y “0' ' '..I.»o—--Q_— .5 .0‘00”...0"002..”C‘-.-9.o-)“.f...m-‘~“MOO-H.Q-wufl”C . ....\ o. ‘9 -o'o..p,‘ . - ~- -‘u.--y‘- "0..n.-.....'.§m~""”.y‘~'.‘n"wquw—-V"YV‘1v."y_“-m.w“ . . REIAflVE STATUS, ANTICIPATED INTERACHON AND r ' SOCIAL FACILITATION AS omnmmmrs or HUMOROUS RESPONSES TO EMBARRASSMENT . n .' I 9 I . o . . . . , .. - . WW ANN Wm - . - O ‘ o -. | . . . ’ '- I ' O ‘ . - ‘ . . '. H t; _' F : . > I'w I .:' " . D. p. - - - > -.', - . 'Q. - n L -S. . - . , I 't' I . ’ ' r "- | O:¢ < - ' -.1 .- I v . -- v w’ ' 7 u 3 ‘ — - - 2'" - ‘ - . ) ..- a_ _ _ . - .- ' ’ ' '. r .9.'- _ . .'.'~" . ' -‘ ’_. r 3. , - . _ . .-_ _ . D .~. — ‘ - - — _— .‘ 2. ' . - . a F );'1 A ‘ '. - - - O '-- - o:- 1 ' ’ _ f ; ’ - ' - .. - ‘ ‘ - d— " . r- .> ‘ . - ': ‘ .'- .14": . -_ S - . ‘ O — ' . ‘ ~ ‘ -~_ . . - ‘ - - _ -- - ‘ -_ " _ ' — V: f 3.- - . ‘ . . -.-. ...'. -_ ;. , .2 .._ -21: ' - .- , ;_.- ‘_ - .-_.. v ;-' -- ctr}..- - ‘ . . - . ' . ' ‘ -‘ ‘_ « : ‘_ . - - ‘L' ‘3’- _ - . -_ 9 ~_ - - . J. -., , . '. A. - _ - - - ~ -_ _ ~ -. ¢*.f; ‘ ' ', ' : ° . . _ v - ~ - . ~ —-‘ .. -‘ : l - _--' ,‘ - ‘ '_ ' -:-. A“... ' ‘o o . - '- : 0 - ' ’ - . . - . 2.. o a I - ‘- - _: r' . _ ‘ _ . _ 5" f - - ' " - ' ' . - . - - - -- -':-7- ~ ‘ ' , l v ‘ n . . ‘ . ' ‘ fi - ' -. - - _. ... 1"; 5 - - - * ‘, ’ ‘ - f' . . - — : C ’. ' 4 0 I. ‘ . ' . r .2- . . - ._ , . - < . . _ ‘ ‘ ‘ '.3 ‘ ‘ ' ’ . . ' ‘ . . - .'.’ - r' . o H ’ ’ ' A '- . ' u - . - - . . _- - l -4 —.‘ _- - - \ é}: : ' . ’ - . - , a - — ' I ‘~ ’ — --.- o o a a. .o‘ l L ' ‘ ' ' ' .. - .? -Ko‘ » o a. - - . - rg' ' c . - - <9 , ~ - ' o ' ‘ .‘ o ~ 0... a p. ' _ - o , : i ' ..: ' - - , . o - a. A - , _ , . . - ’ _ I 4 - '\ l . .m. ‘Q 3:. \ a Q | z ' ~ , a - ‘Jtéh‘ U‘ l ' ‘ ~ . “90.31.“ D :J’ Q .3 . t | . . l . . . . I . . A: ~ v - - '1'. ' figtf‘.‘ofl’ 16‘3‘ D c ~ -ru 1 . - o \ 1 \ \O o i u on A h 3 ' U. ' . U M . . l ‘1 5 u 6 .. o .- § .. 1 fl . ‘ v Q I. . f... I t W“ .' JL, \ M 855 W 31293 00 \ W \ NW “HE-1873 MSU LIBRARIES —_ RETURNING MATERIALS: Place in book drop to remove this checkout from your record. FINES WIII be charged if book is returned after the date stamped beIow. .. I" v f .. » ABSTRACT RELATIVE STATUS, ANTICIPATED INTERACTION, AND SOCIAL FACILITATION AS DETERMINANTS OF HUMOROUS RESPONSES TO EMBARRASSMENT By Barbara Ann Walker Humor is a face-saving teChnique that is frequently used to restructure an interpersonal communication situation when embarrassment has broken down communication patterns. In an embarrassing interaction, it is expected that more humor will be used between persons of relae tively equal status than.when people are of unequal status. It is expected that more humor will be initiated by a person of relatively higher status in an interaction than by a person of lower status. It is hypothesized.that'Hore'humor*will be used.when interactants antici- pate future interaction than when no future interaction is expected, and that more persons observing an interaction act to facilitate humor— our responses to embarrassment. Sixty subjects engaged in an embarrassing interaction with an experimenter by telephone. Results indicate that there is no significant effect for anticipated interaction. MOre laughter between persons of relatively equal status than between persons of unequal status is found. The number of people present has a positive, significant effect on laughter. Results are discussed in terms of the tentative nature of Barbara Ann Walker the communication process during embarrassment, and in terms of Goffman's (1959) concept of derisive collusion. Additional researCh issues concerning the study of humor and embarrassment are raised. Z} l/37ii/é/ RELATIVE STATUS, ANTICIPATED INTERACTION, AND SOCIAL FACILITATION AS DETERMINANTS OF HUMOROUS RESPONSES TO EMBARRASSMENT By Barbara Ann walker A.THESIS Submitted to Midhigan State University in partial fulfillment of the requirements for the degree of MASTER OF ARTS Department of Communication 1975 Accepted by the faculty of the Department of Communication, College of Comrmmication Arts, Michigan State University, in partial fulfillment of the requirements for the Master of Arts degree. - /’ // ‘ /} .-" - 1’ £75] / ,' / J '/ ( /., 0' Ar)”; 4 :5; // _ ‘ f”( /,I"£; :7 . f I Director of Thesn.s A "ii/“'7’, .—’-' l/III // L v " o Guidance Committee: “ " r;- “"5" ,1 " - 4/; [(Chairman ii ACKNOWLEDGMENTS I would like to extend my gratitude to Dr. Edward L. Pink for his criticism.and advise on this thesis, and for his continued support and encouragement as my adviser. I am indebted to him for the knowl- edge and understanding he has given me. I would like especially to thank NOrm.LoPatin and Lon Graves for their conscientious and enthusiastic effort as coders. Without their dependability, this thesis would not have been possible. I would also like to thank Carol Hall and Cheryl Vignali who acted as callers. I would like to thank Dr. Joseph WOelfel for his advice and encouragement. Finally, special thanks goes to my father, for expressing "I'll believe it when I see it." iii Chapter TABLE OF CONTENTS Introduction . . . . An Overview of Theoretic Approaches to Studying Embarrassment . . . An Overview of Theoretic Approadhes to Studying Humor . . Causes of Embarrassment . . . . Consequences of Embarrassment . Humor as a Face—Saving TeChnique . . Conditions Affecting the Use of Humor . II EXPERIMENTAL DESIGN AND METHODS Overview of Experimental Design . . . Stimulus Questions Experimental Manipulations and Independent Variables . . . . . Subjects Dependent Variables and ManipuIation Checks Coder Training . . . . . Interviewer Training . . . . Experinental Procedure . . . . III RESULTS, CONCLUSIONS, AND DISCUSSION Manipulation Checks . . . Reliability and Validity of Measures . . Embarrassment as a Dependent Variable . Humor and Laughter as Dependent Variables . Evidence fOr Hypotheses . . . . . Implications . . . . . . . APPENDICES . . . . . . . A.- EXPERIMENTAL PROTOCOL . . . . B - EXPERIMENTAL SCHEDULE . . . . . . REFERENCES . iv Page 23 26 27 32 35 37 HO HO H0 H3 #5 H8 51 en su 61 63 Table LIST OF TABLES Intercorrelations Among and Descriptive Statistics for Variables Cell Means for Embarrassment (Evaluated by Coders). Canonical Correlations of Independent Variables with Embarrassment Cell Means for Laughter (Evaluated by Coders) Canonical Correlations of Independent Variables with Laughter . Canonical Correlations of Status (Quadratic) and Facilitators with Laughter Cell Means for Humor (Evaluated by Coders) Canonical Correlations of Independent Variables with Humor Page H1 H3 nu H5 H6 u7 H8 H9 CHAPTER I 1.1 Introduction Embarrassment is a threat to any interaction we encounter. The experience of embarrassment quickly spreads from one person to all participants in an interaction. The fumbling and uncomfortable silences resulting from embarrassment are carefully and systematically avoided in most conversations. Since embarrassment is undesirable for all interactants, an embarrassed individual will frequently try to save face - to create an impression that he or she has not lost composure. Face—saving may take many forms, including walking away from the situation, apologizing, becoming aggressive, or using humor and unseriousness. The focus of this research will be on humor and unseriousness as face—saving techniques in an embarrassing interaction. A number of authors discuss humor and unseriousness as responses to an embarrassing interaction. Goffman (1967) describes joshing as a technique used to release tension caused by embarrassment, by creating the impression that nothing serious is happening. Emerson (1970) describes how humor functions in a gynecological exam to sustain the impression that "no one is embarrassed." The literature indicates that humor is a face-saving technique used as a response to an embarrassing interaction. 2 In this chapter, the literature involving huHor'as a response to an embarrassing interaction will be discussed. Both the causes and con— sequences of embarrassment will be discussed. There will be a focus on face—saving as a consequence of embarrassment, and on hunor as a face- saving technique. The theoretic discussion will be organized as follows. First, an overview of the theoretic approaches to studying embarrassment and humor will be presented. Next, the causes and consequences of embarrassment will be discussed. Then, the literature presenting humor as a conse- quence of embarrassment will receive focused attention. Finally, the variables of anticipated fUture interaction and status will be described as variables affecting the use of humor. As discussed, the study of embarrassment has been approached from several theoretic perspectives. The first section of this discussion will present an overview of the theoretic approaches to studying enbarrassment. 1.2 An Overview of Theoretic Approaches to Studying_Embarrassment The occurrence of embarrassment in social interaction has been studied from several perspectives. Empirical research has been conducted from.the point of view of role theory, empathy theory, balance theory and exchange theory. In this section, an overview of these different theo- retic peISpectives will be presented. 1.21 Role theory - Role theory looks at roles as sets of expectations of behavior shared among persons in a social interaction. Embarrassment occurs when an individual becomes aware that he or she is not behaving according to the expectations of others in an interaction. 3 Role theory explanations of embarrassment started with the work of Erving Goffmen (1967) in "Embarrassment and Social Organization." Goffman describes the process by which individuals become embarrassed and how embarrassment Spreads in an interaction. Gross and Stone (196H) expand Goffman's explanation by collecting recollections of embarrassing incidents from over 1000 persons. weinberg (1968), relying on.Goffman's conceptual explication, studied the occurrence of embarrassment in nudist colonies. 1.22 Empathy theory — Empathy theory is concerned with the differential ability of individuals to experience embarrassment. This theory con- siders an individual's response to the social situation as the unit of analysis. The more an individual can empathize with an interactant who is experiencing embarrassment, the more he or she will experience embarrassment. Several authors focus on empathy as a necessary condition for embarrassment, starting with Hellpach in 1913. Sattler, in an empirical study (1965) and a literature review (1965), concludes that the more a person can empathize, the more he or She will experience embarrassment. Medigliani (1966) finds a positive relationship between the ability to empathize and embarrassibility. 1.23 Balance theory -.A balance theory conceptualization of embarrass— ment occurs when there is incongruity between a person's expectations of a social situation and events that take place in the situation. Another person is necessary for someone to be conscious of the incongruity between his or her subjective expectations and social reality. The unit of analysis is an individual's subjective experience of embarrassment. u Heider (1958) first used balance theory to explain embarrass- ment. He describes embarrassment as "the disintegrating effect of self— consciousness produced by the seeming exposure of oneself to the per- ception of another" (1958:7H). Sattler (1965) finds that peOple report that they would experience more embarrassment in an imbalanced state than in a balanced state. 1.2H Exchange theory — Exchange theory conceptualizes embarrassment as a cost in social exchange. Although an individual may be receiving tangible rewards by remaining in an interaction, the cost of embarrass- ment may cause him or her to withdraw from the situation. Studies using this perspective measure the amount of facework an individual engages in to avoid embarrassment. Brown and Garland (1971) pay subjects to sing an embarrassing song, and measure facework as the amount of money lost by reluctance to sing this song to an audience. Garland and Brown (1972) conduct several studies measuring facework as a function of audience competency. Embarrassment is an important concept in social interaction, receiving attention from.severa1 perspectives. Humor, another variable of interest in this study, has also been studied from different theo- retic perspectives. Next, a brief overview of the humor literature will be presented. 1.3 An Overview of Theoretic Approaches to Studying Humor Humor has been studied from a number of theoretic perspectives. Several authors have developed schemes for looking at the general social functions of humor. Other authors analyze how humor functions to socialize peOple for their roles. A few experimental studies of humor 5 have been conducted. Following is an overview of the humor literature. 1.31 Social functions of humor - Stephenson (1951) provides a usefu1 analysis of the functions humor serves in social interaction, dividing humor into conflict and control functions. A.major article about the conflict function of humor is Orbdick's (19H2) article about gallows humor.l .A model for the social function of humor is provided by Martineau, in Goldstein and Md3hee (eds.) (1972). He looks at the social function of humor on three levels: (1) Totally within a group; (2) In an intergroup situation, fecusing on the internal structure of the group; (3) In an intergroup situation, focusing on the interaction between groups. 1.32 Socialization fUnctions of humor*- Olsen and Whittaker (1966) stress the importance of joking in socializing student nurses for their roles. Ruth Coser's (1959) analysis of humor provides one of the most compre- hensive treatments of the social functions of humor. Examining the social structure in a.hospita1 ward, she emphasizes that humor fUnctions to socialize people into different social groups. In a second study, Coser (1960) studies the social functions of humor among the staff in a mental hospital. 1.33 Experimental studies of humor - Davis and Farina (1970) conducted an experimental study of how humor functions as communication, concluding that peOple may choose to communicate by using humor because a direct expression would be socially awkward. Lefcourt et_§1, (197H) measure humor as a response to a word association test, using locus of control as an independent variable. Next, a fecused presentation of the variable of embarrassment will be presented. This discussion will be separated into the causes and 6 consequences of embarrassment. First, a consideration of the different theoretic approaches to the causes of embarrassment will be presented. 1.H Causes of Embarrassment Several authors have looked at the conditions under which embar— rassment occurs. One way of looking at the cause of embarrassment is analyzing how people fail to maintain appropriate roles. we can also fecus on the process of alienation from interaction. Another*ana1ysis of the cause of embarrassment looks at deficiencies in a person's self image. 1.Hl Failure to maintain appropriate roles — Goffman (1967) discusses the role of interactant in a conversation. EaCh person is required to maintain proper involvement and is also required to insure that others maintain their involvement. A.joint involvement emerges, which Goffman describes as "a precariously steady state that is likely at any time to lead the individual into some form of alienation" (1967:117). When the proper role of interactant is not maintained, embarrassment occurs, up— setting the equilibrium.in the interaction. Goffman explains that one of the ways in which embarrassment arises is in situations of role conflict "when the self projected is somehow confronted with another self which, though valid in other con- texts, cannot be here sustained in harmony with the first" (1967:108). .Audience segregation normally keeps people from.exoeriencing embarrass- ment because people play out different roles for different audiences. When audience segregation breaks down, people are called on to plav out conflicting roles. As a means of expressing that a conflict exists, an individual becomes embarrassed. Reflecting a role theory analysis, Gross and Stone (196H) define "embarrassment as occurring whenever some central assumption in a trans— action has been unexpectedly and unqualifiedly discredited fer at least one participant" (emphases in original, 196H:10). .According to their survey of embarrassing incidents, Gross and Stone (196H) hold that one of three ways in which embarrassment occurs is when assumptions people make about each other are disturbed. Weinberg (1968) extends the role theory analysis of Goffman and Gross and Stone. He defines embarrassment as "an experience where there is a leap out of the routine world, an absence of social guidelines and a subsequent discomposure and self consciousness" (1968:382). Embarrass— ment occurs when one's role is called into question, and there is a tran- sition to experiencing one's self, stripped of role. This occurrence of experiencing one‘s self_is the form.of alienation Goffman calls self- consciousness. Authors who have approached embarrassment from.a role theory per— spective describe embarrassment as occurring when an interactant fails to maintain an appropriate role, or when role conflict occurs. Another way of looking at embarrassment focuses on the process of alienation from,interaction. 1.H2 .Alienation from interaction — In "Embarrassment and Social Organ— ization," Goffman discusses the process by whiCh alienated persons become embarrassed. Essentially, the alienated person becomes embarrassed due to the impressions he or she makes on others. .As a requirement for interaction each individual must present himself or herself in a way that 8 is apprOpriate for the occasion and the audience. In order for the interaction to proceed, others must accept the identities presented by each person in the interaction. When identities that persons present are discredited by events that take place in the conversation, embar- rassment occurs. Interaction consciousness and external preoccupation are two other standard ferms of alienative misinvolvement discussed by Goffman. Individuals experience interaction consciousness when they are concerned, more than is normative, with the way an interaction, as an interaction, is proceeding. The painful silences of many conversations illustrate a form of interaction consciousness that usually leads to embarrassment. Self—consciousness is a common form of alienation leading to embarrassment. A self-conscious individual is appraising hOW'Well he or she is performing in a conversation. When an individual believes that he or she is not performing adequately, embarrassment may occur. Individuals in an interaction become alienated when they break norms of involvement. Another way of looking at the cause of embarrass- ment focuses on the individual's self-image, instead of the normative structure. 1.H3 Deficiency in self-image - Modigliani (1966) postulates an inter— vening process of attribution between the deviant act and an experience of embarrassment. During embarrassment, an individual attributes certain negative attributes to the self, which affects the person's self concept. Few enough attribute dimensions are relevant on a specific occasion so that being lower on one makes a distinct difference in self— 9 image. .A person's concept of his or her attributes, and therefore his or her selfeimage, is influenced largely by the most recent information he or she receives. When this recent infOrmation is negative, embarrassment is likely to occur. Image_is important in understanding.Modigliani's concept of embar- rassment. He defines image as the attributes that are salient in a given social encounter. He describes the experience of embarrassment as one of "heightened self—consciousness of a deficient image that is distinctly restricted in content" (1966:28). Analyzing changes in an individual's self—image is one way of con- ceptualizing the cause of embarrassment. As discussed, we can also look at the failure to maintain apprOpriate roles or at the process of alien— ation from involvement to analyze the cause of embarrassment. In the next section of this discussion, the consequences of embarrassment will be considered. 1.5 Consequences of Embarrassment There are several ways of looking at the consequences of embar- rassment. Some authors have looked at the way in which embarrassment incapacitates interactants. Another consequence of embarrassment is an attempt to save face. 1.51 Incapacitates interactants — Gross and Stone (196H) emphasize the consequences of embarrassment for social relationships. They hold that the important social feature of embarrassment is that it incapacitates peOple for role performance. They suggest that every enduring social relationship provides ways of preventing embarrassment. 10 Goffman (1967) describes what happens to an individual who becomes embarrassed, or loses face, in an interaction. "His manner and bearing may falter, collapse, and crumble" (1967z8). The objective signs of embarrassment, including fumbling, stuttering, blushing, quavering speech, and absent-mindedness make it difficult fOr a person to fully participate in a conversation. Embarrassment has consequences for all interactants in a conver— sation. Since an embarrassed individual is not able to fully perfOrm the role of a poised interactant, others in the conversation are unsure of how to respond to the embarrassed person. NOt knowing what to do or how to respond to the embarrassed person, other interactants may experience the flustering and fumbling of embarrassment. If embarrassment has been specificly caused by one individual, he or she may experience embarrass— ment for not maintaining the role of a tactful person. As a consequence of embarrassment, we have looked at how people are incapacitated from participation in an interaction. Another conse- quence of embarrassment is that an individual attempts to save face. Next, face—saving as a consequence of embarrassment will be discussed. 1.52 Face-saving - An individual may attempt to use face-saving tech- niques in response to embarrassment. Goffman (1967) uses the phrase to save face to refer to the process by which a person creates an impression that he or she has not lost face. Face-saving techniques that are employed in any interaction are normative. One face-saving technique used in reSponse to embarrassment is establishing role distance. Goffman.(1967) uses role distance to mean an expression of separateness between an individual and his or her role. 11 By becoming separate from a role, in the interaction, a person shields himself or herself from the disapproval of others. Archibald and Cohen (1971) find that the more role distance subjects establish, the less they are embarrassed by disapproval. Their results indicate that establishing role distance is an effective face-saving technique in preventing embar- rassment. Goffman (1961) discusses unseriousness and humor as strategies used to establish role distance. By behaving in an unserious way, an indi- vidual can project the image that events that are occurring should not be taken as a reflection on him or her. By joking, "the individual makes a plea for disqualifying some of the expressive features of the situation as sources of definitions of himself" (1961:105). Several authors describe how different forms of humor are used in reSponse to embarrassment. By using humor, an individual saves face in a situation where his or her identity as an interactant is threatened. In the next section of this discussion, humor will be analyzed as a face- saving technique. 1.6 Humor as a Face—Saving Technique Several authors describe how different forms of humor are used in response to embarrassment. Joshing, joking, irony, and banter are forms of humor that are used as face—saving techniques. Due to the social structure, individuals repeatedly find themselves in situations of role conflict. By using humor, an individual saves face in these situations. Goffmar1(1967) states that joshing is "a means of releasing the tension either caused by embarrassment or whatever caused embarrassment" l2 (1967:112 footnote). Goffman discusses joking in a situation of role conflict. Due to the way any organization is structured, an individual is frequently called on to play conflicting roles. He suggests that both joshing and embarrassment reduce role conflict by denying reality to the situation. Burns (1953) describes irony and banter as two styles of inter- action used to avoid embarrassment. Embarrassment, he says, occurs most typically in a situation in which two statuses are presented simultan— eously. As an alternative to embarrassment, banter is a "style of inter— action used when two roles are presented to an individual and he decides to retain the status appropriate to both, while, as he must, acting out the role of only one" (1953:655). Banter is used to play at being hostile, distant, and unfriendly while suggesting friendliness. Irony is used when one is playing at being friendly, while intimating rejection or unfriendliness. Thus, instead of becoming embarrassed or allowing a conflict of status to be expressed, an individual uses irony or banter. Emerson's (1970) explanation of the social reality of gyneco- logical exams directly relates humor to embarrassment. The major defini— tion to be sustained in such a situation is that "no one is embarrassed" because the patient is involved in a medical situation, not a sexual situation. Humor provides a way in which concern about the gynecological exam.and the sexual connotations can be expressed indirectly. If expressed directly, the elements joked about would be embarrassing because they are contrary to the definition of the situation. The use of humor allows an individual to save face in an embar— rassing interaction. In reSponse to an embarrassing, or a potentially 13 embarrassing, situation humor releases tension and reduces conflict by defining the situation as unserious or not real. Humor is used to different degrees depending on the interaction in which embarrassment occurs. In the next section of this discussion, conditions effecting the use of humor in an embarrassing interaction will be examined. 1.7 Conditions Affecting the Use of Humor 1.71 Statu§_- One important variable effecting the use of humor in an interaction is the relative status of the participants. Studies of joking relationships in primitive societies indicate that humor functions to maintain social status. Later research suggests that the relative status of interactants determines whether or not a person initiates humor in a conversation. Radcliffe-Brown (l9H0) discusses how joking relationships in primitive societies function to maintain the social order. He dis— tinguishes two main varieties of joking, symmetrical and asymmetrical. In symmetrical relationships, each party teases or makes fun of the other. In the asymmetrical relationship, person A.jokes with person B, but person B does not joke back or jokes only a little. Bradney (1957) maintains that the use of symmetrical or asymr metrical joking bears a definite relationship to status. Studying the joking relationship among members of the staff of a large department store, she fOund, in general, that joking was established more frequently between members of the same status. Between persons of different status, joking may either be symmetrical or asymmetrical. When asymmetrical joking does take place, it is usually initiated by superiors and aimed at subordinates. 1H Coser (1960) studies the distribution of the use of humor by status. If the conflict theories of humor are correct, she argues that those who are low in the status hierarchy would direct fewer witticisms at higher status persons than higher status persons would direct at lower status persons. Higher status persons, because of their legitimate authority, would feel free of the need to agress against persons of lower status. She finds that lower status members used humor le§§_frequently than do persons higher in authority. Coser proposes that "the status structure is supported by down— ward humor" (1960:86). Persons high in authority have more right to use aggressive humor whereas those low in status are not permitted an agres— sive outlet, even in the form of humor. Coser suggests that aggressive feelings experienced by lower status members may be expressed through humor in informal social gatherings among equals, but are not expressed directly to their superiors. Based on the discussions of Bradney and Coser we would expect humor to be initiated more frequently between same status members, or directed at lower status persons by higher status persons. This leads to the following hypothesis: In an embarrassing interaction, there will be less humor generated from a lower status person to a higher status person than when two persons are of the same status or from a higher status person to a lower status person. Among the variables affecting the use of humor in an embarrassing interaction, the greatest theoretic support is found fer hypotheses dealing with status. Another variable of interest, although receiving 15 less research attention, is anticipated future interaction. Next, the variable of anticipated future interaction will be discussed. 1.72 Anticipated future interaction —.A variable of interest that has been theoretically related to face-saving is anticipated future inter- action. Several authors, operating from an exchange theory perspective, measure the effect of anticipated future interaction on face-saving. Brown and Garland (1971) argue that an audience's inability to provide feedback weakens the constraints it imposes, lessening the control it has over a person's behavior. They hypothesize that face— saving will be greater when a person expects to receive evaluative feed- back than when this feedback is not expected. The effect of audience feedback on face—saving was nonsignificant. In a second experiment, Brown and Garland (1971) again tested the effect of anticipated feedback on face-saving, strengthening the feedback manipulation. Whereas in the first experiment subjects were not certain whether the feedback was to be delivered personally or impersonally, in the second experiment subjects anticipated fUture interaction with the audience. Brown and Garland found that subjects sang five times longer when they did not expect to meet strangers than when they did anticipate future interaction. The results of the Brown and Garland (1971) studies indicate that anticipated future interaction increases face-saving. .Although there is not extensive empirical evidence on this variable, anticipated future interaction will be included as an independent variable of secondary interest. Based on the research discussed, the fOllowing hypothesis can be derived: 16 In an embarrassing interaction, an embarrassed individual will use more humor when there is anticipated future interaction than when there is no anticipated future interaction. Another variable of interest that effects the use of humor is the number of people in an interaction. Next, different explanations of the effect of the number of people on the use of humor will be discussed. 1.73 Number of people — Several authors discuss how the number of peOple effects an interaction. Other people may increase an interactant's general level of arousal, which may increase the use of humor. Other peOple may also define humor as appropriate in the interaction. .A greater number of people can affect embarrassment, embarrassment being mitigated as the number of people increases. According to Zajonc (1965), social facilitation theory provides an explanation of why the presence of others may have an effect on the use of humor. According to social facilitation theory, the presence of others may provide cues as to what responses are appropriate or inapprOpriate. Other people may also increase a person's general level of arousal. Increased arousal, or the presence of others who define a situation as humorous may both increase the use of humor. Coser (1960) discusses why audience size might effect the use of humor. According to her analysis, a humorist is a humorist because of the audience. Her definition of humor reflects an audience centered approach, humor being a remark that receives a laughing response. Humor depends on "the collective perception of those to whom it addresses and is therefore defined by the social situation in which it occurs" 17 (1960:81). Goffman (1959) discusses how several participants in a conversation secretly communicate information about a third person. Goffman calls this activity "derisive collusion," and "it typically involves a secret dero— gation of the audience" (159:187). The presence of more persons in.an interaction makes derisive collusion possible. One possible f0rm.of derisive collusion is humor. Participants in a conversation may joke in order to communicate infermation about a third person. Sattler (1966) stresses the importance of the presence of others for embarrassment to occur. Embarrassment implies both the presence of another and the awareness that the other's presence is directed toward oneself. He also states that embarrassment can be mitigated in situations where there is a third person. Homer may be a technique that functions to mitigate embarrassment. Goffman (1967) discusses how audience size has an effect on the behavior of the participants. He states that participants have more license in a large-scale interaction than when there are two or three peOple because "the more participants there are to sustain the proceedings, the less dependent the occasion will be on any one participant" (1967:131). From social facilitation theory and the discussion of Coser (1960) we would expect the presence of a greater number of people to increase the use of humor. If humor is a form of derisive collusion, we would predict that a greater number of people increases the probability of derisive collusion, increasing the use of humor. we have discussed the theoretical considerations of this experi- ment. Next, we will describe the experimental design and methods. CHAPTER II EXPERIMENTAL DESIGN AND METHODS 2.1 Overview of Experimental Design This is an experimental study which was conducted by phone inter— views. Male students in Communication 100 classes were interviewed by female interviewers. Communication 100 is an introductory course entitled "Human Communication I," which deals with basic communication principles. Subjects were asked a series of embarrassing questions and were later asked to rate the interview on a number of dependent vari— ables, including unseriousness or humor. Two trained coders listened to the interview on extension phones. They were seated in separate rooms where they could not hear or see each other. Mouth pieces on the phones were removed so that the subjects could not hear the coders. As soon as the reSpondents were through answering the embarrassing questions, the coders placed the receivers where the rest of the interview could not be heard. The coders then rated the interview on a set of dependent measures. There were two independent variables in this experiment: antici- pated interaction and status. Subjects were told by the interviewer either to anticipate meeting with her in the future or not to anticipate meeting her, creating two levels of the future interaction variable. The interviewer introduced herself as either a college professor, an 18 19 undergraduate or a high school student, creating three levels of the status variable. Thus, there were six experimental cells in this design, which can be diagrammed as follows: ANTICIPATED STATUS INTERACTION YES NO High 3 6 Medium 2 5 Low 1 H The following variables were measured by the coders: humor or unseriousness, laughter, openness, anger, avoidance, loss fer what to say, anxiety and embarrassment. The respondent also indicated the level of these variables as he perceived them in the interaction. The interviewer recorded some additional dependent measures. She counted the number of Speech blocks the respondent used in answering the five questions, and she also recorded the total amount of time the subject spent answering the five questions. The fOIlowing procedure will be used to describe the experimental design. First, the selection and pre—test of embarrassing questions will be described. Then, the two independent variables, status and future interaction, will be examined, and the manipulation of these variables will be described. Next, the sampling of subjects will be discussed. After discussing sampling, the dependent variables and manipula- tion checks will be described, and the measurement of these variables will be discussed. Then, coder training and interviewer training will 20 be described. Now, we will turn to the selection and pre—test of embar- rassing questions. 2.2 Stimulus Questions 2.21 Selection of embarrassing questions — Embarrassing questions were generated from.Taylor and Altman's (1966) Intimacy-Scaled Stimuli fer Use in Studies of Interpersonal Relationships. Items in this report con- sist of statements about various aSpects of self that peOple might dis— cuss when forming interpersonal relationships. These items were scaled according to their intimacy value by two populations, Navy men and college males. Items in this report that were scaled high in intimacy value by college males were considered in generating embarrassing questions. The items that were considered had a mean scale value of 9.00 or more out of 10.00. Theoretically, these items were used since embar— rassment would be likely to occur when an individual discloses informa— tion that is intimate to him.or her. Several kinds of items that were ranked high in intimacy value were not used to generate embarrassing questions. Items dealing with sexual fantasies or masturbation were not used since reSpondents might consider the phone interview a prank call. Items dealing with one's Spouse were eliminated because only a small percentage of undergraduates, who were being used as subjects, were married. The f01lowing items were generated to be used as embarrassing questions in the pre—test: would you describe in detail the most embarrassing Situation you've ever been in. Describe your most positive traits for attracting the opposite sex. 21 Describe your most negative traits, those that turn off the Opposite sex. 2.22 Pretest of embarrassing questions — These questions were pre—tested in several Communication 100 classes. The experimenter interviewed male students individually in a private location while other students were in class. The students' answers to the embarrassing questions were tape- recorded. During the pre-test, students were told that the interviewer was conducting research on interpersonal communication fer her'Master's Thesis, and that she wanted to ask them a few questions about interpersonal relationships. .After being asked the three embarrassing questions, the tape recorder was turned off and the reSpondentS were asked to evaluate the interaction they just had. The following instructions were given: I would like you to evaluate our conversation. I would like to describe the measuring system.to you, Since you might be unfamiliar with it. I will call 10 units the average amount of a trait, say humor, in a conversation. Zero units will be the complete absence of a trait. If this conversation was twice as fUnny as an average con— versation, we would say it had 20 units of humor. If it was half as funny as an average conversation, we would say it had five units of humor. Is this rating system clear? Using the system described, the respondents were asked to rate the amount of the following traits in the conversation they just had: humor, openness, anger, avoidance, embarrassment, anxiety, laughter, cooperation , and loss fer what to say. .After these judgment items, the subjects were asked the following questions: 1. Imagine if a female were to call you on the phone to ask you these questions. would you hang up? 2. Can you think of any queStions that I could ask that would be more embarrassing, but would not cause you to hang up if they were asked on the phone? 22 3. WhiCh of the questions were most embarrassing? H. Which of the questions were the least embarrassing? Students suggested the use of the fOIlowing items because they were considered to be very embarrassing: Describe your last sexual encounter; describe your body; describe your sexual fantasies; describe your sexual adequacy; describe your last date. Items were eliminated that concerned sexual experience, however. A recent survey on campus indicated that many college freshmen had very little sexual experience. Since it is socially desirable to be sexually experienced, we might expect to get deceptive responses to questions about sexual experience. In addition, it conceivably is normative for college age males to exaggerate sexual experiences. If questions about sex were asked, we might be measuring deceiving responses instead of responses to embarrassment. The items that were used for the interview were those that were judged to be the most embarrassing. Items were reworded and new items were added based on suggestions students made during the pre-test. There were five different pre—test sessions that lasted approximately one and one—half hours each. The following items were selected for use in the interview: 1. Describe in detail the most embarrassing situation you've ever been in. 2. Describe your'most positive traits for attracting the opposite sex. 3. Describe your most negative traits for "turning off" the Opposite sex. H. Describe a recent situation with a female where you were at a loss for what to do or you didn't know how to behave. 23 5. Describe your body from.head to fOOt. As you make your descriptions, explain What you like or don't like about each body part. One of the purposes of the pre—test was to see how easily respon— dents could understand the measuring scale. During the pre-test, the experimenter found that all of the respondents were able to use the mea— suring system that was described to them, However, subjects had a tendency to respond as if 10 Units was the upper limit of the reSponse set, instead of an average value. In the final instructions, redundant information was provided about the scaling system. This was to ensure that the respon- dents understood that the typical 1 to 10 Unit scaling system was not being used. The redundant information that was added to the instructions was the f01lowing: Remember that you choose numbers above 10 to indicate more than an average amount of a trait and numbers below 10 to indicate less than an average amount of a trait. 2.3 Experimental Manipulations and Independent Variables Two variables in this experiment were manipulated, status and future interaction. There were three levels of the status relative to the status of the subject. The other variable was anticipated interaction, which had two levels. Subjects either anticipated or did not anticipate future interaction with the interviewer. Below is a description of the two experimental variables. 2.31 Status manipulation — High, medium, and low status was manipulated. Low status indicated that the interviewer was lower in status than the respondent. The number of years of school between the respondent and the interviewer was kept constant for this manipulation. The interviewer 2n introduced herself as being the same year in high school as the reSpon— dent was in college. For example, if the reSpondent was a junior in college, the interviewer introduced herself as a junior in high school. The following introduction was used fOr the low status manipu- lation: Hello, my name is . Although I am only a High School (Freshman, SOphomore, Junior, or Senior), I am conducting research at M.S.U. with the Department of Communication. I am doing this to fulfill a requirement fer my high school Social Studies class. The interviewer introduced herself as "guly_a high Sohool student" to reinforce the low status manipulation. Additional reinforcement for low status was added with the information that the interviewer was ful- filling a high school social studies class requirement. In the medium status manipulation, the interviewer introduced her— self as being the same year in school as the respondent. The introduc— tion for this manipulation was as fellows: Hello, my name is . I am.an under- graduate (Freshman, Sophomore, Junior, Senior), like yourself, conducting research with the Department of Communication. I am doing this to fulfill a require— ment for my undergraduate Social Science class. "Like yourself" was added in the medium status manipulation to reinforce that the interviewer was of equal status. The infOrmation about fulfilling a requirement for an undergraduate social science class was provided for two reasons. First, it was to strengthen the equal status manipulation. Secondly, it was designed to provide equivalent information to the low status condition. In the high status condition, the interviewer introduced herself as being of status higher than the respondent. She introduced hereself 25 as follows: Hello, my name is . I am.a visiting professor at M.S.U., conducting research with the Department of Communication. I am doing this as part of my post—doctoral work in social science. The information about doing post-doctoral work in social science was added to reinfOrce the high status manipulation. It was also pro— vided so that there would be an equivalent amount and kind of informa— tion as in the other two conditions. 2.32 Future interaction - In order to manipulate future interaction, reSpondents were chosen so that interacting with the interviewer in the future would be a plausible interaction. If respondents had been called from the general population, it would have been difficult to provide an incentive for them.to meet with the interviewer in the future. Subjects were chosen from Communication 100 classes to make future interaction plausible. In Communication 100 classes, researchers are frequently coming into class, and conducting personal interviews during class time. Respondents were told, in the anticipated future interaction condition, In addition to talking with you on the phone, I will be talking with you in your Comm. 100 class at the earliest possible time. This will be a personal interview, giving me more information. Your instructor has given me permission to meet and talk with you. Subjects were told that their instructor had granted permission fer the interview in order to strengthen the manipulation. Instructors had been told the content of this manipulation, in case any questions arose in class. 26 In the no anticipated future interaction condition, subjects were told This will simply be a phone interview. I will not need to meet with you or talk with you after this interview. 2.33 Number of people — In section 1.73, the inclusion of the number of people as an independent variable is discussed in detail. This was a measured variable in the experiment. The interviewer asked the subject how many people were with him, 2.H Subjects Subjects chosen were.male students in Communication 100 classes. Communication 100 students were chosen in order to make the anticipated future interaction manipulation plausible. Researchers frequently use Communication 100 students as subjects for research during class time. It would be reasonable for a researcher to conduct a personal interview with a student during class time. Male students were selected as respondents to:maximize embar- rassment. Since the subject matter of the interview was sexual, embar~ rassment would be greatest if females interviewed males than if females interviewed females or if males interviewed males. Males interviewing females was ruled out because the phone call could possibly be viewed as obscene. Subjects were chosen from fifteen sections of Communication 100. The night section of Communication 100 was eliminated Since the instruc- tor promised his class that no more research would be conducted during his class time. The total subject pool was approximately H00 males. P .l {.[I' r.|l tllrll. \ 'l 1". [‘l.l|l||\’|llll 11 III,‘ {(11 il‘ll‘l’l‘llllt 27 Subjects were chosen to be interviewed as fellows. Out of the fifteen sections of Communication 100 that met during the day, seven classes were informed about the experimental procedures and were asked whether their classes could be used for interviews. All instructors who were asked agreed to let their classes be used. The subjects who were actually interviewed were those who could be reached by phone in the evening (see Appendix B fOr the schedule of interviewing). Sixty interviews were completed, consisting of twenty— seven freshmen, fifteen sophomores, ten juniors, seven seniors, and one subject for whom year in school was not recorded. 2.5 Dependent Variables and Manipulation Checks There were ten dependent variables and two manipulation checks. The dependent variables were humor or unseriousness, openness, anger, cooperation, avoidance, laughter, loss for what to say, anxiety, embar- rassment, length of the conversation and number of monosyllabic speech blocks. Below is a brief description of the rationale for measuring these variables. 2.51 Description of dependent variables — The variable humor or unseriousness has been extensively discussed in Chapter I as a reSponse to embarrassment. Laughter was measured to determine the extent to which it covaries with humor. It was expected that laughter would occur with extreme embarrassment, whereas humor would occur with moderate embar— rassment. Openness was measured because it was thought that the more open subjects were, the less likely they would be to experience embarrassment. Anger was measured because it was possible for subjects to become angry Illilllli 28 at the embarrassing nature of our questions or at the intrusion of their privacy, instead of becoming embarrassed. Avoidance was included as a variable because Goffman (1967) and others describe this as a possible response to embarrassment. COOperation was added as a variable to obtain a measure of the extent to which the subjects were willing to participate in the interview. Loss fer what to say and anxiety were included in order to esti— mate the concurrent validity of our measure of embarrassment. Both anxiety and loss for what to say were thought to be domains of content that are measuring a concept similar to embarrassment. The extent to which these variables correlate with embarrassment provides a measure of the concurrent validity of our measure of embarrassment. The variable of length or total time the subject spent answering the questions was included as a variable to provide an estimation of the concurrent validity of our measures of embarrassment. Several authors, including Brown and Garland (1971), Garland and Brown (1972) and Kleck 2:431. (1966) find that the more embarrassed subjects are, the less time they spend in an interaction. If our measures of embarrassment correlate negatively with length of time, we will have concurrent validity for our measures of embarrassment. The number of monosyllabic speech blocks was counted by the inter— viewer. Goffman (1967) discusses that stuttering and speech blocks are objective signs that embarrassment is occurring. ,- 2.52 Measurement of dependent variables - Coders reached consensus on the criteria for coding the dependent variables after listening to a number of taped interviews. (See section 2.6 for a thorough discussion 29 of coder training.) Below is a description of how coders evaluated eight emotions: humor or unseriousness, anger, openness, laughter, avoidance, loss for what to say, anxiety and embarrassment. The vari— able cooperation was eliminated because coders thought that it was measuring the same phenomenon as anger. (1) Unseriousness/Humor was divided into three different cate— gories. The first category was making fun Of self. Exaggeration and understatement were frequently used when individuals made fUn Of theme selves. Secondly, humor or unseriousness was used to make fun Of the interviewer. The third category was humor or unseriousness that was directed at the situation. In all three Of the categories, the coders looked for banter, mock insults, sarcasm, exaggeration, teasing, and understatement. (2) Openness: Coders were able to classify high and low Open- ness. An individual who was very willing to discuss personal experi— ences was judged as being high in openness. Individuals Who were rated highest in openness described their immediate emotional reactions tO the situation, such as "I'm too embarrassed to tell you what happened," or "I'm scared to explain this because I don't know what you'll think." Individuals rated low in Openness, that is, less than average, were those who said as little as possible about their experiences. These individuals used a detached, monotone voice. They sounded as if they were reading an account Of someone else's experiences. (3) Apgep; .Anger was detected by short, abrupt sentences. Angry individuals were those who acted hostile. They Spoke in a raised tone Of voice. 30 (H) Avoidance: Several criteria were developed for avoidance that was higher than average. After listening to all the taped inter- views, cOderS agreed that respondents who refused to answer one question displayed twice as much avoidance (20 Units) as an average interaction. They also agreed that respondents who refused to answer two questions were displaying three times as much avoidance (30 Units) as an average interaction. Another category Of responses considered as high in avoidance were "pseudo—intellectual" or highly intellectualized responses. For example, one subject said "One might use retrospective—introjection in order to answer this question." (5) Laughter: Coders agreed that in the taped interactions they listened to, one laugh was average. If a respondent laughed once, he was given a score Of ten units Of laughter. The coders agreed that subjects who laughed three times showed twice as much laughter (20 Units) as an average interaction. (6) LOSS for what to say: This item.was judged independently Of the content Of the response. Persons judged below 10 were characterized as "bullshitters," i.e. they would continue to talk after they had adequately answered the question. Persons who exhibited a lot Of Speech blocks and who stuttered were judged as above average in this category. (7) Anxiepy: Persons who were rated with a higher than average amount of anxiety stuttered and used Speech blocks such as "um's" and "ah's." Talking exceptionally fast, using a high pitched voice, or having one's voice crack were also judged as indicating higher than average anxiety. 31 (8) Embarrassment: Persons who were rated with higher than average embarrassment stuttered and used speech blocks. They also swallowed and cleared their throats frequently. Subjects who changed the subject Of the conversation frequently and who hesitated a lot when answering questions were judged as being higher than average in embar— rassment. The other two dependent measures were monosyllabic speech blocks and time. Both variables were measured by the interviewer. For a description Of the measurement Of these variables, see section 2.7. 2.53 Manipulation checks — Manipulation checks were added at the end of the interview to insure that anticipated future interaction and status were remembered by the subject. To check the manipulation Of anticipated future interaction, the interviewer asked the respondent: DO you remember if I mentioned that we were going to meet? What did I say? For the status manipulation check, the interviewer asked the reSpondent: DO you remember who I said I was? The interviewer probed until the subject gave her some indication of the status that he perceived her as occupying. The interviewer recorded on the coding fOrm either high, medium, or low status, depending on the status Of the interviewer relative to the subject, as perceived by the subject. (Illilllil’lll‘lll‘ll.l.l[olllllll 32 2.6 Coder Trainipg Males close to the age of the subjects were used as coders. It was expected that male coders would be more empathic than female coders with the male respondents, especially since some Of the questions were sexual in nature. Training Of the coders took place over a three week period, prior tO the phone interviews. Three coders were chosen from.upper level communication courses. They were chosen based on their availability to make a time commitment tO the research project. Before training sessions fOr the coders started, the coders were told that their fUnction was to judge the responses Of subjects to a number Of embarrassing questions. Coders were not told at this time that there would be different experimental conditions, leaving them naive to the research hypotheses. The experimenter stressed to the coders that their judgments were extremely important. They were told that without their careful judgments, the experiment would be meaningless. This point was repeated several times during each training session. Below is a description Of the training sessions which took place over a three week period: 2.61 Training session 1 (2 1/2 hours) — During this session, the system Of measurement was thoroughly explained. Ten units was called the average amount of a trait, and 0 units was called the complete absence Of a trait. Coders were told that they could use any whole numbers they wished to indicate the amount Of a trait in an interaction. Several examples were used to explain the measuring system. For example, if a conversation was 33 twice as fUnny as an average conversation, it would be rated as 20 units Of humor, and if it was half as funny as average it would be rated 5 units Of humor. The coders were then told the 9 judgments they were to make: humor, Openness, cooperation, anger, avoidance, laughter, loss fOr what to say, anxiety, and embarrassment. Questions arose as to what was meant by humor, and the meaning was clarified as the subject's general unserious— ness about the situation. The first taped interaction was then played. When the interaction was finished, the coders recorded their judgments, consisting Of whole numbers, on a sheet Of paper next tO each Of the 9 categories. When all three Of the coders were finished with their judgments, each item.was discussed separately. .As each category or item was discussed, each coder gave reasons for his judgment. If there was a discrepancy Of more than 5 units for an item, the taped interaction was replayed. .After the tape was replayed, the coders were told to discuss the judgments they had made until they reached consensus as to which judgment was most apprOpriate. Two taped interviews were evaluated during the first session. During both interviews, each judgment item.was discussed separately. .At the end Of the session, the coders were told to listen to interactions around them, trying to apply the measuring system being used. 2.62 Training session 2 (3 hours) - Four taped interviews were judged. The two interviews that were judged the first day were judged again, to refresh the coders' memories. Two new taped interviews were judged, and each item was discussed separately by the coders. Using the same 3H procedure used during the first training session, the tape was replayed for items with more than five units discrepancy, and the coders were told to reach a consensus as to what judgment was most accurate. 2.63 Training session 3 (3 hours) - First, each item Of judgment was discussed. Each coder was asked to recall judgments that he had made that were both higher and lower than average. Then the coders discussed, item by item, what characteristics were present in the interactions that they had judged as being both more and less than average. 2.6H Training session H (2 1/2 hours) — One Of the three coders was absent from.the training session. It was decided that the absent coder would not be used in the experiment since it was difficult fOr him to meet regularly with the other coders. Five interviews were judged. The coders were told to pay partic- ular attention to the criteria they were using to make their judgments. 2.65 Training session 5 (3 1/2 hours) — The coders reached a consensus as tO What criteria to use in making each Of the eight judgments. See section 2.5 for a discussion of the criteria used to make each Of the eight judgments. 2.66 Training session 6 (2 1/2 hours) — The criteria developed were applied to five new interviews. TWO coders were used at this training session since the third coder had been eliminated. Judgments where there were more than five units of discrepancy were discussed separately. 2.67 Training session 7 (2 1/2 hours) — One Of the nine judgment cate- gories, OOOperation, was eliminated from the dependent measures. Coders felt that OOOperation was the Opposite of avoidance, and that subjects who were low in avoidance could be considered OOOperative. 35 During this training session, one interview as found where the subject scored average, i.e. 10 units, on all Of the dependent measures. Another tape was found where the subject scored 10 units on all but one Of the measures. These tapes were used as baseline tapes later to remind the coders of the criteria they started out with. Three more interviews were judged during this session. The general procedure for completing telephone interviews was discussed. Times were arranged when both the coders could be available to interview subjects. 2.68 Training session 8 (2 hours) - Three practice phone calls were completed, with the coders listening on extension lines. .After each phone call, the coders compared their responses. The coders then dis- cussed the phone calling procedure. In addition to the training described, the coders completed practice interviews immediately befOre most of the calling session. See Appendix B for the Experimental SOhedule, which includes a description Of the scheduling Of practice calls. In addition tO training the coders, the interviewers were also trained to ensure that all the interviewers were conducting calls in the same way. The following is a description Of how interviewers were trained. 2.7 Interviewer Training Interviewers were two undergraduate upperclass females in addition to the author. They were selected as interviewers because they were poised and self—confident. This was desirable Since the interviewers were going to ask embarrassing questions, and could easily become 36 embarrassed themselves. One week before interviewing started, the interviewers were given a OOpy Of the experimental protocol (see Appendix A). They were instructed to read and practice each of the six introductions, defining the experimental conditions, the written instructions for the dependent measures, and the five embarrassing questions. Before interviewing started, each interviewer listened to 3 taped interviews in order to Obtain an idea of what kind Of responses to expect. Each interviewer interviewed another person while the experimenter Observed. Specifically, the following instructions were given to the interviewers: (1) During the introduction, speak Slowly enough to be understood. Use a pleasant, conversational tone Of voice. (2) Use your own name for the introduction. (3) to not provide any additional information at the beginning of the interview. If the subject wants to know more about the study, repeat the infOrmation already given. Restate that we are conducting research about interpersonal communication, and that we want to know how the subject perceives his relationships with others. (H) In conditions I, II, V or VI (Appendix A) look on the class list to check the subject's year in school. In conditions I and IV, intro— duce yourself as the same year in high school as the subject is in college, according to the protocol for the introduction. In conditions II and V, introduce yourself as the same year in college as the subject, according to the protocol. (5) The interaction is to be timed in the following way. As soon as the subject starts talking, start the stOp watch. As soon as the subject pauses, stop the stOp watch. Only the subject's responses to the five embarrassing questions are to be timed. (6) Monosyllabic Speech blocks are to be timed in the following way. Every time.a.subject used the following sounds: "ah," "uh," "umu" "er," or any other similar one-syllable sound, the interviewer is to make a score on a scratch pad. Only Speech blocks in response to the five embarrassing questions are to be recorded. 37 (7) Interviewers were told to give no feedback to the subject's answers to the five embarrassing questions. If the subjects did not understand the 1n1t1a1 question, the interviewer was told to repeat it. (8) If a respondent indicated that the subject was having a hard time answering a question by reSponding "I don't know if I can answer that," "I can't think Of anything," or "DO I have to answer this?," the inter— viewer was told to reSpond "DO the best you can." (9) If the reSpondent refuses to answer, or indicates that he can not answer after being told to do the best he could, the interviewer was told to go on to the next question. (10) .After asking the embarrassing questions, the interviewer read the evaluation instructions as they are written (Appendix A). The interviewer was instructed to answer all questions thoroughly, making sure that the subject understood the task. (11) .After collecting the subject's judgment about the dependent measures, the interviewer was instructed to read the closing remarks on the last page Of the protocol. They were instructed to honestly answer any questions the subject might have. 2.8 Experimental Procedure The experiment consisted Of a phone interview, where a subject was asked a number Of embarrassing questions. Two coders unobtrusively listened to the interview and coded the subjects' reSponses on a number Of dependent variables, which were described in section 2.7. Below is a description Of the experimental preparations, interview protocol and the coders' protocol. 2.81 Experimental preparations — The following procedures were used to set up the experiment. Several hours before the experiment started the Experimenter checked the copies Of the M.S.U. class lists from which the subjects were taken. The phone numbers Of the male students in the class were Obtained from the M.S.U. student directory, and were written on the OOpy Of the class list. 38 Ten minutes before phone calling started, the instruments were prepared and set up in the appropriate rooms. The following instruments were placed in the room where the interviewer was making calls: a set Of class lists with phone numbers, a scratch pad, a set Of coding fOrms, a copy Of the experimental protocol, and an orthogonal polynomial table. The interviewer was given a whistle. The two coders went to Rooms 525 and 527 South Kedzie Hall. They each took a set Of dependent measures with them, .As soon as they reached the telephones, they unscrewed and removed the mouth pieces. They checked their telephones to ensure that they were listening on the appropriate extension. The Experimenter then told the interviewer Which one Of the six conditions to start with in making the phone calls. The condition that interviewers started with was rotated. Interviewers were told to try to complete interviews in blocks Of six, i.e., to try to complete one inter— view in each condition in each session. 2.82 Interviewer protocol - After preparing the experimental set-up, the interviewer dialed a number on the class list. As soon as She was finished dialing the number, but before the phone started ringing, she blew the whistle. This was a signal fOr the coders tO pick up their phones. If the individual on the class list was not there, the inter— viewer asked fOr a time to call back. The interview was completed as described in the training Of the interviewers, in section 2.7. As soon as the interviewer completed the phone call, the phone number Of the subject was recorded on the coding sheet, in a way that ensured confidentiality. 39 The experimenter was present in the room.where the phone calling was taking place when she was not the interviewer. This was tO ensure that phone calls were being completed in a similar way across inter— viewers. 2.83 COder protocol — After arranging the telephones as described in the experimental set up, the coders recorded caller number and coder number on all Of their coding sheets. When the whistle was blown, they were told to pick up their receivers. The coders listened to the interview up to the point where the interviewer started to explain the dependent measure. At this point, the coders placed the receiver on the desk so that they could not hear the conversation. Then they started to record their judgments Of the eight dependent variables. They waited in their rooms until the interviewer yelled "O.K.," indicating that the interview was over. The coders then placed the telephone receivers back on the hook, and went into the interviewer room. The coders then sat down in two chairs facing the interviewer's desk. She read the coded phone number and the year in school to the coders, who recorded these on their coding forms. The interviewers then looked on top Of the protocol sheet fOr the condition number that the interviewer had used. They Similarly recorded "Condition." The inter- viewer and the coders then compared the five digits they had recorded as condition, to eliminate possible coding errors. An experimental schedule Of calling is provided in Appendix B. We have now considered the theory and methodology of our experi- ment. Next, we will discuss the results and conclusions. CHAPTER III RESULTS, CONCLUSIONS, AND DISCUSSION 3.1 Manipulation Checks The correlation of the manipulation Of status and recalled status is .580. The correlation Of our manipulation Of anticipated interaction and recalled anticipated interaction is .8H5. Both Of these correlations are significant beyond the .001 level. Hence, both manip— ulations seem.tO be highly successful. 3.2 Reliability and Validipy Of Measures 3.21 Reliability - The reliability between coders is the correlation be- tween their estimates Of each dependent variable. Table 1 contains the intercorrelations among all variables to be discussed. The intercoder reliabilities for the dependent measures Of interest are from .8H to .99; this is clear evidence for the high reliability of all measures. 3.22 Validipy — we expect embarrassment to correlate highly with anxiety and speech loss because both variables measure similar domains Of content. we also expect embarrassment to correlate negatively with the length Of the conversation. An examination Of ten appropriate correla- tions in Table 1 shows all to be significant at least at p<.01, indicating the validity of the embarrassment measures. H0 H1 Ho. v a v." mo. ..V av." smegma mmoew Sam .H havoc so emeoo who; =H= Saxons mmHHmHhm> .mm H z x new .Nm u z eoHaz hoe .Hepmcmav me Hamoxm mHamHom> sumo pom om u 2a »m .m .380 NE :N: nodes Low 0 AmLOpOpflmHOmmv m NN.- HN.- ON.- mH.- «SN. No.- mo. HH.- HH.- mH.- HNN.mv mm.m N geese Hmn9m> NHx HH.- NH.- :H.u mo.- mH. 0H,- mo.- Ho. mo.- NH.- HHm.mV mm.m H gossm Hmaeo> me mo.- so. so.- mo.- NN. mo. mo.- HN.- 30.- mo. Hmo.mv Nm.NH N ewsma me mo.- mo. No.- so. mH. Ho. mo.- NN.- so.. so. Hmm.mv mm.NH H awsma :Hx «SSS.-««NS.-««SS.-««mm.-e«H:. mH. so. OH.- No. Ho.- Hmm.NmHvs:.NmH nausea me News. «New. «Ham. «was. Hom.u mm.- NH.- NH. mH. No. HNm.NV NH.mH N meH eoomam NHx «ems. «Sam. «How. «New. «HN.- Ho.- 30.- mH. HH. mo.- ANo.mV m:.mH H meH aommam HHx H Seam. «Sam. «Sam. «om.- mo.- NN.- Ho.- :N. mo. Hom.NV Nm.mH N sHmec< on H «Ham. «era. «Sam.- HH.- HH.- No. NH. so. ASS.NV Nm.mH H sHmeq< ax H «ems. «Nm.- HH.- HN.- No. HN. so. Amm.NV NN.mH N HcQEmmmonmaem ax H emN.- No.- mH.- HH. mH. OH. Hms.Nv NH.mH H HamemmmeHmaem ex H mo. Ho.- mN.- NQ.- Ho. ANS.HV NN.H mooomeHHHome ax H o o a o Hm:.HV oo.N coHHomembcH om x «.mx H o o o ANS. V oo.N cOHHomomoaH Hm x_< 3x H o o Hm:.Ho oo.N Home OHHmHumso .msompm mx H o ANS. V SS.N HHmV emmcHH .msHmbm Nx H AHo.HO oo.H a so...“ moeemflfimpm o>flporuomom pom. wag meMeOHmSHOOcHoeoH .H pang. "III?! II illlulllllllll‘l || llllf}' -‘ 'IIIIII H2 .N omeoo so =N= noses pom “AmHOHMPeHsommV m amoome.mmooo wow .H omeoo so cocoa moms =H= smegma mmHomHoms .mm x no... .3 n z 623 o8 .sfimcod Ho.JV o ”.3.“ 8.4 o 9.." z 2o max voooxm oaomwhm> romp 90% em m QQESm Honsm> ea «smm. «are. same. mm. semm.l :m.l x H «ems. «was. «SN. esN.- mH.: H posse Hmooms on H «was. mo. mH.- mo.- N owsmo me H No.- mH.- mo.- H omsmo :Hx H «smm.- Homm.- opwcmo me H «Ham. N mmoo oomoom NHx H H mmoo oommom HHx me me :Hx me NHx HHx H.o.mv z form of face—saving -————> (time 2) On the other hand, if one uses unseriousness to establish distance from an interaction to prevent embarrassment, unseriousness may be negatively related or unrelated to embarrassment. Among all subjects, then, the relationship between unseriousness and embarrassment may be spuriously influenced by the proportion Of subjects differing in these uses Of unseriousness, and by the time Span Of the interaction and Of the measurement procedure. Secondly, all relations in this discussion have been tested statistically with models that are linear in form; other effects have been examined by apprOpriate coding SO that the relations could be examined within the general linear model. Since the dependent variables are measured in ways akin to psychophysics, perhaps some form of power law might be more appropriate to these data (see, e.g., Hamblin, 197H). Thirdly, the evaluation of the subjects' responses to the inter— action was made by coders in the included analyses. Clearly, these coders lacked any visual means for assessing the subject's response; 53 even if the visual component were available, the coders' estimates may not be equivalent to the subject's own assessment. This is the issue of self—perception vs. person-perception by Observers, and is worth fUrther investigation. The three above issues require additional research. The results we have reported concern the nature Of interaction, and deal with verbal humor and laughter as interactional techniques employed to maintain self and the process Of interaction. Humorous responses tO embarrassing interactions give us important infOrmation on how communication patterns are restructured once communication breakdown begins. we find that a key determinant Of the use Of laughter is the sheer number of others present in the communication situation. Further research specifying the nature Of the causal link between failure Of the communication system and interpersonal strategies for its recovery are clearly needed to expand our understanding Of social interaction. IIII ll 'Illlillu' llll'lnl ill! .II. IIIL’.‘ ‘II II" III N I! ll." '1‘. I I'll ‘ APPENDIX A EXPERIMENTAL PROTOCOL APPENDIX A EXPERIMENTAL PROTOCOL CONDITION I HELLO, MY NAME IS . ALTHOUGH I AM ONLY A HIGH SCHOOL , I AM CONDUCTING RESEARCH AT M.S.U. WITH THE DEPARTMENT OF COMMUNICATION. I AM DOING THIS TO FULFILL A REQUIREMENT FOR MY HIGH SCHOOL SOCIAL STUDIES CLASS. WE ARE CONDUCTING RESEARCH ABOUT INTERPERSONAL COMMUNICATION. WE WANT TO KNOW HOW YOU PERCEIVE YOUR RELATIONSHIPS WITH OTHER PEOPLE. YOUR HELP CAN MAKE A CONTRIBUTION TO WHAT WE KNOW ABOUT INTERPERSONAL RELATIONSHIPS. CAN I ASK YOU A FEW QUESTIONS? IN ADDITION TO TALKING WITH YOU ON THE PHONE, I WILL BE TALKING WITH YOU IN YOUR COMM. 100 CLASS AT THE EARLIEST POSSIBLE TIME. THIS WILL BE A PERSONAL INTERVIEW, GIVING ME MORE INFORMATION. YOUR INSTRUCTOR HAS GIVEN ME PERMISSION TO MEET AND TALK WITH YOU. CONDITION II HELLO,MYNAMEIS . IAMANUNDER— GRAIUAT'E , LIKE YOURSELF, CONDUCTING RESEARCH WITH THE DEPARTMENT OF COMMUNICATION. I AM DOING THIS TO EULFILL A REQUIREMENT 5H 55 FOR MY UNDERGRADUATE SOCIAL SCIENCE CLASS. WE ARE CONDUCTING RESEARCH ABOUT INTERPERSONAL COMMUNICATION. WE WANT TO KNOW HOW YOU PERCEIVE YOUR RELATIONSHIPS WITH OTHER PEOPLE. YOUR HELP CAN MAKE A CONTRIBUTION TO WHAT WE KNOW ABOUT INTERPERSONAL RELATIONSHIPS. CAN I ASK YOU A.FEW QUESTIONS? IN ADDITION TO TALKING WITH YOU ON THE PHONE, I WILL BE TALKING WITH YOU IN YOUR COMM. 100 CLASS AT THE EARLIEST POSSIBLE TIME. THIS WILL BE A PERSONAL INTERVIEW, GIVING ME MORE INFORMATION. YOUR INSTRUCTOR HAS GIVEN ME PERMISSION TO TALK WITH YOU. CONDITION III HELLO, MY NAME IS . I AM.A VISITING PROFESSOR AT'M.S.U., CONDUCTING RESEARCH WITH THE DEPARTMENT OF COMMUN- ICATION. I AM DOING THIS AS PART OF MY POST-DOCTORAL WORK IN SOCIAL SCIENCE. WE ARE CONDUCTING RESEARCH ABOUT INTERPERSONAL COMMUNICATION. WE WANT TO KNOW HOW YOU PERCEIVE YOUR RELATIONSHIPS WITH OTHER PEOPLE. YOUR HELP CAN MAKE A CONTRIBUTION TO WHAT WE KNOW ABOUT INTERPERSONAL RELATIONSHIPS. CAN I ASK YOU A FEW QUESTIONS? IN ADDITION TO TALKING WITH YOU ON THE PHONE, I WILL BE TALKING WITH YOU IN YOUR COMM. 100 CLASS AT THE EARLIEST POSSIBLE TIME. THIS WILL BE.A PERSONAL INTERVIEW, GIVING ME MORE INFORMATION. YOUR INSTRUCTOR HAS GIVEN ME PERMISSION TO TALK WTTH YOU. ..tl.lo 5 6 CONDITION IV HELLO, MY NAME IS . ALTHOUGH I AM ONLY A HIGH SCHOOL , I AM CONDUCTING RESEARCH AT M.S.U. WITH THE DEPARTMENT OF COMMUNICATION. I AM DOING THIS TO FULFILL A REQUIREMENT FOR MY HIGH SCHOOL SOCIAL STUDIES CLASS. WE ARE CONDUCTING RESEARCH ABOUT INTERPERSONAL COMMUNICATION. WE WANT TO KNOW HOW YOU PERCEIVE YOUR RELATIONSHIPS WITH OTHER PEOPLE . YOUR HELP CAN MAKE A CONTRIBUTION TO WHAT WE KNOW ABOUT INTERPERSONAL RELATIONSHIPS. CAN I ASK YOU A FEW QUESTIONS? THIS WILL SIMPLY BE A PHONE INTERVIEW. I WILL NOT NEED TO MEET WITH YOU OR TALK WTTH YOU AFTER THIS INTERVIEW. YOUR INSTRUCTOR HAS GIVEN ME PERMISSION TO TALK WITH YOU. CONDITION V HELLO, MY NAME IS . I AM AN UNDERGRADUATE , LIKE YOURSELF, CONIIJCTING RESEARCH WTTH THE DEPART- MENT OF COMMUNICATION. I AM DOING THIS TO FULFILL A REQUIREMENT FOR MY UNDERGRADUATE SOCIAL SCIENCE CLASS. WE ARE CONDUCTING RESEARCH ABOUT INTERPERSONAL COMMUNICATION. WE WANT TO KNOW HOW YOU PERCEIVE YOUR RELATIONSHIPS WTTH OTHER PEOPLE. YOUR HELP CAN MAKE A CONTRIBUTION TO WHAT WE KNOW ABOUT INTERPERSONAL RELATIONSHIPS. CAN I ASK YOU A FEW QUESTIONS? THIS WILL SIMPLY BE A PHONE INTERVIEW. I WILL NOT NEED TO MEET WITH YOU OR TALK WITH YOU AFTER THIS INTERVIEW. YOUR INSTRUCTOR HAS 57 GIVEN ME PERMISSION TO TALK WITH YOU. CONDITION VI HELLO, MY NAME IS . I AM A VISITING PROFESSOR AT M.S.U. , CONDUCTING RESEARCH WITH THE DEPARTMENT OF COMMUNICATION. I AM DOING THIS AS PART OF MY POST—DOCTORAL WORK IN SOCIAL SCIENCE. WE ARE CONDUCTING RESEARCH ABOUT INTERPERSONAL COMMUNICATION. WE WANT TO KNOW HOW YOU PERCEIVE YOUR RELATIONSHIPS WTTH OTHER PEOPLE. YOUR HELP CAN MAKE A CONTRIBUTION TO WHAT WE KNOW ABOUT INTERPERSONAL RELATIONSHIPS. CAN I ASK YOU A FEW QUESTIONS? THIS WILL SIMPLY BE A PHONE INTERVIEW. I WILL NOT NEED TO MEET WITH YOU OR TALK WITH YOU AFTER THIS INTERVIEW. YOUR INSTRUCTOR HAS GIVEN ME PERMISSION TO TALK WITH YOU. 1. FOR THE FIRST QUESTION, I WOULD LIKE YOU TO DESCRIBE IN DETAIL THE MOST EMBARRASSING SITUATION YOU'VE EVER BEEN IN. 2. DESCRIBE YOUR MOST POSITIVE TRAITS FOR ATTRACTING THE OPPOSITE SEX. 3. DESCRIBE YOUR MOST NEGATIVE TRAITS FOR "TURNING OFF" THE OPPOSITE SEX. 58 H. DESCRIBE A RECENT SITUATION WTTH.A FEWALE WHERE YOU WERE AT A LOSS FOR WHAT TO DO, OR YOU DIDN'T KNOW HOW TO BEHAVE. 5. I WOULD LIKE YOU TO DESCRIBE YOUR BODY FROM HEAD TO FOOT. AS YOU MAKE YOUR DESCRIPTIONS, EXPLAIN WHAT YOU LIKE OR DON'T LIKE ABOUT EACH BODY PART. I WOULD LIKE YOU TO EVALUATE THE CONVERSATION WE JUST HAD. IN ORDER TO MAKE THIS EVALUATION, I WOULD LIKE TO DESCRIBE THE MEASURING UNITS. WE ARE GOING TO CALL 10 UNITS THE AVERAGE.AMOUNT OF.A TRAIT, SAY HUMOR, IN.A CONVERSATION. 0 UNITS WILL BE THE COMPLETE ABSENCE OF HUMOR. IF THIS CONVERSATION WAS.H5EEpAS FUNNY AS AN AVERAGE INTER— ACTION, YOU WOULD RATE IT AS 5 UNITS OE'EUNNINESS. IF IT WAS TWO TIMES AS FUNNY AS AN AVERAGE CONVERSATION, IT WOULD BE RATED AS 20 UNITS OF FUNNINESS. YOU CAN USE ANY WHOLE NUMBERS YOU WISH. REMEHBER THAT YOU CHOSE NUMBERS ABOVE 10 TO INDICATE MORE THAN AVERAGE AMOUNT OF.A TRAIT, AND NUMBERS BELOW 10 TO INDICATE LESS THAN AVERAGE AMOUNT OF A.TRAIT. PHONE NO: CALLER: YEAR SCHOOL: CONDITION: CODER: 1. IF 10_UNITS IS THE AMOUNT OF HUMOR OR UNSERIOUSNESS IN AN AVERAGE INTERACTION, AND 0 IS THE COMPLETE.ABSENCE OF THAT TRAIT, HOW HUMOROUS OR UNSERIOUS WAS THIS CONVERSATION? UNITS 59 2. IE 10 UNITS IS THE AMOUNT OF OPENNESS IN THE AVERAGE INTERACTION, AND 0 IS THE COMPLETE ABSENCE OF THAT TRATT, HOW OPEN WERE YOU (WAS THE SUBJECT) DURING THE INTERACTION? UNITS 3. IF 1_(_)_ UNITS IS THE AMOUNT OF ANGER IN AN AVERAGE INTERACTION, AND 0_ IS THE COMPLETE ABSENCE OF THAT TRAIT, HOW ANGRY WERE YOU (WAS THE SUBJECT) DURING THIS CONVERSATION? UNITS H. IF l_(_)_ UNITS IS THE AMOUNT OF AVOIDANCE IN THE AVERAGE INTERACTION, AND 0 IS THE COMPLETE ABSENCE OF AVOIDANCE, HOW MUCH DID YOU WANT TO (DID THE SUBJECT) AVOID THIS INTERACTION? UNITS 5. IF 1_0_ UNITS IS THE AMOUNT OF LAUGHTER IN AN AVERAGE CONVERSATION, AND 0 IS THE COMPLETE ABSENCE OF LAUGHTER, HOW MUCH DID YOU (DID THE SUBJECT) LAUGH? UNITS 6. TE 10 UNTTS IS THE AMOUNT THAT PEOPLE ARE AT A LOSS FOR WHAT TO SAY IN AN AVERAGE CONVERSATION, AND 0 UNITS IS THE ABSENCE OF THAT TRAIT, HOW MUCH WERE YOU (WAS THE SUBJECT) AT A LOSS FOR WHAT TO SAY? UNITS 7. IE 10 UNTTS IS THE AMOUNT OF ANXIETY IN AN AVERAGE INTERACTION, AND 0 IS THE COMPLETE ABSENCE OF THAT TRAIT, HOW ANXIOUS WERE YOU (HAS THE SUBJECT) DURING THIS CONVERSATION? UNTTS 8. IF _l__0_ UNTTS IS THE AMOUNT OF EMBARRASSMENT IN AN AVERAGE INTER— ACTION AND _0_ IS THE COMPLETE ABSENCE OF THAT TRATT, HOW EMBARRASSED WERE YOU (WAS THE SUBJECT) DURING THIS CONVERSATION? UN ITS 60 9. LENGTH OF CONVERSATION. SEC. 10. DO YOU REMEHBER IF I MENTIONED THAT WE WERE GOING TO MEET? WHAT DID I SAY? 11. DO YOU REMEMBER WHO I SAID I WAS? HIGH MEDIUM LOW 12. NUMBER OF UM'S AND.AH'S, INCLUDING ALL MONOSYLLABTC SPEECH BLOCKS: 13. HOW MANY PEOPLE ARE WITH YOU? ' THANK YOU VERY MUCH FOR YOUR ANSWERS. (AS I INDICATED EARLIER, I WILL NOT BE COMING INTO YOUR COMM. 1oo CLASS TO TALK WITH YOU.) I WOULD APPRECIATE IT IF YOU WOULD NOT DISCUSS THIS INTERVIEW WITH ANYONE IN YOUR CLASS. DO YOU HAVE ANY QUESTIONS. (IF NOT) THANK YOU VERY MUCH FOR YOUR HELP. APPENDIX B D