.l :3. "Pr-4‘: A. x ‘i :2... v r w .I 3: ) ‘J~ wig-mafia m. Y..- ‘ . .g V “12"). r: .Jh’dauflnnd _ _ X . .1‘.h.. : .n. .1; .n “I 20 LIBRARY 20 8 Michigan State mversity This is to certify that the dissertation entitled ON THE HESITATION TO SHARE BAD NEWS: THREE EMPIRICAL STUDIES presented by JAYSON LEE DIBBLE has been accepted towards fulfillment of the requirements for the Ph.D. degree in COMMUNICATION 29—..A Major Professor’s Signature JULY 24, 2008 Date MS U is an affirmative-action, equal-opportunity employer PLACE IN RETURN BOX to remove this checkout from your record. TO AVOID FINES return on or before date due. MAY BE RECALLED with earlier due date if requested. DATE DUE DATE DUE DATE DUE 5/08 K:IPro;/Acc&Pres/CIRCIDateDuemdd ON THE HESITATION To SHARE BAD NEWS: THREE EMPIRICAL STUDIES By Jayson Lee Dibble A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Communication 2008 ABSTRACT ON THE HESITATION TO SHARE BAD NEWS: THREE EMPIRICAL STUDIES By Jayson Lee Dibble The MUM effect refers to the robust research finding that people are reluctant to share bad news. Three studies were conducted to illuminate the nature of the MUM effect. First, an experiment tested two reasons for the MUM effect in the interactions of close friends and strangers. Next, communication behaviors of bad news revelation are explored. A third study experimentally tested the direction of the MUM effect. Study 1 experimentally investigated whether the reluctance to share bad news to a close friend results because of charitable concerns about the other’s feelings or because of self-presentation concerns. Participants (N = 165) shared good or bad news (i.e., high or low score on a phony test) with a close friend or a stranger, and the delay (in seconds) before the sender revealed the news was measured. Senders of bad news waited longer to engage the recipient than senders of good news. This replicated the MUM effect. However, no main effect emerged for friend/stranger condition nor was there a significant interaction. Study 2 utilized as its data source the videotaped interactions from the previous experiment and coded communication behaviors associated with good or bad news delivery. Variables examined included teasing the recipient, greeting the recipient, message elaborateness, and whether the sender overtly questioned the validity of the test. Senders teased recipients when delivering bad news to a close friend, and senders questioned the test’s validity only when the score was low. No associations were observed between news valence and greetings, nor between news valence and message elaborateness. The third study again employed a false feedback test design to uncover the direction of the MUM effect (i.e., whether MUM effect results from reluctance to share bad news or eagerness to share good news). Participants (N = 114) shared good, neutral, or bad news (i.e., high, medium, low score) to a stranger. News valence, delay before response, and reluctance were measured. Senders viewed high scores as more positive than medium scores, which were viewed as more positive than low scores. Delay before response as well as reluctance also appeared to be linearly related to test score. When scores were high, senders experienced minimal reluctance and shortest delays. When scores were low, reluctance was higher and delay was greater. Taken together, these studies suggest bad news delivery to involve face threat considerations which may be affected by the nature of the relationship between the interlocutors. Perhaps developing rapport with recipients will benefit both recipients as well as those who routinely deliver bad news. Additionally, these study raise questions regarding current methods associated with conducting bad news research. Cepyfight by JAYSON LEE DIBBLE 2008 For Gramma and Grampa Schultz and my parents. ACKNOWLEDGMENTS Having grown up in the port city of Muskegon, Michigan, nautical imagery springs readily to mind. Hence, if completing a dissertation can be compared to moving a sailboat along some charted course, then the rudder of this ship was undoubtedly under the skillful helm of my advisor, Dr. Tim Levine. Dr. Levine, your patience, advice, reassurance, and encouragement were invaluable. I will strive each day to emulate your model of excellence when it comes to research, teaching, and advising my own students. A million thank yous are due also to the remainder of my doctoral committee: Dr. Steve McComack, Dr. Bill Donohue, and Dr. Larry Schiamberg. I also wish to specially recognize Dr. John Sherry—my new gold standard of liberation and revolution. You have set the bar in my mind as to how a researcher- teacher-scholar can also be a devoted husband and father. I am indeed better for knowing you and your family. Stooge City Liberation Front forever! Many thanks go out to Austin Lee, the “HTML wizard” who created the computer screens used in the second experiment. Moreover, I would be remiss to not thank Marge Barkman. Spending four years in the office directly across the hall from yours was the smartest move of my entire education. Finally, I wish to thank my family. Mom never let up after she first told me at age 5 that I could become a doctor. I love you, dad, Ryan, and Jessica for pushing me and encouraging me through thick and thin. Equally deserving is my last honoree—Amy, I don’t know what I ever did to deserve you, but I promise I’ll work each day to repay your support and love. vi TABLE OF CONTENTS LIST OF TABLES .......................................................................................................... ix LIST OF FIGURES .......................................................................................................... x CHAPTER 1: INTRODUCTION ........................................................................................................... 1 CHAPTER 2: LITERATURE REVIEW ................................................................................................. 4 Bad News .............................................................................................................. 4 Relational Closeness ............................................................................................. 8 Emotional Congruence ......................................................................................... 9 CHAPTER 3: THE MUM EFFECT AND RELATIONSHIP CLOSENESS ...................................... 12 Method ............................................................................................................... 16 Participants ............................................................................................. 17 Design ..................................................................................................... l 7 Procedures .............................................................................................. l 8 Measures ................................................................................................ 22 Results ................................................................................................................ 25 Discussion ........................................................................................................... 29 CHAPTER 4: THE MUM EFFECT AND COMMUNICATION BEHAVIOR .................................. 34 Method ................................................................................................................ 35 Communication Variables ..................................................................... 36 Results ................................................................................................................ 38 Teasing ................................................................................................... 38 Greetings ................................................................................................ 40 Questioning Test’s Validity ................................................................... 42 Elaborateness of the Scorer’s Message .................................................. 44 Discussion .......................................................................................................... 45 CHAPTER 5: DIRECTION OF THE MUM EFFECT ......................................................................... 55 Method ................................................................................................................ 59 Participants .............................................................................................. 60 Procedures .............................................................................................. 60 Measures ................................................................................................ 67 Results ................................................................................................................ 69 Induction Checks ................................................................................... 69 Sex Effects ............................................................................................. 70 vii Test of Hypotheses and Research Question ........................................... 70 Discussion .......................................................................................................... 74 CHAPTER 6: GENERAL DISCUSSION ............................................................................................ 79 NOTES .......................................................................................................................... 84 APPENDICES ............................................................................................................... 85 APPENDIX A: Study 1 Pre-Questionnaire ....................................................... 86 APPENDIX B: Study 1 Scorer Post-Questionnaire .......................................... 90 APPENDIX C: Study 1 Taker Post-Questionnaire ............................................ 95 APPENDIX D: Social Perception Task ............................................................. 99 APPENDIX E: Scorer’s Instruction Sheet ....................................................... 102 APPENDIX F : Study 3 Scorer Questionnaire ................................................. 103 APPENDIX G: Study 3 Taker Questionnaire .................................................. 109 REFERENCES ............................................................................................................. l 16 viii LIST OF TABLES TABLE 1: Means, Standard Deviations, and Significance Tests for News Valence Induction Check Indicators as a Function of Feedback Condition ................... 27 TABLE 2: Mean Times to Response (in Seconds) As a Function of News Valence and Friends/Strangers ...................................................................................... 28 TABLE 3: Frequency of Teasing by Experimental Condition ...................................... 39 TABLE 4: Frequency of Greeting by Experimental Condition ..................................... 42 TABLE 5: Frequency of Test Validity Rejection by Experimental Condition ............. 43 TABLE 6: Mean Scorer Message Length (Words Per Message) As a Function of News Valence and Friends/Strangers ........................................................................ 45 TABLE 7: Means and Standard Deviations for News Valence Induction Check Indicators As a Function of Valence Category ................................................... 69 TABLE 8: Zero-order Correlations Among Primary Variables .................................... 71 ix LIST OF FIGURES FIGURE 1: Screen Utilized by Senders to Input Test-taker’s Scores ........................... 62 FIGURE 2a: Screen Shot of Randomly Generated Superior Score ................................ 64 FIGURE 2b: Screen Shot of Randomly Generated Average Score ................................ 65 FIGURE 2c: Screen Shot of Randomly Generated Inferior Score ................................. 66 FIGURE 3: Delay Before Feedback as a Function of Test Score .................................. 72 FIGURE 4: Sender Reluctance as a Function of Test Score .......................................... 73 FIGURE 5: Sender’s Perceptions of News Valence As a Function of Test Score ........ 74 CHAPTER 1: INTRODUCTION At some point in their lives, individuals have found or will find themselves in the position of having to relay some negative information to another. In lay speak, this is referred to as delivering bad news. Rosen, Tesser, and colleagues (Johnson, Conlee, & Tesser, 1974; Rosen & Tesser, 1970; Rosen & Tesser, 1972; Tesser & Rosen, 1975; Tesser, Rosen, & Conlee, 1972) observed that, in general, individuals display some degree of hesitation to deliver bad news to the person for whom the news is bad. Rosen and Tesser term this reluctance the MUM effect (keeping “Mum” about Undesirable Messages). Research documents MUM effects manifesting in various ways. For example, workers at an agency for aid to the disabled took longer to communicate the agency’s decision if that decision was to deny aid than if it was to grant aid (Tesser, Rosen, & Tesser, 1971). Senders will often distort negative information in a positive direction so as to soften its negative tone (Fisher, 1979). Sussman and Sproull (1999) observed a MUM effect to occur in that, given the option, senders opted to deliver bad news via electronic means instead of face-to-face. More recently, Yariv (2006) found that elementary school principals experienced a large degree of hesitation when faced with delivering feedback to poorly-perforrning teachers. In each of these studies, the outcome is ascribed to the sender having somehow hesitated or experienced unpleasantness when faced with delivering bad news. Two plausible explanations for the MUM effect can be found in the literature. The first argues that people are loath to transmit bad news because they anticipate psychological discomfort associated with bringing about negative affect in the receiver (Johnson, Conlee, & Tesser, 1974). The other reason holds that people hesitate to share 1 bad news not for the primary benefit of the receiver, but because they fear the receiver will look upon them unfavorably (Rosen & Tesser, 1972). Several studies have undertaken the task of evaluating the validity of both explanations, and their results have been generally consistent with the second explanation: that any reluctance to transmit bad news is a self-serving public display motivated by fear of being evaluated negatively (Bond & Anderson, 1987; Johnson et al., 1974; Rosen & Tesser, 1972; Uysal & Oner- Ozkan, 2007). A feature common to studies supporting the self-presentation explanation is that most examined the relay of bad news between strangers. It might be questioned if this explanation for the reluctance to transmit bad news holds if people are sharing bad news with someone they know personally, such as a close friend. It is reasonable to expect that both explanations could operate at different times, depending on the situation. Self- presentation concerns might be more salient with strangers while concern for others might be stronger with friends. The results of three studies are presented. The primary goal of the first study is to address the boundary conditions of the two explanations for why individuals might hesitate to share bad news with a close friend within the context of a live interaction episode. The second study attempts to identify some of the communication behaviors that accompany the MUM effect. Specific verbal behaviors were treated as dependent variables to observe what, if any, behavioral differences exist when people share good news or bad news with either a close fi'iend or a stranger. Further, it is the goal to note whether any behavioral differences are more or less consistent with either of the hypothesized explanations for the MUM effect. 2 The third study aims to further elucidate the nature of the MUM effect. Past experiments conclude a reluctance to share bad news on the basis of a mean difference between bad news versus good news conditions. Curiosity remains as to whether the MUM effect obtains because people dislike sharing bad news, or whether people are particularly eager to share good news. Prior research that observed bad news to be shared much slower than good news would be consistent with either possibility. This second experiment adds a “neutral news” condition to which both bad news and good news sharing can be compared. To begin, the concepts of bad news and relational closeness are briefly defined and discussed in Chapter 2. Chapter 3 reports the first experiment. Both explanations for the reluctance to share bad news are explicated, with arguments being forwarded as to how each might operate within close relationships. Chapter 4 presents the results of the second study which explores verbal behaviors that accompany and possibly enlighten the reasons for the MUM effect. The third study, an experiment which seeks empirical support for the direction of the MUM effect, is discussed in Chapter 5. CHAPTER 2: LITERATURE REVIEW Bad News Rosen and Tesser’s (1972) seminal work with undesirable messages is generally assumed to mark the beginning of bad news research. Bad news has since been studied in a variety of contexts. A popular setting in which bad news continues to be examined involves the delivery of bad news from physicians to their patients (e. g., Carter, Nutt, & Carter, 2007; Casarett & Quill, 2007; Dworkind, 2006; Eggly et al., 2006; Greenberg, Ochsenschlager, O’Donnell, Mastruserio, & Cohen, 1999; Strauss, Sharp, Lorch, & Kachalia, 1995; Widerman, 2002). Health care practitioners routinely deliver bad news in the course of their duties, and are thus eager for insight regarding the process of breaking bad news. Most of this research is targeted toward optimizing the bad news delivery process for the benefit of the patient. However, physicians reported that the stress they experienced from delivering bad news frequently lingered from several hours to three or more days (Ptacek, Ptacek, & Ellison, 2001). The healthcare setting alone justifies the investigation of the bad news delivery process from the sender’s perspective (instead of focusing exclusively on the recipient). Other contexts in which bad news delivery has been examined include organizational settings when supervisors must relay bad news to employees (Wagoner & Waldron, 1999), and in non-interpersonal settings such as radio (Galician & Vestre, 1987) and television newscasts (N ewhagen & Reeves, 1992; Veitch & Griffitt, 1976). The published articles that have defined bad news come mainly from the medical domain. For example, Fallowfield and Jenkins (2004) consider bad news to be “any information that produces a negative alteration to a person’s expectations about their 4 present and future...” (p. 312). Other authors consider news to be bad if it results in persistent cognitive, behavioral, or emotional deficits in the receiver (Ptacek & Eberhardt, 1996) Obj ectively defining bad news is problematic because what counts as bad news for one person might not be perceived as bad news by another (Eberhardt-McKee & Ptacek, 2001; Eggly et al., 2006; Fallowfield & Jenkins, 2004; Ptacek & Eberhardt, 1996). The Fallowfield and Jenkins definition seems to capture the lay understanding of bad news in the sense that people do tend to continue labeling a piece of information “news” even when that information is technically no longer novel (e. g., “I received some bad news earlier this week.”). However, this definition might be overly inclusive in the sense that intrapersonally derived information (e.g., a frightening dream resulting in a negative memory) becomes indistinguishable from news that is received from an outside source. Ptacek and Eberhardt’s (1996) definition comes closer to reserving the label bad news only for the negative information as it is being conveyed for the first time. Also, this definition reflects the view that the news must be transmitted socially (Eberhardt- McKee & Ptacek, 2001). The current paper builds on the previous definitions to conceptually define bad news as a message communicating information that is previously unknown to the receiver, is anticipated to be personally relevant to the receiver, and is perceived by the delivery agent to be negatively valenced by the receiver. Several characteristics follow from this definition and apply to the types of bad news messages studied thus far. These deal with the message’s valence, novelty, extremity, personal relevance, and links to perceived outcomes. The remainder of this section discusses these characteristics in greater detail as they set bad news apart from other types of messages. 5 The expression “bad news,” as generally used, describes a particular type of message. Individuals subjectively interpret these messages (Eberhardt-McKee & Ptacek, 2001; Eggly et al., 2006; Fallowfield & Jenkins, 2004; Ptacek & Eberhardt, 1996). Thus, the message itself and its subsequent interpretation by individuals serve as the basic units of analysis. A number of studies have examined bad news at the individual message level (e. g., Beach, 2002; Dibble & Gabriel, 2004; Freese & Maynard, 1998; Wagoner & Waldron, 1999). However, with the exception of Beach and Freese and Maynard, these studies also relied on self-report data given by deliverers alone. Because every bad news message must have a recipient, it also seems useful to examine the process of bad news delivery and interpretation at the transactional level, taking into account the sender and the receiver. The current studies examine feedback delivery in the context of a live interaction. “Bad news” is a compound expression. If the two words making up this compound term are taken separately, the first word refers to the message’s valence, the second its novelty. Heath (1996) describes valence as whether the message represents desirable or undesirable outcomes. Bad news is negatively valenced because, by definition, it indicates undesirable outcomes. Ptacek and Eberhardt (1996) characterize these undesirable outcomes as cognitive, behavioral, or emotional deficits. Moreover, if a message conveys undesirable outcomes, it is likely to evoke negative affect in those perceiving the undesirable outcomes. Heath also notes that certain messages can be more negatively valenced than others. This refers to the message’s extremity, and it essentially represents a descriptive attribute of the valence along a continuum. In other words, all bad news messages convey negatively valenced information, but some information is 6 perceived to be more negative (i.e., more extreme) than other information. As Fallowfield and Jenkins (2004) put it, there are gradations of bad news. For example, participants in Heath’s study found news of 1,100 muggings each year to be more negative than news reporting 10 muggings each year. Similarly, the news of a rape-murder is arguably more negative than learning that a friend is canceling a lunch date. The second half of the compound expression, “news,” indicates that the information contained within the message is indeed novel (i.e., that the information is previously unknown to the receiver). This report defines bad news as signaling undesirable outcomes. These outcomes can be real, but the mere perception of undesirable outcomes is sufficient to judge a message as bad news. In Bond and Anderson’s (1987) study, the perceived undesirable outcome was low intelligence, and this was indicated by a failing score on an intelligence test. Participants in Johnson et al.’s (1974) study had to relay bad news that involved the perceived receipt of an electric shock. Rosen and Tesser (1972) manipulated their participants to perceive the news of threatened harm toward an experimental confederate. Each of these investigators was able to instantiate a judgment of bad news in the participants, even though the undesirable outcomes were not real. Another dimension of bad news is its degree of personal relevance. Rosen and Tesser (1972) use the term involvement interchangeably with relevance. The amount of relevance can vary (Heath, 1996). For example, Veitch and Griffitt (1976) operationalized bad news in the form of a news story about the forecast of substantial increases in local food prices. An average television viewer may rightly view rising grocery prices in a different state as bad news, but this information would be less personally relevant than, say, bad news tolling the rise of grocery prices more locally (e. g., in the viewer’s home state). The following example clarifies the difference between extremity and relevance. Most people would agree that the news of a mother’s stillborn baby is extreme in nature (Fallowfield & Jenkins, 2004). Given that most people—even people not related to the baby—would view this information as extremely negative, this news will be more relevant to the actual mother who lost her baby than to the average watcher of the nightly news. Thus, relevance and extremity are distinct attributes of bad news messages. In general, the negative impact of a message should increase concomitantly with increases in either extremity or relevance (although not necessarily to equal degrees). Having briefly defined and discussed some of the dimensions that appear to characterize bad news, this paper now turns to introduce the second concept of importance, relational closeness. Relational Closeness Often equated with intimacy (Floyd & Parks, 1995; Wood & Inrnan, 1993) and characterized in lay usage as an overall emotional bond, the level of relational closeness is marked by the degree of mutual dependence between two people, the frequency of their impact on one another, and the strength of impact per occurrence. Mutual dependence manifests itself through a variety of physical and psychological attributes. Thus, behaving close and feeling close are distinct indicators of closeness (Aron, Aron, & Smollan, 1992). Moreover, defining closeness as occurring between two people indicates that closeness transcends relationship types. For example, persons in same-sex friendships, romantic relationships, and parent-child relationships experience closeness (Afifi & 8 Schrodt, 2003; Aron, Aron, Tudor, & Nelson, 1991; Berscheid, Snyder, & Omoto, 1989). Closeness is also conceptualized as a continuous concept as opposed to an all-or-nothing phenomenon. It occurs in gradations (Aron & Fraley, 1999). Finally, it should be mentioned that this definition accommodates the view that closeness is multiply determined. That is, variables predicting closeness for one relationship do not always predict closeness for another relationship, and several different variables can produce closeness. Researchers have attempted to link some of the main variables to the particular types of relationships for which those variables predict closeness (e. g., Floyd & Parks, 1995; Wood & Inman, 1993; Yaughn & Nowicki, Jr., 1999). Yet, Maxwell (1985) concludes that mutual dependence might be the most useful indicator of closeness because, especially in long-standing relationships, the upper limits of plausible predictors such as liking, attraction, and commitment may have already been reached. Expanding further on the idea of closeness as mutual dependence, Aron and colleagues (e. g., Aron, Aron, & Smollan, 1992; Aron, Aron, Tudor, et al., 1991; Aron & Fraley, 1999) capture this idea well by viewing closeness as including other in the self. According to this perspective, “in a close relationship the individual acts as if some or all aspects of the partner are partially the individual’s own” (Aron, Aron, & Smollan, 1992, p. 598). This suggests a blurring of the boundaries between those aspects that characterize each individual (Aron, Aron, Tudor, et al., 1991 ). Emotional Congruence A final concept bearing on the explanations for the MUM effect is the ability for individuals to adopt the emotions of others. Often this is considered to be the result of empathic processes (Doherty, 1997). Cognitive models of empathy assert that people 9 listen to the description of another’s emotional experience, and that those descriptions trigger memories of similar experiences which generates an emotional response similar to that of the other. On the other hand, Hatfield and her colleagues (e. g., Hatfield, Cacioppo, & Rapson, 1992) argue that emotional congruence can and does occur without conscious cognitive effort. She describes emotional contagion as the primitive tendency of one person to “catch” another person’s emotions. Essentially, emotional contagion stems from people’s unconscious and automatic tendency to mimic the movements, expressions, postures, and vocalizations of another person. This tendency to mimic the expressions of another is theorized to serve as a heuristic by which people gauge the feelings of others. Thus, by behaving as another does, one comes to feel as that person does. Regardless of whether active cognitive empathy or primitive emotional contagion provides the mechanism for the way people come to feel as others feel, several studies document emotional congruence as occurring across a variety of relationship types (e.g., Gump & Kulik, 1997; Howes, Hokanson, & Loewenstein, 1985). Scheflen (1964) contends that people are particularly likely to mimic those they care for. Scheflen’s contention makes sense. People should be most concerned with the emotions of their significant others because significant others, by definition, have the greatest impact on people’s lives. Further, the most coordinated mimicry will occur between persons who like and/or love one another, and that emotional contagion should be stronger in situations where people construe themselves in terms of their interrelatedness to each other than if they view themselves more as independent or unique (Hatfield et al., 1992). Research is consistent with this View, finding that emotional 10 congruence effects become stronger as individuals grow closer (e. g., Anderson, Keltner, & John, 2003; Kenny & Kashy, 1994). In this chapter bad news has been defined and explicated; relational closeness has been conceptualized as a degree of mutual dependence, not limited to romantic relationships; and both concepts appear to be linked to a certain degree of emotional similarity between relational partners. These concepts inform the propositions to be tested, and they are treated in the chapters that follow. 11 CHAPTER 3: THE MUM EFFECT AND RELATIONSHIP CLOSENESS The first study seeks to determine which of two explanations for the MUM effect is more consistent with the tendency of an individual to hesitate to share bad news with a close friend. These explanations are described below in greater detail, and they essentially reduce to concerns about self versus concerns primarily about the other. Arguments are made below for their mutually exclusive predictions. According to the first explanation, individuals are reluctant to transmit bad news because they empathize with the receiver and do not wish to evoke negative affect in the receiver. Tesser and Rosen (1975) characterize this as the “recipient’s emotionality hypothesis” (p. 213). Essentially, this hypothesis implies that people assume good news has a positive effect on the other’s affect and bad news has a negative effect on the other’s affect. Senders prefer not to put receivers in a negative affective state; thus, they are less likely to communicate bad news than good news. A sender’s desire to avoid inducing a state of negative affect in a receiver may relate to the sender’s wish to avoid experiencing negative affect himself or herself (Bond & Anderson, 1987). For example, senders might feel bad for the receiver’s situation or empathize with the receiver‘s distress. Regardless of the actual emotion felt by the sender, this explanation holds that people feel badly about causing negative affect in others, and consequently, people hesitate to share bad news. On the other hand, individuals could be more concerned about casting themselves in a negative light. This is the concern underpinning the lay expression “Don’t shoot the messenger.” According to this second explanation, termed the “fear of negative evaluation” hypothesis by Rosen and Tesser (1975, p. 203), individuals are primarily 12 motivated to maintain their own face and self-image, and that they hesitate to share bad news in order to avoid being evaluated negatively by the receiver (Bond & Anderson, 1987; Tesser & Rosen, 1975). Goffrnan (1967) defines face “. . .as the positive social value a person effectively claims for himself by the line others assume he has taken during a particular contact” (p. 5). Society expects people to be considerate of others’ faces. As Goffman puts it, “The person who can witness another’s humiliation and unfeelingly retain a cool countenance of himself is said in our society to be heartless” (p. 11). Bad news is face-threatening to the receiver. Thus, by minimizing face threats to others (e.g., withholding bad news) one avoids being viewed as heartless. In the end, “lest they seem blithe to others’ misfortune, lest they seem callous and cruel, people keep mum” (Bond & Anderson, p. 177).I Several studies have tested these explanations (Bond & Anderson, 1987; Johnson et al., 1974; Rosen & Tesser, 1972), and their data, for the most part, were consistent with the negative evaluation explanation. Bond and Anderson’s study illustrates the typical method of testing both explanations. Participants in their experiment gave feedback to a confederate receiver who either “excelled at” or “failed” what was assumed to be an intelligence test. Hence, this feedback constituted the news valence induction (i.e., good or bad news). News valence was also crossed with participant visibility in that participants were told that either the test-taker could see the participant, or that the participant was invisible to the test-taker throughout the entire experiment. Bond and Anderson found that the participants who delivered failure feedback and were visible to the test-taker took the longest to give feedback, as compared to the other three cells, and the remaining three cell means did not significantly differ from each other. Bond and 13 Anderson concluded that self presentation concerns motivated the hesitation to transmit bad news because, if concern for inducing negative affect in the receiver were truly motivating reluctance, participants who were delivering failure feedback should have hesitated approximately equally, whether or not they were visible to the test-taker. However, Bond and Anderson (1987) and the other cited studies that tested both explanations for the MUM effect did so using participants who were strangers to one another. Consequently, it is not surprising that the negative evaluation explanation prevailed. As Johnson et a1. (1974) note, situational factors may influence the particular motive that becomes salient. For example, Johnson et a1. reasoned that if the negative evaluation hypothesis holds, a sender might be more fearful of a negative evaluation from an attractive recipient than from an unattractive recipient; therefore, he or she would be less inclined to share bad news with the attractive recipient. Yet, Rosen, Johnson, Johnson, and Tesser (1973, cited in Johnson et al., 1974) found that senders relayed more bad news to an attractive recipient. Further, Bond and Anderson crossed news valence with whether the receiver was visible or invisible to the sender. If senders and receivers cannot see each other, emotional congruence cannot occur as strongly. Taken together, these issues imply that it would be premature to cast blanket rejection upon the recipient emotionality hypothesis, having not investigated the hypotheses in other contexts. The current study is designed such that senders and receivers are visible to one another during the news-sharing interaction, which should provide the opportunity for empathy and/or emotional contagion to operate. Moreover, this study adds a fiiends condition to compare with the strangers condition. 14 The research hypothesis in the first study simply predicts a replication of the MUM effect, observed originally by Rosen and Tesser (1970). Given the robustness of this finding in previous research, the following hypothesis is forwarded. H1: Time to response will be greater for those who share bad news than for those who share good news. This investigation is primarily concerned with whether bad news will be shared less readily with a fiiend or with a stranger. Both patterns of results are plausible, depending on whether the reluctance to share bad news is related to negative evaluation concerns or to concerns about hurting the other. First consider the negative evaluation hypothesis with respect to close relationships. This hypothesis deals with negative self impressions, arguing that self-presentation concerns motivate people to keep quiet about bad news. If closeness is conceptualized and defined as including other in the self, an individual’s self-presentation concerns should decrease as closeness increases because the self-other distinction is less salient (Schlenker, 1984). If self-presentation concerns decrease as individuals grow closer to one another, there should be a corresponding reduction in motivation to keep quiet about sharing bad news when the reluctance is attributable exclusively to self-presentation concerns, and a person should share bad news more often (and more quickly) with a close fiiend than with a stranger. Moreover, it seems that closeness would also compel more honesty, which should also be consistent with quicker disclosure of the bad news. If, on the other hand, the reluctance to share bad news is a function of concern for the recipient’s emotions, one would expect an individual to be less likely to transmit bad news to a close friend than to a stranger. Beach (2002) analyzed a series of telephone 15 conversations between a father and his son, as the father revealed the news of mother’s cancer to the son for the first time. This is a closer relationship than between strangers, and Beach noted the father’s hesitation and internal discomfort with actually relaying the news to the son, detailing the coping strategies the father employed in order to remain focused on his duty to inform the son. Beach’s study might typify a context in which concern for the son’s emotions is the motivation for the reluctance to transmit bad news. Recall that bad news, by definition, evokes negative affect in the receiver. Given that negative affect is evoked in the receiver, emotional congruence suggests that the sender, who is in a close relationship with the receiver, might come to experience a similar emotional state (Tesser & Rosen, 1975). Essentially, the sender would be creating in himself or herself a state of negative affect. Negative affect, by definition, is aversive. Thus, an individual, wishing to avoid negative affect, should be more reluctant to share bad news with a close friend than with a stranger.2 Still, it is plausible that a sender could be simultaneously concerned for oneself, as well as the receiver. In this case, it is difficult to predict the direction of any difference in reluctance when sharing with a stranger versus a close fiiend. Because both explanations are plausible, the following research question is forwarded. RQl: Do people wait longer to share bad news with a stranger or with a close friend? Method Under the guise of a study on “communication in relationships,” female participants delivered feedback to either a fiiend or a stranger who they believed either succeeded or failed on a “social perception test.” The overall design strategy is similar to 16 that used by Bond and Anderson (1987). It should be recalled, however, that Bond and Anderson’s participants consisted of stranger dyads. The current study addressed the boundary conditions for each explanation for the MUM effect within the context of female friendships. Participants Female undergraduate students (N = 330) from a large Midwestern university served as participants. The decision to use only female participants stemmed from the desire to eliminate any possible confounding effects of sex. Participants were predominantly European American, and ranged in age from 17 to 31 years, with an average age of 19.38 years (SD = 1.43). Participants indicated that they had known the person with whom they reported to the laboratory for an average of 3 years and 9 months (SD = 52.59 months). The lengths of these friendships ranged from 1 month to 21 years and 2 months, with 68 participants reporting a fiiendship of 6 months or less. Preliminary analyses on MUM effect research indicated that effect sizes are medium in magnitude (r averages .3 0-40). A priori power analysis revealed that, in order to achieve a power of at least .80, it was necessary to recruit approximately 100 dyads (Cohen, Cohen, West, & Aiken, 2003). Design This experiment utilized a 2 X 2 independent groups design wherein two levels of news valence (good or bad) were fully crossed with two levels of closeness (stranger or close fiiend). The basic procedural paradigm involved one member of a dyad (the “scorer”) scoring and relaying performance feedback on a “test” taken by the scorer’s partner (the “test-taker”). The test’s score was manipulated (without either partner being 17 made aware of the manipulation) to reflect superior performance (good news) or poor performance (bad news). The primary dependent measure was reluctance to transmit feedback, operationalized as time to response (Ahmed & Bi gelow, 1993). Time to response was defined as the number of seconds from the moment the scorer received the score sheet until her first verbal utterance to the test-taker after having entered the test- taker’s room. Procedures Participants were instructed to bring along a close fiiend of at least 6 months. Thus, participants arrived at the laboratory in pairs. Two pairs of friends reported at one time, and each participant provided informed consent. Random assignment was used to determine all experimental conditions. Those assigned to the friends condition underwent all experimental procedures with the friend they brought to the lab. Those assigned to the stranger condition were partnered with a member of the other pair who reported to the laboratory at the same time. Once the pairings had been created, participants were made aware of their partners, and all four participants then completed an initial packet of questionnaires. A privacy screen was used to prevent participants from viewing each other’s questionnaire responses. This packet included pretest measures of mood, closeness, liking, and general empathic concern. After participants completed the initial questionnaires, the within-dyad roles (i.e., scorer or test-taker) were assigned by coin toss. Scorers would provide performance feedback to the test-takers. After the dyads learned their roles, they were told about the “social perception test.” As the induction of news valence, this test had no real purpose other than to provide 18 a plausible task about which scorers could provide feedback to the test-takers. Participants were told that the social perception test assesses one’s ability to read the nonverbal signals of others. Moreover, all participants were told that those who score high on this test were excellent at reading nonverbal cues, and that those who scored low might be socially awkward. The social perception test consisted of twenty pairs of photographed faces. For each pair of faces, test-takers were asked to determine whether (a) the two people pictured were romantic partners, (b) the two people pictured were platonic fiiends, or (c) the two people pictured were strangers to one another. Scorers were told that they would be scoring their partner’s tests once the test-taker was finished taking the test. Although two dyads reported for each hour-long session, space limitations only allowed one dyad to be tested at a time. A coin toss determined which dyad would go first. The dyad that went first remained in the experimental rooms while the other dyad was led to a waiting area, out of earshot from the experimental rooms. The members of all of the second dyads were kept separated from each other to prevent the strangers from getting to know one another. After leading the second dyad to their waiting areas, the experimenter returned to administer to the first dyad the social perception test activity. The test-taker was taken to a nearby testing room across a small hallway while the scorer remained in the room where the initial questionnaires were completed. From their respective rooms, test-takers and scorers began their respective tasks. Test-takers completed the bogus social perception test and recorded their answers on Scantron-style bubble sheets. Using bubble sheets provided a response recording format that minimized the opportunity for scorers to 19 recognize their partner’s handwriting (which would compromise the experimental manipulation). As the test-taker was working on her test, the scorer reviewed instructions for scoring the test and providing feedback to the test-taker. These instructions came in the form of one printed page, and they included an answer key, a percentage guide sheet to assist in calculating the percent correct, bogus scoring norms (indicating that the average score was 70%), and instructions regarding what to do once she was finished scoring her partner’s test. Both the test-taker and the scorer were left alone while they undertook their respective assignments. Test-takers were not told in advance that they would be receiving their performance feedback from their partners. The decision to surprise test-takers with their partners’ feedback was made in order to boost the perceived importance of the task and hopefirlly strengthen the news valence induction. Upon completing her social perception test, test-takers informed the experimenter. The experimenter appeared in the test-taker’s room, collected the test-taker’s answer sheet, and said, “Thank you. I will now get your results. Please make yourself comfortable while you wait.” The experimenter then left the room with the test-taker’s real answer sheet. After the experimenter exited the test-taker’s room, but before he arrived at the test scorer’s room (i.e., while in the small hallway), the experimenter replaced the test-taker’s actual score sheet with a phony score sheet that had been pre- filled to indicate either 6 out of 20 correct responses (3 0%), or 17 out of 20 correct responses (85%). This manipulation of the scoring sheets served as the induction of the independent variable news valence. It was anticipated that a 30% score would be 20 perceived as bad news, and a score of 85%, by comparison, would be viewed as good news. When the experimenter arrived at the scorer’s room with the phony scoresheet, he said to the scorer, “I have your partner’s scoresheet right here. Do you have any questions about how to score the test before you begin scoring?” After any procedural questions were addressed, the scorer was given either a 30% scoresheet or an 85% scoresheet, according to the experimental condition assigned. As mentioned before, scorers were provided with a sheet of paper on which was printed a phony normative index of results with which scorers could interpret their partner’s results. This index mentioned that results on this particular test hold implications for the test-taker’s social skills. The index reported that scores above 70% tend to indicate high nonverbal perceptual ability, and lower scores indicate that a person might be socially awkward. Just like the social perception test, this normative index was equally bogus and served merely to enhance the importance of the task to the participants by getting them to hopefully attach potentially real and personal relational consequences. Scorers were left alone to score the test as instructed, and to provide performance feedback to the test-taker when scorers were finished scoring the test. (Note: The point at which the scorer received the phony scoresheet marked the start of the timer which measures time to response.) The experimenter closed the scorer’s door and sat in a chair in the hallway just outside the test-taker’s room. From this chair he could observe both the scorer’s exit from her room and her interaction with the test-taker in the test-taker’s room. The experimenter was fully visible to the scorer as she entered the test-taker’s room to deliver the feedback. 21 When the time came for the scorer to relay the result, scorers walked to the test- taker’s room and were given the opportunity to share the result. No specific instructions were given as to what the scorer should say to the test-taker. Participants were allowed to freely interact and say anything they wished. Moreover, participants were given as much time as they needed. This interaction was videotaped for later coding of communication behavioral indicators of reluctance (the results of which are presented in Chapter 4). For purposes of measuring time to response, the moment at which the scorer made her first verbal utterance to the test-taker served as the endpoint for the stopwatch. The scorer then returned to her original scoring room, and both dyad members completed a second packet of questionnaires from their separate rooms. These questionnaires included posttest measures of mood (positive affect, negative affect), closeness, and liking, as well as manipulation check items designed to assess the news valence induction. Once both dyad members completed their posttest questionnaires, they were led to the waiting area while this process was repeated for the second dyad. After the second dyad was finished, all participants were fully debriefed, thanked for their participation, asked not to discuss this study with anyone, and then dismissed. Measures Closeness. The primary independent variables of interest in this study are relational closeness and news valence, and the dependent variable is time to response. Although fiiend-stranger condition was experimentally varied, a measure of closeness was given to serve as an induction check. The items used to assess closeness originated from a variety of sources. The Inclusion of Other in the Self scale is a one-item series of seven Venn-like diagrams (Aron et al., 1992). Respondents circled the pair of 22 overlapping circles that best described their relationship. Items adapted fi'om the Close Relationships Questionnaire (Maxwell, 1985) included, “My relationship with my partner 99 6‘ is close, When we are apart, I miss my partner a great deal,” and “My partner and I disclose important and personal things to each other.” Two items, “My partner and I want to spend time together,” and “I’m sure of my relationship with my partner,” were taken from Walker and Thompson (1983). The items, “When I have free time I choose to spend it alone with my partner,” and “My relationship with my partner is important in my life,” were taken from Miller and Lefcourt’s (1982) Social Intimacy Scale. One item, “I consider my partner when making important decisions in life,” was adapted from Berscheid, Snyder, and Omoto’s (1989) Relationship Closeness Inventory. The remaining items, “My partner and I have a strong connection,” “My partner is a priority in my life,” “My partner and I do a lot of things together,” and “I think about my partner a lot,” were developed by the authors. Each item featured a 7-step Likert-type response format (1 = Strongly Disagree, 7 = Strongly Agree). Cronbach’s alpha for these items taken together was .99 (n = 323). News Valence. News valence was measured as an induction check of whether participants viewed an 85% score or a 30% score as good news or bad news, respectively. Similar to the induction check conducted by Bond and Anderson (1987), scorers were asked to respond to items such as, “The test-taker performed well,” and, “The test-taker performed poorly.” News valence was assessed from the test-taker’s perspective as well with items of the type, “My feedback indicated that I performed well,” and, “My feedback indicated that I performed poorly.” Respondents indicated their response using a 6-step, Likert-type scale (1 = Strongly Disagree, 6 = Strongly Agree). Items were 23 reverse coded where appropriate so that numbers 4 and greater reflect perceived good news. It was hoped that the exclusion of a scale midpoint would alleviate the need to determine whether a midpoint score constitutes good news or bad news. These items were highly intercorrelated for both test-takers and scorers (r range 81-85), thus, they were collapsed and averaged as a single factor. Reliabilities were high for both test-takers (a = .96, n = 165) as well as for scorers (a = .96, n = 165). Empathic Concern. Six items originally inspired by Davis (1983) and utilized by Stiff, Dillard, Somera, Kim, and Sleight (1988) tapped participants’ general levels of concern and regard for the welfare of others. Although the empathic concern of test- takers was assessed, only that of the scorers is conceptually related to this study. Consequently, the demonstrated reliability of this scale results from including scorers responses only (a = .74, n = 164). Sample items included, “When I see someone being taken advantage of, I feel kind of protective toward them,” “I would describe myself as a pretty soft-hearted person,” and, “I often have tender, concerned feelings for people less fortunate than me.” Participants responded using a 7-step Likert-type scale anchored by 1 (strongly disagree) and 7 (strongly agree), where higher numbers were taken to mean higher levels of empathic concern. Time to Response. Time to response was operationalized as the time in seconds from the point the scorer received the phony answer sheet to the instant she began making verbal utterances to the test-taker after having already entered the test-taker’s room. Time was manually kept by the experimenter using a handheld stopwatch. The two experimental rooms were situated such that the experimenter was able to view both the scorer’s exit from her own room, as well as her interaction with the test-taker in the test- 24 taker’s room, and thus was able to start and stop the watch accordingly. These behavioral cues were adequately overt and objective so as to ensure more precise chronograph start and stop points. Descriptive analysis revealed the distribution of this variable to be slightly right-skewed, but for the most part to approximate a normal distribution in shape. Results The induction check analyses were conducted using the individual as the unit of analysis. Those analyses that specifically dealt with the hypotheses and research question utilized the scorers only because time to response is a completely between dyads variable (i.e., within a dyad both scorer and test taker have the same time to response, but time to response varies from dyad to dyad). Induction Checks Checks ascertained whether or not experimental differences had been induced with respect to the closeness and news valence independent variables. An independent samples t-test that assumed non-equal variances revealed a statistically significant difference between the fiiends group (n = 164, M = 5.18, SD = 1.25) and the strangers group (n = 159, M: 1.18, SD = 0.40), t(196.76) = 38.94,p < .01, r = .94. Participants assigned to the fiiends condition indeed reported feeling closer to their partners than did participants in the strangers condition. Means, standard deviations, t-test results, and effect sizes for the news valence induction check indicators can be viewed in Table 1. Significant differences were found between the 85% group (Test-takers M: 5.56, SD = 0.53; Scorers M = 5.79, SD = 0.36) and the 30% group (Test-takers M = 2.17, SD = 0.87; Scorers M= 2.29, SD = 0.75) for both test-takers as well as scorers. Compared to 30% score, the 85% score was clearly 25 seen as better by both test-takers [t(l46.58) = 30.47, p < .01, r = .92] and scorers [t(128.70) = 39.23,p < .01, r = .95]. To check the induction of news valence further, one-sample t-tests were utilized to compare scorer’s news valence ratings to the scale midpoint (3.5 was used as the test value because the respondents scaled on a 1-6 metric). Scorers delivering 85% scores (M = 5.79, SD = 0.36, n = 77) were significantly above the midpoint [t(76) = 56.12, p < .01 , r = .99], and those delivering 30% scores (M = 2.29, SD = 0.75, n = 88) reported valences that fell significantly below the midpoint [n = 88, t(87) = -15.28,p < .01, r = .85]. Within the fiiends condition alone, one-sample t-tests comparing both 85%- sharing fiiends and 30%-sharing fiiends with the scale midpoint (3.5) were also significant. Scorers relaying 85% feedback to a fiiend (M = 5.90, SD = 0.21, n = 39) rated the test score as significantly more positive than the midpoint [t(3 8) = 71.96, p < .01, r = .99], and those sharing 30% scores to a fiiend (M = 2.42, SD = 0.72, n = 46) reported news valences that were significantly more negative than the midpoint [t(45) = -10.18, p < .01, r = .83]. Table 1. Means, Standard Deviations, Significance Tests, and Effect Sizes for News Valence Induction Check Indicators as a Function of Feedback Condition. Good News Bad News Variable M SD M SD t(df) r T’s perceptions of news valence 5.56 .53 2.17 .87 30.47(l46.58)* .92 S’s perceptions of news valence 5.79 .36 2.29 .75 39.23(128.70)* .95 Notes. T = Taker, S = Scorer. Tests assumed non-equal variances. *p < .01. Tests of Hypothesis and Research Question A 2 X 2 factorial analysis of variance, using closeness condition and news valence as independent variables, was conducted on scorers’ response times. Time to response means and standard deviations for the four ANOVA cell combinations are reported in Table 2. H1 predicted that time to response would be greater for scorers delivering 30% feedback (bad news) than for those delivering 85% feedback (good news). This analysis revealed a statistically significant and substantial main effect for news valence, F (1, 161) = 44.76, p < .01, n2 = .22, r = .47. As predicted, scorers who delivered 30% feedback (M = 136.64 seconds, SD = 41.51) took longer to deliver that feedback than did scorers who delivered 85% feedback (M = 100.16 seconds, SD = 25.82). Thus, the MUM effect seems 27 to have been replicated in this study. No significant main effect for closeness condition was found, F(1, 161) = 2.06, p = .15, n2 = .01, nor was there a significant interaction between news valence and closeness condition, F(1, 161) = 1.12, p = .29, n2 = .01. Table 2. Mean Times to Response (in Seconds) As a Function of News Valence and Fri ends/S tran gers. Bad News Good News Marginal mean Friend 129.85 99.13 115.75 (40.37) (27.68) (38.17) Stranger 143.43 101.18 123.36 (42.05) (24.09) (40.50) Marginal mean 136.64 100.16 G = 119.44 (41.51) (25.82) (39.38) Notes. N = 165 dyads. Standard deviations are reported in parentheses under cell means. G = Grand mean. RQl asked whether scorers who delivered bad news would differentially wait before delivering their feedback if the test taker was a friend or a stranger. A pairwise comparison was made between the bad news-stranger and the bad news-friend cell means. Scorers were slower by an average of 13.58 seconds to deliver bad news to a stranger than to a fiiend, but this difference was not statistically significant, t(86) = -1.55, p= .13, r= .16. 28 Other Analyses Although no a priori hypothesis was made, a post hoc correlation was calculated between scorers’ empathic perception ratings and their times to response. A significant and positive correlation might be taken as preliminary evidence of a potential relationship between a scorer’s concern for the taker and her reluctance to deliver feedback. Across all scorers, this correlation was statistically significant and positive, but small, r(162) = .16, p < .05. The more empathic a scorer was in general, the greater was her delay in delivering feedback to the test-taker. The correlation between these same two variables was calculated independently for scorers who delivered bad news to friends and for scorers who delivered bad news to strangers to address the post hoc question of whether empathy and response time are linked more strongly when scorers delivered bad news to a friend as opposed to a stranger. When scorers shared bad news with a fiiend, this correlation was even larger and significant, r(44) = .31, p < .05. When scorers delivered bad news to a stranger, this correlation failed to reach significance, r(40) = .17, p = .30. It might appear that empathy plays a larger role when the feedback is given to a friend, but the difference between these two correlations was not statistically significant, Z = 0.67, p = .75. Discussion The purpose of this first study was to further understanding of the interpersonal delivery of bad news. The specific focus was on the nature of a sender’s reluctance to transmit the bad news to a recipient, depending on whether this news was shared with a close fiiend or a stranger. 29 A hypothesis and a research question addressed these issues. H1 predicted that scorers who were about to tell test-takers that takers had earned 30% (and were thereby implied to be socially awkward) would delay transmission of the feedback longer than would scorers who told takers that takers had earned 85% (and were ostensibly skilled at reading nonverbal cues). This was a predicted replication of the MUM effect (Rosen & Tesser, 1972), and the data were consistent with this prediction. Scorers who had to deliver feedback to a taker that “scored” 30% indeed took an average of 36 seconds longer to share the feedback than did scorers who relayed feedback to a taker who “scored” 85%. RQl concerned the nature of the reluctance to transmit bad news and any differential transmission to a friend or stranger. If scorers took longer to deliver bad news feedback to strangers than to friends, it might indicate that scorers were primarily concerned about their self-presentation. If scorers shared feedback more rapidly with strangers than with their close fiiends, this might indicate that scorers were more concerned about hurting the feelings of the taker. Scorers were about 14 seconds slower to deliver 30% feedback to strangers than to friends. Although this difference was not statistically significant, it is in the direction one would expect if a sender’s primary concern was maintaining one‘s own face. Thus, the mean difference alone might be taken as most consistent with the negative evaluation explanation. However, it is still reasonable to expect that a sender’s reluctance can be driven by the desire to avoid hurting the receiver. A failure to demonstrate a link between empathic concern and time to response would certainly cast doubt on the recipient‘s emotionality explanation. This study revealed a statistically significant correlation 30 between a scorer’s general level of empathic concern and how long that scorer delayed sharing the news. Admittedly, this result cannot be taken as direct evidence of concern for the receiver’s emotions because a correlation says nothing of causation, and this correlation might even be spurious. Because the link between empathic concern and time to response was not falsified, however, the possibility remains open that, in certain instances, senders hesitate to share bad news with receivers because senders are afraid to hurt the receivers’ feelings. A number of limitations might have influenced the outcomes observed. First, concern may be raised as to whether or not bad news was actually instantiated in the mind of the scorers. Across each of the induction check comparisons, compared to an 85% score, a score of 30% was clearly seen as worse. Moreover, if respondents can be said to have taken the scale response options of 3-4 to indicate some sort of valence inflection region, then the one-sample t-tests also suggest that those delivering 85% feedback perceived themselves to be delivering good news, and scorers bearing 30% scores perceived themselves to deliver some sort of bad news. Further, this pattern of results held even within the friends condition alone. The induction of bad news here seemed sufficient for purposes of testing the general MUM effect. It should also be recalled that the means used in the induction checks resulted from news valence measurement stems that were worded in such a way that the scorer was inferring how the test-taker would view the news. Therefore, based on private self reports, it would appear that scorers perceived themselves to be delivering good or bad news even when the target was a fiiend. Nonetheless, developing a bad news induction that raises the stakes by boosting other attributes of the news (e. g., relevance, extremity) 31 should be preferable and might provide a better context in which the recipient’s emotionality motive might operate. Procedural decisions may have contributed to a less than optimal induction of bad news. For example, the bogus social perception test featured photographs of 40 individuals, artificially paired, but in a way that the individuals paired together were not in the same photograph. Test-takers audibly questioned their own ability to perform well, raising comments like, “How am I supposed to be able to read their nonverbals if they’re not even in the same picture?” Consequently, it is possible that many test-takers considered themselves lucky to have scored even a 30%, and consequently did not perceive their feedback to constitute bad news. This study was mainly interested in how the scorer perceived the news she was sharing. To avoid arousing a similar perception in scorers, the decision was made to not show the scorer the actual social perception test packet. It was hoped that scorers would base their valence judgments on the manipulation information given (e. g., scoring norms and meaning given to certain scores). The testing task might have been more plausible had the test items pictured two people in the same photograph. Another issue might be raised with respect to the size of the difference between the bad news group and the good news group. In order to provide uniform start and stop points for the chronograph the decision was made to leave the watch running while the scorer was actually scoring the Scantron. Thus, scoring time is necessarily bound up with any reluctance or “contemplation time.” Admittedly, this arrangement is not optimal for capturing reluctance and nothing but reluctance. However, the task of scoring every Scantron consisted of comparing each bubbled-in item to an answer key. Even though 32 scorers in the bad news condition were indeed marking more items as “incorrect,” there is no reason to suspect that it takes more than 30 seconds to place check marks next to a few more “incorrect” items. Thus, although it would be helpful to devise a way to uniformly start the chronograph after the scorer has learned the test-taker’s “result” but before she began contemplating her response to the test-taker, a more than 30-second difference between good news and bad news groups still seems robust enough to conclude that a MUM effect is occurring. In short, further evidence for the MUM effect seems to have obtained. However, in the experimental scenario carried out, the closeness of the participants did not appear to shed light on the matter of whether any scorer reluctance experienced was due to self- presentation concerns or to concerns about bringing about negative emotions in the test- taker. The nature of the dependent variable (i.e., time) may have contributed largely to these results. Therefore, the next study proposes to test the same propositions utilizing communication behaviors as dependent variables. 33 CHAPTER 4: THE MUM EFFECT AND COMMUNICATION BEHAVIOR Consistent with prior MUM effect research, the data fiom Chapter 3 demonstrated that individuals will delay the transmission of negative feedback longer than when delivering positive feedback. Whether the target was a fiiend or stranger, however, did not appear to impact this delay to any important degree. Perhaps the closeness between the sender and target impacts the bad news delivery process in other ways. The search for insight with regard to this issue is one of the goals for this second study. What follows is an exploration of the communicative differences that may or may not exist between deliverers of good news and deliverers of bad news. Here, verbal communication differences provide the primary focus for study. At least two sources bolster the expectation that individuals will vary their communication behavior according to the valence of the information they expect to share. First, Mehrabian’s (1971) immediacy principle asserts that people are drawn toward things they like or prefer, and they distance themselves from things they dislike or disprefer. Essentially, people enact immediacy behaviors in the presence of things about which they feel positive, and they behave in nonimmediate ways in situations of discomfort or anxiety (e.g., when delivering bad news). Because certain communication environments preclude participants from physically carrying their bodies toward or away from the target, people will often use abbreviated forms of approach or avoidance congruent with their evaluation of the situation. These abbreviated approach or avoidance signals are routinely manifested through verbal and nonverbal behavior. Second, conversation analysts have documented linguistic differences between those who bear good tidings versus those who bear bad tidings. Freese and Maynard 34 (1998) examined over 100 naturally occurring two-person conversations in which good or bad news was relayed. They observed the portion of each conversation dealing with the actual delivery of the news to occur in a four-tum patterned fashion, the news delivery sequence (NDS). Of particular relevance to this study is the finding that during the NDS senders would tailor their linguistic prosody to the valence of the message. Freese and Maynard reasoned this was done in order to communicate emotion consistent with good or bad news (i.e., joy or sorrow). If senders adjust their communicative behaviors according to their degree of liking for the communication situation (Mehrabian, 1971 ), and if these types of adjustments can take the form of variations in the prosodic features of senders’ language according to the positivity or negativity of the message they carry (Freese & Maynard, 1998), perhaps other lingusitic differences might be exhibited. The variables under consideration here are teasing, greetings, message elaboration, and repudiating the social perception test’s validity. Method The videotaped interactions collected during the experiment reported in Chapter 3 served as the initial data source for this investigation. Unfortunately, the video recording equipment failed to operate for two dyads (1.2% of total), leaving 163 dyadic interactions to be analyzed. Moreover, of these 163 dyads, the audio portion of the video recording was not sufficiently intelligible for two. However, this precluded these dyads from only certain tests. Thus, the actual number of usable cases is listed separately near each analysis. 35 First, each scorer’s dialogue was transcribed verbatim as it pertained to the delivery of the test performance feedback. For some dyads, once the topic of the test results had been covered, conversations drifted toward irrelevant topics (e.g., class schedules, common acquaintances). In a few of these cases, the experimenter had to interrupt the couple in the interest of time. Thus, although the transcripts do not always reflect everything that was said in its entirety, every effort was made to capture the scorer’s dialogue from the moment she entered the taker’s room until she either left to return to her room or the talk became irrelevant to the context of delivering test feedback. Communication Variables Once transcripts had been produced, they were read through several times in order to familiarize the researcher with the nature of the interactions. The complete videotaped interactions were also viewed multiple times so as to identify conversation variables that might be meaningful. Patterns of behavior that seemed to vary from one dyad to the next received consideration as potential variables for coding. In the interest of maximizing reliability, focus was kept to variables that could be measured fairly easily. Variables identified for later coding included teasing, greetings, message elaboration, and questioning the validity of the social perception test. Each variable is described below. Teasing. As mentioned in the discussion section of the previous chapter, scorers were sometimes observed to tease or make fun of the test-taker in the course of delivering their message. Two coders independently viewed all 163 dyadic interactions and for each dyad made a simple dichotomous judgment as to whether or not they believed (subjectively) the scorer made fun of the test-taker. Inter-rater reliability was calculated using Cohen’s (1960) kappa because, unlike raw percent agreement calculations, kappa 36 corrects for agreement achieved merely by chance. The obtained kappa value was .90 which, according to the conventions outlined by Landis and Koch (1977), indicates very high agreement. Cases for which the coders initially disagreed (n = 6) were discussed by the coders until agreement was achieved.3 Greetings. Based on an inspection of the transcripts, scorers appeared to differ about whether or not they explicitly greeted the test-taker during the interaction. When a greeting did occur, it took the form of Hello, Hi, or Hey.4 Other scorers would frequently begin their news delivery encounter by offering an utterance to the effect of Okay, or All right before proceeding to inform the test-taker about her test result. Okay and All right are attention-getting devices placed at the beginning of the communication episode, but Sacks (1975, 1995) regards these as different from greetings. In order to be felicitous, greetings must occur according to rules which are different from other classes of utterances that may be used to mark the beginning of talk. For example, unlike Okay and All right, the relevance for greetings evaporates after the beginning of the conversation. If a greeting comes later in the talk, it will have already likely gone noticeably absent, which means negative inferences will have already been prompted. Coding greetings was therefore straightforward, as the transcripts could be inspected for scorers who began their news delivery interaction with either Hello, Hi, or Hey. Elaborateness of Message. Scorers’ messages to test-takers also appeared to vary in their degree of elaborateness. Objective word count served as the proxy for message elaborateness. Transcripts were reformatted and screened to verify that they only contained the scorer’s talk. Then, the word count feature through Microsoft Word was used to obtain the message length for each scorer. 37 Questioning Test Validity. Cursory reading of the transcripts revealed that scorers at times appeared to reframe the meaning of the test result such that the scorer seemed to be repudiating the validity of the social perception test. Exemplar utterances include “It’s probably rigged,” “I don’t think you’re socially awkward,” “I don’t think it reflects your personality,” and “I don’t think it’s true.” The choice to measure this variable is predicated on a striking and contradictory piece of logic, at least on the face of it. If the social perception test is presented as a rigorously-developed scientific instrument for measuring nonverbal perception skill, and if scorers honestly believe this to be the case, then it seems rather peculiar that scorers would make an utterance inconsistent with their belief in the presence of the test-taker. Again two coders read through the transcripts and made a dichotomous judgment as to whether or not they sensed the scorer said something to indicate her disagreement with the test result. Coders demonstrated perfect agreement in categorizing each case (kappa = 1, n = 161). Results Teasing During the running of the original experiment fi'om which these videotapes and transcriptions emerged, the investigator observed an unexpected phenomenon of interest. The assumption underlying these studies is that, in general, people hesitate to transmit bad news. However, once experimental trials began, some scorers were witnessed actually making fun of the test-taker about the test-taker’s social perception test score. In other words, not only did the scorer display zero behavioral evidence as to her reluctance to deliver the bad news, she rushed into the test-taker’s room, eager to relish a joke at the test-taker’s expense. 38 Moreover, these jokes, when they occurred, varied in their degree of subtlety. For example, some scorers employed little more than a mildly disparaging and sarcastic tone while concurrently informing the test-taker that the test-taker was socially awkward. This frequently took the form of the scorer mock congratulating the test-taker for a low score. At the more extreme end came pejorative jabs such as “You got a thirty percent. You got fourteen out of twenty questions wrong! You’re stupidl”, “Why did you get thirty percent! We gotta work on you. You weird!”, and “Some bad news. You’re an idiot!” Of 163 usable dyads, 37 scorers were judged to have teased the test-taker (22.6%). Table 3 lists the frequencies with which scorers teased the test-takers according to experimental condition. Table 3. Frequency of Teasing by Experimental Condition. Bad News Good News No Teasing Teasing No Teasing Teasing Stranger 38 4 38 0 (43.7%) (4.6%) (50.0%) (0.0%) Friend 18 27 32 6 (20.7%) (31.0%) (42.1%) (7.9%) Notes. Numbers in parentheses are percentages of the total number of usable scorer cases within each of the valence conditions (Bad News, 11 = 87; Good News, n = 76). 39 Strikingly, most of the teasing occurred when scorers were delivering bad news (n = 31) as opposed to good news (11 = 6). Equally striking is that of the 37 total scorers who teased the test-taker, only 4 teased the test-taker who was also a stranger. Follow up chi- square tests were statistically significant for an association between teasing and both fiiends/strangers, 380, n = 163) = 28.05,p < .01, (p = .42, and news valence, 380, n = 163) = 17.79, p < .01, (p = -.33. By and large, scorers who chose to make fun of the test- taker did so when the test-taker was both a close friend and was receiving bad news. Greetings Studying the videotapes and transcripts of the news delivery interactions revealed that some scorers greeted the test-taker while entering (or shortly after having entered) the test-taker’s room and some scorers did not. Those who greeted the test-taker by and large did so with Hello, Hi, or Hey. When a greeting was not offered, scorers routinely began the interaction with another attention-getting device such as Okay, Alright, or So. With regard to news valence, scorers might be observed to use fewer greetings when delivering bad news as opposed to good news. Bad news represents negative information by virtue of it signaling undesirable consequences. Negative information is thought to require more cognitive effort to process it than does positive information (Ito, Larsen, Smith, & Cacioppo, 1998). If additional cognitive capacities are devoted to processing the negative information, there may be fewer faculties with which to attend to customary conversational norms such as greetings. Consequently, the greeting is passed over while a person cognitively attends to other matters. In other words, thinking about how they will transmit the bad news might preoccupy scorers to the point that they forget to greet the test-taker as per the linguistic convention when beginning a conversation. 40 Still, bad news might be met with fewer greetings than good news for another reason. Speakers sometimes deliberately forgo greetings to show that they are specifically not beginning a conversation with the bearer (Sacks, 1995). Rather, another type of utterance is used to begin the interaction (e.g., excuse me, pardon me). Sacks calls these initial utterances “tickets,” and they are to convey that the speaker has an express purpose for the interaction as well as a sense of what that purpose is, but that the encounter will not become a ratified conversation (Vol II, p. 195). It is easy to apply these ideas to the context of delivering bad news. Because delivering bad news is face threatening to both the sender and the receiver, scorers may wish to execute their duties in the quickest way possible. Whether the scorer fears a slight to her own or the test-taker’s face and emotions, she may opt to begin the encounter by signaling her desire that this interaction not be treated as a conversation with all the conventions, obligations, and turn-taking rules that ordinarily pertain. In other words, it would not be surprising to see scorers attempt to avoid establishing a lengthy conversation. A common way people do this is by forgoing the greeting (Sacks, 1995). The frequencies with which scorers greeted test-takers can be viewed in Table 4. These data revealed that 25.3% (n = 22) of scorers sharing bad news used greetings, compared to 36.8% (n = 28) of the scorers who shared good news. However, a chi-square test failed to show an association for news valence and greeting, 12(1, 11 = 163) = 2.55, p = .11. Thus, it is not possible to conclude that scorers who relay bad news greet test- takers less frequently than do those who bring good news. A significant chi-square did obtain for fiiends/strangers and greetings in that scorers greeted strangers more frequently than test-taking friends, x2(1 , n = 163) = 8.88, p < .01, (p = -.23. 41 Table 4. Frequency of Greeting by Experimental Condition. Bad News Good News No Greeting Greeting No Greeting Greeting Stranger 26 16 20 17 (32.9%) (20.3%) (25.3%) (21.5%) (100%) Friend 39 6 28 11 (46.4%) (7.1%) (33.3%) (13.1%) (100%) Notes. Numbers in parentheses are percentages of the total number of usable scorer cases within each of the closeness conditions (Friends, n = 84; Strangers, n = 79). For friend/stranger and greeting: x20, 11 = 163) = 8.88, p < .01, two-tailed. For news valence and greeting: {(1, n = 163) = 2.55,p = .11, two-tailed. Questioning Test ’s Validity In the course of delivering feedback to the test-taker, some scorers made at least one explicit utterance consistent with their rejection of the validity of the social perception test. Obvious examples include the following: “It’s probably rigged” and “I don’t think the test is accurate.” Other scorers expressed their judgment by disagreeing with the diagnosis implied by the low test result through utterances like “I don’t think you’re socially awkward,” “I don’t think it reflects your personality,” and “You’re not awkward to me.” Table 5 lists the frequencies with which scorers uttered a rejection of the validity of the social perception test. Of 161 usable cases (two were excluded because the audio 42 was not sufficiently intelligible to code for this variable), only 11 featured scorers who somehow questioned the validity of the social perception test. Table 5. Frequency of Test Validity Rejection by Experimental Condition. Bad News Good News No Rejection Rejection No Rejection Rejection Stranger 37 5 37 0 (46.8%) (6.3%) (46.8%) (0%) (100%) Friend 38 6 38 0 (46.3%) (7.3%) (46.3%) (0%) (100%) Notes. Numbers in parentheses are percentages of the total number of usable scorer cases within each of the closeness conditions (Friends, n = 82; Strangers, n = 79). For friend/stranger and rejection: 780, n = 161) = 0.06, p = .80, two—tailed. For news valence and rejection: x20, n = 163) = 10.30, p < .01, two-tailed. However, it is interesting to note that all 11 of these were cases in which scorers were delivering bad news. No scorer who shared good news was heard to repudiate the test’s validity. Thus, a significant association between news valence and repudiation obtained, 380, n = 161) = 10.30, p < .01, (p = -.25. No association can be inferred between friends/strangers and repudiation, {(1, n = 161) = 0.06, p = .80. Thus, although scorers only questioned the test results when the result was low, those who did reject the results did so approximately equally between fiiends and strangers. 43 Elaborateness of the Scorer ’5 Message A cursory inspection of the scorer message transcripts indicated a degree of variance in message length from one scorer to another. Message length was inferred by a simple word count, obtained using the “word count” feature in Microsoft Word. With regard to news valence, venturing a prediction about the nature of any association would be difficult because arguments might be made for the expectation of a mean difference in either direction. Intuition might suggest that people having to share negative information will be more cognizant of the face threat involved, and will thus work harder to minimize that face threat. If this work came in the form of oral communication, then speakers should create more verbose messages when having to share bad news than when passing along good news. On the other hand, recalling the rationale for why harbingers of bad news might wish to avoid any additional conversation with the target of their ill tidings, it could be observed that people sharing bad news will say the minimum necessary to relay the pertinent information. Indeed similar instances occurred in the data when scorers entered the test-taker’s room, stated the test score and nothing else, then left. Thus, the more elaborate messages would accompany good news instead of bad news. These hypotheses are admittedly post hoc, but had they been offered a priori, the data collected here do not point one way or the other. As Table 6 reflects, scorers sharing bad news formed messages that were an average of 26.19 (SD = 21.23) words in length. This compares to an average word count of 23.57 (SD = 13.63) for scorers sharing good news. Fully crossing friends/strangers with news valence in a 2 X 2 ANOVA using word count as the dependent variable revealed no main effect for news valence, F (1 , 159) = .88, p = .35, 172 = .01. A main effect for fiiends/strangers did obtain, however, with 44 scorers sharing lengthier messages with friends (M = 29.48, SD = 21.18) than with strangers (M: 20.33, SD = 12.74), F(1, 159) = 10.63,p < .01, n2 = .06. Although scorers talked more to their fiiends, whether they shared good news or bad news appeared to bear little on the elaborateness of the messages they constructed. Table 6. Mean Scorer Message Length (Words Per Message) as a Function of News Valence and F fiends/Strangers. Bad News Good News Marginal Mean Stranger 21.00 19.57 20.33 (14.08) (11.16) (12.74) Friend 31.26 27.47 29.48 (25.59) (14.78) (21.18) Marginal Mean 26.19 23.57 (21.23) (13.63) Notes. Standard deviations are in parentheses. For news valence, F(1, 159) = .88, p = .3 5, n2 = .01. For fiiends/strangers: F(1, 159) = 10.63, p < .01, n2 = .06. Discussion The goal of this second investigation was to document verbal communication behaviors associated with sharing good or bad news. This attempts an answer to the call of Eberhardt-McKee and Ptacek (2001) to increase understanding about what givers do when delivering bad news. Communicative actions such as teasing, greeting, reframing 45 the face threat by invalidating the source of the negative information, and varying the length of sender messages were explored. Each will be discussed in turn. Teasing Probably the most surprising finding to emerge from these data related to scorers having made fun of the test-takers when delivering bad news. Such an act runs counter to everything that has been said about the MUM effect thus far. This effect predicts that compared to good news, people will be more reluctant to share bad news. However, nearly 36% of the scorers who shared bad news did so by teasing their partner about the low test score. This certainly did not reflect any reluctance or hesitation to deliver the bad news. On the contrary, it would appear these scorers enjoyed adding insult to injury. Further, the overwhelming majority of these bad news teases occurred between fiiends (only 4 scorers made fun of a stranger). If the MUM effect holds and people are indeed reluctant to transmit bad news, then what explains the almost eagerness on the part of these scorers to joke about their partner’s plight? In other words, under what conditions will senders not only fail to deliver their message with contrition, but will do so while having a laugh at the hearer’s expense? An answer might be constructed by applying insights from politeness theory (Brown & Levinson, 1987) and Uncertainty Reduction Theory (Berger & Calabrese, 1975) In Brown and Levinson’s (1987) model of politeness, individuals are motivated to reduce threats to two types of face: positive face and negative face. Positive face refers to a person’s desire to be approved of in social interaction, while negative face is a person’s desire to be free from imposition. This model further assumes that as one interacts with 46 others, he or she rationally estimates the face-threatening potential of any actions he or she is about to take (i.e., face-threateing acts = FTA). Subsequently, one selects (again rationally) a communication strategy that maximizes the positive faces of the self and the other, while minimizing transgression of the negative faces of self and other. Facework is done so as to preserve the expressive order of the communication encounter (Goffrnan, 1967). The practice of doing facework in this manner is termed politeness.5 One of the factors postulated to contribute to the magnitude of the face threatening act is the social distance between the interlocutors (Brown & Levinson, 1987). According to Brown and Levinson, social distance is based on “. . .an assessment of the frequency of interaction and the kinds of material or non-material goods (including face) exchanged...” between the speaker and the hearer (p. 77). In general, the greater the social distance between the two people, the greater the weightiness or seriousness of the FTA. Recall that most of the teasing in this study occurred between close fiiends (i.e., low social distance). Brown and Levinson do not speculate as to the relative impact of social distance versus other predictors of the seriousness of the FTA, but the positive relationship between social distance and PTA seriousness should predict that one can tease her close fiiend while delivering bad news simply because she does not view the threat to face as serious. As Zajdman (1995) notes, there are conditions under which both the speaker and the bearer “. . .agree that face demands be suspended for the sake of the other interest, which is to get a laugh” (p. 326). Moreover, a summary of teasing literature (Kowalski, Howerton, & McKenzie, 2001) recognizes socialization and boosting affiliation as motives for teasing. In short, teasing one another about issues that are not overly important is simply part of being fiiends. 47 Berger and Calabrese (1975) presented a theory that attempts to account for the communication behaviors enacted when two strangers first encounter one another. This has come to be known as Uncertainty Reduction Theory (URT). URT predicts that when strangers come together in a communication situation, they are motivated to reduce uncertainty about one another so as to increase the reliability of predictions they can make with regard to how each should and will behave. It is further hypothesized that uncertainty is reduced through communication. Combining the politeness perspective’s ideas about social distance with URT, one would expect social distance and uncertainty to be positively related (because both should decrease concomitantly with relational closeness). That is, if uncertainty is reduced, social distance should also be reduced. Thus, for a person to tease another while delivering bad news, which would under normal circumstances be highly face threatening, the sender must hold a relatively low degree of uncertainty about the target. A target who is a close friend would afford such a low level of uncertainty. This is consistent with Brown and Levinson’s (1987) writing that increased closeness between speaker and bearer affords the speaker with license to enact behaviors that would appear to increase in their face-threatening nature. Of course, the argument so far assumes that the sender who teases the target while delivering bad news does so, in effect, because she has judged the F TA to be minimally threatening. Brown and Levinson (198 7) also mention that humor can accompany an FTA as an intentional strategy oriented toward minimizing the damage to face. H ere, the FTA is still assumed to be threatening, and the humor becomes less an end in itself (i.e., to get a laugh) and more a means to an end (i.e., to allow the target to save face by diffusing the threat in the humorous absurdity of the situation). Whereas the reasoning in 48 the previous paragraph predicts that face concerns are temporarily suspended for purposes of sharing a laugh, here face concerns remain in the forefront, and any joking is done ostensibly for the target’s benefit. The four scorers who dared to tease a stranger about her bad news might be explained in this way. Of course, methodological considerations do not permit a formal test of these explanations using these data. Greetings With respect to greetings, the failure to find a difference between sharers of good news versus sharers of bad news is more difficult to explain. People may indeed feel some sort of psychological reluctance to pass along the bad news. However, the simple fact that the news was negative might not be enough to overwhelm one’s sense of conversational decorum. Sacks (1975) contends that greetings are ahistorically relevant, meaning there is no exclusion rule that would dictate greetings are no longer needed after a certain number of conversations have taken place. This is not to imply that greetings are always present at the beginning of every conversation. Sacks’ contention stems from the observation that even couples who have been married more than thirty years will frequently greet each other when beginning new conversations. Perhaps this trend for greetings continues to operate in the social workings of individuals independent of the valence of the topic of conversation. Explaining the difference between scorers greeting their fiiends less often than strangers is more straightforward. First, Uncertainty Reduction Theory (Berger & Calabrese, 1975) would predict that strangers (i.e., uncertainty is high) would more closely follow scripted behavior, which might involve their remaining vigilant to observe the conventions associated with striking up a formal conversation. To the extent that 49 greetings are included in these conventions, they should more fiequent at the beginning of conversations between strangers (all other things being equal). Sacks’ (1995) comments on greetings point to another possibility. He notes that when greetings do occur, they occur at the beginning of the beginning of conversations. If the conversation is somehow interrupted, even by lengthy time intervals, the conversation partners might fail to greet one another when they next talk. Yet, the greeting is not noticeably absent. The lack of greeting in this situation implies that the succeeding instance of talk is merely a continuation of a previously interrupted conversation. Thus, a new relevance for greeting had not been created, and no greeting was necessary. Because pairs of fiiends reported to the laboratory together, it is not hard to imagine their prior conversation being interrupted by an hour-long experiment that required them to separate temporarily (i.e., the current experiment in which they participated). Then, upon reuniting during the news revelation encounter, in the scorer’s course of sharing the news she and her fiiend simply picked up the talk where they left off. Sharing news with a stranger would count as a brand new conversation, which would increase the relevance for a greeting. Questioning Test ’s Validity There seems to be a clear effect of news valence on the scorer’s verbal communication as manifested by an apparent need to reframe the face threatening action. Although only eleven scorers went on record (Brown & Levinson, 1987) as disputing the validity of the social perception test, all of these did so in the process of delivering bad news to the test-taker. This is consistent with an anecdotal observation noted during the running of the experiment that generated these data: Scorers delivering bad news reported 50 having double checked the answers much more often than did scorers delivering good news. Once again, it seems, a bias emerged where scorers rarely or never second guessed the validity of the test when the test-taker “did” well. In fact, several scorers who delivered good news were observed to have explicitly indicated their confidence in the validity of the test through comments like “I knew you’d be good at reading people,” and “You got seventeen out of twenty correct. So you are exceptionally skilled at reading people.” In both good news and bad news conditions, scorers compared a scantron to an answer key and marked incorrect answers with a pencil. There is no reason to expect that scorers would make fewer errors simply because the good news scantrons contained more correct answers. Yet, scorers seeing an eighty-five percent rarely questioned their test- correcting abilities, much less the validity of the overall social perception test instrument. At least one explanation might account for why scorers publicly repudiated the test’s validity as well as their apparent failure to double check their work when correcting an eighty-five percent result. Error Management Theory (EMT, Haselton & Buss, 2000) was developed to explain the evolution of biases. EMT holds that people are biased toward avoiding the more costly error. Haselton and Buss used EMT to explain the findings that men appeared to overestimate a woman’s sexual interest in him, while women were biased toward underestimating a man’s relational commitment. Perhaps good and bad news, by virtue of their implications for both the speaker and the hearer’s face, represent yet another domain in which biases would have evolved so as to avoid the more costly error. H ere, there are two possible correct actions and two possible errors. First, a scorer can correctly relay 30% or 85% feedback when the score was indeed 30% or 85% (i.e., no error is made). With respect to the potential errors, the 51 scorer could either inform a test-taker that she is socially awkward when her test indicates otherwise, or the scorer could inform the test-taker that she is not socially awkward when she might be. To commit the first error is more egregious because of the face threats it represents during the interaction. If people are motivated to avoid threats to face as Goffman (1967) and others assert, this implies there is a substantial cost incurred by performing an FTA. To commit the second error (assuming the error remains undiscovered by the test-taker) involves no threat to face and, hence, less cost. Moreover, in this study, people might derive further reassurance by the idea that there is likely a high correlation between what the scorer indicates as the test score and the true test score. Thus, if a bias develops, it should be against relaying bad news when the news is actually not bad. In this experiment, such a bias might explain why scorers would be more apt to double check their work when the score is low, as well as why they attempt to mitigate the face threat implied by a low score by verbally rejecting the validity of the test without doing so when the score is high. As attractive as EMT might be however, one of its drawbacks is that, although it parsimoniously explains biases that are observed, it renders no predictions about when a bias should occur. If a face-threat bias exists with such a low-stakes encounter like the bogus social perception test, why should it not be the case that fiiends and strangers would differentially disagree with the test result? The finding here that scorers appeared to disparage the test’s validity equally among friends and strangers forestalls the conclusion that the closeness between partners impacts any tendency toward using this verbal technique to reframe the face threat. Perhaps the nature of the bad news has more 52 to do with this behavior than the nature of the relationship between the sender and the receiver. Elaborateness of Message Finally, this study provided no evidence that scorers produce messages of different lengths as a fimction of news valence. A difference was found, however, between fiiends and strangers. Extreme caution must be exercised when interpreting the data with regard to this variable as a number of limitations cast doubt on their reliability. First, although every effort was made to capture the scorer’s messages in their entirety, there were times when the video recording (and, by extension, the transcription) ended before the end of the actual encounter. Moreover, there were times during the actual experiment where, always after the scorer had executed her news delivery duties, the experimenter interrupted the dyads from their discussion because the banter had clearly wandered off topic and time was becoming short. In both cases, these decisions were made because it was felt that the dialogue of importance had already been recorded. Nonetheless, using crude word count as a proxy of message elaboration necessitates that proper comparisons can only be made when the interaction episodes are complete. Of course an argument could be made for unitizing the messages in some other way, but that was not the case for this study. Second, these interactions were live and, as such, test- takers provided much of the talk to which scorers responded. Test-taker comments were deleted from the transcripts prior to analysis, but the influence of the test-taker’s talk remains necessarily confounded with any effects of news valence. Any relationship between news valence and message elaboration clearly awaits future investigation. 53 Taken together, these data enrich the picture of the bad news delivery process over traditional methods of inquiry (i.e., inferring MUM effects from time variables). Moreover, whereas no effect was found for friend/stranger on time to response from Study 1, the relationship between the interactants did appear to influence the bad news process with respect to teasing such that senders teased receivers when the receiver was a friend and when the news was bad. The idea that closeness might have given senders a different way of perceiving the face threat, and thereby different ways of dealing with it, warrants firrther inquiry. If the nature of the relationship between the sender and receiver can influence how the sender perceives the consequences of the news for the receiver and, by extension, the net threat to face, a possible point of intervention is introduced for those looking to improve the process by which they deliver bad news. Researchers should continue to explore the possible role played by relational characteristics in the context of delivering bad news. 54 CHAPTER 5: DIRECTION OF THE MUM EFFECT Newly reported evidence suggests that the reluctance to share bad news might be a process distinct from any eagerness to share good news (Uysal & Oner-Ozkan, 2007). This is consistent with Ito et al. (1998) who argue that the human brain processes negative information in a cortical area separate from where it processes positive information. Moreover, an inspection of the frequency distributions for each experimental condition in the Chapter 3 study revealed that regardless of closeness condition, the good news groups exhibited much less variation than the bad news groups. Uysal and Oner-Ozkan (2007) contend that the study of the bad news process might be enhanced if researchers do not assume that what motivates people to withhold bad news is the opposite of what spurs them to share good news. For example, when faced with having to share negative information, additional cognitive resources are recruited to address the potential face-threatening act. Because good news is not face- threatening, these same cognitive faculties are not needed, leaving other motivations to possibly become salient. The emotion literature indicates that people will transmit good news in order to boost the mood of the recipient (Carlsmith & Gross, 1969). Thus, although both withholding negative information and revealing positive information can potentially benefit senders and receivers, the degrees to which participants are benefitted as well as the processes that underlie the motivations to behave either way are ripe for further study. Nonetheless, a clear empirical difference exists between good news sharing and bad news sharing. However, the direction of this difference remains to be tested. The purpose of the third study is to further elucidate the nature of the mean difference 55 commonly taken as evidence of the MUM effect. Toward this end it is necessary to include news valences that are viewed to be more neutral. Research on the MUM effect asserts that differences in the way people share bad news versus good news are the result of a felt reluctance to transmit negative information. The typical experiment gives participants the opportunity to share good news (in various operationalizations) or bad news (in various operationalizations) to some invested recipient. The outcome measure is often either a simple dichotomous judgment of whether or not the news was communicated, or a measure of time delay before the onset of the news. In any event, outcomes are compared from only two groups: the “good news” group and the “bad news” group. Mean differences that show decreased (or slower) transmission of the bad news compared to the good news are interpreted as a psychological reluctance of senders to share bad news (i.e., a MUM effect). As already noted, withholding bad news is not the only strategy by which people manage information sharing when communicating with others. Situations drive people to eagerly relay positive information as well. Carlsmith and Gross (1969) observed people to transmit good news with a pronounced measure of dispatch. Carlsmith and Gross’s rapid responders were motivated to reveal the good news quickly because they desired to boost the mood of the bearer. Two scenarios thus seem consistent with the pattern of results observed in extant MUM effect research that only pits the sharing of a supposedly positive message against the sharing of a supposedly negative message. These explanations can be unconfounded by comparing good/bad news transmission to the transmission of news that is neutral in valence. Neutral news would take the form of messages containing information that is 56 previously unknown to the receiver, but that neither signals desirable nor undesirable consequences. That is to say, the news is neither good nor bad; it is just something the hearer did not know ahead of time. Regardless of the reason, if senders are motivated to avoid delivering bad news, senders should delay sharing decidedly negative information more than they will delay sharing information that is neither good nor bad (i.e., neutral). Thus, the following hypothesis is forwarded. H2: Senders will take longer to share bad news than neutral news. If bad news transmission and good news transmission are indeed distinct processes, then it is entirely possible that people could also be motivated to share good news with more rapidity (for the reason proposed by Carlsmith and Gross (1969) or for some other reason) than they would share news that is neither good nor bad. If, for example, senders are especially urged to share good news as a means of improving the hearer’s mood, then results should obtain which are consistent with the following hypothesis. H3: Senders will share good news more rapidly than neutral news. Any desire to avoid sharing bad news has already been suggested as possibly being a process separate from that which governs any eagerness to share good news. The data gathered here might also inform the mechanism by which people determine news valence and how they evaluate what it signals for themselves and for others. If news valence is a single, continuous dimension, it should be possible to locate valence values that would be judged as positive, negative, and varying intermediates. The interesting questions surround the size of any “neutral region” and how people might conceive of it. 57 Senders no doubt make predictions about how the receiver will interpret the news (Miller & Steinberg, 1975). What is unclear, however, is how senders represent the valence of the news in their minds. When faced with having to share a message containing novel information to a bearer, do sharers immediately assign a default judgment of good or bad, or is there a valence range that elicits neither good news judgments nor bad news judgments? A continuous-scored test might serve to illustrate. Consider an exam anchored by 100% (perfect score) or 0% (failure to mark any items correct). Low test values are typically seen as bad, and high values are viewed as good. That is, under prevailing U.S. educational standards, most people would be happy to achieve a 96%, and they would likely feel disappointment with a 25%. Curiosity surrounds how people would judge 55%, 64%, 70%, and so on. Do people base their valence judgments on some sort of internal personal threshold such that anything above the threshold is judged to be good and anything below that threshold bad? Or, perhaps there are two thresholds, one for segmenting good news and another for delimiting bad news. If the heuristic does involve two thresholds, then would the values between the two thresholds be viewed as neutral (i.e., neither good nor bad)? These processes have implications for any reluctance to share bad news, eagerness to share good news, and the behaviors associated with news transmission in general. The plotted relationship between news valence and time to response might yield clues as to the nature of how people conceptualize the valence of information they share with others. If people employ a singular cognitive threshold-type heuristic, they are making a wholesale prediction about how the receiver will View the news. That is, in the 58 sender’s mind, the receiver is either going to view the news as good news or as bad news. Such a relationship would be consistent with a scatterplot that emerges in the form of a step or ogival function. A certain range of values are all viewed as bad news, and a certain range above those are all viewed as good news. Any piece of information whose valence falls below threshold renders a judgment of bad news, and senders will behave in ways already known to be consistent with bad news delivery. If, on the other hand, people view news valence as occurring along more subtle gradations, there could be a range of valence values for which senders either believe the news to be neutral, or for which the sender is unsure about how message will be perceived by the receiver. The behavior of senders should be affected by this “neutral range” in ways that are distinct from both their good news delivery behavior and their bad news delivery behavior. A scatterplot yielding a linear function would indicate the salience of mid-range valence values which could prompt behavior that is different fi'om that expected if senders were sharing decidedly positive information or decidedly negative information. Because both scenarios are possible, another research question is posed. RQ3: What appears to be the form of the relationship between news valence and transmission rapidity? Method A new experiment was conducted wherein news valence was represented as a continuous variable. As before, time to response served as the primary outcome measure. This experiment again utilized the procedural paradigm where senders relay false test feedback to hearers in the context of a live interaction. 59 Participants A total of 228 undergraduate (female = 127, male = 101) students from a large Midwestern university served as participants. The average age of the participants was 19.95 years (SD = 1.46), and ages ranged from 18 to 29. The race-ethnic makeup of the participants was predominantly European American (70.9%), followed by Afiican American (14.8%), Asian (3.1%), Latino/Latina (1.8%), and Native American (1.3%). Approximately 8.1% of the participants indicated they were multiracial or of some other race-ethnicity. As in the first experiment (Chapter 2), participants were paired. This created 114 dyads. The sex composition of each dyad was also recorded. Here, 35 dyads featured female scorers and test-takers, 32 dyads featured female scorers and male test-takers, 25 dyads had male scorers and female test-takers, and 22 dyads had male scorers sharing with male test-takers. Procedures The experimental procedures largely resemble those from the first experiment (Chapter 3), although type of relationship (i.e., fiiends vs. strangers) was not induced this time. Participants again arrived at the laboratory in pairs, and informed consent was properly obtained. Roles (i.e., test-taker, sender) were randomly assigned as before, and participants were again led to believe the social perception test holds implications about whether or not the test-taker is skilled at reading the body language of others (i.e., socially awkward). Several modifications fi'om the first experiment were made. First, the bogus social perception test items were constructed such that both faces appeared in the same 60 photograph. The prior study revealed that test-takers might not have expected themselves to score very well at the outset. The perceived ridiculousness of having to infer nonverbal cues from people who were not even pictured together might have reduced the felt relevance of the test (and, hence, any bad news). In other words, the change was made to boost the perception that test-takers stood a fair chance at getting more items correct because test-takers could more plausibly infer nonverbal reactions from the people pictured in the test items. More importantly, it was hoped that senders (previously referred to as scorers) would be less able to attribute poor performance to an overly difficult test. Second, senders were not required to actually tally the test-taker’s score themselves as before. This was to sharpen the time to response measure such that it more precisely captured reluctance/rumination time (because scoring is no longer included in the timed delay). As before, test-takers worked on the test in a separate but nearby room. Once the test-taker finished, he or she carried the scoresheet to the sender who was waiting in the scoring room, sitting at a small desk behind a laptop computer. Test-takers would sit patiently across from the senders while the senders keyed the test-taker’s answers into the computer. Figure 1 depicts the screen display senders viewed while keying in the test-taker’s answers. 61 01 6 8 I. '9 9 V 8 Z I O O O O O O O O O > O O 0 ll 0 O O O O O ' O “'O O O 0 E$afi§afiaaa§3 0000000000) 0000000000- 0 o o o oo do on . O Figure I . Screen utilized by senders to input test-taker’s scores. Note “Get Score” button in lower middle portion of screen. 62 Participants were told that this computer does the scoring calculations. In reality, the computer was programmed with a random number generator such that by clicking on a certain on-screen button, the sender activated the display of a random number between 90 and 10 (see Figures 2a-2c). Senders were told ahead of time that this number corresponded to the test-taker’s scoring percentage. Additional graphics and scoring norms displayed onscreen so as to boost the credibility of the testing procedure as well as the sender’s ability to interpret the score. Senders were free to reveal the test-taker’s result by saying anything they liked. A digital video camera recorded the sender’s behavior during the interaction. The researcher activated the camera from behind one- way glass using a remote control. Remote activation of the camera enabled the researcher to remain completely invisible to the participants during the course of the interaction, another procedural gain over the first experiment. 63 can <2: «33 9.5032. 35...... 2330. 323 .3032. 353 mu. $438..“ 3....132 ., ,. G 111 men menu?" mm... \. we» a ..:...¢Q ” score once “Get Screen shot example of a randomly generated “superror Figure 2a rs activated. Score” button t we..._s_122..::c: Em. a a... a... . lac. .. afinmmmrhm 859020.. ,. Act-.33 > neutral > bad. With respect to the psychological reluctance that senders experience, good < neutral < bad. Because the neutral results equaled neither the positive nor the negative results, these data cannot be taken to clearly favor either of the proposed directions for the MUM effect. Of course, neither do these data necessarily refute either an eagerness to share good news or an unwillingness to relay bad news. There is the possibility that some cognitive latitude of neutrality could exist along with either the eagerness driver or the unwillingness mechanism (or both). From a simple test score and some prompting as to what the scores mean, people certainly seem able to discern a wide range of valence possibilities. Moreover, those mid-range valence possibilities do seem to associate with delay to response and reluctance in systematic ways. That is, mid-range scores were associated with mid-range outcomes. Although the design considerations for this study are substantially improved over the previous study, which should permit greater confidence in the data, certain limitations 75 have nevertheless lingered. Despite a strong induction as evidenced by both the news valence items as well as the reluctance items, one cannot be absolutely certain that absolute levels of bad news or reluctance were instantiated. As before, although lower scores were viewed as more negative than average or high scores, this does not automatically translate to judgments of “bad news” on the part of the sender. However, considering that this experiment was conducted using a testing procedure and metric well understood and internalized by undergraduates (i.e., high percentages are good, low percentages are bad), senders tended to rate news valences and reluctances well below the scale midpoint when appropriate, and senders behaved (i.e., delay their feedback) consistent with a lower or higher score, interesting counterpoints are raised to this line of argument. If senders are not taking these low scores to imply some shade of bad news, the observed results become incredibly difficult to explain. Other limitations exist with respect to the time to response variable. Measuring the times from the camera’s internal chronograph seems to have improved the precision over the timing from the previous study (where a handheld stopwatch captured delays). Again, however, the decision was made to use the moment the sender begins his or her first utterance to the test-taker (after activating the score screen). Perhaps senders begin vocalizing to recipients as a means of expressing their reluctance and (possibly) buying themselves more time with which to formulate and package their negative message. For example, instead of following the initial pause with words communicating the test result, senders could simply say “Yeah,” then take a few more moments of silence before expressing the actual test result. Using the first uttering of “Yeah” as the timer stop point necessarily precludes the ability to capture what might be a conceptually meaningful 76 pause that occurs after the sender makes such an utterance. Although it is unfortunate to miss out on these situations, capturing the delay as was done in this study was uniform across cases. Moreover, a failure to capture full pauses should serve to attenuate the effects, thereby obscuring meaningful effects. It should be noted that most of the effects obtained with respect to the use of time to response as a dependent variable were nonetheless significant. External validity is also somewhat limited. These data were collected using university undergraduates who averaged 20 years of age. Further news of only one type was included. Questions might be raised as to whether the MUM effect and its associated effects are impacted as individuals age and/or gain experience with delivering valence- sensitive information. The impact of sex on the bad news delivery process remains ambiguous. No specific hypotheses were made because sex was not the focal variable in this study. Nonetheless, these data are clearly not consistent with any expectation of sex effects. The trend for high scores to be rated as more positively valenced, to be shared faster, and to be associated with less reluctance than lower scores did not depend on either the sex of the sender or test-taker. Perhaps this finding is consistent with the larger body of sex effect research which finds relatively few true psychological differences between the sexes. One of the goals of this study was to determine whether, when faced with delivering some novel information, senders will assume a default judgment of good or bad news, or if they would view news valences more along a continuum that included mid-range possibilities. Although the nature of the relationship between test score and 77 any of the dependent variables did not answer this question definitively, it appears that people do internalize and react to mid-range valences in ways that are different from clear-cut positive or negative valences. What remains unclear is whether these differences are merely magnitude differences along a single valence continuum, or if people really internalize two different metrics when calculating the overall valence signaled by a piece of information. Although the goal in this study was to instantiate a valence-flee range of test scores (i.e., neutral news), the resultant data do not clearly indicate that this happened. It is possible that scorers saw the mid-range values not as being devoid of all valence in any absolute sense, but as signaling a certain degree of negativity and a certain degree of positivity. When the readings on the two metrics are mentally compared, the stronger value nets the resultant judgment of the overall news valence. In other words, a score of 60% might be judged to be moderately high on positivity (i.e., the test-taker was clearly not socially awkward) but still registers a bit of negativity (i.e., the test-taker did not quite achieve superior performance status). In the end, however, the sender calculates the good to outweigh the bad and delivers the news in a way that is quicker than, and less reluctant than, that she would if the news was overwhelmingly bad. This line of reasoning should be explored in the future. 78 CHAPTER 6: GENERAL DISCUSSION Three studies were undertaken in order to better understand the process of delivering bad news from the perspective of the sender. The results of the first experiment (Chapter 3) were consistent with prior MUM effect research in that people differentially delay the news depending on the valence of the news they plan to share. Chapter 4 revealed that the nature of the relationship between the sender and receiver can impact the news delivery process, and that this impact is observable through various verbal communication behaviors (e.g., greetings, recasting the negative impact of the news). Those for whom delivering bad news is routine might be advised to do what they can to develop a closer rapport with the recipient. Doing so not only enhances the receipt of the news by the bearer, but a sense of intimacy also seems to expand the repertoire of communication reactions available to the sender. Senders who are closer to the recipients appear to possess unique ways of refiaming the face threats associated with delivering bad news, which may ease the burden of having to break the news. Finally, Chapter 5 demonstrates that people can and do comprehend news valence along a continuum that associates in a linear fashion with outcomes that accompany the delivery of news of that valence. Unfortunately, little evidence emerged from which to conclude a winner with respect to the reason why people would be reluctant to share bad news. The first experiment failed to find a significant difference between fiiends and strangers who shared bad news. Here, time to response was the dependent variable. Chapter 4 considered four dependent variables, but failed to show an effect for news valence on two of the four (greetings and elaborateness of message). Of the remaining two that were 79 affected by news valence (teasing and repudiation of test validity), only teasing also showed a difference between friends and strangers who shared bad news such that teasing most often occurred when the news was bad and the receiver was a friend. Nonetheless, if the recipient’s emotionality explanation operates, then senders hesitate to share bad news out of an intrapsychic concern for the recipient’s feelings. If this were the case, one would expect to see little or no teasing of the recipient at all. For a sender to make fun of the recipient about her misfortune would hardly communicate empathy with the recipient. If the negative evaluation explanation holds, then senders hesitate to share bad news because they fear the recipient will look upon them unfavorably. For this to be the case, teasing (if it occurs) should only occur when the relationship between the sender and receiver is such that the sender can reliably predict the receiver’s reaction to hearing the news. Thus, the observation that friends teased other friends about their bad news is likely more consistent with the negative evaluation explanation. Researchers point to the possibility that sharing bad news is a process distinct from sharing good news. Uysal and Oner-Ozkan (2007) note that the current explanations for the reluctance to share bad news cannot also plausibly explain the eagerness of some people to share good news. In their view people eagerly pass along good news in active anticipation of a favorable evaluation, whereas the avoidance of bad news transmission is directed defensively toward mitigating the negative evaluation. Further evidence for the separateness of bad news delivery from good news delivery is gleaned when the results of the first and second experiments are compared. However, whereas typical social scientists concern themselves with means, mean 80 differences, and other analyses utilizing measures of central tendency, it is the measures of dispersion that appear to manifest the effect of interest here. In both experiments, there seems to be a discrepancy between the variances associated with sharing good news versus the variances associated with sharing bad news. The wider variances accompany bad news delivery whereas the variances for good news outcomes tend to be relatively narrow. The data from the first experiment demonstrate this pattern (see Table 1). After judging the information to be good news, senders have little to worry about with regard to how they deliver the news. Interestingly, senders sharing good news seem to do so in fairly uniform ways (at least with respect to the variables measured here). However, when the novel information to be shared is evaluated as bad, the face-threatening potential must be assessed, and differences in face threat calculations would give rise to greater variance in the way the bad news is communicated. Surprisingly, the literature seems to be silent with regard to this variance discrepancy between. good news sharing and bad news sharing. Studies that have compared the two valences in a dichotomous fashion seem to have done so without regard to the hetereogeneity of variance issue. For example, Bond and Anderson (1987) perform several ANOVA analyses, but they neither discuss their attention to AN OVA’s statistical assumptions nor do they report the standard deviations along with their means. Almost every test reported in this collection of studies for which news valence was treated observed a heterogeneity of variance between valence groups. One wonders as to the pattern of variances observed in prior studies. Attending to this issue is important regardless of the true nature of good/bad news. If good/bad news delivery are indeed 81 bipolar anchors of the same continuum, then a possible failure to observe basic statistical assumptions leaves the inferential integrity in question. On the other hand, if good news delivery and bad news delivery are indeed distinct processes, then to blindly compare a “good news group” to a “bad news group” is to commit a gross conceptual misspecification. The present report stands in agreement with Uysal and Oner-Ozkan (2007) that the study of bad news delivery would be enhanced by examining it separately from the good news delivery process. Such a move should address the statistical problem of heterogeneity of variance while moving more in line with these processes as they occur in vivo. A final concern involves the way in which MUM effect research has been carried out in the past. The traditional social scientific practice of categorizing variables so as to fit the ANOVA design has begotten a line of bad news research that is not necessarily congruent with reality. The scatterplots from the second experiment are strongly consistent with a continuous representation of news valence. Fitting good news conditions against bad news conditions as in the past obscures the potentially meaningful middle range of valence and its corresponding consequents. Whether or not this middle range truly represents neutral news remains up for debate. Nonetheless, the abandonment of dichotomous thinking with regard to the process of delivering news represents an important step if news delivery research is to proceed most fruitfully. Perhaps the nature of the news plays a larger role in determining the character of the bad news delivery process. Future research should focus on the various dimensions of bad news itself, developing ways to experimentally manipulate dimensions like relevance and extremity. This would firrther imply that different types of bad news be investigated. 82 Taking age and experience into consideration for their potential impact on the bad news delivery process would broaden scholarly understanding and improve generalizability. It should be recalled that past research focused primarily on recipient’s concerns. The present studies attended to issues associated with the sender. Future research should attempt to examine the effects of bad news delivery on both senders and recipients in the same study. In this way, the experiences of the recipient can be more reliably mapped onto the experiences of the sender, and researchers come closer to the true interactive nature of this important communication event. 83 NOTES I It should be noted that the ability to identify with another’s emotions plays a role here as well, but in this case the ability to gauge the receiver’s emotions is marshaled toward avoiding being viewed as heartless. 2It should be noted that these arguments are derived primarily on logical grounds, and that they represent the ideal situation. These predictions should hold if all other things are equal. In reality, however, other situational factors probably contribute to the actual motivation behind the reluctance to share bad news. For example, there is reason to believe that the type of bad news to be transmitted might dictate the reason for reluctance, even within a close relationship. Within romantic pairings alone, Dibble and Gabriel (2004) categorize at least four types of messages, and Wagoner and Waldron (1999) identify just as many types within superior-subordinate settings. These message types primarily differ with respect to who was responsible for the bad news consequences. If one considers a man ruminating over whether he will reveal to his girlfiiend that he had kissed another woman during a Spring Break getaway, it follows that he may very well have a legitimate fear of being evaluated negatively by his girlfiiend. (In fact, he should probably expect to be evaluated negatively by his girlfriend!) However, in the case of a wife breaking to her husband the news of the death of her husband’s close friend, she may truly feel badly for her husband’s loss, which might align more with the recipient emotionality hypothesis. When considered in the presence of other factors, both explanations might be operating at different times. 3With regard to the 6 cases about which the coders initially disagreed, the disagreement was primarily due to inconsistent hearing of all of the scorer’s dialogue, not so much because the coders disagreed as to what they take to be teasing. In most cases, replaying the videotaped interaction was sufficient to resolve the discord. 4One scorer did greet the test-taker by saying Holla, and, because this is a current cultural variation of Hello, it was regarded as a greeting. 5It is recognized that certain scholars (e.g., Watts, 2003) take issue with the idea of equating facework with politeness. The thrust of their contention is that speaker-oriented models blindly assume that the speaker’s communication strategies will be heard as polite, when this is not necessarily the case. (The ranges listed here were presented to senders as a heuristic with which to interpret the test score. However, because the random number generator was limited to values from 10 to 90 inclusive, these cutoffs cleanly divided the possible field of scores into three ranges, each containing 27 values: good news, 90-64; neutral news, 63-37; and bad news, 36-10. This enabled news valence to be analyzed categorically as well as continuously. 84 APPENDICES 85 APPENDIX A COMMUNICATION IN RELATIONSHIPS PRE-QUESTIONNAIRE Instructions: Throughout this questionnaire, any use of the word “partner” refers to the person with whom you’ve been paired for this study. (Note: This may not be the person with whom you reported to the laboratory.) Think about your relationship with your partner. Please circle the picture below that best describes your relationship with that person. (a (a Instructions: The following questions again refer to your relationship with your partner. Please think about your relationship with your partner when responding to the following questions. Please respond to the following statements using this scale. Simply print the number of your choice in the space preceding each item. 1 2 3 4 5 6 7 Strongly Strongly Disagree Agree 2. My relationship with my partner is close. 3. When we are apart, I miss my partner a great deal. 4. My partner and l disclose important personal things to each other. 5. My partner and I have a strong connection. 6. My partner and I want to spend time together. 7. I’m sure of my relationship with my partner. 8. My partner is a priority in my life. 86 9. My partner and I do a lot of things together. 10. When I have free time I choose to spend it alone with my partner. 1 1. I think about my partner a lot. 12. My relationship with my partner is important in my life. 13. I consider my partner when making important decisions. Instructions: The next scale consists of a number of words that describe different feelings and emotions. Read each item and then print the appropriate answer in the box that precedes that word. Indicate to what extent you feel this way right now. Use the following scale to record your answers. 1 2 3 4 5 Very slightly or A little Moderately Quite a bit Extremely not at all Interested Irritable Distressed Alert Excited Ashamed Upset Inspired Strong Nervous Guilty Determined Scared Attentive Hostile Jittery Enthusiastic Active Proud Afraid Instructions: Please respond honestly to the following statements using the scale below. Remember: The words “my partner” refer to the person with whom you’ve been paired for this study. This is not necessarily the person with whom you reported to the laboratory. 1 3 4 5 6 7 9 Disagree Agree Completely Completely 34. When I am with my partner, we are almost always in the same mood. 35. I think that my partner is unusually well-adjusted. 87 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. I would highly recommend my partner for a responsible job. In my opinion, my partner is an exceptionally mature person. I have great confidence in my partner’s good judgment. Most people would react very favorably to my partner after a brief acquaintance. I think that my partner and l are quite similar to each other. I would vote for my partner in a class or group election. I think that my partner is one of those people who quickly wins respect. I feel that my partner is an extremely intelligent person. My partner is one of the most likable people I know. My partner is the sort of person whom I myself would like to be. It seems to me that it is very easy for my partner to gain admiration. Instructions: Please respond to the following statements using this scale. Simply print your response in the space preceding each item. l 2 3 4 5 6 7 Strongly Strongly Disagree Agree 47. When I see someone being taken advantage of, I feel kind of protective toward them. 48. When I see someone being treated unfairly, I sometimes don’t feel much pity for them. 49. I often have tender, concerned feelings for people less fortunate than me. 50. I would describe myself as a pretty soft-hearted person. 51. Other people’s misfortunes do not usually disturb me a great deal. 88 52. I am often touched by the things that I see happen. Thank you for your honest responses. Someone will be along shortly to provide further instructions. 89 APPENDIX B COMMUNICATION IN RELATIONSHIPS POST-QUESTIONNAIRE For Test Scorers Instructions: Please respond honestly to the following statements using this scale. Simply print the number of your choice in the space preceding each item. 1 2 3 4 5 6 Strongly Strongly Disagree Agree __ 1. The test-taker did fairly well. __ 2. The test-taker did fairly poorly. __ 3. The test-taker likely viewed my feedback as bad news. _ 4. The test-taker likely viewed my feedback as good news. __ 5. The test-taker will likely feel positively about my feedback. __ 6. The test-taker will likely feel negatively about my feedback. Instructions: Please respond honestly to the following statements using this scale. Indicate your response by writing the number of your choice in the space that precedes each item. 1 2 3 4 5 6 7 Strongly Strongly disagree agree 7. I was happy to give the test-taker her result. 8. I made sure to double check the test-taker’s answers. 9. I am worried about what the test-taker will think of me. 10. I am concerned that the test-taker will consider me to be insensitive. 9O 1 1. I am worried about how these results will affect the test-taker emotionally. 12. I believe I have created a painful experience for the test-taker. Instructions: From this point on, any use of the word “partner” refers to the person with whom you’ve been paired for this study. (Note: This may not be the person with whom you reported to the laboratory.) Think about your relationship with your partner. Please circle the picture below that best describes your relationship with that person. (0) (o 6 Instructions: The following questions refer to your relationship with your partner. Please think about your relationship with your partner when responding to the following questions. Please respond to the following statements using this scale: 1 2 3 4 5 6 7 Strongly Strongly Disagree Agree 14. My relationship with my partner is close. 15. When we are apart, I miss my partner a great deal. 16. My partner and I disclose important personal things to each other. 17. My partner and I have a strong connection. 18. My partner and I want to spend time together. 19. I’m sure of my relationship with my partner. 20. My partner is a priority in my life. 91 21. My partner and I do a lot of things together. 22. When I have free time I choose to spend it alone with my partner. 23. I think about my partner a lot. 24. My relationship with my partner is important in my life. 25. I consider my partner when making important decisions. Instructions: The next scale consists of a number of words that describe different feelings and emotions. Read each item and then mark the appropriate answer in the space that precedes that word. Indicate to what extent you feel this way right now. Use the following scale to record your answers. 1 2 3 4 5 Very slightly A little Moderately Quite a bit Extremely or not at all Interested Irritable Distressed Alert Excited Ashamed Upset Inspired Strong Nervous Guilty Determined Scared Attentive Hostile Jittery Enthusiastic Active Proud Afraid Instructions: Please respond honestly to the following statements using the scale below. Note: The words “my partner” refer to the person with whom you’ve been paired for this study—not the person with whom you reported to the laboratory. 1 2 3 4 5 6 7 9 Disagree Agree Completely Completely 46. When I am with my partner, we are almost always in the same mood. 47. I think that my partner is unusually well-adjusted. 92 48. I would highly recommend my partner for a responsible job. 49. In my opinion, my partner is an exceptionally mature person. 50. I have great confidence in my partner’s good judgment. 51. Most people would react very favorably to my partner after a brief acquaintance. 52. I think that my partner and I are quite similar to each other. 53. I would vote for my partner in a class or group election. 54. I think that my partner is one of those people who quickly wins respect. 55. I feel that my partner is an extremely intelligent person. 56. My partner is one of the most likable people I know. 57. My partner is the sort of person whom I myself would like to be. 58. It seems to me that it is very easy for my partner to gain admiration. Instructions: Please complete the following demographic information. 59. 60. 61. 62. Your age: years Please circle below the label that best describes your relationship with your partner: Stranger Acquaintance Casual friend Close Friend Best Friend Romantically Involved This time, circle the label below that best describes your relationship to the person with whom you originally reported to the lab: Stranger Acquaintance Casual fiiend Close Friend Best Friend Romantically Involved How long have you known the friend with whom you reported to the lab? Years Months 93 63. How long have you been at the status of relationship indicated in question 61 with your fiiend? Years Months Upon finishing, please bring completed questionnaire to experimenter in the hallway. Thank you again for your participation. 94 APPENDIX C COMMUNICATION IN RELATIONSHIPS POST-QUESTIONNAIRE For Test Takers Instructions: Please respond honestly to the following statements using this scale. Simply print the number of your choice in the space preceding each item. l 2 3 4 5 6 Strongly Strongly Disagree Agree l. I did fairly well. 2. I did fairly poorly. 3. The feedback I received was bad news. 4. The feedback I received was good news. 5. I feel positively about the feedback I received. 6. I feel negatively about the feedback I received. Instructions: From this point on, any use of the word “partner” refers to the person with whom you’ve been paired for this study. (Note: This may not be the person with whom you reported to the laboratory.) Think about your relationship with your partner. Please circle the picture below that best describes your relationship with that person. (it) (a c) 95 Instructions: The following questions refer to your relationship with your partner. Please think about your relationship with your partner when responding to the following questions. Please respond to the following statements using this scale: 1 2 3 4 5 6 7 Strongly Strongly Disagree Agree 8. My relationship with my partner is close. 9. When we are apart, I miss my partner a great deal. 10. My partner and I disclose important personal things to each other. 1 1. My partner and I have a strong connection. 12. My partner and I want to spend time together. 13. I’m sure of my relationship with my partner. 14. My partner is a priority in my life. 15. My partner and I do a lot of things together. 16. When I have free time I choose to spend it alone with my partner. 17. I think about my partner a lot. 18. My relationship with my partner is important in my life. 19. I consider my partner when making important decisions. Instructions: The next scale consists of a number of words that describe different feelings and emotions. Read each item and then mark the appropriate answer in the space that precedes that word. Indicate to what extent you feel this way right now. Use the following scale to record your answers. 1 2 3 4 5 Very slightly A little Moderately Quite a bit Extremely or not at all 96 Interested Irritable Distressed Alert Excited Ashamed Upset Inspired Strong Nervous Guilty Determined Scared Attentive Hostile Jittery Enthusiastic Active Proud Afraid Instructions: Please respond honestly to the following statements using the scale below. Note: The words “my partner” refer to the person with whom you’ve been paired for this study—not the person with whom you reported to the laboratory. 1 2 3 4 5 6 7 9 Disagree Agree Completely Completely 40. When I am with my partner, we are almost always in the same mood. 51. My partner is the sort of person whom I myself would like to be. 97 41. I think that my partner is unusually well-adjusted. 42. I would highly recommend my partner for a responsible job. 43. In my opinion, my partner is an exceptionally mature person. 44. I have great confidence in my partner’s good judgment. 46. I think that my partner and I are quite similar to each other. 47. I would vote for my partner in a class or group election. 49. I feel that my partner is an extremely intelligent person. 50. My partner is one of the most likable people I know. 45. Most people would react very favorably to my partner after a brief acquaintance. 48. I think that my partner is one of those people who quickly wins respect. 52. It seems to me that it is very easy for my partner to gain admiration. Instructions: Please complete the following demographic information. 53. Your age: years 54. Please circle below the label that best describes your relationship with your Linn—61': Stranger Acquaintance Casual fiiend Close Friend Best Friend Romantically Involved 55. This time, circle the label below that best describes your relationship to the person with whom you originally reported to the lab: Stranger Acquaintance Casual fiiend Close Friend Best Friend Romantically Involved 56. How long have you known the fiiend with whom you reported to the lab? Years Months 57. How long have you been at the status of relationship indicated in question 55 with your friend? Years Months Upon finishing, please bring completed questionnaire to experimenter in the hallway. Thank you again for your participation. 98 APPENDIX D SOCIAL PERCEPTION TASK Instructions: Each of the following 20 items features a picture of two people. For each item, indicate your assessment of the relationship between the people pictured by selecting one of the options below. Please record your answer by filling in the letter of your choice on the bubble sheet provided. To assist us in scoring, please do not make any other marks on the bubble sheet. For each pair, bubble in (A) if you think the pair are strangers, (B) if you think the pair are apquaintances, or (C) if you think the pair are romantic partners. 99 100 3.. 1 15. 17. l9. APPENDIX E SCORING THE SOCIAL PERCEPTION TEST Thank you for assisting us with scoring the Social Perception Test. Your partner will be bringing his or her completed scoresheet for you to process. Please direct the test-taker to have a seat in the cushy chair (please take care not to move any furniture). The computer in front of you will calculate the result, but we need you to enter your partner’s answers. Please do not move the computer as you process the scoresheet electronically as follows: 1. For each of the 20 items, use the mouse to key your partner’s responses into the appropriate. Simply click over the letter of the response. All items must be answered before you can obtain the score. 2. Click GET SCORE near the bottom right of the screen. This will bring up your partner’s score. Scores should be interpreted according to the following criteria: Score range Result 64%-100% Superior performance 37%—63% Average performance 0%-3 6% Inferior performance Those who receive superior scores (64%-100%) seem to be exceptionally skilled at reading other people. Inferior scores (0%-36%) generally indicate people who are socially awkward because they are less skilled at reading the nonverbal signals of others. Interpreting nonverbal cues is important for building and maintaining healthy relationships. After you have completed scoring the test, we would like you to communicate the result to the test-taker in person. Please share the results of her performance. You are free to take as long as you need. When you are finished sharing your partner’s results, please direct the test-ta_ker to return to his or her original room. You’ll remain in your room. Please make yourself comfortable. The experimenter will be in shortly. 102 The experimenter will bring one final questionnaire for each of you to complete. APPENDIX F COMMUNICATION AND SOCIAL PERCEPTION SCORER QUESTIONNAIRE Please print Test-Taker’s score here Instructions: Please respond honestly to the following statements using this scale. Simply print your response in the space preceding each item. 1 2 3 4 5 6 7 Strongly Strongly Disagree Agree 1. _ The test-taker did fairly well. 2. __ The test-taker did fairly poorly. 3. __ The test-taker likely viewed my feedback as bad news. 4. _ The test-taker likely viewed my feedback as good news. 5. __ The test-taker will likely feel positively about my feedback. 6. _ The test-taker will likely feel negatively about my feedback. Instructions: The next scale consists of a number of words that describe different feelings and emotions. Read each word and then print the appropriate answer in the box that precedes that word. Indicate to what extent you feel this way right now. Use the following scale to record your answers. l 2 3 4 5 Very slightly A little Moderately Quite a bit Extremely or not at all Interested Irritable Distressed Alert Excited Ashamed Upset Inspired Strong Nervous Guilty Determined Scared Attentive Hostile Jittery 103 Enthusiastic Active Proud Afraid Instructions: Please respond honestly to the following statements using this scale. Simply print your response in the space preceding each item. 1 2 3 4 5 6 7 Strongly Strongly Disagree Agree 27. I didn’t want to share my partner’s score with them. 28. __ I was eager to speak to my partner about their result. 29. _ I felt reluctant to tell my partner how they did. 30. __ I wasn’t looking forward to giving my partner their result. 31. _ I wasn’t sure what I was going to say to my partner. 32. __ I experienced no hesitation to share my partner’s score with them. Instructions: This section contains statements about personality traits. Please choose the member of each pair of statements that you believe is more descriptive of yourself. Circle the letter (either “A” or “B”) that corresponds to the statement that is most like you. 33. A. I am not concerned with looking good at all times. 34. 35. 36. 37. 38. B. I am concerned with looking good at all times. A. I tend to exaggerate my achievement so others will accept and respect me. B. I try not to exaggerate my achievements, but sometimes my pride appears bragging. A. When I experience criticism and defeat, I learn about areas which need improvement. B. When I experience criticism and defeat, it reinforces my weaknesses and inadequacies. . When 1 give something my all but come up short, others still appreciate it. A B. When I give something my all but come up short, others don’t appreciate it. A . I usually go out of my way to make sure that a person knows that I helped him or her. : . I like to help others even if they don’t know that I did it. B A. Even when I am by myself, I rarely feel alone and uneasy. B . Even when I am in a group of friends, I often feel alone and uneasy. 104 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. A. It would be nice to earn the attention and admiration of others. B. It is very important that others pay attention to and admire what I do. A. When friends play a joke on me, 1 try to enjoy it with them. B. When friends play a joke on me, I wonder if they are laughing at me. A. I foresee the day when life will be ideal. B. I foresee the day when life will be better for me. A. It is partially my responsibility to be aware of others’ needs. B. It is my responsibility to meet my own needs and nobody else’s. A. My life seems to have a normal number of ups and downs. B. My life seems to have more than its share of downs. A. I enjoy being with just a few select fiiends at parties. B. I enjoy being the center of every one’s attention at parties. A. It is better to be yourself than to be popular. B. It is better to be popular than to be yourself. A. Given the opportunity, I will take advantage of others without feeling guilty. B. I find it difficult to take advantage of others. A. I try to appreciate it when others have what I lack. B. I resent others having what I lack. A. I do not expect others to go out of their way for me. B. I do expect others to go out of their way for me. A. When I have been irresponsible, I try to lie my way out of it. B. When I have been irresponsible, I am prepared to face the consequences. A. How I appear to people is more important than howl feel. B. How I appear to people is not as important as how I feel. A. Impressing others is important to get ahead. B. Impressing others is okay only when it is truthfirl self-expression. A. I do not deceive others in order to obtain something. B. I may deceive others by being friendly when I really dislike them in order to obtain something. A. I tend to become involved in relationships in which I alternatively adore and despise my partner. B. I tend to become involved in stable, long-lasting relationships. 105 54. A. I am bound by the principle of fairness only when it is to my benefit. B. I am always bound by the principle of fairness. 55. A. Life has been fairly exciting, but am still not satisfied with it. B. Life has been fairly exciting, and I am content with it. 56. A. Defeat or disappointment usually make me try harder. B. Defeat or disappointment usually shame or anger me, but I try not to show it. Instructions: Please respond honestly to the following statements using this scale. Simply print your response in the space preceding each item. l 2 3 4 5 6 7 Strongly Strongly Disagree Agree 57. _ It is important for me to look good in front of other people. 58. _ Maintaining a positive self-image is important to me. 59. __ Other people having a good impression of me is important to me. 60. __ Making a good impression is important to me. 61. __ It is important to be respected by others. 62. __ I want my privacy to be respected. 63. __ I don’t want my personal life discussed. 64. __ I want other people to stay out of my business. 65. __ My boundaries should be respected. 66. __ I prefer to be able to do what I want, when I want. 67. __ It’s important not to put myself in a situation where I would feel indebted to someone. 68. __ I don’t like to be obligated to do anything. 69. __ I prefer to have control over my personal time and space. 70. __ I don’t like having to owe anyone anything. 106 71. I try to help other people feel good about themselves. 72. _ Helping others maintain a positive image of themselves is important to me. 73. __ I make an effort to not let others embarrass themselves. 74. _ It is important for me to show respect for others. 75. __ It is important for me not to make other people look bad. 76. __ When in public, I hold back my criticism of others so that they will not look bad. 77. __ Even if someone annoys me, I keep my thoughts to myself so as to avoid hurting their feelings. 78. __ I try to avoid telling other people what they should do. 79. __ I try not to impose on others. 80. _ It is important to respect other people’s right to do what they want to do. 81. _ I try not to invade other people’s personal matters. 82. _ I try not to tell others what to do. Instructions: Please respond honestly to the following statements using this scale. Simply print your response in the space preceding each item. 1 2 3 4 5 6 7 Strongly Strongly Disagree Agree 83. Before criticizing somebody, I try to imagine how I would feel if I were in their place. 84. If I’m sure I’m right about something, I don’t waste much time listening to other people’s arguments. 85. I sometimes try to understand my friends better by imagining how things look from their perspective. 107 86. I believe that there are two sides to every question, and I try to look at them both. 87. I sometimes find it difficult to see things from the other person’s point of view. 88. I try to look at everybody’s side of a disagreement before I make a decision. 89. When I’m upset at someone, I usually try to put myself “in that person’s shoes” for awhile. 90. __ When I see someone being taken advantage of, I feel kind of protective toward 316?“:— When I see someone being treated unfairly, I sometimes don’t feel very much pity for them. 92. __ I often have tender, concerned feelings for people less fortunate than me. 93. __ I would describe myself as a pretty soft-hearted person. 94. __ Sometimes I don’t feel sorry for other people when they are having problems. 95. __ Other people’s misfortunes usually do not disturb me a great deal. 96. __ I am often quite touched by things that I see happen. Just a few more. These last items ask for some basic demographic information. Please indicate your response in the appropriate space. 97. Your age in years 98. Your sex: (Circle one) Male Female 99. The Test-Taker’s Sex: (Circle one) Male Female 100. Race/ethnicity: (Check all that apply) __ African American Asian Native American __ Latino/Latina __ European American (White) _ Other (Please list) 108 APPENDIX G COMMUNICATION AND SOCIAL PERCEPTION TEST-TAKER QUESTIONNAIRE Please print your score here Instructions: Please respond honestly to the following statements using this scale. Simply print your response in the space preceding each item. 1 2 3 4 5 6 7 Strongly Strongly Disagree Agree 1. __ I did fairly well. 2. __ I did fairly poorly. 3. __ The feedback I received was bad news. 4. _ The feedback I received was good news. 5. __ I feel positively about the feedback I received. 6. __ I feel negatively about the feedback I received. Instructions: The next scale consists of a number of words that describe different feelings and emotions. Read each word and then print the appropriate answer in the box that precedes that word. Indicate to what extent you feel this way right now. Use the following scale to record your answers. 1 2 3 4 5 Very slightly A little Moderately Quite a bit Extremely or not at all Interested Irritable Distressed Alert Excited Ashamed Upset Inspired Strong Nervous Guilty Determined Scared Attentive Hostile Jittery 109 Enthusiastic Active Proud Afraid Instructions: Please respond honestly to the following statements using this scale. Simply print your response in the space preceding each item. l 2 3 4 5 6 7 Strongly Strongly Disagree Agree 27. __ The Scorer engaged in more eye contact with me than most other people. 28. __ The Scorer’s body was more tense than most other people’s. 29. __ The Scorer seemed eager to talk to me. 30. __ The Scorer smiled more than most other people. 31. __ The Scorer seemed more distant fi‘om me than most other people. 32. __ The Scorer seemed more vocally expressive than most other people. 33. __ The Scorer had a more relaxed body position than most other people. 34. _ The Scorer directed his/her body position less toward me than most people usually do. 35. __ The Scorer seemed reluctant to talk to me. Instructions: This section contains statements about personality traits. Please choose the member of each pair of statements that you believe is more descriptive of yourself. Circle the letter (either “A” or “B”) that corresponds to the statement that is most like you. 36. A. I am not concerned with looking good at all times. B. I am concerned with looking good at all times. 37. A. I tend to exaggerate my achievement so others will accept and respect me. B. I try not to exaggerate my achievements, but sometimes my pride appears like bragging. 38. A. When I experience criticism and defeat, I learn about areas which need improvement. B. When I experience criticism and defeat, it reinforces my weaknesses and inadequacies. 39. A. When I give something my all but come up short, others still appreciate it. B. When I give something my all but come up short, others don’t appreciate it. 110 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. 55. A. I usually go out of my way to make sure that a person knows that I helped him or her. B. I like to help others even if they don’t know that I did it. A. Even when I am by myself, I rarely feel alone and uneasy. B. Even when I am in a group of friends, I often feel alone and uneasy. A. It would be nice to earn the attention and admiration of others. B. It is very important that others pay attention to and admire what I do. A. When friends play a joke on me, I try to enjoy it with them. B. When friends play a joke on me, I wonder if they are laughing at me. A. I foresee the day when life will be ideal. B. I foresee the day when life will be better for me. A. It is partially my responsibility to be aware of others’ needs. B. It is my responsibility to meet my own needs and nobody else’s. A. My life seems to have a normal number of ups and downs. B. My life seems to have more than its share of downs. A. I enjoy being with just a few select fiiends at parties. B. I enjoy being the center of every one’s attention at parties. A. It is better to be yourself than to be popular. B. It is better to be popular than to be yourself. A. Given the opportunity, I will take advantage of others without feeling guilty. B. I find it difficult to take advantage of others. A. I try to appreciate it when others have what I lack. B. I resent others having what I lack. A. I do not expect others to go out of their way for me. B. I do expect others to go out of their way for me. A. When I have been irresponsible, I try to lie my way out of it. B. When I have been irresponsible, I am prepared to face the consequences. A. How I appear to people is more important than how I feel. B. How I appear to people is not as important as how I feel. A. Impressing others is important to get ahead. B. Impressing others is okay only when it is truthful self-expression. A. I do not deceive others in order to obtain something. B. I may deceive others by being friendly when I really dislike them in order to obtain something. 111 56. A. I tend to become involved in relationships in which I alternatively adore and despise my partner. B. I tend to become involved in stable, long-lasting relationships. 57. A. 1am bound by the principle of fairness only when it is to my benefit. B. I am always bound by the principle of fairness. 58. A. Life has been fairly exciting, but am still not satisfied with it. B. Life has been fairly exciting, and I am content with it. 59. A. Defeat or disappointment usually make me try harder. B. Defeat or disappointment usually shame or anger me, but I try not to show it. Instructions: Please respond honestly to the following statements using this scale. Simply print your response in the space preceding each item. 1 2 3 4 5 6 7 Strongly Strongly Disagree Agree 60. __ It is important for me to look good in front of other people. 61. __ Maintaining a positive self-image is important to me. 62. __ Other people having a good impression of me is important to me. 63. _ Making a good impression is important to me. 64. __ It is important to be respected by others. 65. _ I want my privacy to be respected. 66. _ I don’t want my personal life discussed. 67. __ I want other people to stay out of my business. 68. __ My boundaries should be respected. 69. _ I prefer to be able to do what I want, when I want. 70. __ It’s important not to put myself in a situation where I would feel indebted to someone. 112 71. I don’t like to be obligated to do anything. 72. _ I prefer to have control over my personal time and space. 73. _ I don’t like having to owe anyone anything. 74. __ I try to help other people feel good about themselves. 75. __ Helping others maintain a positive image of themselves is important to me. 76. __ I make an effort to not let others embarrass themselves. 77. __ It is important for me to show respect for others. 78. __ It is important for me not to make other people look bad. 79. __ When in public, I hold back my criticism of others so that they will not look bad. 80. __ Even if someone annoys me, I keep my thoughts to myself so as to avoid hurting their feelings. 81. _ I try to avoid telling other people what they should do. 82. __ I try not to impose on others. 83. _ It is important to respect other people’s right to do what they want to do. 84. __ I try not to invade other people’s personal matters. 85. _ I try not to tell others what to do. Instructions: Please respond honestly to the following statements using this scale. Simply print your response in the space preceding each item. 1 2 3 4 5 6 7 Strongly Strongly Disagree Agree 86. Before criticizing somebody, I try to imagine how I would feel if I were in their place. 113 87. If I’m sure I’m right about something, I don’t waste much time listening to other people’s arguments. 88. I sometimes try to understand my fiiends better by imagining how things look from their perspective. 89. __ I believe that there are two sides to every question, and I try to look at them both. 90. __ I sometimes find it difficult to see things from the other person’s point of view. 91. __ I try to look at everybody’s side of a disagreement before I make a decision. 92. __ When I’m upset at someone, I usually try to put myself “in that person’s shoes” for awhile. 93. __ When I see someone being taken advantage of, I feel kind of protective toward them. 94. __ When I see someone being treated unfairly, I sometimes don’t feel very much pity for them. 95. __ I often have tender, concerned feelings for people less fortunate than me. 96. __ I would describe myself as a pretty soft—hearted person. 97. __ Sometimes I don’t feel sorry for other people when they are having problems. 98. __ Other people’s misfortunes usually do not disturb me a great deal. 99. __ I am often quite touched by things that I see happen. Just a few more. These last items ask for some basic demographic information. Please indicate your response in the appropriate space. 100. Your age in years 101. Your sex: (Circle one) Male Female 102. The Scorer’s sex (Circle one) Male Female 103. Race/ethnicity: (Check all that apply) __ Afiican American Asian 114 Native American Latino/ Latina European American (White) Other (Please list) Thank you so much for participating in this study! 115 REFERENCES 116 REFERENCES Afifi, T. D., & Schrodt, P. (2003). “Feeling caught” as a mediator of adolescents’ and young adults’ avoidance and satisfaction with their parents in divorced and non-divorced households. Communication Monographs, 70, 142-173. Ahmed, S. M. S., & Bigelow, B. (1993). Latency period as affected by news content. The Journal of Social Psychology, 133, 361 -364. Andersen, J. F., Andersen, P. A., & Jensen, A. D. (1979). The measurement of nonverbal immediacy. Journal of Applied Communication Research, 7, 153-180. Anderson, C., Keltner, D., & John, O. P. (2003). Emotional convergence between people over time. Journal of Personality and Social Psychology, 84, 1054-1068. Aron, A., Aron, E. N., & Smollan, D. (1992). Inclusion of other in the self scale and the structure of interpersonal closeness. Journal of Personality and Social Psychology, 63, 596-612. Aron, A., Aron, E. N., Tudor, M., & Nelson, G. (1991). Close relationships as including other in the self. Journal of Personality and Social Psychology, 60, 241 - 253. Aron, A., & Fraley, B. (1999). Relationship closeness as including the other in the self: Cognitive underpinnings and measures. Social Cognition, I 7, 140-160. Beach, W. A. (2002). Between dad and son: Initiating, delivering, and assimilating bad cancer news. Health Communication, 14, 271-298. Berger, C. R., & Calabrese, R. J. (1975). Some explorations in initial interaction and beyond: Toward a developmental theory of interpersonal communication. Human Communication Research, 1, 99-112. Berscheid, E., Snyder, M., & Omoto, A. M. (1989). The relationship closeness inventory: Assessing the closeness of interpersonal relationships. Journal of Personality and Social Psychology, 5 7, 792-807. Bond, C. F., Jr., & Anderson, E. L. (1987). The reluctance to transmit bad news: Private discomfort or public display? Journal of Experimental Social Psychology, 23, 176-187. Brown, R, & Levinson, S. (1987). Politeness: Some universals in language usage. Cambridge, England: Cambridge University Press. 117 Burgoon, J. K. (1976). The Unwillingness-to-Communicate Scale: Development and validation. Communication Monographs, 43, 60-69. Carter, J. H., Nutt, J. G., & Carter, E. H. (2007). Breaking bad news: Patients’ perspectives on being given the diagnosis of Parkinson’s disease. Neurology, 68(12), A230-A231. Casarett, D. J ., & Quill, T. E. (2007). “I’m not ready for hospice”: Strategies for timely and effective hospice discussions. Annals of Internal Medicine, 146(6), 443-449. Cohen, J. (1960). A coefficient of agreement for nominal scales. Educational and Psychological Measurement, 20, 37-46. Cohen, J ., Cohen, R, West, S. G., & Aiken, L. S. (2003). Applied multiple regression/correlation analysis for the behavioral sciences. Mahwah, NJ: Lawrence Erlbaurn Assoc. Davis, M. H. (1983). Measuring individual differences in empathy: Evidence for a multidimensional approach. Journal of Personality and Social Psychology, 44, 113-126. Dibble, J. L., & Gabriel, S. M. (2004). “I can ’t make our date Exploring a taxonomy of bad news messages in romantic relationships. Paper presented at the annual conference of the Central States Communication Association, Cleveland, OH. Doherty, R. W. (1997). The emotional contagion scale: A measure of individual differences. Journal of Nonverbal Behavior, 21, 131-154. Dunnet, C. W. (1980). Pairwise multiple comparisons in the unequal variance case. Journal of the American Statistical Association, 75, 796-800. Dworkind, M. A. (2006). Giving bad news: How to give hope in a hopeless situation. Journal of Palliative Care, 22(3), 221. Eberhardt-McKee, T. L., & Ptacek, J. T. (2001). I’m afraid I have something bad to tell you: Breaking bad news from the perspective of the giver. Journal of Applied Social Psychology, 31, 246-273. Eggly, S., Penner, L., Albrecht, T., Cline, R. J. W., Foster, T., Naughton, M., et al. (2006). Discussing bad news in the outpatient oncology clinic: Rethinking current communication guidelines. Journal of Clinical Oncology, 24, 716-719. 118 Fallowfield, L., & Jenkins, V. (2004). Communicating sad, bad, and difficult news in medicine. The Lancet, 363, 312-319. Floyd, K., & Parks, M. R. (1995). Manifesting closeness in the interactions of peers: A look at siblings and friends. Communication Reports, 8, 69-76. Fisher, C. D. (1979). Transmission of positive and negative feedback to subordinates: A laboratory investigation. Journal of Applied Psychology, 64, 533-540. Freese, J, & Maynard, D. W. (1998). Prosodic features of bad news and good news in conversation. Language in Society, 27, 195-219. Galician, M., & Vestre, N. D. (1987). Effects of ‘good news’ and ‘bad news’ on newscast image and community image. Journalism Quarterly, 64, pp. 399-405, 525. Goffman, E. (1967). Interaction ritual. New York: Pantheon Books. Greenberg, L. W., Ochsenschlager, D., O’Donnell, R., Mastruserio, J ., & Cohen, G. J. (1999). Communicating bad news: A pediatric department’s evaluation of a simulated intervention. Pediatrics, 1 03, 1210-121 7. Gump, B. B., & Kulik, J. A. (1997). Stress, affiliation, and emotional contagion. Journal of Personality and Social Psychology, 72, 305-319. Haselton, M. G., & Buss, D. M. (2000). Error management theory: A new perspective on biases in cross-sex mind reading. Journal of Personality and Social Psychology, 78, 81-91. Hatfield, E., Cacioppo, J. T., & Rapson, R. L. (1992). Primitive emotional contagion. In M. S. Clark (Ed.), Review of Personality and Social Psychology, 14, 151-177. Howes, M. J ., Hokanson, J. E., & Loewenstein, D. A. (1985). Induction of depressive affect after prolonged exposure to a mildly depressed individual. Journal of Personality and Social Psychology, 49, 11 10-1113. Ito, T. A., Larsen, J. T., Smith, N. K., & Cacioppo, J. T. (1998). Negative information weighs more heavily on the brain: The negativity bias in evaluative categorizations. Journal of Personality and Social Psychology, 75, 887-900. Johnson, R. E., Conlee, M. C., & Tesser, A. (1974). Effects of similarity of fate on bad news transmission: A reexamination. Journal of Personality and Social Psychology, 29, 644-648. 119 Kenny, D. A., & Kashy, D. A. (1994). Enhanced co-orientation in the perception of fiiends: A social relations analysis. Journal of Personality and Social Psychology, 67, 1024-1033. Kowalski, R. M., Howerton, E., & McKenzie, M. (2001). Permitted disrespect: Teasing in interpersonal interactions. In R. M. Kowalski (Ed.), Behaving badly (pp. 177- 202). Washington, DC: American Psychological Association. Landis, J. R., & Koch, G. G. (1977). The measurement of observer agreement for categorical data. Biometrics, 33, 159-174. Maxwell, G. M. (1985). Behaviour of lovers: Measuring the closeness of relationships. Journal of Social and Personal Relationships, 2, 215-238. Mehrabian, A. (1966). Immediacy: An indicator of attitudes in linguistic communication. Journal of Personality, 34, 26-34. Mehrabian, A. (1971). Silent messages. Belmont, CA: Wadsworth. Miller, G. R., & Steinberg, M. (1975). Between people: A new analysis of interpersonal communication. Chicago, IL: Science Research Associates. Miller, R. S., & Lefcourt, H. M. (1982). The assessment of intimacy. Journal of Personality Assessment, 46, 514-518. Newhagen, J. E., & Reeves, B. (1992). The evening’s bad news: Effects of compelling negative television news images on memory. Journal of Communication, 42, 25-41. Ptacek, J. T., & Eberhardt, T. L. (1996). Breaking bad news: A review of the literature. JAMA, 276, 496-502. Ptacek, J. T., Ptacek, J. J ., & Ellison, N. M. (2001). “I’m sorry to tell you...” Physician’s reports of breaking bad news. Journal of Behavioral Medicine, 24, 205-217. Rosen, S., & Tesser, A. (1970). On reluctance to communicate undesirable information: The MUM effect. Sociometry, 33, 253-263. Rosen, S., & Tesser, A. (1972). Fear of negative evaluation and the reluctance to transmit bad news. The Journal of Communication, 22, 124-141. Sacks, H. (1975). Everyone has to lie. In M. Sanches & B. G. Blount (Eds), Sociocultural dimensions of language use (pp. 57-79). New York: Academic Press. 120 Sacks, H. (1995). Lectures on conversation. Cambridge, MA: Blackwell Publishers. Schlenker, B. R. (1984). Identities, identifications, and relationships. In V. J. Derlega (Ed.), Communication, intimacy, and close relationships (pp. 71-104). Orlando, Fl: Academic Press. Stiff, J. B., Dillard, J. P., Somera, L., Kim, H., & Sleight, C. (1988). Empathy, communication, and prosocial behavior. Communication Monographs, 55, 198- 213. Strauss, R. R, Sharp, M. C., Lorch, C., & Kachalia, B. (1995). Physicians and the communication of “bad news”: Parent experiences of being informed of their child’s cleft lip and/or palate. Pediatrics, 96, 82-89. Sussman, S. W., & Sproull, L. (1999). Straight talk: Delivering bad news through electronic communication. Information Systems Research, 10, 150-166. Tesser, A., & Rosen, S. (1975). The reluctance to transmit bad news. In L. Berkowitz (Ed.), Advances in experimental social psychology (V 01. 8, pp. 193- 232). New York: Academic Press. Tesser, A., Rosen, S., & Conlee, M. C. (1972). News valence and available recipient as determinants of news transmission. Sociometry, 35, 619-628. Tesser, A., Rosen, S., & Tesser, M. (1971). On the reluctance to communicate undesirable messages (the MUM effect): A field study. Psychological Reports, 29, 651-654. Trees, A. R., & Manusov, V. (1998). Managing face concerns in criticism: Integrating nonverbal behaviors as a dimension of politeness in female fiiendship dyads. Human Communication Research, 24, 564-583. Uysal, A., & Oner-Ozkan, B. (2007). A self-presentational approach to transmission of good and bad news. Social Behavior and Personality, 35, 63-78. Veitch, R., & Griffitt, W. (1976). Good news—bad news: Affective and interpersonal effects. Journal of Applied Social Psychology, 6, 69-75. Wagoner, R., & Waldron, V. R. (1999). How supervisors convey routine bad news: Facework at UPS. The Southern Communication Journal, 64, 193-210. Walker, A. J ., & Thompson, L. (1983). Intimacy and intergenerational aid and contact among mothers and daughters. Journal of Marriage and the Family, 45, 841 -849. 121 Watts, R. J ., (2003). Politeness. Cambridge, UK: Cambridge University Press. Widerman, E. (2002). Communicating a diagnosis of cystic fibrosis to an adult: What physicians need to know. Behavioral Medicine, 28, 45-52. Wood, J. T., & Inman, C. C. (1993). In a different mode: Masculine styles of communicating closeness. Journal of Applied Communication Research, 21, 279- 295. Yariv, E. (2006). “Mum effect”: Principals’ reluctance to submit negative feedback. Journal of Managerial Psychology, 21, 533-546. Yaughn, E., & Nowicki, Jr., S. (1999). Close relationships and complementary interpersonal styles among men and women. The Journal of Social Psychology, 139, 473-478. Zajdman, A. (1995). Humorous face-threatening acts: Humor as strategy. Journal of Pragmatics, 23, 325-339. 122