Bereaved adults' evaluations of grief management messages: Effects of message person centeredness, recipient individual differences, and contextual factors more

Death Studies, 32: 399–427, 2008 Copyright # Taylor & Francis Group, LLC ISSN: 0748-1187 print/1091-7683 online DOI: 10.1080/07481180802006711 BEREAVED ADULTS’ EVALUATIONS OF GRIEF MANAGEMENT MESSAGES: EFFECTS OF MESSAGE PERSON CENTEREDNESS, RECIPIENT INDIVIDUAL DIFFERENCES, AND CONTEXTUAL FACTORS JESSICA J. RACK, BRANT R. BURLESON, GRAHAM D. BODIE, AMANDA J. HOLMSTROM, and HEATHER SERVATY-SEIB Purdue University, West Lafayette, Indiana, USA This study identifies grief management strategies that bereaved adults evaluate as more and less helpful, assesses whether the person centeredness of these strategies explains their helpfulness, and determines whether strategy helpfulness varies as a function of demographic, personality, and situational factors. Participants (105 bereaved young adults) assessed the helpfulness of 16 grief management strategies; these strategies were coded for their degree of person centeredness. Strategy person centeredness was strongly correlated with helpfulness. Strategy helpfulness varied as a function of participant gender and the disruptiveness of the decedent’s death, but not as a function of need for cognition or decedent closeness. Can I see another’s woe And not be in sorrow too? Can I see another’s grief, And not seek for kind relief? –William Blake For most of us, grief is as inevitable as death. Although erroneous, both everyday beliefs and the media tell us there are no words that can help a person through the grieving process. Moreover, most people feel ill-prepared to help another person deal with a loss ( James & Friedman, 1998), even though they really Received 4 April 2007; accepted 15 July 2007. This research was supported by a Research Incentive Grant awarded to Brant Burleson by the College of Liberal Arts at Purdue University. We gratefully acknowledge Jennie Gill, Lisa Hanasono, and Jerilyn Mincy for their assistance with data collection. A version of this article was presented at the annual meeting of the International Communication Association, San Francisco, May 2007. Address correspondence to either Jessica Rack or Brant Burleson at the Department of Communication, Purdue University, West Lafayette, IN 47907-2098. E-mail: jjrack@ purdue.edu or brantb@purdue.edu 399 400 J. J. Rack et al. want to help. Zunin and Zunin (1991) captured the essence of this dilemma: Often, those who sincerely wish to console find their desire to help tightly bound by a sense of helplessness. As much as we care for a grieving friend or family member, we cannot bring back the dead and we cannot take the bereaved’s grief away. There simply are no stock phrases or pat answers. On the contrary, when confronted by the anguish of another’s grief, words seem to slip through our fingers like sand. (p. xiii) Nevertheless, when experiencing grief, most people turn to others in their social networks for comfort and support, and many of these appear to benefit from doing so. What should helpers say to help comfort the bereaved? And, just as important, what should they avoid saying? The purpose of the present study was to discover what grief management strategies are evaluated as more and less helpful by adults coping with the recent death of another. We conceptualize grief management as one type of emotional support somewhat similar to comforting. Burleson (1994) conceptualizes comforting messages as those intended to help others who are dealing with a variety of everyday stressors (e.g., getting a poor grade; having an argument with a friend). However, Burleson distinguished such everyday comforting from ‘‘strategies used to cope with extreme feelings of depression or grief arising from extraordinary events (such as the loss of a spouse)’’ because ‘‘some research (e.g., Brockoop, 1973; Lindemann, 1965) suggests that these more intense emotional experiences require responses qualitatively different from those used to manage everyday emotional upsets’’ (1994, pp. 136–137). Thus, the present study sought to (a) identify the types of messages that recently bereaved adults evaluate as more and less helpful in assisting them cope with their grief, (b) assess whether the concept of person centeredness can be used to characterize more and less helpful grief management strategies, and (c) determine whether evaluations of the helpfulness of grief management efforts systematically vary as a function of selected demographic, personality, and situational factors. To further develop the foundation for this study, subsequent sections of this rationale explain why scholars should be concerned with identifying more and less effective grief management strategies, what we currently know Evaluations of Grief Management Messages 401 about the qualities of such strategies, and why demographic, personality, and situational factors might influence evaluations of these strategies. Why Study Grief Management Strategies Grief is a universal social problem that is always hurtfully disruptive, sometimes extensively so. Further, there are periods in the life cycle (e.g., the college years, middle age) when losses are especially common. For example, at any given time, 23% of college undergraduates are in the first year of grieving a family member’s death, and this increases to 47% of undergraduates if the time since death encompasses two years (Balk, 1997). Moreover, the effects of grief can be quite damaging. Grief in childhood is associated with depression, sleep and behavior disturbances, post-traumatic stress disorder, and lack of interest in school (Dowdney, 2000). For adolescents, grief may be linked to depression, chronic illness, guilt, anxiety, disturbances in self-esteem, and difficulty in school and in relationships (Balk & Corr, 2001). In adults, grief is associated with shock, anger, guilt, anxiety, sadness, sleep and appetite disturbances, apathy, social withdrawal, and substance abuse (Littlewood, 1992). Recent scholarship suggests that grief may be better conceptualized as a social process rather than a private, internal event. For example, Kastenbaum (2004) held that ‘‘mourning is a significant process of interaction between survivors and their society’’ (p. 359, emphasis added). People rarely grieve alone. Family members may collectively grieve at the death of one of its own, as may friends and other associates of the deceased. Thus, it is important to examine the communicative strategies that ordinary people use in the effort to manage the grief experienced by others in their social networks. In addition, study of the communicative management of grief may provide insight about fundamental communication processes. For example, grief, along with its associated thoughts, feelings, and behaviors, is one of the most powerful emotions that humans experience, so research examining how communication affects grief may afford considerable insight about how messages can influence emotional states, as well as how emotions influence the processing and outcomes of messages (Burleson & Planalp, 2000; Nabi, 2003). 402 J. J. Rack et al. Despite the significant social consequences of grief and its character as a theoretically rich social phenomenon, surprisingly little empirical work has examined how supportive communication from social network members can affect the grieving process. The few studies that have examined messages intended to manage grief have used diverse approaches and methods. For example, some research has examined ritualistic forms of communication that generally address large audiences, such as funeral orations and eulogies (e.g., Dennis & Kunkel, 2004; Kunkel & Dennis, 2003). Pennebaker (1997) has examined the griever’s communicative behavior by exploring the psychological and physical benefits of writing about traumatic experiences in a journal or narrating them to a tape recorder. Some studies (e.g., Pennebaker, Mayne, & Francis, 1997) have found that producing such monologues is a helpful way of coping with grief, whereas other studies (e.g., Range, Kovac, & Marion, 2000) have found no effects for such monologues ´ (see review by Smyth, 1998). Other research (e.g., Rime, ´ Finkenauer, Luminet, Zech, & Philippot, 1998; Rime, Philippot, Boca, & Mesquita, 1992) has examined how socially sharing emotions (such as grief) might lead to emotional recovery, or reduction in the arousal elicited when an emotional memory is recalled ´, (Pennebaker, Zech, & Rime 2001). Still other research has examined how grievers seek support from others (e.g., Eckenrode & Wethington, 1990; Lepore, Silver, Wortman, & Wayment, 1996). Although there are some benefits to diverse foci and methods in a research area, Shapiro (2001) argued that the current interpersonal perspective on grief management is fragmented and lacks systematic conceptualization. Given the present state of knowledge, we believe that a particularly appropriate focus for research lies in delineating the features of messages that distinguish more and less helpful efforts to assist the bereaved manage their grief. The few studies directed at this goal have used one of two methods, the naturalistic paradigm or the message perception paradigm. Studies using the naturalistic paradigm (Dunkel-Schetter, Blasband, Feinstein, & Herbert, 1992) typically have research participants recall a recent circumstance in which they were grieving a loss and then describe both helpful and unhelpful comments they received from various support providers. For example, Davidowitz and Myrick (1984) asked 25 bereaved adults to indicate how others had responded to their grief, as well as which responses were Evaluations of Grief Management Messages 403 helpful and unhelpful. Of the received responses, 80% were regarded as unhelpful and only 20% as helpful. The most to least helpful categories of statements were feeling-focused, clarification= summary, question, reassurance=support, interpretation=analysis, and advice=evaluation. Hogan and DeSantis (1994) surveyed 140 bereaved adolescents on what they perceived as helping or hindering them cope with grief resulting from the death of a sibling. The most helpful form of social support was reported to be ‘‘people being there for me’’ whereas the least helpful action was reported to be ‘‘people not being there for me.’’ In an influential study using this paradigm, Lehman, Ellard, and Wortman (1986) asked people who had lost a spouse or a child in an automobile accident to recall helpful and unhelpful things that were said to them during their bereavement. Lehman et al. developed a typology of 21 grief support strategies based on participants’ descriptions of helpers’ behaviors. Expressing concern for the welfare of the bereaved, offering presence, and providing the opportunity for the bereaved to ventilate feelings were among the strategies rated most helpful. Minimizing the bereaved’s feelings and giving unsolicited advice were rated as least helpful. Two-thirds of the respondents indicated that at least one person said something unhelpful, even though they recognized that most of these messages were intended to be helpful. Furthermore, most of the unhelpful responses were made by relatives, friends, and acquaintances. Studies using the message perception paradigm (Burleson & MacGeorge, 2002) present participants with lists of grief management strategies written by the researcher and then ask participants to rate these strategies for qualities such as helpfulness, sensitivity, and effectiveness. For example, Range, Walston, and Pollard (1992) had a sample of college students evaluate the helpfulness of 30 statements that might be made to a bereaved person. These researchers found that expressions of willingness to help or listen were consistently rated as most helpful; however, the perceived helpfulness of several strategies varied significantly as a function of the decedent’s cause of death (i.e., natural, accident, suicide, homicide), a finding replicated in subsequent research (Knight, Elfenbein, & Messina-Soares, 1998). Building on the findings of Lehman et al. (1986), Marwit and Carusa (1998) identified 14 message strategies that people may use when trying to support a 404 J. J. Rack et al. grieving other. These researchers asked people who had experienced the death of a parent during adolescence to rate the helpfulness of each message strategy. Among the most helpful strategies were offering presence, allowing the bereaved to ventilate, expressing concern, and complimenting the deceased; among the least helpful were giving advice and minimizing the feelings of the bereaved. A limitation of studies using both research paradigms is an absence of theoretical accounts explaining why some message strategies are rated as more helpful than others. Thus, although these studies tell us what strategies are evaluated as more and less helpful, they do not explain why particular strategies are experienced as more or less helpful. Servaty-Seib and Burleson (2007) sought to address this limitation by investigating whether the personcenteredness of grief management strategies is correlated with their perceived helpfulness (also see Angell, 1998). In support contexts, person-centeredness is the extent to which a message validates, recognizes, and=or acknowledges the recipient’s feelings and experiences (see Burleson, 2003). Thus, messages low in person centeredness deny the other’s feelings and perspective by criticizing these feelings, challenging the legitimacy of these feelings, or telling the other how he or she should act and feel (e.g., ‘‘You can’t let something like this get to you; you have to get on with your own life and not take this so hard. Just stay busy’’). Messages that exhibit a moderate degree of person centeredness afford an implicit recognition of the other’s feelings by distracting attention from the troubling situation, offering expressions of sympathy and condolence, or presenting non-feeling-centered explanations of the situation intended to reduce the distress (e.g., ‘‘There is a purpose for everything, even if you don’t see it at the time. He was a really good person, and is no longer suffering. Why don’t you join us for dinner tonight?’’). In contrast, highly person-centered comforting messages explicitly recognize and legitimize the other’s feelings, help the other to articulate those feelings, elaborate reasons why those feelings might be felt, and assist the other to see how those feelings fit in a broader context (e.g., ‘‘I am so sorry; I can only guess how sad you must feel about his passing; I know the two of you were very close. I really care about you and how you are doing, and I’m available anytime you want to talk about things or just want company. Times like this can be rough, and I’m here to listen’’). Evaluations of Grief Management Messages 405 Servaty-Seib and Burleson (2007) determined the personcenteredness of each of Marwit and Carusa’s (1998) 14 message strategies (by expert coding) and then had adolescents who had experienced a death in the previous two years rate the helpfulness of each message strategy. Perceived message helpfulness was strongly correlated with message person centeredness, suggesting that message person centeredness may explain message helpfulness. Evaluations of strategy helpfulness were also found to vary as a function of demographic, personality, and situational factors, suggesting that these should be explored in future work. Focus of the Present Study The present study sought to replicate and extend in several ways previous findings regarding the properties of more and less helpful grief management strategies. First, much of the extant research has focused on the efficacy of strategies with adolescents (e.g., Hogan & DeSantis, 1994; Marwit & Carusa, 1998; Servaty-Seib & Burleson, 2007). However, considerable research indicates that college students frequently experience bereavement and need helpful responses to their grief by caring peers (Balk, 1997, 2001). The present study thus sought to assess the usefulness of Marwit and Carusa’s strategy typology with a young adult (college student) sample, as well as determine the extent to which young adults and adolescents evaluate grief management strategies similarly. We posed the two following research questions: RQ1: What grief management strategies do young adults view as more and less helpful? RQ2: To what extent do evaluations of strategy helpfulness obtained from young adults coincide with those obtained in previous research from other age groups? Second, we expanded the 14-category typology of message strategies developed by Marwit and Carusa (1998) by including two additional strategies. Politeness theory (Brown & Levinson, 1987) predicts that people will use politeness forms when interacting with others to preserve positive and negative face. Marwit and Carusa’s typology includes the strategy of complimenting the deceased, but helpers may also compliment the bereaved in an 406 J. J. Rack et al. effort to enhance their face and make them feel good about themselves (e.g., ‘‘You were so helpful during his illness’’; ‘‘I know your mom was proud of you’’). In addition, ‘‘Pollyanna’’ theory (Goodhart, 1985; Sears, Stanton, & Danoff-Burg, 2003) suggests that one way to cope effectively with an upset or loss is to highlight the positives of the situation (i.e., find the silver lining in the clouds; e.g., ‘‘She is no longer in pain’’; ‘‘He is no longer suffering’’). Thus, we posed the following research question about complimenting the bereaved and highlighting the positives of the situation: RQ3: How do young adults evaluate the helpfulness of grief management strategies that compliment the bereaved and highlight the positives of the situation? Third, we sought to replicate with our adult sample the key finding obtained by Servaty-Seib and Burleson (2007) that the person-centeredness of grief management messages explains participants’ evaluations of message helpfulness. We hypothesized: H1: The person-centered quality of grief management strategies will be positively associated with young adults’ evaluations of the helpfulness of these strategies. Existing research indicates that responses to grief management strategies vary as a function of demographic factors such gender of the recipient, situational factors such as the cause of death, and personality factors such as the extent to which recipients generally see social support as available to them. Burleson, Bodie, and their colleagues (Bodie & Burleson, in press; Burleson, in press) recently developed a dual-process model for the reception and outcomes of supportive messages in an effort to provide a comprehensive explanation for multiple factors that affect the processing and effects of these messages. Like dual-process models for persuasion, such as the Elaboration Likelihood Model (ELM; Petty, Rucker, Bizer, & Cacioppo, 2004) and the Heuristic-Systematic Model (HSM; Todorov, Chaiken, & Henderson, 2002), this model maintains that people will cognitively elaborate on supportive messages to a greater or lesser extent depending on their motivation and ability to do so. When motivated and able, the recipients of grief management messages should systematically process these messages and, therefore, distinguish more sharply among better and Evaluations of Grief Management Messages 407 worse message forms than recipients less motivated and=or able to process these messages. This general hypothesis was evaluated in the present study by assessing how several factors believed to influence the motivation and ability to process grief management messages affected the extent to which participants differentiated between more and less person-centered messages in terms of their perceived helpfulness. Both men and women view highly person-centered comforting messages as more helpful than less person-centered messages (Burleson & Kunkel, 2006). However, women evaluate highly person-centered messages somewhat more positively than men, and men evaluate low person-centered messages someone more positively than women (e.g., Kunkel & Burleson, 1999). Moreover, previous research indicates gender differences in evaluations of grief management messages (e.g., Knight et al., 1998; Servaty-Seib & Burleson, 2007). As a group, women exhibit higher levels of cognitive complexity (e.g., Samter, 2002), empathy (e.g., Trobst, Collins, & Embree, 1994), and emotional intelligence (e.g., Brackett, Mayer, & Warner, 2004) than do men. Because these abilities appear to influence the capacity to systematically process support messages, especially highly sophisticated messages (see Burleson & Caplan, 1998), women should be better able than men to process grief management messages, and thus distinguish between better and worse forms of these messages. In addition, some scholars (e.g., Tannen, 1990; Wood, 1996) believe that men and women constitute different cultures and use distinct standards in evaluating the quality of support strategies, with women valuing to a greater extent than men the explicit discussion of thoughts and feelings. Both of these lines of research suggest gender differences in the evaluation of more and less person-centered grief management messages. Hence, we hypothesized H2: In their ratings of message helpfulness, women will distinguish more sharply than men in their evaluations of more and less person-centered messages, evaluating highly person-centered grief management messages more positively than men and evaluating low person-centered messages less favorably than men. Need for cognition (NFC) ‘‘refers to an individual’s tendency to engage in and enjoy effortful cognitive endeavors’’ (Cacioppo, 408 J. J. Rack et al. Petty, & Feng Kao, 1984, p. 306). Previous research indicates that NFC motivates more extensive processing of persuasive messages (see Petty et al., 2004). It seemed reasonable to suppose that NFC would motivate more systematic processing of grief management messages, resulting in those high in NFC distinguishing more sharply between better and worse forms of these messages than those low in NFC. Hence, we hypothesized: H3: In their ratings of message helpfulness, high NFC participants will distinguish more sharply than low NFC participants in their evaluations of more and less person-centered messages, evaluating highly person-centered grief management messages more positively than low NFC participants and evaluating low person-centered messages less favorably than low NFC participants. Previous research indicates that evaluations of the helpfulness of grief management messages can be influenced by situational factors such as the cause of the decedent’s death and the recipient’s relationship to the decedent (e.g., Knight et al., 1998; Servaty-Seib & Burleson, 2007). Contextual factors such as these can be conceptualized as influencing the degree of emotional upset experienced, and thus the motivation to process grief management messages. In the current study, we examined how two aspects of the bereavement context—the bereaved’s closeness to the decedent and the degree of disruption caused by the death—influenced responses to grief management messages. Both of these contextual factors might be viewed as heightening the relevance of the decedent’s death, thereby increasing the motivation to process grief management messages. More specifically, those close to the decedent probably experience more intense grief than those less close, and those whose lives are more disrupted by the death of the decedent also probably experience more intense grief. Presumably, those experiencing higher levels of grief will be more motivated to process grief management messages than those experiencing lower levels of grief. Thus, we hypothesized: H4: In their ratings of message helpfulness, participants close to the decedent will distinguish more sharply than participants less close to the decedent in their evaluations of more and less person-centered messages, evaluating highly personcentered grief management messages more positively than Evaluations of Grief Management Messages 409 less close participants and evaluating low person-centered messages less favorably than less close participants. H5: In their ratings of message helpfulness, participants whose lives were substantially disrupted by the decedent’s death will distinguish more sharply than participants whose lives were less disrupted in their evaluations of more and less person-centered messages, evaluating highly person-centered grief management messages more positively than less disrupted participants and evaluating low person-centered messages less favorably than less disrupted participants. Finally, we sought to determine how evaluations grief management strategies varied as a joint function of participant gender, participant need for cognition, closeness to the decedent, and disruptiveness of the decedent’s death (RQ4). Method Participants Participants were 105 college students (65.7% female, 77.1% White; mean age ¼ 20.3 years) attending a large Midwestern university who had experienced a death loss in the previous two-year period (10%: 0–3 months, 16%: 4–6 months, 24%: 7–12 months, 25%: 13–18 months, and 26%: 19–24 months prior). Type of loss included extended family members (e.g., grandparent, cousin; 59%), peers (e.g., friends, classmates; 26.7%), immediate family members (e.g., mother, brother, spouse; 6.7%), very close friends or romantic partners (4.8%), and those not fitting the aforementioned categories (2.9%). Causes of death included prolonged or terminal illness (e.g., cancer, Alzheimer’s; 32.4%), accidents (e.g., automobile, domestic; 23.8%), old age (17.1%), and sudden illness (e.g., heart attack, stroke; 11.4%); the remainder of participants indicated that the death was due to some other cause (e.g., murder, suicide) or was unknown (15%). Procedure The study was announced in undergraduate communication courses and participants came to a data collection session to complete a questionnaire. Participants were initially asked if they had experienced a death loss within the previous two years, and if so, 410 J. J. Rack et al. if they were willing to think about it and complete a measure regarding the experience. If the participants had not experienced a death loss or were not comfortable thinking about a death loss, they were given the opportunity to participate in an alternative study. Instrumentation DEMOGRAPHIC INFORMATION Participants reported information regarding age, gender, ethnicity, class rank, and academic major. LOSS EXPERIENCES SURVEY (LES) The LES consisted of an 18-item questionnaire designed to assess details surrounding the death and the relationship of the bereaved to the decedent. Seven questions assessed the contextual characteristics of the death, including the bereaved’s relationship to the decedent, the time since the death, the cause of the death, and other circumstances; responses to these items were used for descriptive purposes. To measure closeness to the decedent, participants responded to an additional seven 5-point items (1 ¼ strongly disagree to 5 ¼ strongly agree) that tapped perceived interpersonal distance (e.g., ‘‘I felt close to this person,’’ ‘‘I felt I could share my most intimate feelings with this person’’). These items possessed obvious face validity as a measure of closeness and are similar in content to other measures of closeness and intimacy (e.g., Parks & Floyd, 1996). Internal consistency of this measure, as assessed by Cronbach’s alpha, was .90. To assess disruptiveness of the death, participants then responded to four 5-point items that tapped the degree of disorder experienced by the participant as a result of the death (e.g., ‘‘After this person died, I found it difficult to pay attention at school or work,’’ 1 ¼ strongly disagree, 5 ¼ strongly agree; ‘‘How upset were you at the time of the death?’’ 1 ¼ not at all upset, 5 ¼ very very strongly upset). This scale possesses face validity and exhibited good internal consistency (Cronbach’s a ¼ .85). MODIFIED SUPPORT-INTENDED STATEMENTS SCALE (SISS) Marwit and Carusa’s (1998) SISS originally consisted of 42 items, 3 items for each of 14 grief management strategies. The version of the SISS used in the present study consisted of 64 message Evaluations of Grief Management Messages 411 statements that instantiated 16 grief management strategies (four statements for each of the 16 strategies; see Table 1). In addition to the 14 strategies previously identified by Marwit and Carusa, two strategies (complimenting the living and highlighting the positive) were introduced in the present study. Thirty-nine of the statements were taken from the 42 developed by Marwit and Carusa; 3 other items were reworded due to poor internal consistencies reported by Servaty-Seib and Burleson (2007). In addition, 22 additional messages were added to the scale. For each of the 14 original strategies, 1 additional message was introduced, creating a total of 4 messages for each strategy. Four messages were generated for each of the 2 new strategies. Participants rated each message for helpfulness on a 5-point scale (1 ¼ very harmful to 5 ¼ very helpful, with the mid-point of 3 ¼ neither). Internal consistency analyses were performed for each of the 16 sets of four items; the results of these analyses are reported in Table 1. Internal consistencies ranged from a low of .60 to a high of .92 and averaged .79 for the 16 strategies. Scores for the items were averaged to generate a mean helpfulness rating for each strategy. Each of the 14 previously identified strategies had been coded for its degree of person centeredness in Servaty-Seib and Burleson (2007). For the two new strategies, two coders were provided with the four (unlabeled) items representing each strategy and were instructed to identify the level of person centeredness exhibited by the strategy. The coders used the 9-level hierarchy of comforting messages developed by Applegate (1980) and Burleson (1982), which consists of three major levels with three sublevels within each major level. Both coders agreed on the person-centered level of both strategies. The coded levels for the person centeredness of the 16 strategies are reported in Table 2. NEED FOR COGNITION (NFC) The Short-Form Need for Cognition Scale (NCS; Cacioppo et al., 1984) is a self-report measure designed to measure ‘‘an individual’s tendency to engage in and enjoy effortful cognitive endeavors’’ (p. 306). Cacioppo et al. noted that research on this individual difference is predictive of the manner in which people deal with information. Participants responded to 18 items using a 5-point scale (1 ¼ extremely uncharacteristic of me to 5 ¼ extremely characteristic of me). Scores were averaged across the 18 items to 412 J. J. Rack et al. TABLE 1 Items and Internal Consistency Coefficients for 16 Support Strategies Strategy=items Offer presence (‘‘being there’’) I am here for you. If you need company, I am here for you. I would be happy to stay with you if you’d like company. I am nearby if you need me. Express willingness to listen; provide opportunities to express feelings Would you like to talk about it? If you want to talk, I will listen. I am a good listener if you need one. I really want to know how you are doing. Express care and concern I care about what happens to you. I really care about how you are doing. I am concerned about how you feel. I really want to know how you are doing. Include in social activities Would you like to come with us to the football game? We usually go to the movies on Friday’s, why don’t you come? How about a movie? Do you want to come to a friend’s party with me? Compliment the deceased He=she always had a kind word for everyone. He=she was fun to be around. I really liked your (e.g., parent, sibling). I really enjoyed spending time with him=her. Compliment the living I am proud of you for being so brave. You have done a nice job of looking after everyone. You were so helpful during this difficult time. You are being so strong. Discuss being reunited Someday you will reconnect. Your spirits will always be together. In time, your spirits will reunite. Your souls will always be united. Highlight the positive He=she is no longer suffering. He=she is no longer in any pain. Was he=she in much pain? He=she no longer has any worries. a coefficient .72 .79 .76 .89 .84 .72 .89 .73 (Continued ) Evaluations of Grief Management Messages TABLE 1 Continued Strategy=items Provide a religious perspective It is God’s will. We are not always meant to understand God’s purpose. The Lord works in mysterious ways. Put your faith in God. Identify with other’s feelings I know what it is like. I know how you feel. I can imagine what it might be like for you. I know it must really hurt right now. Provide a philosophical perspective Everything that happens is for a purpose, you just don’t see it. Everything has a time and a place and a purpose. Going through this pain makes us appreciate the good in life. Some things we just can’t control in this world. Offer contact with similar others I have some friends who this happened to; would you like to talk with them? I can arrange for you to meet with someone in your position. Would you like to talk to someone who has been through this? I know a group that meets to discuss loss. Would you like to talk with them? Discuss memories of deceased Do you remember how it was before this happened? What was it like before? Do you remember when. . .? I remember when he=she did . . . (some event). Provide tangible support Do you need help cleaning out possessions? Do you have errands I can run for you? Would you like to come to dinner so you don’t have to cook? Would you like me to run to the store for you? Give advice You should keep busy. When you feel bad you should exercise. You should read self-help books. You must get on with your life. Minimize feelings; forced cheerfulness Do not take it so hard. It cannot be that bad. It should not affect you. You shouldn’t let this get you down. 413 a coefficient .91 .75 .69 .92 .81 .74 .60 .86 414 J. J. Rack et al. TABLE 2 Mean Helpfulness Rating and Person-Centered Coding for 16 Support Strategies Person-centeredness codingÃà Support strategy Offer presence (‘‘being there’’) Express willingness to listen Express care and concern Compliment the deceased Compliment the bereaved Discuss being reunited Highlighting the positive Discuss memories of deceased Provide tangible support Provide a religious perspective Identify with other’s feelings Include in social activities Provide a philosophical perspective Offer contact with similar others Give advice Minimize feelings; forced cheerfulness Helpfulness ratingà 4.05 3.78 3.76 3.70 3.60 3.28 3.19 3.17 3.12 3.01 2.95 2.90 2.80 (.56) (.63) (.61) (.78) (.60) (.85) (.78) (.81) (.68) (1.0) (.70) (.95) (.81) Sublevel 8 8 7 5 6 6 4 4 4 6 7 4 6 6 3 2 Major level III III III II II II II II II II III II II II I I 2.53 (.82) 2.18 (.53) 1.45 (.59) Note. Coefficients in parentheses are standard deviations. à By participants, N ¼ 105. Ãà By expert coders. yield an index for need for cognition (a ¼ .92). Cacioppo and Petty (1982) reported evidence for the convergent, discriminant, content, and predictive validity of this construct. Results Means and standard deviations for participants’ ratings of support strategy helpfulness appear in Table 2. RQ1 asked what grief management strategies young adults viewed as more and less helpful. Similar to findings obtained by Marwit and Carusa (1998) and Servaty-Seib and Burleson (2007), the most helpful strategies included offering one’s presence (‘‘being there’’), expressing the willingness to listen, and expressing care and concern. The least Evaluations of Grief Management Messages 415 helpful strategies included giving advice and minimizing of the other’s feelings. To address RQ2, evaluations of the grief management strategies obtained from the young adults in the current study were compared with those by the adults in Marwit and Carusa’s (1998) study and those by the adolescents in Servaty-Seib and Burleson’s (2007) study. We correlated the mean helpfulness ratings for the 14 strategies rated in all three studies. The pattern of helpfulness ratings for the 14 strategies obtained in the present study was strongly correlated with the helpfulness ratings obtained by both Marwit and Carusa’s, r ¼ .84, p < .001, and Servaty-Seib and Burleson, r ¼ .92, p < .001. Thus, in answer to RQ2, the support strategies were evaluated quite similarly in terms of their helpfulness by three distinct samples. RQ3 asked how participants evaluated the two strategies introduced by the present study, compliment the bereaved and highlight the positives of the situation. The mean helpfulness rating for compliment the bereaved was 3.60 (SD ¼ .60), 5th most helpful of the 16 strategies. The mean helpfulness rating for highlight the positive was 3.19 (SD ¼ .78), 7th most helpful of the 16 strategies. H1 predicted that the person centeredness of the SISS strategies (as determined by expert coders) would be positively associated with participants’ evaluations of the helpfulness of these strategies. This hypothesis was tested by correlating evaluations of strategy helpfulness with the major level of strategy person centeredness for the 16 strategies. A strong, positive correlation between strategy person-centeredness and perceived helpfulness provided support for H1, r ¼ .77, p < .001. We next assessed whether evaluations of strategy helpfulness were moderated by participant gender (H2), need for cognition (H3), closeness to the decedent (H4), or disruptiveness caused by the decedent’s death (H5). Median splits were used to create low and high groups for need for cognition, closeness, and disruptiveness. Planned comparisons were used to test these hypotheses. Power to detect the mean differences specified in H2–H5 (at a ¼ .05) was excellent for large (d ¼ .80) and moderate (d ¼ .50) effects sizes (.98 and .70, respectively), but was modest (.17) for small effects (d ¼ .20). The planned comparisons evaluating H2 contrasted, at each level of message person centeredness, men’s and women’s 416 J. J. Rack et al. TABLE 3 Means and Standard Deviations for Evaluated Helpfulness of Low, Moderate, and High Person-Centered Grief Management Strategies as Moderated by Participant Gender, Need for Cognition, Closeness to the Deceased, and Disruptiveness of the Death Level of message person centeredness Low Independent variables Gender Men (n ¼ 34) Women (n ¼ 69) Need for cognition Low (n ¼ 52) High (n ¼ 53) Closeness to decedent Low (n ¼ 50) High (n ¼ 55) Disruptiveness of death Low (n ¼ 52) High (n ¼ 53) M SD Moderate M SD M High SD 2.02 1.71 1.86 1.77 1.84 1.80 1.93 1.71 .54 .45 .48 .51 .58 .42 .55 .43 3.11 3.12 3.14 3.12 3.10 3.16 3.11 3.15 .48 .44 .43 .49 .50 .43 .46 .46 3.46 3.71 3.65 3.62 3.63 3.64 3.64 3.63 .50 .45 .44 .51 .53 .42 .48 .47 evaluations of message helpfulness (see Table 3). These analyses indicated, as predicted, that women evaluated low person-centered messages as less helpful (M ¼ 1.71) than did men (M ¼ 2.02), t(101) ¼ 3.04, p < .01, whereas women evaluated highly personcentered messages as more helpful (M ¼ 3.71) than did men (M ¼ 3.46), t(101) ¼ 2.59, p < .05. Men (M ¼ 3.11) and women (M ¼ 3.12) did not differ in their evaluation of messages exhibiting moderate levels of person centeredness, t(101) ¼ .12, ns. The planned comparisons for H3 contrasted, at each level of message person centeredness, evaluations of message helpfulness by participants with low and high levels of need for cognition. No support was obtained for H3; need for cognition did not influence the perceived helpfulness of messages either low in person centeredness, t(103) ¼ .92, ns, or high in person centeredness, t(103) ¼ .41, ns (see Table 3). The planned comparisons evaluating H4 contrasted, at each level of message person centeredness, evaluations of message helpfulness by participants reporting low or high closeness to the decedent. No support was obtained for H4; closeness to the Evaluations of Grief Management Messages 417 decedent did not influence the perceived helpfulness of messages exhibiting either low-person centeredness, t(1, 103) ¼ .43, ns, or high-person centeredness, t(103) ¼ .02, ns (see Table 3). The planned comparisons evaluating H5 contrasted, at each level of message person centeredness, evaluations of message helpfulness by participants who experienced low or high levels of disruptiveness due to the decedent’s death. Partial support was obtained for H5; consistent with predictions, participants who experienced greater disruptiveness evaluated low person-centered messages as less helpful (M ¼ 1.71) than participants who experienced less disruptiveness (M ¼ 1.93), t(103) ¼ 2.18, p < .05. However, contrary to predictions, participants who experienced greater and lesser disruptiveness did not differ in evaluations of highly person-centered messages, t(103) ¼ .21, ns (see Table 3). To explore whether evaluations of message helpfulness were affected by any higher-order interactions among the independent variables (RQ4), we conducted a 2  2  2  2  3 analysis of variance (ANOVA) in which the between-groups factors were participant gender, need for cognition, closeness, and disruptiveness, the within-subjects factor was message person centeredness (low, moderate, and high), and the dependent variable was the rated helpfulness of the support strategy. Because there was low power to detect significant four-way interactions among the betweengroups factors and the significant five-way interaction between the between-groups factors and the repeated factor, the sums of squares for these higher-order interaction terms were pooled with the error term. Power to detect two-way and three-way interactions among the between group variables was .26, .95, and 1.00, respectively for low, moderate, and large effects. The omnibus F test revealed a significant main effect for message person centeredness, F(2, 87) ¼ 172.18, p < .001, g2 ¼ .80. Significant main effects were not detected for gender, F(1, 103) ¼ 1.14, ns, need for cognition, F(1, 103) ¼ 1.28, ns, closeness, F(1, 103) ¼ 0.26, ns, or disruptiveness, F(1, 103) ¼ 1.23, ns. Other than effects previously tested with planned comparisons, the ANOVA detected only one additional effect that approached significance: There was a near-significant three-way interaction among gender, need for cognition, and closeness to the decedent, F(1, 88) ¼ 3.65, p < .06, g2 ¼ .04. Decomposition of this interaction indicated that for women, need for cognition 418 J. J. Rack et al. FIGURE 1 Perceived helpfulness of grief management messages by men as a function of the interaction between need for cognition (NFC) and closeness to the decedent. had no effect on message evaluations at either level of closeness to the deceased. In contrast, when men were close to the deceased, those high in need for cognition evaluated the grief management messages less positively (M ¼ 2.71) than did men low in need for cognition (M ¼ 3.01), t(14) ¼ 2.08, p < .057; however, need for cognition had no influence on men’s evaluations of the grief management messages when they reported a low level of closeness to the deceased (Ms ¼ 2.91 and 2.90 for low and high need for cognition, respectively). Put another way, men high in need for cognition were much more critical of the grief management messages than men low in need for cognition, but only when they were close to the decedent (see Figure 1). Discussion Although grief is a universal phenomenon with profound personal and social consequences, little research has sought to identify message strategies that helpers can use to assist the bereaved in coping effectively with their grief. Thus, the present study was conducted in the effort to identify the grief management strategies that bereaved young adults evaluate as more and less helpful, determine if the person-centered quality of these messages could explain judgments about their helpfulness, and explore whether Evaluations of Grief Management Messages 419 evaluations of message helpfulness are moderated by individual and situational factors. Replicating findings obtained with samples of adults (Marwit & Carusa, 1998) and adolescents (Servaty-Seib & Burleson, 2007), we found that our sample of young adults most positively evaluated the grief management strategies of offering one’s presence (‘‘being there’’), expressing the willingness to listen, and expressing care and concern, whereas the least positively evaluated strategies included giving advice and minimization of the other’s feelings. There were also a few interesting age-related differences across the studies. For example, Servaty-Seib and Burleson found that adolescents positively evaluated the strategy of including the bereaved in social activities (rated 4th of 14 strategies for helpfulness), whereas we found that our young adults viewed this strategy as rather unhelpful (rated 12th of 16). Including the bereaved in social activities is a way of distracting their attention from the hurtful loss; adolescents appear to appreciate this strategy more than do (what we assume to be) cognitively and emotionally more sophisticated young adults. Further, our young adults indicated that the strategy of discussing memories of the deceased was somewhat helpful (8th of 16), perhaps because such discussions help process emotions about the other’s loss (Pennebaker et al., 2001; Rime et al., 1998). In contrast, adolescents evaluated this strategy as somewhat unhelpful (11th of 14), perhaps because they are less emotionally developed than adults and thus less interested in exploring their thoughts and feelings about the decedent. Agerelated differences such as these should be explored in future research which includes measures that permit the identification of the specific mechanisms underlying developmental changes in responses to grief management efforts. Despite these age-related differences, correlational analyses indicated very similar patterns of evaluation for the SISS strategies across all three studies, indicating that people of different ages have largely similar ideas about which grief management strategies are more and less helpful. Extending the results of earlier research, we found that young adults’ evaluations of message helpfulness were largely explained by the person-centered quality of the grief management strategies. Servaty-Seib and Burleson (2007) previously found that adolescents’ evaluations of the helpfulness of the SISS grief management strategies were strongly associated with the person-centeredness of 420 J. J. Rack et al. these strategies. Moreover, Angell (1998) found that judgments made by expert helpers (doctoral students in counseling psychology) about the appropriateness of grief management strategies largely coincided with the coded person-centered quality of these strategies. Thus, the theoretical dimension of person centeredness can explain judgments about the helpfulness of grief management messages made by both experts and diverse groups of ordinary actors. These findings supplement the extensive research showing that the person-centered quality of mundane comforting strategies ¨ explains naıve actors’ judgments about the helpfulness of these strategies (see review by Burleson et al., 2005). Building upon a recently developed dual process model for the reception of supportive messages (Bodie & Burleson, in press), we reasoned that bereaved individuals who processed the SISS grief management strategies more systematically would distinguish more sharply between better and worse (i.e., high and low personcentered) strategies than bereaved individuals who processed these messages less systematically. We argued that women (on average) should have a greater ability than men (on average) to process these messages due to their generally higher levels of social-cognitive and emotional development. Further, we suggested that the motivation to systematically process grief management messages, and thus the tendency to distinguish between better and worse forms of these, should be increased by both certain personality variables (need for cognition) and situational factors (closeness to the decedent; disruptiveness of the decedent’s death). Only partial support was obtained for these hypotheses. Consistent with expectations, women distinguished more carefully than men between grief management messages that exhibited low and high levels of person centeredness, evaluating highly person-centered strategies as more helpful and low personcentered strategies as less helpful. This is quite similar to the pattern of gender differences observed in evaluations of mundane comforting strategies (see review by Burleson & Kunkel, 2006). We also found, as expected, that those whose lives were more disrupted by the death of another evaluated messages exhibiting low person centeredness as less helpful than did those whose lives were less disrupted by the other’s death. People who experienced a disruptive death found messages low in person centeredness to be more offensive than those whose lives had been less disrupted, Evaluations of Grief Management Messages 421 which suggests that they may have processed these messages somewhat more deeply. However, contrary to our expectations, evaluations of highly person-centered messages did not vary as a function of disruptiveness. Moreover, evaluations of the grief management strategies were not moderated by need for cognition or closeness to the deceased. Taken together, these findings suggest that factors associated with the ability to process grief management messages (i.e., participant gender) lead to more extensive processing of these messages, and thus sharper distinctions between better and worse forms of these messages, than do factors associated with the motivation to process these messages (i.e., need for cognition, closeness to the decedent, disruptiveness of the death). In retrospect, it appears that virtually all of the bereaved in the present study were highly motivated to process the grief management messages, and thus rather sharply distinguished between the better and worse forms of these messages. The person-centered quality of the grief management messages accounted for a very large amount of the variability in the judged helpfulness of these messages (g2 ¼ .80), which suggests that participants actively attended to these messages (i.e., processed them systematically). After all, the participants in the present study were bereaved—they were experiencing (or recently had experienced) the very powerful emotion of grief. The experience of that emotion should certainly motivate attention to grief management messages, especially if the decedent was known at least somewhat well, as was the case in the present study. Other factors influencing the motivation to process supportive messages (such as need for cognition and closeness to the decedent) may exert a rather subtle effect, especially in comparison to the relatively large effect attributable to the experience of grief. One limitation of the present study is that it did not include measures of message elaboration or processing (e.g., thought listing); clearly, future work will need to include such measures to more directly evaluate claims that variation in depth of message processing leads to differences in responses to supportive messages. In addition, the present study assumed that women had greater ability than men to process grief management messages due to their generally superior levels of social-cognitive and emotional development (e.g., Samter, 2002; Trobst et al., 1994). Measures of social-cognitive and emotional skills were not included 422 J. J. Rack et al. in the current study but should be in future research so that their effects on message processing, as well as their role as mediators of gender differences, can be evaluated directly. Third, our study relied on retrospective self-reports of those who had suffered a loss sometime within the preceding two-year period. Responses to grief management messages may well vary systematically at different points in the grieving process; this issue should also be addressed in future research. We detected an unexpected three-way interaction among gender, closeness to the decedent, and need for cognition. Decomposition of this interaction indicated that men high in need for cognition evaluated grief management messages less positively than men low in need for cognition when grieving the death of someone who had been close; however, there was no difference in men’s evaluations of messages as a function of need for cognition in the case of a less close decedent. Men high in need for cognition typically have a strong desire for the world to make sense, so the death of a close other may undermine their assumptions that the world is just, benevolent, and sensible (JanoffBulman, 1992), producing intense grief. Coupled with violated assumptions about the world, this intense grief may lead these men to devalue all grief management efforts. Obviously, this post hoc account needs to be evaluated in future research. We sought to improve the SISS by (a) adding 2 new strategies (compliment the bereaved and highlight the positives of the situation) to the 14 originally proposed by Marwit and Carusa (1998); (b) replacing items found to exhibit problematic reliability in previous research; and (c) adding new items so that four, rather than three, items were used to assess each strategy. These modifications resulted in a more inclusive measure of grief management strategies and greater internal consistency for each of the strategies assessed. Although one strategy (give advice) exhibited problematic internal consistency, the internal consistencies for the 15 other strategies ranged from acceptable to excellent. Pragmatic Implications The findings of the current study, in conjunction with those obtained in other research (Knight et al., 1998; Lehman et al., 1986; Marwit & Carusa, 1998; Range et al., 1992; Servaty-Seib & Evaluations of Grief Management Messages 423 Burleson, 2007), can inform specific prescriptions regarding how to help a grieving other. It appears that giving advice or minimizing the bereaved’s feelings are almost never evaluated as helpful, whereas offering one’s presence, expressing the willingness to listen, and expressing care and concern are consistently evaluated as helpful. Thus, across different participant samples (adolescents, young adults, adults), instrumentation, and research paradigms, there is remarkable consistency in the grief management strategies that people find more and less helpful. Given the prevalence of grief, and its potentially serious personal and social consequences, these findings are important. People are likely to experience others grieving frequently in their lives, and it is essential for people to know how to console a grieving other. There is every reason to believe that people can be trained in how to use helpful grief management strategies and avoid using unhelpful strategies. Principles for training more general comforting skills have been identified (e.g., Burleson, 2003), and it should be possible to adapt these to the special circumstances of consoling the bereaved. Although it appears safe to conclude that people judge certain grief management strategies to be particularly helpful and others unhelpful, several important questions remain to be addressed that have significant pragmatic implications. For example, what are the long-term effects, as well as the short-term effects, of various message strategies? Do some strategies lead to long-term emotional change whereas others only result in short-term emotional change? Can providing helpful support to the bereaved influence their long-term coping abilities, perhaps even assisting them to deal effectively with the grief arising from the deaths of others in the future? And how do grief management efforts affect the quality of the interpersonal relationship between the helper and recipient, both in the short-term and over longer periods? Research that addresses these questions should contribute substantially to the theory and practice of grief management. References Angell, L. R. (1998). Communication comforting strategies and social bereavement: Verbal and nonverbal planning and appropriateness. Journal of Personal and Interpersonal Loss, 3, 271–283. Balk, D. E. (1997). Death, bereavement, and college students: A descriptive analysis. Mortality, 2, 207–220. 424 J. J. Rack et al. Balk, D. E. (2001). College student bereavement, scholarship, and the university: A call for university engagement. Death Studies, 25, 67–84. Balk, D. E. & Corr, C. A. (2001). Bereavement during adolescence: A review of research. In M. S. Stroebe, R. O. Hansson, W. Stroebe, & H. Schut (Eds.), Handbook of bereavement research: Consequences, coping, and care (pp. 199–218). Washington, DC: American Psychological Association. Bodie, G. D. & Burleson, B. R. (in press). Explaining moderators of the effects of supportive messages: Can a dual-process model account for extant findings? In C. Beck (Ed.), Communication yearbook 32. Mahwah, NJ: Erlbaum. Brackett, M. A., Mayer, J. D., & Warner, R. M. (2004). Emotional intelligence and its relation to everyday behaviour. Personality and Individual Differences, 36, 1387–1402. Brockoop, G. W. (1973). Crisis intervention: Theory, process, and practice. In D. Lester & G. W. Brockoop (Eds.), Crisis intervention and counseling by telephone (pp. 3–29). Springfield, IL: C. C. Thomas. Burleson, B. R. (1994). Comforting messages: Features, functions, and outcomes. In J. A. Daly & J. M. Wiemann (Eds.), Strategic interpersonal communication (pp. 135–161). Hillsdale, NJ: Erlbaum. Burleson, B. R. (2003). Emotional support skill. In J. O. Greene & B. R. Burleson (Eds.), Handbook of communication and social interaction skills (pp. 551–594). Mahwah, NJ: Lawrence Erlbaum Associates. Burleson, B. R. (in press). What counts as effective emotional support? Explorations of individual and situation differences. In M. T. Motley (Ed.), Studies in applied interpersonal communication. Thousand Oaks, CA: Sage. Burleson, B. R. & Caplan, S. E. (1998). Cognitive complexity. In J. C. McCroskey, J. A. Daly, M. M. Martin, & M. J. Beatty (Eds.), Communication and personality: Trait perspectives (pp. 230–286). Cresskill, NJ: Hampton Press. Burleson, B. R. & Goldsmith, D. J. (1998). How the comforting process works: Alleviating emotional distress through conversationally induced reappraisals. In P. A. Andersen & L. K. Guerrero (Eds.), Handbook of communication and emotion: Research, theory, applications and contexts (pp. 245–280). San Diego, CA: Academic Press. Burleson, B. R. & Kunkel, A. W. (2006). Revisiting the different cultures thesis: An assessment of sex differences and similarities in supportive communication. In K. Dindia & D. J. Canary (Eds.), Sex differences and similarities in communication (2nd ed., pp. 137–159). Mahwah, NJ: Erlbaum. Burleson, B. R. & MacGeorge, E. L. (2002). Supportive communication. In M. L. Knapp & J. A. Daly (Eds.), Handbook of interpersonal communication (3rd ed., pp. 374–424). Thousand Oaks, CA: Sage. Burleson, B. R. & Planalp, S. (2000). Producing emotion(al) messages. Communication Theory, 10, 221–250. Burleson, B. R., Samter, W., Jones, S. M., Kunkel, A. W., Holmstrom, A. J., Mortenson, S. T., & MacGeorge, E. L. (2005). Which comforting messages really work best? A different perspective on Lemieux and Tighe’s ‘‘receiver perspective.’’ Communication Research Reports, 22, 87–100. Cacioppo, J. T. & Petty, R. E. (1982). The need for cognition. Journal of Personality and Social Psychology, 42, 116–131. Evaluations of Grief Management Messages 425 Cacioppo, J. T., Petty, R. E., & Feng Kao, C. (1984). The efficient assessment of need for cognition. Journal of Personality Assessment, 48, 306–307. Davidowitz, M. & Myrick, R. D. (1984). Responding to the bereaved: Analysis of ‘‘helping’’ statements. Death Education, 8, 1–10. Dennis, M. R. & Kunkel, A. D. (2004). Fallen heroes, lifted hearts: Consolation in contemporary presidential eulogia. Death Studies, 28, 703–731. Dowdney, L. (2000). Annotation: Childhood bereavement following parental death. Journal of Child Psychology and Psychiatry, 41, 819–830. Dunkel-Schetter, C., Blasband, D., Feinstein, L., & Herbert, T. (1992). Elements of supportive interactions: When are attempts to help effective? In S. Spacapan & S. Oskamp (Eds.), Helping and being helped: Naturalistic studies (pp. 83–114). Newbury Park, CA: Sage. Eckenrode, J. & Wethington, E. (1990). The process and outcome of mobilizing social support. In S. Duck (Ed.), Personal relationships and social support (pp. 83–103). London: Sage. Goodhart, D. E. (1985). Some psychological effects associated with positive and negative thinking about stressful event outcomes: Was Pollyanna right? Journal of Personality and Social Psychology, 48, 216–232. Hogan, N. S. & DeSantis, L. (1994). Things that help and hinder adolescent sibling bereavement. Western Journal of Nursing Research, 16, 132–153. James, J. W. & Friedman, R. (1998). The grief recovery handbook: The action program for moving beyond death, divorce, and other losses. New York: Harper Collins Publisher. Janoff-Bulman, R. (1992). Shattered assumptions: Towards a new psychology of trauma. New York: Free Press. Kastenbaum, R. J. (2004). Death, society, and human experience. Boston: Allyn & Bacon. Knight, K. H., Elfenbein, M. H., & Messina-Soares, J. A. (1998). College students’ perceptions of helpful responses to bereaved persons: Effects of sex of bereaved persons and cause of death. Psychological Reports, 83, 627–636. Kunkel, A. W. & Burleson, B. R. (1999). Assessing explanations for sex differences in emotional support: A test of the different cultures and skill specialization accounts. Human Communication Research, 25, 307–340. Kunkel, A. D. & Dennis, M. R. (2003). Grief consolation in eulogy rhetoric: An integrative framework. Death Studies, 27, 1–38. Lehman, D. R., Ellard, J. H., & Wortman, C. B. (1986). Social support for the bereaved: Recipients’ and providers’ perspectives on what is helpful. Journal of Consulting and Clinical Psychology, 54, 438–446. Lepore, S. J., Silver, R. C., Wortman, C. B., & Wayment, H. A. (1996). Social constraints, intrusive thoughts, and depressive symptoms among bereaved mothers. Journal of Personality & Social Psychology, 70(2), 271–282. Lindemann, E. (1965). Symptomatology and management of acute grief. In H. J. Parad (Ed.), Crisis intervention: Selected readings (pp. 7–21). New York: Family Services Association of America. Littlewood, J. (1992). Aspects of grief: Bereavement in adult life. London: Tavistock= Routledge. 426 J. J. Rack et al. Marwit, S. J. & Carusa, S. S. (1998). Communicated support following loss: Examining the experiences of parental death and parental divorce in adolescence. Death Studies, 22, 237–255. Nabi, R. L. (2003). Exploring the framing effects of emotion: Do discrete emotions differentially influence information accessibility, information seeking, and policy preference? Communication Research, 30, 224–247. Pennebaker, J. W. (1997). Opening up: The healing power of expressing emotions. New York: Guilford Press. Pennebaker, J. W., Mayne, T. J., & Francis, M. E. (1997). Linguistic predictors of adaptive bereavement. Journal of Personality and Social Psychology, 72, 863–871. ´ Pennebaker, J. W., Zech, E., & Rime, B. (2001). Disclosing and sharing emotion: Psychological, social, and health consequences. In M. S. Stroebe, R. O. Hansson, W. Stroebe, & H. Schut (Eds.), Handbook of bereavement research: Consequences, coping, and care (pp. 517–544). Washington, DC: American Psychological Association. Petty, R. E., Rucker, D. D., Bizer, G. Y., & Cacioppo, J. T. (2004). The elaboration likelihood model of persuasion. In J. S. Seiter & R. H. Gass (Eds.), Perspectives on persuasion, social influence, and compliance gaining (pp. 65–89). Boston: Allyn & Bacon. Range, L. M., Kovac, S. H., & Marion, M. S. (2000). Does writing about the bereavement lessen grief following sudden, unintentional death? Death Studies, 24, 115–134. Range, L. M., Walston, A. S., & Pollard, P. M. (1992). Helpful and unhelpful comments after suicide, homicide, accident, or natural death. Omega: Journal of Death and Dying, 25, 25–31. ´ Rime, B., Finkenauer, C., Luminet, O., Zech, E., & Philippot, P. (1998). Social sharing of emotion: New evidence and new questions. In W. Stroebe & M. Hewstone (Eds.), European review of social psychology (Vol. 9, pp. 145– 189). Chichester, UK: John Wiley & Sons. ´ Rime, B., Philippot, P., Boca, S., & Mesquita, B. (1992). Long-lasting cognitive and social consequences of emotion: Social sharing and rumination. In W. Stroebe & M. Hewstone (Eds.), European review of social psychology (Vol. 3, pp. 225–258). Chichester, England: John Wiley & Sons. Samter, W. (2002). How gender and cognitive complexity influence the provision of emotional support: A study of indirect effects. Communication Reports, 15, 5–16. Sears, S. R., Stanton, A. L., & Danoff-Burg, S. (2003). The yellow brick road and the Emerald City: Benefit finding, positive reappraisal coping, and posttraumatic growth in women with early-stage breast cancer. Health Psychology, 22, 487–497. Servaty-Seib, H. L. & Burleson, B. R. (2007). Bereaved adolescents’ evaluations of the helpfulness of support-intended statements: Associations with person centeredness and demographic, personality, and contextual factors. Journal of Social and Personal Relationships, 24, 207–223. Shapiro, E. R. (2001). Grief in interpersonal perspective: Theories and their implications. In H. Stroebe & S. Stroebe (Eds.), Handbook of bereavement research (pp. 301–327). Washington, DC: American Psychological Association. Evaluations of Grief Management Messages 427 Smyth, J. M. (1998). Written emotional expression: Effect sizes, outcome types, and moderating variables. Journal of Consulting and Clinical Psychology, 66, 174–184. Tannen, D. (1990). You just don’t understand: Women and men in conversation. New York: William Morrow. Todorov, A., Chaiken, S., & Henderson, M. D. (2002). The heuristic-systematic model of social information processing. In J. P. Dillard & M. Pfau (Eds.), The persuasion handbook: Developments in theory and practice (pp. 195–211). Thousand Oaks, CA: Sage. Trobst, K. K., Collins, R. L., & Embree, J. J. (1994). The role of emotion in social support provision: Gender, empathy, and expressions of distress. Journal of Social and Personal Relationships, 11, 45–62. Wood, J. T. (1996). She says=he says: Communication, caring, and conflict in heterosexual relationships. In J. T. Wood (Ed.), Gendered relationships (pp. 149–164). Mountain View, CA: Mayfield. Zunin, L. M. & Zunin, H. S. (1991). The art of condolence. New York: HarperColllins.