great
financial burden, but Mary is still not sure about having
the baby. She has other interests in her life and would
like to pursue them; on the other hand, she is not complete
ly adverse to being a mother. What do you think Mary should
do? How do you feel about Bob's position on the matter?
- 70-

-71-
Discussion Problem 2
(Living-Together Problem)
Sandra has just gotten home from work and finds her
boyfriend John just ending a telephone conversation. John
tells her that he was speaking with his parents and that
they are arriving the next morning for his birthday. He
then tries to convince Sandra that she will have to make
it look like she is not living there, but is merely
spending the day. While John feels very committed to Sandra
now after three months of living together, he is afraid his
parents would not react very well to finding out about them
at this time. Sandra, however, is of a different opinion,
and states that she does not need to pretend for anyone.
What do you think Sandra should do? How do you feel about
John's position in the matter?

APPENDIX D
COVERBAL BEHAVIOR MEAN VALUES
INTERPERSONAL ORIENTATION MEAN SCORES

-73-
Coverbal Behavior Mean Values
Turn
Length
Facing
Movement
Touching
Head
Tilts
5.8
.38
.20
. 70
.22
cl
Turn Length is the mean time in seconds of all turns. All other behavior means
represent the mean value of the proportion of each turn during which the
respective behavior was occurring.
Interpersonal Orientation Mean Scores
FIRO-B
SOOT
WI
El WC
EC
SE
Cent
Comp
5. 7
5.1 3.4
3.8
26.9
2.0
51. 8

REFERENCES
Adler, A. Social interest. New York: Putnam, 1939.
Allport, G.W. Pattern and growth in personality. New York:
Holt, Rinehart and Winston, 1961.
Argyle, M., and Dean, J. Eye contact, distance, and
affiliation. Sociometry. 1965 2_8, 289- 304.
Argyle, M. Social Interaction. New York: Atherton Press,
1969.
Ausubel, D. Ego development and the personality disorders.
New YrokT Grue and Stratton, 1952.
Bateson, G., Jackson, D., Haley, J. and Weakland, J. Toward
a theory of schizophrenia. Behavioral Science, 1956,
1, 251-264.
Cervin, V. Relationship of ascendant-submissive behavior
in dyadic groups of human subjects to their emotional
responsiveness. Journal of Abnormal and Social
Psychology, 1957, 5j4, 241-249.
Chappie, E.D. Quantitative analysis of the interaction of
individuals. Proceedings of the National Academy of
Science, 1939, 2j>_, 58-67.
Chappie, E.D. Personality differences as described by
invariant properties of individuals in interaction.
Proceedings of the National Academy of Science, 1940,
26, 10-16.
Chappie, E.D. The interaction chronograph: Its evolution
and present application. Personnel, 1949 2_5 2 85- 307.
Chappie, E.D. The standard experimental (stress) inter
views as used in interaction chronograph investigations.
Human Organization, 1953, 1_2, 23- 32.
Chappie, E.D., and Arensberg, C.M. Measuring human rela
tions: An introduction to the study of the interaction
of individuals. Genetic Psychological Monographs,
1940, 22, 3-147.
- 74-

- 75-
Chapple, E.D., Chappie, M.F., and Repp, J.A. Behavioral
definitions of personality and temperament character
istics. Human Organization, 1954 1_3> 34- 39.
Cherry, C. On Human Communication: A Review, a Survey,
and a Criticism. New York: Sciences Editions, 1961.
Dittmann, A.T. The relationship between body movements and
moods in interviews. Journal of Consulting Psychology,
1962, 26, 480.
Duncan, S.D., Jr. Some signals and rules for taking speak
ing turns in conversation. Journal of Personality and
Social Psychology, 1972 23, 283-292.
Ekman, P., and Friesen, W.V. Hand movements. Journal of
Communication, 1972 2_2_, 353- 374.
Ellsworth, P.C., and Ludwig, L.M. Visual behavior in
social interaction. Journal of Communication, 1972,
2_2 375-403.
Exline, R.V. Visual interaction: The glances of power and
preference. In J.K. Cole (Ed.), Nebraska Symposium on
Motivation (Vol. 19). Lincoln: University of
Nebraska Press, 1971.
Fast, J. Body language. New York: M. Evans and Company,
1970.
Freedman, N., and Hoffman, S. Kinetic behavior in altered
clinical states: Approach to objective analysis of
motor behavior during clinical interviews. Perceptual
and Motor Skills, 1967, 2_4 52 7- 539.
Frey, S. Tonic aspects of behavior in interaction. In
A. Kendon, R.M. Harris, and M.R. Key (Eds.), Organiza
tion of behavior in face-to-face interaction. The
Hague: Mouton, 1975.
Fromm, E. Escape from freedom. New York: Rinehart, 1941.
Fromm, E. Man for himself. New York: Rinehart, 1947.
Goldman-Eisler, F. A study of individual differences and
of interaction in the behavior of some aspects of
language in interviews. Journal of Mental Science,
1954, 100, 177-197.
Goldman-Eisler, F. Psycholinguistics. London: Academic
Press, 1968.

-75-
Gottschalk, L., Winget, C., and Gleser, G. Manual of
Instructions for using the Gottschalk-Gleser Content
Analysis Scales^ Berkeley: University of California
Press, 1969.
Hays, W.L. Statistics. New York: Holt, Rinehart and
Winston, 1963.
Kelly, F., Beggs, D. and McNeil, A. Multiple regression
approach. Carbondale, Illinois: Southern Illinois
University Press, 1969.
LaFrance, M., and Mayo, C. Racial differences in gaze
behavior during conversations: Two systematic
observational studies. Journal of Personality and
Social Psychology, 1976, 3_3, 547-552.
Lorenz, M., and Cobb, S. Language behavior in manic
patients. Archives of Neurology and Psychiatry, 1952,
67, 763-770.
Mahl, G.F. Disturbances and silences in patients' speech
in psychotherapy. Journal of Abnormal and Social
Psychology, 1956 5J5, 1-15.
Markel, N.N. (Ed.). Psycholinguistics: An introduction
to the study of speech and personality. Homewood,
Illinois: Dorsey Press, 1969.
Markel, N.N. Coverbal behavior associated with conversa
tion turns. In A. Kendon, R.M. Harris, and M.R. Key
(Eds.), Organization of behavior in face-to-face
interaction. The Hague: Mouton, 1975.
Markel, N.N., Bein, M.F., Campbell, W.W., and Shaw, M.E.
The relationship of self-rating of expressed inclusion
and speaking time. Language and Speech, in press.
Markel, N., Bein, M., Phillis, J. The relationship between
words and tone-of-voice. Language and Speech, 1973,
IS, 15-21.
Maslow, A.H. A theory of metamotivation: the biological
rootery of the value of life. Journal of Humanistic
Psychology, 1967 7_, 93- 127.
Maslow, A.H. Toward a psychology of being. (2nd ed.).
Princeton^ Van Nost rand, 1968.
Matarazzo, J.D., Wiens, A.N., Matarazzo, R.G., and Saslow,
G. Speech and silence behavior in clinical psycho
therapy and its laboratory correlates. In J.M. Shlien
(Ed.), Research in psychotherapy, Vol. 3. Washington:
American Psychological Association, 1968.

-77-
Mehrabian, A. A semantic space for nonverbal behavior.
Journal of Consulting and Clinical Psychology, 1970,
15, 248-257.
Mehrabian, A. Nonverbal communication. In J.K. Cole (Ed.),
Nebraska Symposium on Motivation (Vol. 19). Lincoln:
University of Nebraska Press, 1971.
Mehrabian, A., and Ksionzky, S. Models for affiliative
and conformity behavior. Psychological Bulletin, 1970,
_74 110- 126.
Morrison, D.F. Multivariate Statistical Methods. New
York: McGraw-Hill, 1967.
Norwine, A., and Murphy, 0. Characteristic time intervals
in telephonic conversation. Bell System Technical
Journal, 1938 _17 281- 291.
Osgood, C.E., Suci, G.J., and Tannenbaum, P.H. The
measurement of meaning. Urbana, Illinois: University
of Illinois Press, 1957.
Prebor, L.D. The natural history of a conversation.
Doctoral dissertation, University of
Florida, 1972.
Sapir, E. Speech as a personality trait. American Journal
of Sociology, 1927, 82, 895.
Scheflen, A.E. Quasi-courtship behavior in psychotherapy.
Psychiatry, 1965, 2_8, 245-257.
Scheflen, A.E., and Scheflen, A. Body language and social
order communication or behavioral control. Englewood
Cliffs, N.J.: Prentice-Hall, Inc., 1972.
Schtz, W.C. The interpersonal underworld. Palo Alto,
California: Science and Behavior Books, 1966.
Schtz, W.C. The FIRO scales manual. Los Angeles:
Consulting Psychologists Press, Inc., 1967.
Sullivan, H.S.
New York:
The interpersonal theory of psychiatry.
Norton, 1953.
Sullivan, H.S.
New York:
The fusion of psychiatry and social science.
Norton, 1964.
Vargas, R.A. A study of certain personality characteristics
of male college students who report frequent positive
experiencing and behaving. Doctoral
dissertation, University of Florida, 1968.

-78-
Verzeano, M., and Finesinger, J.E. An automatic analyzer
for the study of speech in interaction and in free
association. Science, 1949, 110, 45-46.
Woolf, H.B. (Editor in Chief). Webster's New Collegiate
Dictionary. Springfield, Massachusetts: G.C. Merriam
Co., 1973.
Ziller, R.C. The social self.
Inc., 1973.
New York: Pergamon Press

BIOGRAPHICAL SKETCH
William Wallace Campbell was born on September 28,
1939. After graduating from Murrah High School in Jackson,
Mississippi, he attended Millsaps College for two years.
After working in Washington, D.C. for two years and owning
a business in Atlanta for four years, he returned to college
in January, 1968. He received his B.A. degree from Georgia
State University in Atlanta, Georgia. After working
briefly for the Fulton County Family and Children Services
in Atlanta he entered graduate school at the University of
Florida in September, 1970 to begin his studies in clinical
psychology. He was awarded the M.A. degree in December,
1971. He completed his course work in clinical psychology
in the summer of 1973. Mr. Campbell completed a one-year
internship in clinical psychology at the Department of
Clinical Psychology in September, 1974. Shortly after
completing his internship he began working for the Jackson
ville Drug Abuse Program in Jacksonville, Florida while
continuing to work on his doctoral dissertation. He is
presently still employed at the Drug Program in Jacksonville.
During his first year of graduate study he was a Graduate
School Fellow. From the second year of graduate study
through his internship he was a United States Public Health
- 79-

-80-
Service trainee.
Mr. Campbell married the former Jeannie Clay in
December, 1968. He is the father of one daughter, Brooke
Elaine, and one son, Ross Bartley.

I certify that I have read this study and that in my
opinion it conforms to acceptable standards of scholarly
presentation and is fully adequate, in scope and quality,
as a dissertation for the degree of Doctor of Philosophy.
y )'] }'?1 CUt
Norman N. Markel, Ph.D.,
Chairperson
Professor of Speech, Anthro
pology, and Psychology
I certify that
opinion it conforms
presentation and is
I have read this study and that in my
to acceptable standards of scholarly
fully adequate, in scope and quality,
as a dissertation for the degree of Doctor of Philosophy.
Robert C. Ziller, Ph.D.
Professor of Psychology
I certify that
opinion it conforms
presentation and is
as a dissertation fo
I have read this study and that in my
to acceptable standards of scholarly
fully adequate, in scope and quality,
r the degree of Doctor of Philosophy.
Wiley Q. Rasbury, Ph.\
Assistant Professor of
Clinical Psychology

I certify that
opinion it conforms
presentation and is
as a dissertation fo
I have read this study and that in my
to acceptable standards of scholarly
fully adequate, in scope and quality,
r the degree of Doctor of Philosophy.
Robert H. Hornberger, Ph.D.
Assistant Professor of
Clinical Psychology
I certify that
opinion it conforms
presentation and is
as a dissertation fo
I have read this study and that in my
to acceptable standards of scholarly
fully adequate, in scope and quality,
r the degree of Doctor of Philosophy.
1 11 Ulll CL O U U U 111 C j ill LJ
Assistant Professor of
Speech
This dissertation was submitted to the Graduate Faculty of
the Deaprtment of Psychology in the College of Arts and
Sciences and to the Graduate Council, and was accepted as
partial fulfillment of the requirements for the degree of
Doctor of Philosophy.
December, 1976
Dean, Graduate School



-23-
The two interpersonal orientation instruments. FIRO-B,
utilizing the verbal self-report technique, attempts to
measure directly the individual's interpersonal orientation.
FIRO-B solicits information from the person as to his
behavior with other people. Through divulging these be
havioral preferences, the level of the person's needs in
different interpersonal areas is assessed. Since these
ratings are based on the person's expressed behavioral
preferences, the ratings should prove to be good predictors
of how the person will behave in an interpersonal situation.
A source of error in such predictions lies in the extent
to which the person honestly responded to the instrument.
SOOT, while tapping aspects of the personality which
are similar to those tapped by FIRO-B, is in some important
ways looking at the person with a different approach. SOOT,
except in the case of the complexity scale, avoids the use
of words in the assigned tasks. Even with the complexity
scale, words are not used in the same sense as in the
FIRO-B. A verbal self-report is not involved with the
complexity scale, rather it is an adjective checklist.
Throughout SOOT there is the attempt to avoid 'good-bad',
'right-wrong', 'acceptable-unacceptable' type choices on
the part of the responder. This is accomplished through
the extensive use of symbols rather than words. Another
important difference between SOOT and FIRO-B lies in the
interpersonal characteristics being measured. SOOT measures
characteristics which are shaped through interpersonal


-52-
of the variables (facing was the weakest of all coverbal
behaviors, and expressed inclusion and wanted inclusion fell
fifth and sixth in order of magnitude of the seven inter
personal orientation measures), there was no support for
this hypothesis.
In sum, this study provided statistical support for
the relationship between coverbal behavior and interpersonal
orientation. The most salient interpersonal orientation
characteristic of this relationship was the interpersonal
need for control as measured by the Fundamental Inter
personal Relations Orientation-Behavior scale (Schtz,
1966, 1967). The most salient of the coverbal behaviors
of this relationship were head tilts and body touching.
The overall personality types that emerged are clearly
the abdicrat and the autocrat as described by Schtz (1966)
and the dominance dimension as described by Scheflen and
Scheflen (1972). The main coverbal behavioral manifesta
tions of the abdicratic (or submissive) type is a large
amount of head tilting, little body touching, and short
speaking turns. The coverbal manifestations of the auto
cratic (or dominant) type is the mirror image: little head
tilting, a large amount of body touching, and long speaking
turns.
Methodological Considerations
Preserving the naturalness of the experimental
environment was emphasized throughout the study. Subjects


INTERPERSONAL ORIENTATION CORRELATES OF NONVERBAL
BEHAVIOR IN CONVERSATIONAL INTERACTION
By
WILLIAM WALLACE CAMPBELL
A DISSERTATION PRESENTED TO THE GRADUATE COUNCIL OF
THE UNIVERSITY OF FLORIDA
IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE
DEGREE OF DOCTOR OF PHILOSOPHY
UNIVERSITY OF FLORIDA
1976


-71-
Discussion Problem 2
(Living-Together Problem)
Sandra has just gotten home from work and finds her
boyfriend John just ending a telephone conversation. John
tells her that he was speaking with his parents and that
they are arriving the next morning for his birthday. He
then tries to convince Sandra that she will have to make
it look like she is not living there, but is merely
spending the day. While John feels very committed to Sandra
now after three months of living together, he is afraid his
parents would not react very well to finding out about them
at this time. Sandra, however, is of a different opinion,
and states that she does not need to pretend for anyone.
What do you think Sandra should do? How do you feel about
John's position in the matter?


self-complexity (Comp) from SOOT.
The following hypotheses were proposed:
1. There is an overall relationship between inter
personal orientation and nonverbal behavior.
2. F and HT are positively related to WI and in
versely related to El.
3. TL and M are positively related to WC and Cent.
4. High HT and low TL are associated with high WC
and low EC.
5. High T and low TL are associated with low SE and
low EC.
6. Comp is not correlated with nonverbal behavior.
The canonical correlation analysis was the statistical
procedure employed in this study. Hypothesis 1 was statis
tically supported (p < .004). The other five hypotheses
were evaluated through examination of the canonical coef
ficients. Hypotheses 4 and 6 appeared strongly supported.
Hypotheses 3 and 5 seemed partially supported. Hypothesis
2 was not supported.
Statistical support for hypothesis 1 was viewed as
support for further study of the relationships between
interpersonal orientation and nonverbal behavior. The
overall personality types which emerged were the abdicrat,
the autocrat, and the dominant personality. The main be
havioral manifestations of the abdicratic (or submissive)
type were a large amount of head tilting, little body
touching, and short speaking turns. The behavioral
vi 1


-27-
Hypotheses
Utilizing five coverbal behavior measures obtained
from 24 videotaped conversations, and seven interpersonal
orientation scores, four from FIRO-B and three from SOOT,
this study attempts to support the existence of a relation
ship between coverbal behavior and interpersonal orientation.
Naturalness of the conversations was maximized as subjects
were placed in a comfortable environment and asked to
discuss relevant social discussion problems with virtually
no time constraints placed on the length of their conver
sations .
Canonical correlation analysis is applied to the data.
The canonical correlation is conceptually more complex than
a t-test or a pairwise correlation such as the Pearson
product moment correlation. Similarly, its interpretations
are more complex. To predict an overall relationship
between the two groups of data is straightforward enough,
and that prediction is directly testable with the canonical
correlation analysis. However, to look inside the data
groups and make specific, directional relationship pre
dictions involving several variables involves risks beyond
the obvious risk of being wrong by predicting a relation
ship that does not exist. There exists the risk of failing
to support an hypothesis, not necessarily because the
hypothesized relationship does not hold true, but because
of the way the variables were combined in the computation


-56-
naturalistic conversational interaction seems to be an
excellent setting within which to study them.
Taking a more long-range viewpoint, the application
of the more refined and more clearly defined relationships
has important implications for humans. Scheflen and
Scheflen (1972) pointed out in quite some detail the per
vasive extent to which coverbal behavior is a vital part
of our communication. They presented impressive evidence
of the fact that this subtle means of communication can be,
and in fact is, extensively utilized to control members of
society. Not only is coverbal behavior a powerful means of
control, but it is also so subtle that it can be used out
side the awareness of the controlled person. Further
research will lead to a better understanding of the
'coverbal language'. The subliminal message is a difficult
one to which to respond. As heightened awareness is
accomplished, more rational, adaptive responses should
follow. This may lead to a new freedom for the individual
from undesirable control imposed by society. In a more
positive way, this increased awareness can result in in
creased effectiveness on the part of those individuals who
are attempting to bring more freedom and individuality for
the person, for example in the work of the better trained
psychotherapist.


-42-
Table 2
Pearson Correlation Coefficients
for Reliability Data
Behavior
R
Turn Length
.98
Facing
.96
Left Hand Movement
.92
Right Hand Movement
. 79
Left Hand Body Touching
.99
Right Hand Body Touching
.98
Head Tilt
. 89
Correlations were computed on duration of behavior as
originally recorded and as recorded by a different
monitor during 30-second intervals for each behavior of
each conversation.


APPENDIX D
COVERBAL BEHAVIOR MEAN VALUES
INTERPERSONAL ORIENTATION MEAN SCORES


-55-
Future Research and Application
Adequate evidence was presented to encourage and
warrant further investigation of the relationships between
coverbal behavior and interpersonal orientation. Of the
interpersonal orientation measures studied, Schtz' control
dimensions appeared especially salient. Ziller's self
esteem and self-centrality measures also appeared promising
in terms of further research. Following the identification
of these personality characteristics as related to coverbal
behavior, future research can focus more specifically on
these behaviors and their relationships with coverbal
behavior. Self-complexity was supported as a measure of
personal adjustment and should be explored further in that
context.
For the researcher interested in coverbal behavior,
there was no evidence pointing to the elimination of any
of the five coverbal behaviors for consideration in future
research. Turn length and facing, two often studied and
apparently important behaviors, appeared as the least
powerful of the coverbal behaviors. It may be that the
correct combination of behaviors and personality characteris
tics has not yet been identified in the case of these two
variables. This study hoped to identify basic relation
ships for further study. Which of those relationships are
selected is partly a function of the interest of the
researcher. Whichever relationships are investigated, the


-17-
reason, only the inclusion and control scales were used
from FIRO-B. Schtz describes interpersonal needs for
inclusion and control as follows (Schtz, 1966):
1) Interpersonal need for Inclusion (I): the need
to establish and maintain a satisfactory relation
with other people with respect to interaction and
association, and
2) Interpersonal need for Control (C): the need to
establish and maintain a satisfactory relation
with other people with respect to control and
power.
Schtz (1966) refers to each person's behavior in each
of these dimensions in terms of the behavior expressed (E)
towards others and the behavior wanted (W) from others.
Emerging then are four dimensions of the individual's
interpersonal orientation: Wanted Inclusion (WI), Expressed
Inclusion (El), Wanted Control (WC), and Expressed Control
(EC). Schtz (1966, 1967) developed the FIRO-B as a self-
report means of tapping each of these dimension. The
FIRO-B is presented in Appendix A.
Schtz (1966) describes individuals whose behavior
is consistently involved in satisfying his needs in the
different dimensions. He describes the high I person with
terms such as, "interact, communicate, and attend" (p. 21).
The person with high C needs is described as displaying
behavior such as, "dominance and control" (p. 22),
while the low C person is seen as,


TABLE OF CONTENTS
Page
ACKNOWLEDGEMENTS ii
LIST OF TABLES v
ABSTRACT vi
CHAPTER
I INTRODUCTION 1
Purpose of the Study 4
Coverbal Behavior 5
Interpersonal Orientation 8
Review of the Research Literature .... 9
Research Rationale 12
Coverbal Behaviors 12
Turn length 13
Facing 13
Hand movement 14
Body touching 15
Head tilts 15
Derived Coverbal Measures 15
Interpersonal Orientation 16
Fundamental Interpersonal
Relations Orientat ion-Behavior. 16
Self-Other Orientation Tasks. 19
The two interpersonal orien
tation instruments 2 3
Statistical Analysis 24
Hypotheses 27
II METHOD 31
Subjects 31
Equipment 31
Discussion Problems 32
Interpersonal Orientation Instruments . 33
Procedures 35
Data Reduction 36
Derived Coverbal Measures 38
Data Analysis 39
i i i


-6-
the verbal message. When a father speaks comforting words
to his frightened child while holding the child in a
gentle, caring manner the father is transmitting comple
mentary verbal and coverbal messages. The man at the beach
who tells his lover he would never look at another woman,
while he is watching every bikini-clad lovely in sight,
is sending verbal and coverbal messages which are contra
dictory. The college professor, expounding on a certain
theory to his class, while he still has shaving cream under
his ear, is probably sending a more prominent message about
himself (from the listeners' points of view he may be
sending messages about his grooming, the fact that he
overslept that morning, or whatever) with his coverbal
message (the shaving cream) than with his verbal message
concerning the theory at hand. In this example, the co
verbal message probably does not complement or contradict
the verbal message, most likely it is independent of the
verbal message. Nonetheless, the coverbal message is a
prominent, possible dominant, component of the communica
tion.
Bateson, Jackson, Haley and Weakland (1956) presented
a now well-known case for the need for agreement between
the verbal and coverbal message, proposing their theory
that through consistent conflict between the verbal and
coverbal messages the speaker creates what they termed the
'double-bind'. Their theory went on to suggest the
double-bind as a prime factor in the creation of the


-12-
Research Rationale
Coverbal Behaviors
The point was made earlier that there is always co
verbal behavior present during conversation. There is, in
fact, so much behavior occurring that it becomes an
important task of the researcher to choose just what
behavior will be selected out for analyzing. Four ques
tions were asked in this research so that the final
behavior selection could be accomplished. Is the behavior
detectably recorded when a dyadic conversation is video
taped? Can the relatively untrained observer reliably
record the onset and termination of the behavior as the
videotape is monitored? Are the behaviors easily observed
during normal conversational interaction? Are the
behaviors related to the interpersonal orientation of the
speaker?
Previous research and pilot studies prior to this
research answered the first three questions affirmatively.
The outcome of this research will have to answer the fourth
question. Since the results of this study were obviously
not available when the behaviors were selected, selection
was based on previous research which asked similar questions
(e.g., Chappie, 1940; Chappie and Arensberg, 1940; Chappie
et al., 1954; Lorenz and Cobb, 1952; Norwine and Murphy,
1938). The coverbal behaviors selected for investigation
Turn length, Facing, Hand movement, Body
were:


-46-
the relative importance of each variable is inferred from
the individual coefficients.
Looking at the coefficients of the coverbal behaviors,
Head Tilt (HT) with its coefficient of +1.00 provides the
greatest input into C^,. Touching (T) and Movement (M) with
coefficients of -.61 and -.54 respectively are of con
siderable importance. Turn Length (TL) with a coefficient
of -.45 is of some importance. The coefficient of -.23
assigned to Facing (F) indicates that facing behavior was
not meaningful in the context of this particular canonical
relationship.
The strongest single variable from the interpersonal
orientation data was Expressed Control (EC) with its
coefficient of -1.00.Wanted Control (WC) with a rather
large coefficient of +.76 is the next strongest variable
entering into 10 Self-Esteem (SE), Self-Centrality
(Cent), and Expressed Inclusion (El) are of considerable
importance with their respective coefficients of -.60, -.60,
and -.55. Wanted Inclusion (WI) and Complexity (Comp) with
their respective coefficients of +.05 and -.03 are of
virtually no importance in the canonical variable 10 in
this relationship.
The second canonical correlation was not significant
(R^ = .78, = 34.41, df = 24, p > .05). Thus it was
concluded that there were no other linear combinations of
the two groups of data which would yield canonical variables
which were significantly correlated and which were


- 75-
Chapple, E.D., Chappie, M.F., and Repp, J.A. Behavioral
definitions of personality and temperament character
istics. Human Organization, 1954 1_3> 34- 39.
Cherry, C. On Human Communication: A Review, a Survey,
and a Criticism. New York: Sciences Editions, 1961.
Dittmann, A.T. The relationship between body movements and
moods in interviews. Journal of Consulting Psychology,
1962, 26, 480.
Duncan, S.D., Jr. Some signals and rules for taking speak
ing turns in conversation. Journal of Personality and
Social Psychology, 1972 23, 283-292.
Ekman, P., and Friesen, W.V. Hand movements. Journal of
Communication, 1972 2_2_, 353- 374.
Ellsworth, P.C., and Ludwig, L.M. Visual behavior in
social interaction. Journal of Communication, 1972,
2_2 375-403.
Exline, R.V. Visual interaction: The glances of power and
preference. In J.K. Cole (Ed.), Nebraska Symposium on
Motivation (Vol. 19). Lincoln: University of
Nebraska Press, 1971.
Fast, J. Body language. New York: M. Evans and Company,
1970.
Freedman, N., and Hoffman, S. Kinetic behavior in altered
clinical states: Approach to objective analysis of
motor behavior during clinical interviews. Perceptual
and Motor Skills, 1967, 2_4 52 7- 539.
Frey, S. Tonic aspects of behavior in interaction. In
A. Kendon, R.M. Harris, and M.R. Key (Eds.), Organiza
tion of behavior in face-to-face interaction. The
Hague: Mouton, 1975.
Fromm, E. Escape from freedom. New York: Rinehart, 1941.
Fromm, E. Man for himself. New York: Rinehart, 1947.
Goldman-Eisler, F. A study of individual differences and
of interaction in the behavior of some aspects of
language in interviews. Journal of Mental Science,
1954, 100, 177-197.
Goldman-Eisler, F. Psycholinguistics. London: Academic
Press, 1968.


-25-
variable of interest. With the multiple regression analysis
equation not only is there a single indicator of the
relationship between the single variable and the linearly
combined group of variables, but we also are supplied with
the coefficients of the variables themselves as they appear
in the regression equation. These coefficients are the
weights assigned to each of the original variables in the
group. The coefficients are indications of the magnitude
(numerical value of the coefficient) and the direction
(algebraic sign with the coefficient) of the contribution
made by each of the original variables as they were com
bined. Thus, we have an idea of to what extent and in
what direction each variable makes its contribution.
The canonical correlation goes still one step beyond
the multiple regression analysis. The canonical correla
tion linearly combines variables from two groups of data
so that a pair of canonical variables emerge, one from each
group of data. The correlation between these newly created
canonical variables is computed and then it is possible to
address the question of whether the two groups of data are
related. The level of the correlation indicates the
strength of the relationship, that is, it is an indication
of how much common data is contained in the two groups.
Similarly to the regression equation, we are supplied with
coefficients which indicate the magnitude and direction of
the influence of each original variable on the canonical
variable.