THE EXPRESSMK 0F AFFECT 55% WE xCQUAWTAMCE PROCESS Tnesis far ”the Degree :31 M A. MECHEGAN STATE MNEVERSW JANET M. BULLOCK 196-9 ABSTRACT THE EXPRESSION OF AFFECT IN THE ACQUAIETANCE PROCESS by Janet M. Bullock This study attempted to characterize the expression of affect in the acquaintance process and to generate hypotheses for further research. Data were obtained from the first, second, seventh and eighth session tapes of meetings between previously unacquainted college students who were recruited by Conley (1968) and randomly assigned to male-female dyads. The results from the nine dyads studied demonstrated none of the expected changes in the apprOpriateness, tense, direction and quality of affect over time. Responses were highly variable at first, but stabilized into regular patterns over time. Hence. it would seem that people have very definite and regular ways of expressing themselves. There were significant sex differences in that males talk more than females overall: they express more inappropriate affect Janet M. Bullock than females; and they express less affect about themshlves than do females. It was suggested that males are more anxious in the dyadic relation and this inhibits them. A sex x sessions inter- action in the use of the present tense suggests. too, that a shift to greater intimacy causes the male proportion of inappro- priate affect to increase further, indicating both greater anxiety in the more intimate situation and greater ambivalence about the relationship. The data suggest a variety of further studies with student, non-student and therapeutic populations and with same-sex dyads to test the consistency of the findings. Refinements in the scoring system and an additional scoring were suggested to examine more closely the overall content of conversation with and without affect. THE EXPRESSION OF AFFECT IN THE ACQUAINTANCE PROCESS By Janet M. Bullock A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF ARTS Department of Psychology 1969 § {8353’ /0/2?/49‘? To Terry TABLE OF CONTENTS Page ACKI\.ONLEDGI‘iEI“JTSO0...OOOOOOOOOOOOO00.0.0.0...0.000.000.000111 LIST OFTABLESOOOO0......0.00.0.0...OOIOOIIOOOOOOOOOIOOOOV LIST OFAPPEI'JVDICESOOOOOOO0.0.0....0.0.0....OOOOOOOOOOOOOOV1 Chapter I. INTRODUCTION.......................................1 Statement of the Problem.......................1 History of the Problem and Bationale...........1 Research Methods...............................h Research Concerning Expressions of Affect......6 Summary........................................9 II 0 I‘lErI‘HODOLOGY. O O O O O O O O I O O O O O O O O O I O O O O O O I O O O O O C O O O O O O O 12 overv1eW' O O I O O I O I O O O O O O O O O O I O O O O O O O O O O O O O O I O O O I 12 subJeCtS O O O O O O O O O O O O O O O O C O O O O O O O O O O O O O O O O O O O O C O 12 Procedure. 0 O O O O O I O O O O O O O O O O O 0 C O O O O O O O O O C O O O C O O O 12 III 0 RESULTS. 0 O O O O l O O O O 0 O I O O 0 O O I O I O O O I O O O O O O 0 I O I O 0 O O O O O O 15 IV. DISCUSSIOIJ. . O O 0 O O O O O O O O O O O O O O C O O O O O O I O O O I O I O I O O I C O .2” V. SUI‘EDIIMIOOOIOOOOO0.00.00.00.00...0.00.00.00.0000000028 APPEI‘JDICESOOOO0.00.00.00.00.00....O...0.0.00.00000000000030 .-'n ACKNO NLEDGMEN T S I wish to thank Dr. Gary Stollak. Dr. Andrew Barclay and Dr. Lawrence Messe for their help and direction. 111 LIST OF TABLES Table Page 1. Male, Female and Dyad Means Across Sessions............16 2. Dyad Means and E for One Way AOV.......................18 3. Sex x Sessions AOV: Proportion of Utterances Expressing Affect Toward the Self......................20 h. Sex x Sessions AOV: Proportion of Utterances Expressing Affect in the Present Tense.................22 .I .3 LIST OF APPENDICES Appendix Page A. List Of HypOtheseS TesteflOOOOOOOOOOOOOOOOOOOOOO00.00.30 B. sample scoring SheetOIOOOOOIOOIOIOOOOOOOOOOOOOOOOI.0033 C. AO‘JTables.OIOIOOOOOOOOIOO.00.0.00...00.0.0000000000034 vi THE EXPRESS ON OF AFFECT IN THE ACQUAINTANCE PROCESS I. INTRODUCTION ‘§tatement of the Problem This study grew out of a combined interest in the acquaint- ance process and the process of psychotherapy. It was designed to examine the expression of affect in the acquaintance process and to generate hypotheses on which further research could be founded. Tapes of nine eXperimentally contrived dyads were scored with regard to how much talking was done by each member of the dyad, how often there was silence and how much of the talking consisted of expressions of affect. Also eXplored were sex differences. changes across time in the type of affect, the manner in which it was expressed and the subject matter related to the affect. History of the Problem and Rationale Even when psychotherapy was in its infancy, the importance of getting the client to express his feelings was stressed (Breuer and Freud, 1895). But instead of deve10ping into a cohesive body of theory and techniques, psychotherapy has since generated a proliferation of theories and systems. At the same .time, the application of the medical model to "problems of living" has lent respectability to the mental health profession (Szasz. 1961) and there has been a sharp increase in the number of people seeking professional help with their problems (Schofield, 196M) with consequent pressure on psychotherapists to provide more and better services. Hence. there is a need to examine present theories and techniques to determine which are most effective and efficient for dealing with troubled people. Eysenck (1952). surveying recovery and improvement rates for pe0p1e treated by psychoanalytic therapy. eclectic therapy, and custodial care. found no differences and concluded that no psychotherapy system was better than any other. Even behavioral therapies which have enjoyed very high recovery rates in the past, find their efficiency challenged as they are confronted with a wider range of patients (Klein. et.a1.. 1969). Taking a somewhat different approach. Fiedler (i950a) sought to establish whether the theoretical differences between schools created genuine differences in what respective therapists attempted to achieve or whether the differences were primarily semantic. Using Q-sort ratings. he found that therapists of all schools-and non-therapists described the ideal therapeutic rela- tionship in much the same terms; moreover. (Fiedler. 1950b) found that the "expert" therapists of the psychoanalytic, Bogerian and Adlerian schools, who approximated the ideal relationship in therapy. behaved more like each other than like the neophytes of their respective schools. The most significant factor in psychotherapy seems to be the warm and accepting atmosphere created by the therapist (irrespective of his theoretical convic- tions) which encourages the client to express his feelings freely. On this basis, several authors (Fiedler. 1950a; Rogers, 1957; Schofield. 1964; Stollak. Gureney and Rothbery, 1966) have suggested that the therapeutic relationship is just a specific case of good interpersonal relations in general. Schofield (196A) goes so far as to assert that many people who enter therapy are really only looking for a good friend and Rogers (1957} suggests that the chance for an effective therapeutic relationship~would be improved if therapist and client were matched beforehand on the basis of personality Q—sorts. This suggests that in order to learn about the psychotherapy relationship it would be profitable to investigate the nature of "good" and ”normal" interpersonal relationships outside the therapeutic situation. Another line of reasoning which complements this is that of recent authors including Berne (196i) and Haley (1963). They have emphasized that virtually all human behavior takes place in the presence of human observers and is therefore communicative. Moreover. behavioral information cues are not presented as iso- lates but in series, for communication is ongoing. Any rela- tionship, therapeutic or otherwise. is a communication process. The study of any acquaintance process can provide normative data about communications which can be applied to the therapy process in several ways. First, the client in therapy is being encour- aged to communicate at great length and the therapist must sift through all the cues with which he is presented; normative data can serve to alert him to communications that are atypical, and created by the therapist (irrespective of his theoretical convic- tions) which encourages the client to express his feelings freely. On this basis, several authors (Fiedler. 1950a; Rogers. 1957; Schofield, 1964; Stollak, Gureney and Rothbery, 1966) have suggested that the therapeutic relationship is Just a specific case of good interpersonal relations in general. Schofield (1964) goes so far as to assert that many people who enter therapy are really only looking for a good friend and Rogers (1957) suggests that the chance for an effective therapeutic relationship would be improved if therapist and client were matched beforehand on r the basis of personality Q-sorts. This suggests that in order to learn about the psychotherapy relationship it would be profitable to investigate the nature of “good" and "normal” interpersonal relationships outside the therapeutic situation. Another line of reasoning which complements this is that of recent authors including Berne (i961) and Haley (1963). They have emphasized that virtually all human behavior takes place in the presence of human observers and is therefore communicative. Moreover. behavioral information cues are not presented as iso- lates but in series, for communication is ongoing. Any rela- tionship, therapeutic or otherwise, is a communication process. The study of any acquaintance process can provide normative data about communications which can be applied to the therapy process in several ways. First, the client in therapy is being encour- aged to communicate at great length and the therapist must sift through all the cues with which he is presented; normative data can serve to alert him to communications that are atypical, and to forms of expression which he might otherwise miss. Second, insofar as the client is sulfering from ”problems 01 living“ (Szasz. 1961). or inadequate interpersonal relations. the norms ‘provide an objective goal and model for the therapy process. Third. knowledge of norms for good interpersonal relations may suggest techniques to facilitate the achievement of a good therapeutic relation in which the client can express his feelings. figsearch Methods Knowing the problem to be investigated does not answer the question of how to go about studying it. As early as 1933, (Thomas. 1933) there was an impetus for research on human behavior in open-ended situations. but recent widespread use of tape recorders and video tapes has greatly expended the possi- bilities for research on interpersonal relations as processes. Barker (1963), in speaking for more extensive study of ongoing behavior, recognizes the greater complexity and reduced control in such studies, but argues that psychological research, in seeking a truer picture of man, must study him in natural situa- tions. Within this framework he notes that two types of measures can be used: ”behavior tesserae" which are designed to fit the experimenter's hypothesis and "behavior units" which occur naturally and can be described with little specific training. The former type of measure is convenient, for it enables the experimenter to make a direct statement regarding a theoretical formulation; however. it may ignore or distort the true nature of the interaction in process. The use of behavior units reduces eXperimental control and demands larger chunks of data. but it is advantageous insofar as it is not limited in application to any particular theoretical base. It can, then, provide a more objective picture of what actually transpires, physically and verbally, as people get acquainted.and interact. Most of the researchers on the acquaintance process have emphasized a specific variable which they feel is responsible for the formation and maintenance of a relationship. Newcomb (1961), for example, emphasized the perceived similarity of pre-acquaint- ance attitudes and interestl as measured by several question- naires. Winch (1952) defines attraction as a function of complementary needs, and Thibaut and Kelley (1959) assert that relationships are maintained only insofar as their rewards excede the costs of maintaining them. Most of these studies deal with "behavior tesserae” and, while these theories are informative, they raise the old questions about who (if anyone) is right, what variables are most important, and to what extent are all the theorists talking about the same things in different vocabularies. Little work has been done toward compiling normative data concerning ongoing verbal behavior. Soskin and John (1963) selected the utterance and the ”conversational episode” as behavior units for study. These units were examined on the bases of both structure and content. Utterances were categorized into eight types which were judged either "functional" (informa- tional) or "relational" (emotional and/or concerned with the partner or dyadic unit). The Soskin and John study suggests ways to objectively examine lexically manifested affect, but many cues used in inter- personal relationships, and particularly those used by therapists, are non-lexical, suprasegmental feature of language. McQuown (1957), pittenger, Hockett and Danehy (1960), and Dittman and Wynne (1961) have demonstrated that the methods of descriptive linguistics can be applied to both conversation and to the therapeutic interview reliably. Mahl (1956, 1959a, 1959b, and 1963) uses linguistic descriptions of suprasegmental features of language and some grammatical and content-related cues to score for the presence of anxiety. A combination of these methods, with more emphasis on the lexical aspects, is used here to investigate more general types of affect, the manner of expressing it, the topics regarding which it is expressed, and whether these change in the course of interaction processes. Research Concerning the EXpressions g£_Affect in Acquaintance and Therapngrocess One such study was done by Seeman (19h9). In evaluating tape-recorded successful therapeutic interviews, he found that positive expressions about self and others increased and negative expressions decreased over sessions. In regard to the acquaint- ance process, this would also suggest an increased amount of positive affect expressed over time. Thibaut and Kelley (1959) remarked the importance of the capacity of the partners in a relationship for saying nice things to each other. The expres- sion of negative affect might depend on variables other than the mutual liking of the partners, so no change in that can be posted on the basis of Seeman's study. Unfortunately, few studies are concerned with the direct manifestation of affect as was Seeman's (19M9). The emphasis is usually on latent, unconscious or indirectly expressed feelings (Gottschalk and Hambridge, 1955: Mahl, 1956, 1959a 1959b, 1963; and Pittenger, Hockett and Danehy, 1960). Those areas which contributed most to the thinking in this thesis can be subsummed under the tOpics of research on anxiety, research on formality in interpersonal relations, and studies on warmth. Anxiety. Mahl (1956) demonstrated that the initial stages of a relationship are characterized by greater anxiety and this *is expressed indirectly. In terms of manifest expressions, this might lead to the use of more inapprOpriate or contextually dis- cordant affect initially. Moreover, anxiety would impose self- consciousness on the speaker and consequent restraints on the expression of affect and the direction of affect expressed; specifically, the affect expressed initially would be more ambivalent or qualified and it would be directed to less personal, stereotyped conversation topics. Sarbin (1954) says that women are more willing than men to make inferences from social cues. This would imply that women are less anxious in the social situation and would therefore exhibit less inappropriate affect than males. Less anxiety and consequent self-consciousness would permit more expression of affect about themselves for females, too. Formality. Closely related to the findings on anxiety are those derived from studies on levels of formality in interactions.' Thibaut and kelley (1959) cite several studies indicating that initial meetings in a relationship are characterized by greater formality - this includes more rigid postures and more stereo- typed conversation gambits. ho statements made are about the expression of affect, but it could be eXpected that easing the restraints of formality over time would lead to more talking- i.e. an increase in the number of utterances and a decrease in the amount of silence - and that while one member of the dyad might dominate the conversation initially, they would tend to share the conversation load more equally as the relationship progressed. Reduced formality would also be reflected in the structure and content of utterances. At the risk of equating formality with limitations in forms of expression, it might also be suggested that initially there would be less variety in tenses used to express affect, but that more variety would accompany the passage of time. Soskin and John (1963) found that their dyads expressed more relational utterances when they were alone together and presumably more relaxed and informal. We might hypothesize, then, that less affect would be expressed in the initial, formal stages of a relationship and when it was expressed. it would be less intense (be ambivalent or qualified) and be directed to the subjects of the stereotyped conversation gambits. Thibaut and Kelley (1959) refer to these stereotyped gambits as the means by which people protect themselves from premature commitment to an intimate relationship. In the course of the acquaintance process, then, we would expect the formality to be reduced and intimacy to increase. This would imply the expression of more affect overall, and particularly the expression of affect about non- stereotyped, personal tOpics. A reduction in the amount of ambivalent and qualified affect and an increase in the amount of negative affect would be expected to accompany the relaxation of social prohibitions. I Warmth. Greater intimacy is intuitively related to greater warmth - more positive affect and more concern with the immediate relatinnship - so more positive affect and more concern with self and partner are implied as the relationship progresses. -It has been demonstrated that females show more warmth in the acquaint- ance process (Conly, 1968) so it would be expected that women would express more affect overall and that the affect they expressed would be less negative than that expressed by males; females could also be expected to produce less ambivalent and qualified affect. It might also be implied that greater warmth would suggest that females would express more affect toward their partners than would males. Summary It was pointed out that psychotherapy is a very diversified profession under pressure from heavy work,loads to unite under a maximally productive standard. Research has indicated that theories differ little in practice and that further research can most profitably be directed to objective descriptions of relation- ships, therapeutic and otherwise, in process. In this study, 10 attention is focused on the expression of feelings. Methods for the study of ongoing verbal behavior were discussed, including the lexical and non-lexical aspects of the expression of the affect. Research findings concerned directly or indirectly with the expression of affect were presented and from these a group of fairly specific hypotheses was derived concerning the behavior of male-female dyads. 1. On the basis of research on formality, it was hypo- thesized that the amount of talk would increase over time and that, while one person might dominate the conversation initially, the conversation load would be shared more equally over time. 2. It was hypothesized that there would be changes over time in the amount, the appropriateness, the tense, the direction and the quality of affect expressed. Specifically, it was suggested that increasing informality would be accompanied by greater use of tenses other than the present, less use of ambeva- lent and qualified affect, greater use of negative affect, and less use of stereotyped gambits. Decreasing anxiety would be accompanied by similar consequences as well as a decrease in the use of inappropriate affect. Greater warmth and intimacy would promote the expression of more positive affect. 3. It was hypothesized that there would be sex differences in the amount, the appropriateness, the tense, the direction and the quality of affect. The greater anxiety suggested for males in the social situatinn would mean they would use less affect overall but more inappropriate affect and more ambivalent and qualified affect than females. Males would also use the present 11 tense more and direct more affect toward stereotyped gambits; while females, showing more warmth, would show more positive affect and less negative affect than males and would express more affect toward their partners. II. METHODOLOGY OverView The basic conditions of the study involved scoring taped meetings between male-female dyads who had met for nine half hour sessions. The tapes of the first, second, seventh and eighth sessions were secored as to number of utterances with and without affect and the appropriateness, tense, direction and quality of affect expressed by males and females over time. Subjects The subjects were those obtained by Conley (1968). She recruited 64 unmarried students between 18 and 22 years of age from psychOIOgy courses at Michigan State University. She informed the subjects that this was a study of how people got acquainted. There was no preselection of subjects except that they were required to sign up with a stranger of the opposite sex. Eggcedure Each dyad met for nine half hour sessions over a period of five weeks. These sessions were all tape recorded and the tapes of nine dyads were analyzed in this study. RATERS: The tapes of the first, second, seventh and eighth sessions of nine dyads were rated by four trained undergraduate 12 13 raters. Raters were trained to a criterion of 90 per cent agreement with the experimenter and with each other by practicing on tapes other than those to be used in the final analysis. There was no discussion of hypotheses nor of expected results. SCORING PROCEDURE: The tapes were scored at fifteen second intervals. Every utterance including complete sentences, under- standable partials and vocal segregates was scored separately. If an utterance lasted beyond the bounds of a given interval, it was scored only for the time period in which it was begun. SCORING CATEGORIES: Silence: If no utterances were made during a fifteen second interval, a silence was tallied for that time period. No Affect: Utterances which expressed no affect were simply tallied as such. An expression of affect was judged to exist in statements of opinions, feeling or states of being, in wishes and in indirect expressions of intonation and so forth. Appropriateness: Appropriate affect was defined as that which was reciprocal in the context of the conversation or was otherwise deemed "normal" by the raters: inappropriate affect was defined in terms of exaggerated intonation and/or absence of contextual justification for the affect expressed. Tense: Tense was defined grammatically in direct reference to the tense of the verb in the utterance as past, present, future, present conditional or past conditional. Direction: The direction or object of the expressed affect was also defined by grammatical context. Affect was scored as directed to self: to an object, place or thing: to a situation 11+ or event: to a generalized other person - "they"; to a specific other person who is not present; or to the partner. Quality: Quality of affect was rated as positive, or favorable: negative or unfavorable; ambivalent, undecided or uncertain: or qualified positive or qualified negative, wherein the statement is modified by exceptions or reduced in intensity by expressions like "sort of" or ”pretty much". III. RESULTS Analyses For each variable, the scores for the male members of the pairs were correlated with those of the females to examine the extent to which they were independent. Since none indicated a significant relationship, a sex x sessions analysis of variance, for repeated measures across sessions one, two, seven and eight was performed to investigate the effect of sex, the effect of time, and possible interaction effects for each dependent measure. In addition, one way analyses of variance with repeated measures on sessions one, two, seven and eight were performed to examine the effects of time on the number of fifteen second silences and the difference scores for the amount of talking done by each member of each dyad. Tests for Differences Across Time None of the §_values obtained by the sex 1 sessions analysis of variance was significant. There was a marginally significant E for expressing affect about the partner (F = 2.34, df I 3, 48, .05(p<.10) but as Table 1 indicates, the change over time is not a consistent increase or decrease. 15 16 Sessions 1 2 7_ 8 Male 267 204 258 247 Total Utterances Female 172 169 170 153 Dyad 439 373 429 400 Utterance Expressing Male .32 .31 .25 .26 Affect Female .38 .40 .28 .29 Dyad .35 i36 .26 .29 Affect Expressed Male .79 .70 .80 .77 in Present Tense Female .75 .80 .71 .68 pyad .72 .25 .26 .23 Inappropriate Male .20 .23 .25 .28 Affect Female .15 .15 .17 .17 Dygd .18 .19_ .21 .22 Directed to Male .39 .30 .32 .31 Non-Stereotyped Female .37 .41 .40 .40 Objects Dyad .38 .36 L36 .36 Directed to Self Male .08 .05 .09 .12 Female .11 .18 .13 .12 Dyad .10 .11 .11 .12 Directed to Male .17 .14 .10 .11 Specific others Female .15 .15 .13 .15 Dyad .1§ .15 .12 .13 Directed to Partner Male .20 .11 .13 .08 Female .10 .08 .14 .12 pyad .;5 .99 .13 .10 Ambivalent and Male .42 .38 .42 .40 Qualified Affect Female .38 .38 .34 .30 Dyad .40 .38 .38 .35_ Positive Affect Male .34 .39 .39 .32 Female .31 .28 .49 .52 Dyad .32 .34 .44 .42 Qualified Negative Male .23 .25 .23 .34 And Negative Affect Female .25 .25 .24 .20 Dwad 124 .25 124, 127 Table 1. Male, Female and Dyad Means Across Sessions OTH' 17 In one way analyses of variance for the number of fifteen second silences and for the difference scores, F‘(1.0. (Table 2) Tests fogTSex_Qifferences No significant differences were demonstrated between the sexes regarding the use of the present tense to express affect:L the proportion of affect expressed: the direction of affect to non-stereotyped objects, the partner or specific others: the ex- pression of ambivalent or qualified affect, negative affect or positive affect. {*4 As hypothesized, males tended to express proportionately more inappropriate affect than females (F = 2.09, df = 1, 16, 20 (p( .10) indicating their greater anxiety in the social situ- ation. Also males tended to direct less affect toward themselves than did females (F = 2.03. df = 1, 16, .2o(p (.10). There was an unhypothesixed, but highly significant dif- ference between sexes in the total number of utterances. As demonstrated in Table 1, males dominate the conversation consist- ently (F = 9.04, df = 1, 16 p (.0001).. Previous authors have mentioned that one person will tend to dominate the conversation consistently, (Bales and Sorgatta, 1955: Borgatta and Bales, 1955; Dinoff, et.al., 1962: and Thibaut and Kelley, 1959), but none mentioned any sex differences. Tests for Integations The sex 1 sessions analysis of variance revealed no inter- action effect for the number of utterances expressed: for the preportinn of utterances expressing affect: the proportion of 18 Sessions 1 2 7 8 E Difference Scores .22 .10 .20 .24 (1.0 '“ Silences 1.44 1.11 2.22 4.77 (1.0 Table 2. Dyad Means and §_for One Way AOV 19 inappropriate affect; or the proportion of utterances expressing affect toward non-stereotyped objects, the partner or specific others. There was a marginally significant interaction for the pro- portion of utterances expressing affect toward the self (F = 2.57, df = 3, 48, p (.10) as shown in Table 3. Referring to Table 1. it would appear that males express more affect about themselves as time goes by while a fairly stable 13% of the females' affect 3% it directed toward themselves. . The interaction in the proportion of utterances expressing ambivalent and qualified affect was significant with F‘(1.0. The means given in Table 1 would indicate an interaction for the expression of positive affect, as the female preportions increase from .31 to .52 over time, while the male proportions range only from .32 to .39 and show no consistent time effects. However, the sex x sessions interaction was not significant (F = 1.44, df e 3, 48) and it must be concluded, as with many of the main effects, that there was a great deal of individual variance which rendered the results insignificant. In the sex x sessions analysis of variance for the expres- sion of negative affect, F = 2.16. This is not interaction significant, (df = 3,48, .10 (;»(.20) but the means for males and females (Table 1) would not even indicate this small difference. The means are virtually identical until session eight when the male proportion of utterances expressing negative affect show a sudden increase. As the absolute difference is less than the one seen regarding positive affect, it may be inferred that there is less variance between individuals in the eXpression of negative 20 SV 33 df ES F t~4 Sex .0606 1 .0606 2.03 Subjects .4770 16 .0298 Sessions .0085 3 .0028 (1.0 Sex x Sessions .0362 3 .0121 2.57* Error .2258 48 .0047 p<.10 Table 3. Sex x Sessions AOV: Proportion of Utterances Expressing Affect Toward the Self 21 affect than in the expression of positive. This is probably due to the ubiquity or social mores regarding the eXpression and inhibition of negative affect; since positive feelings are not so rigidly controlled, there is more individual variance in the expression of positive affect. The only indisputably significant interaction was in the use of the present tense to express affect (Table 4). Table 1 r indicates that males express more feelings in the present tense .2 over time and females express less. Pest Hoc Analyses Examination of Table 1 suggested that the means varied more from sessions 1 to 2 than from sessions 7 to 8. Difference scores were therefore computed for sessions 1 and 2 and sessions 7 and 8: these were compared using the t-test for independent measures and t = 18.81 (df = 28, p ( .0005). This suggests that the expression of affect is more variable at first, and levels off over time. A reduction in variability over time was also found by Newcomb (1961) who stated that relationships tend to stabilize into well-defined patterns as they progress. The tendency to stabilize is, the, more marked than any of the originally hypothesized changes over time. Summagyiof Results None of the hypothesized changes over time was significant, but it was demonstrated that the patterns of expressing affect stabilized over time. 22 SV SS df MS F Sex .0168 1 .0168 (1.0 Subjects .4301 16 .0269 Sessions .1081 3 .0060 (1.0 Sex x Sessions .1081 3 .0360 3.60* Error .4788 48 .0100 p<-05 Table 4. Sex Sessions AOV: Proportion of Utterances Expressing Affect in the Present Tense 23 It was shown that males express proportionately more inappropriate affect than females and that they expressed less affect about themselves than females. It was also shown that males talk more than do females. A significant sex x sessions interaction was noted on the use of the present tense to express affect: males use the present tense more and females use it less over time. There was also r? a tendency for males to express affect more about themselves over ” time while females expressed affect about themselves at a stable level. re IV. DISCUSSION The greatest differences noted were sex differences and the overall weight of these was to indicate that males are more anxious than females in the dyad. First, they use proportionately more inappropriate affect, and while anxiety was not the only form of inappropriate affect subsumed by the scoring definitions used in this study, it was virtually the only kind encountered. Also, that males express less affect about themselves suggests they are more self-conscious than females. In the light of this evidence, the male domination of conversation can be interpreted as an attempt to hide their anxiety, rather than as a by-product of the males' culturally determined dominant role: to have males talk more is actually at odds with the cultural stereotypes of a chattery female and a strong, silent male. The sex x sessions interaction regarding the present tense sees the females, as hypothesized, using the present tense less over time, but the males are using it more. It was originally hypothesised that a movement to greater intimacy in the rela- tionship would presume the discussion of less stereotyped, more personal topicgg discussion of childhood experiences or personal goals and ambitions would lead to the expression of affect in tenses other than the present. Unfortunately, our analysis does not illuminate what is happening at this point. 24 in“ 25 Assuming that the increasing use of the present tense is an index of anxiety and noting the increase in apprOpriate affect for males over time (Table 1), it can be inferred that the males are becoming more uncomfortable over time. Thibaut and Kelley (1959) found that ambivalence to the relationship increased when the relationship became more intimate and the individuals involved felt that their independence was being threatened. Cultural norms and statistics on sex differences in college drOp-out rates would suggest that entering an emotionally dependent relationship would definitely be more threatening for the college male than his coed counterpart. This is also seen in the proportions of affect being expressed about the partner: the male appears to be re- treating over time (Table 1). Hence, males would be expected to show more discomfort initially as well as over time. It might be suggested that the topics of conversation change over time, there being a shift to non-stereotyped gambits: this contributes to the males' discomfort: and males do not express as much affect re- garding the personal topics as do females, though they do discuss them without expressions of affect. These idealecould be investigated by a rescoring of the same tapes. It would be necessary to score for the tense and topic of utterances expressing no affect and some changes in the direction of affect categories would have to be made to elucidate the differences between stereotyped and non-stereotyped gambits. It would appear in retrospect that the ”situation", "object", and ”specific other" categories used here can subsume both and are therefore too general as they stand. Still, on the basis of the findings in this study, it may be that while the topics of Fir-.- ’3 37—5-.» n 26 utterances without affect change over time, the directions of affect will retain stable frequencies. After all, the pronounced absence of significant changes over time is one Of the more astonishing results of this study. For nine out of thirteen analyses of changes over time, F ratios are less than one. The primary process variable found was one of stabilization of patterns for expressing affect. The implica- tion is, as suggested by Newcomb (1961), that people enter rela- tionships with characteristic patterns and frequencies of self- expression and a new relationship does little to alter this pattern. Taken at face value, such a finding bodes ill for the therapy process in which the goal is change. However, this is not a therapeutic population, and the initial frequencies of the expression of affect overall may be very different.¥ Mahl's (1959b) proportions of anxiety expressed in therapeutic populations are higher than those found for all inappropriate affect in this study. And Seeman (1949) speaks of the decrease in expressed negative affect as a goal of therapy: presumably, a "normal" p0pu1ation such as the one in this study, would not need to change. The "normal" pOpulatinn is used to provide normative goals for the therapy process. The question arises, then, how representative is this popu- lation and how general are the findings taken from a population of student, male-female dyads? First, it must be noted that students exist in a subculture that encourages the free expres- sion of feelings: they may, therefore, express more affect overall than non-student dyads would. Also the students, as members of an homogeneous culture, are better acquainted when 27 they meet than would be members of a non-student population or, of course, a client and therapist: stabilization might take longer in another population and changes over time would again have to be hypothesized. Finally, it must be asked whether the results would be consistent in same-sex dyads. If the greater amount of male anxiety is related to feelings that their independence is being *3 threatened, it does not seem reasonable that as much anxiety b? would be present in a male-male dyad. Also, small groups of females are colloquially noted for gossip and concern with each _ other's appearance and manners: higher proportions of utterances i (with and without affect) regarding non-stereotyped topics might therefore be hypothesized for female—female dyads. Such research is necessary if norms are to be applied to the therapy process because the therapeutic relationship can be male-male, female-female, or male-female. If behavior is different in each situation for the acquaintance process, it is reasonable to assume that therapy will differ and the therapist must be aware of this. V. SUMMARY This study attempted to characterize the expression of affect in the acquaintance process and to generate hypotheses for further research. Data were obtained from the first, second, seventh and eighth session tapes of meetings between previously unacquainted college students who were recruited by Conley (1968) and randomly assigned to male-female dyads. The results from the nine dyads studied demonstrated none of the expected changes in the appropriateness, tense, direction and quality of affect over time. Responses were highly variable at first, but stabilized into regular patterns over time. Hence, it would seem that people have very definite and regular ways of expressing themselves. There were significant sex differences in that males talk more than females overall: they express more inappropriate affect than females: and they express less affect about themselves than do females. It was suggested that males are more anxious in the dyadic relation and this inhibits them. A sex x sessions inter- action in the use of the present tense suggests, too, that a shift to greater intimacy causes the male proportion of inappro- priate affect to increase further, indicating both greater anxiety in the more intimate situation and greater ambivalence about the relationship. The data suggest a variety of further studies with student, 28 ‘_\ In: 29 non-student, and therapeutic populations and with same-sex dyads to test the consistency of the findings. Refinements in the scoring system and an additional scoring were suggested to examine more closely the overall content of conversation with and without effect. fin.” ‘ l J APPENDIX A HYPOTHESES TESTED 1. It: would 11 2: It over t1 3- It Proport dRid wc HYPOTHESES 1. It was hypothesized that the total number of utterances would increase over time. F"? 2. It was hypothesized that periods of silence would decrease E over time. 3. It was hypothesized that the difference scores between the proportions of total utterances emitted by each member of the dyad would decrease over time. 4. It was hypothesized that the proportion of utterances ex- pressing affect would increase over time. 5. It was hypothesized that females would produce proportionately more utterances expressing affect than would males. 6. It was hypothesized that the pr0portion of utterances ex- pressing inappropriate affect would decrease over time. 7. It was hypothesized that males would produme proportionately more expressions of inappropriate affect than would females. 30 31 8. It was hypothesized that the proportion of utterances ex- pressing affect in the present tense would decrease over time. 9. It was hypothesized that males would produce proportionately more utterances expressing affect in the present tense than would females. 10. It was hypothesized that the proportion of utterances ex- pressing affect about non-stereotyped topics would increase over time. 11. It was hypothesized that females would produce proportion- ately more utterances expressing affect on non-stereotyped tOpics than would males. 12. It was hypothesized that the proportion of utterances ex- pressing affect about thhmselves would increase over time. 13. It was hypothesized that the proportion of utterances ex- pressing affect about themselves would be higher for females than for males. 14. It was hypothesized that the proportion of utterances ex- pressing affect about specific other persons would increase over time. 15. It was hypothesized that females would produne proportion- 32 ately more expressions of affect about specific other persons than would males. 16. It was hypothesized that the proportion of utterances ex- pressing affect about the partner would increase over time. 17. It was hypothesized that females would produce proportion- F"? ately more utterances expressing affect about their partners than I would males. 18. It was hypothesized that the proportion of ambivalent and E? qualified affect expressed would decrease over time. 19. It was hypothesized that males would express prOportion- ately more ambivalent and qualified affect than would females. 20. It was hypothesized that the proportion of utterances ex- pressing positive affect would increase over time. 21. It was hypothesized that females would produce proportion- ately more positive affect than would males. 22. It was hypothesized that the proportion of utterances ex- pressing negative and qualified negative affect would increase over time. 23. It wasyhypothesized that miles would produce proportionately more negative and qualified negative affect than would females. APPENDIX B SCORING SHEET o>Hpmmoa oodmaamsq o>apamom ooawaamza psoam>dns¢ obdumwoz obapdmom mompo ooNfiHmHoGoo poonno sodpmSpam Hospamm henna camaooam mamm .usoo pmmm .esoo anemonm magma oaspsm omsoe pmmm emcee anemonm upmaaaoaaamsH opmaaaonaa< poohhm oz moozoaam % .. .ll .- OI \: APPENDIX C AOV TABLES l SV SS df MS F Sex 124,168 1 124,168 9.04* Subjects 219,829 16 13.739 Sessions 15,164 3 5,054 1.07 Sex X Sessions 15,378 3 5,126 1.09 Error 226,367 48 4,716 * p (.001 Sex x Sessions AOV: Total Number of Utterances Expressed SV SS df MS F Sex .0203 1 .0203 (1.0 Subjects .5824 16 .0364 Sessions .0617 3 .0206 1.02 Sex x Sessions .0581 3 .0194 (1.0 Error .9679 48 .0202 Sex x Sessions AOV: Pr0portion of Utterances Expressing Affect SV SS df MS f Sex .1058 1 .1058 2.09* Subjects .8119 16 .0507 Sessions .0237 3 .0079 < 1.0 Sex x Sessions .0075 3 .0025 (1.0 Error .4033 48 .0084 * .10 (p (.20 Sex x Sessions AOV: Proportion of Utterances Expressing; InapprOpriate Affect SV SS df NS F Sex .0861 1 .0861 (1.0 Subjects 1.0844 16 .1128 Sessions .0072 3 .0024 (1.0 Sex x Sessions .0451 3 .0150 (1.0 Error 1.1268 48 .0235 Sex x Sessions AOV: Proportion of Utterances Directing Affect Toward Non-stereotyped Topics SV SS df MS F Sex .0076 1 .0076 1.0 3? Subjects .4197 16 .0262 Sessions .0623 3 .0208 2.34* Sex x Sessions .0291 3 .0097 1.09 Error .4285 48 .0089 * .05(p(.10 Sex x Sessions AOV: Proportion of Utterances Directing Affect Toward the Partner SV 88 df ms F Sex .0043 1 .0043 (1.0 Subjects .3581 16 .0224 Sessions .0208 3 .0069 (1.0 Sex x Sessinns .0098 3 .0033 (1.0 Error .5657 48 .0118 Sex x Sessions AOV: Proportinn of Utterances Directing Affect Toward Specific Others 35 SV SS df MS F Sex .0618 1 .0618 1.02 Subjects .9674 16 .0605 Sessions .0208 3 .0069 (1.0 Sex x Sessions .0164 3 .0055 (1.0 Error .4405 48 .0092 Sex x Sessions AOV: Proportion of Utterances Expressing Ambivalent and Qualified Affect SV SS df IS F Sex .0696 1 .0696 1.03 Subjects 1.0829 16 .0676 Sessions .0208 3 .0069 (1.0 Sex x Sessions .0554 3 .0184 1.44 Error .6183 48 .0128 Sex x Sessions AOV: PSOportion of Utterances Expressing Positive Affect SV SS df MS F Sex .0153 1 .0153 (1.0 Subjects .9010 16 .0563 Sessions .0104 3 .0034 (1.0 Sex x Sessions .0991 3 .0263 2.15% Error .5862 48 .0122 *.1o(p(.m> Sex x Sessions AOV: Proportion of Utterances Expressing Negative or Qualified negative Affect 36 BIBLI OGRAPHY BIBLIOGRAPHY Auld, F. and Murray, E. "Content Analyses Studies of Psycho- therapy," Psychol. Bull., 62, 377, 1955. Bales, H. and Eorgatta, E. "Size of Group as a Factor in the Interaction Profile," Small Groups. (Hare, Borgatta and Sales: Eds.) Alfred Knopf: New York, 1955. Parker, R. "The Stream of Behavior as an EXperimental Problem," ghe Stream of Behavior. (Barker: Ed.) Appleton-Century- Crafts: New York, 1963. berne, E. Transactional Analysis in Psychotherapy. Grove Press, Inc.: New York. 1961. Eorgatta, E. and Sales, R. "The Consistency of Subject Behavior and the Reliability of Scoring in Interaction Process Analysis," Small Groups. (Hare, Borgatta and Bales:Eds.) Alfred Knopf: New York, 1955. Ereuer. J and Freud, S. "Studies on Hysteria," The Standard Edition of the Complete Psychological Works of Segmund Freud! Vol. I. (Strachey: Ed.) The Hogarth Press: London, 1955. Conley, S. Stratigies of Maneuvers Duringgthe Acquaintance Process. Unpublished Doctoral Dissertation, 1968. Dinoff, M.. Row, J.. Rickard. H.. and Timmons. E. "The Stability of Group Verbal Behavior," Psychol. Bull., 12, 323, 1962. Dittman, A.. and Wynne. ”Linguistic Techniques and the Analysis of Emotionality in Interviews," J. Abn. Soc. Psych., 63, 201. 1961. H Eysench. H. "The effect of Psychotherapy: An Evaluation,” 1 J. Congnit. Psychol.. 16, 319, 1952. ‘ Fiedler, F. "The Concept of an Ideal Therapeutic Relationship," 1? J. Consult. Psycholp, i4. 239, 1950(3), pi . "A Comparison of Therapeutic Relationships in Psychoanalytic, Non-Directive and Adlerian Therapy,“ ngonsult. Psychol.. 14, 436, 1950 (b). Gottschall. L. and Hambridge, G. "Verbal Behavior Analysis: A Systematic Approach to the Problem of Quantifying Psychological Processes," J. Proj. Tech., 19, 387, 1955. Haley, J. “Strategies of Psychotherapy. Grune and Stratton: New York. 1963. Jennings, H. "Individual Differences in the Social Atom," Small Groups. LHare, Borgatta and Bales: Eds.) Alfred Knopf: New York. 1955. Klein, M.._Dittman, A., Parloff. M. and Gill. M. "Behavior Therapy: Observatins and Reflections," J. Consult. Clin. Psych., 33. 259. 1969. Knapp. P. (Ed.). Expression of the Emotions in Man. International Universities Press, Inc.: New York, 1960. Mahl, G. ”Disturbances and Silences in the Patient's speech in Psychotherapy," J. Abn. Son. Psych., 53, 1, 1956. . “Measuring the Patient's Anxiety During Interviews from Expressive Aspects of his Speech.“ Transactions 9; the kew York Academy of Sciences, 21, 249, 1959(a). . ”Exploring Emotional States by Content Analysis," ends in Content Analysis. University of Illinois L Press: Urbana, 1959(b). . "Some Observations About Research on Vocal Behavior," Disorders of Communication., 42, 466, 1964. Marsden, G. "Content Analysis Studies of Therapeutic Interviews: 1954-64,” Psychotherapy Research. (Stollack, Guerney & Bothberg: Eds.) Rand McNally: Chicago, 1966. McQuown, N., "Linguistic Transcription and the Specification of Psychotherapy Interview Materials," Psychiatry, 20, 79, 1957- Newcomb, T. The Acquaintance Process. Holt, Binchart and Winston: New York, 1961. Pittenger, 3., Hockett, C., and Danehy, J. The First Five Minutes: A Sample of Microscopic Interview Analysis. Paul Martineau: Ithaca, N. Y., 1960. Knapp. P. (Ed.). Expression of the Emotions in Man. International Universities Press, Inc.: New York, 1960. Mahl, G. "Disturbances and Silences in the Patient's speech in Psychotherapy," J. Abn. Sot. Psych., 53, 1, 1956. . "Measuring the Patient's Anxiety During Interviews from Expressive Aspects of his Speech," Transactions 9; the New York Academy of Sciences, 21, 249, 1959(a). . "Exploring Emotional States by Content Analysis," Igendg in Content Analysis. University of Illinois Press: Urbana, 1959(b). . "Some Observations About Research on Vocal Behavior," Disorders of Communicatipn., 42, 466, 1964. Marsden, G. "Content Analysis Studies of Therapeutic Interviews: 1954-64,” Psychotherapy Research. (Stollack, Guerney & Rothberg: Eds.) Rand McNally: Chicago, 1966. McQuown, N., “Linguistic Transcription and the Specification of Psychotherapy Interview Materials," Psychiatry, 20, 79, 1957- Newcomb, T. The Acquaintance Process. Holt, Rinchart and Winston: New York, 1961. Pittenger, R., Hockett, C., and Danehy, J. The First Five Minutes: A Sample of Microscopic Interview Analysis. Paul Martineau: Ithaca, I. Y., 1960. Rogers, C. "The Necessary and Sufficient Conditions of Thera- peutic Personality Change," J. Consult. Psychol. 21, 905. 1957. Sarbin, T. "Role Theory," Handbook of Social PsychOIOgy. (Lindzey: Ed.) Addison-Wesly: Cambridge, 1954. ex Schofield, w. Psychotherapy: The Purchase of Friendship. a Prentifle-Hall: New Jersey, 1964. Seeman, J. "The Process of Non-Directive Therapy,” J. Consult. F Psychol., 13, 169, 1949. Soskin, W. and John, V. "The Study of Spontaneous Talk," The Stream of Behavior. (Barker: Ed.) Appleton-Century- Crafts: New York, 1963. Szasz, T., The Myth of Mental Illness. Hoeber-Haper: New York, 1961. Thibaut, J. and Kelley, H. The Social Psychology of Groups. John Wiley and Sons: New York, 1959. Thomas, D. Observational Studies of Social Behavior. Yale University Press: New Haven, 1933. Winch, R. The Modern Family. Holt, Rinehart and Winston: New York, 1952. A061 51989 "I7'1)I)IIIINELI'I'II)I“