.Q ...-.n ...-o '.a‘ - £30.... I... 3.1:. . r 1...... g. .3} .........-..o. ‘3 :9. :1... -. Q 2 .- 1 ..... .o' ...». O... o . 0.: ...-.00: ... an .. o: .- nu. ...! .7 . ‘J.‘)....-I :o-- .u U uhi. ...... av: , u. 3... IF»... ..».u...... .15, .... _. -. P .1910? ...“. on... .. o-znv...) .fioui; 1.2.... .... .q. .t .. .3 . ...... .3, 3.2L: ..QO) -- 02$. 'V-‘luon .104. . v... ‘..~. I‘- 9 o. :30...— .... 9.3.. .1 o ... 5:..C...c. .. ...... ... o . . ... oo~ . I ....w. .23.... .. .o .q. . . \l .... . 3 . ..‘a .. :1 ...!) ”us“. .u o 0‘.”n‘ p. 1;. .... l... ... . v v '0. “v ' ...... $3.. . .o .. n .3. ...}IQ: . ......21. 9-3:. . i. '0- .,‘ u. a. ,40'5-0 .. v0. , do...!.--. 11.. v... .1." gnu-ll . . . .v. I .....a. ...-Os ...! 1.. . . . 0-3....1 ...... . .. . .....§ ‘0... O ....Co-v‘in a . ....O-‘u‘ . .. awn .. . ._ M?.x¢£p Co ‘.3 .7... .. . .... l 5:. . . I... .. Auau an. . .I an. .-. o.£s.‘-.—«7wfl.l\s. '39-“. Q. .1 m... o...,.£.‘-4 .... :32731; m. v2.1.4“... .... 3‘1 .... .. ..m. ... . . . . , .. . . .hr...w...f so. . . . . . . . . p o '0.- ..t 7 o . . . . . .. . . ... . . . . . ‘T-‘N-“Ovfi “I... . . O)! .’....ov..l . .|: 1Q? .lfltctu’. . a. . )8 h. . \r-u. ($.31! . \ u . ..rfihu...d discr. . 4“ t lWJJJunu .- :o%\fla|.a1§h~.~l.fii 95“.”. 3.21.. u‘.‘fi.‘ (a 3%. ‘0‘ |I¢.o3..’.$1‘;¢.':1’ n... Qufiog é .... .. ... .. . _ . . a}: . . . . . . . . . . . . . . . . a . . . . . . ...! ....Oo.\h§.s.—"fi—\!Cti!gt¢1 .o “W“ q . .. . . . . . C . . . ... ”vat 35w...t.o\.10tilc tori!!- oo dico’v‘i‘. .....Otz‘ a. ‘. ...K‘Io: OIQ .‘tlfgt. a I .. ..no.I-.3-${' «O 0". o... 1%, \Y$?S§uu.mwm‘ ‘1.’ \u.v:c . . v.u..~..9§or..§'ww.§.ur.1¢:3.t I .9 . a... man-9s . u... c . . I . *2.) .... out: . o , ... out.» .1033: . ... . .v. v. u. . . . ..u... . .3 .... . . ~51: ..t). ... 135:6: 3‘3339 I 4. ...... 1‘.‘o ...‘Q . 7.. ... ‘3. ‘I. I." Q . .. o. :3. ...... U: . I . ..‘o .‘.X..~ ~..;‘. 2 5......» u. ... o. . II...- ......o: 0 .1 .u. . - ..l . ..«\ 5.... ." ... . CI... . . . 0‘ .0. u. .AO ...-.... 1‘. O c o o ( . .. .c . . , ... .v. .n ...o, e. . . . .... . . _ . .... 0: . . . . .1 . . ... .r c 4. . o . I. . .. a . . . . _ . .. .. .. 5.4... . 3.3.... . 1.: . . 21.1.. .... .. . ...a... . .... ...n... . ... . .... . _ . .. . . . ... . . ... .. ..ml’. 0 155—..13310 .. . ‘00, ‘L i o “‘1‘“... d. .c hard!— 4 u .. .... .. ... c. L ..fit‘io...¢ “I. L...r.L.h:.I-n.s “I‘MUNIJZoithfit 6.... . . c. . , .40 o . .’ ‘Iv. o . . “u. I . 2 A on. .4 9.... ... ./ IVE-.35.. i I: u 5 oh“. wCu:\l.." oonvc ......r? ’o‘. ... .ND‘. gt». v.0.m .. so . .. .9. .. . ua . .....VIS. I...u$..\..“l._.ll. a "...... ‘9‘: c . . . . .. . ... . §23T o. .....I. 2. ...!‘S |Q1(\!1‘ . . . .o . . . . . )Q‘. o. .. A1.iv.\..’§‘\§!\.nu%..1§ .. .q : .. . .. , I .. ..I-.!o...31|~.2; l‘at..lt1\ ‘5chv _ .n . o . . . p. . v o. t. . 1.0.50 .....u-noZ ‘3‘. ’49....a“. (‘1‘qu . .... . . o .. ......l... . , ...:nl‘339‘.$.1.v§\a3v1“1 ‘ . o u.- .‘1 ¢. . . . . . . . . . . . . . .. . ‘0’.b..‘..:43. ..“\.fl . . . . . . . . . . . . ., ....6ti\W$-...¢| . . . . . . .. . , ...-ti; '- . . .5 V... 1.3!. Q . I.’ .‘925 ....l‘ . , .- ....o: . . H. . .. . \l. :0... . . . .. . . ... c. ...o . ..S . .3,..9von .. . . u a. T- .. . . . . . . . ... n. o O ‘ a. I» I .c . £- 0.» \ . . o v a. . . . . ... ‘ t . . o v .31.... .... u . o u .t o . . . I . . .. . ... $0.61....“ . . . . 0‘2””:L. ....’ . _ . . . . . . ... . . . . . . . .‘ 3015‘ . o . . I . . ... . . . .... . (....V‘o .I‘ . . $- . ”No.00. a ...-o“ . .chq-y 22-. . .31.. ‘l .‘r- .o‘.o‘.uh ..cvo‘.“- fix. ...-0....... . . . Wfl-‘Wyn u I ' r.‘ . O . I vi.- :. a3:- ; - .. . r x . n “ . ,. l . v u v n . .. 9 . n o A 9 to .... .. Inca... o .. . .~ ... u. . ..w. .. ..l ; . . n . I . i o . . ... .. .. . . . . a. ..v.....n. ... . Q .u....‘cni..‘ ' . . u I ...I ‘N. . Q. ‘."1\‘ ..... . . ,o L “'3 "',"“!V§”qg '2‘ ' :3 .... . . c. . L.|..\ . . . .... 4. . . . o ... ...... . . ... u . , s . u ...... .c . . ... _ . . a .H . . . ... ...... _ . . . . .. -- ... . . . . . I. - .... . .. .. n . . v u. .. n .. n . .. . .. , . .. p. . .. ... ... . . . .. . . . . hi1... . . . . . . .,...ll.o.. - c , ' a. ..- I? . 3o: . Q. ...'. . 06?. )9. nl. .. . ‘ ...op..._ . .. .Iyl. .. . I.~IQ| u.-t-,3.. .... .. . 110.... .3‘ 13!.- . ). i!‘.-A. 0‘ .mtww, j . § . 7‘2 ..‘ n i o- 0 a Y . vcu . ...-7‘...- ..f I. ...: I. . .o t:..‘ . . , D . .. .v- . ' 'V . ’0 0" .‘l s x'. . .‘ . o . o . 0.1 .. ..- ...s 1... .. . O. , . . . o . _ - . "u, . .u v A . ..oo- .0. .I o . . _ .. a v. . .. . . kid I . . . 1 _ . ...o. , . V . ‘ . . .\ . I ...-I I n. . . ...... ....a cl 0‘... . .. ......O.¢o..,' . _ . . .. . ...: 3 . . . o . a . I? 4 . u. . .. n. .. .\ .c . .1 . . _ . o , .. ......u‘. .0 :u . . c . a. .. . , 3 p. . . u$.o T _ a . . .. v . ¢ ..t. . . . .. ...... . .o . ,. .. .0. . . . . u. .m. ....l..|n . . . . ... . .. . ... . .. .. .. . . . . . . . .. . . . ... ......nI . . . . . . . . .. . . . . .... u a... . ... .. . . .. . . . ..va‘IQ...a. . . . . . . . . ..--l ‘I. 9'. 0:30... . .7! -..!ch D... .‘ '- O a . v . . a.\.‘t - ‘— Q o o..- . .. . . . . . . .. I... .Q‘.(. . ..T- u ...... . \- . u . o , . "0W '--.. a? - ‘. .‘3. ... ' a i: _ : ‘2‘. _ ; ..- o ... 3 .. - ' I '2 '. . . - C .u . . . u. .9 .. .. I , . o . . t . u . o t . ._ ' l .— . - D... o 1 1.58... coo 04:... .....nlitk . ..2. ......o...” . t..lt.... , . .. ... - .o‘.¢ . .a . .. .q . . o o . o “b b g ,. o . . .. _- .- . . 04.- 1 .‘n. QLDH.“Hp4 “11...; Ian... O 0 1.31,... . . .. ... ,u. . A o .. . ’13.. . . ... .. . r \ a J o 1 ..o.‘. .A . . ... . - nus 3... ..... u .15. . 4&1“... t .a.. ...-tngaicw‘nwfipwz . ..o. ..3 . .. ... .. .1!“ . ... . . 3.. . 3v . .... . . . .ru 3 4.5 . . .. . . . . 0.4:? u. . . _ ... .<. . . . . . . . ... t L . ..tr.‘ .\4. . . , . .. . .... a.) o.... _ . . h“ . m . .x .a. .1 a ,. :5 ...... . .. ... ..‘Q .uod.'1 v .3 .I-.. ”a {3.3.3:3. . .. M. .....hwfimw “Hiding-...” . «Wm. 13“.. :r ... .vfilu.1.n...¢..r \VFQ. ”swhuWL thfl'v‘nortwfiua .Mu ... ‘QIDM‘%. ..- g i . . . . ... . . .. ... r . 3-2.1)», ’. . v 30.1“...3. c. I... .‘l‘fiho’ . .. . a... 959.3} 0-. . . . 9P. t . \ . | . . 39 7-1.. 2‘.- u a . Q 37...“ {'13 ‘1“..5 . . . . . t ,‘ZQ! This is to certify that the thesis entitled ATTRIBUTION BASED CAPITALIZATION RESPONSE MESSAGES; ASSESSMENTS OF THEIR EFFECTIVENESS AND RELATIONSHIP TO SELF-ESTEEM presented by REBECCA J ROTH has been accepted towards fulfillment of the requirements for the MA degree in COMMUNICATION /T Maj rofés'sor’s Signature QDI ‘ ‘ Date MSU is an Affinnative Action/Equal Opportunity Employer LIBRARY Mlchlgan State Unlverslty 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 5108 K:/Prolecc&Pres/CIRCIDateDue indd ATTRIBUTION BASED CAPITALIZATION RESPONSE MESSAGES: ASSESSMENTS OF THEIR EFFECTIVENESS AND RELATIONSHIP To SELF- ESTEEM By Rebecca J Roth A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF ARTS COMMUNICATION 2010 ABSTRACT ATTRIBUTION BASED CAPITALIZATION RESPONSE MESSAGES: ASSESSMENTS OF THEIR EFFECTIVENESS AND RELATIONSHIP TO SELF- ESTEEM By Rebecca J Roth Capitalization, or the act of telling another person about one’s success, has received little attention from the communication discipline. This thesis examines the skillfulness of capitalization messages from the perspective of attribution theory. Previous research has suggested that self-esteem may moderate perceptions of capitalization messages. Thus, this study also examined if self-esteem influences likelihood of capitalization or moderates liking for capitalization messages differing in levels of skillfulness, using hypotheses derived from self-verification theory and self- enhancement theory. Participants (N = 312) were asked to complete self-esteem measures, indicate their frequency of sharing positive events with others, and rate capitalization messages in an online survey. Results show that messages making internal, stable, and global attributions were liked better than messages making external, unstable and specific attributions for the positive event. In support of self-verification theory, self- esteem was positively associated with the likelihood of capitalization. In support of self- enhancement theory, both high self-esteem and low self-esteem individuals liked skillful messages better than unskillful messages. The study’s implications as well as limitations and fiiture directions for research are discussed. ACKNOWLEDGEMENTS This thesis would not have been possible without the help and dedication of many influential people over the past two years. First, I must extend my deepest appreciation to my advisor, Amanda Holmstrom. Mandy was always there to offer support for both the stressful times and the good. Her guidance, dedication, enthusiasm and experience made this thesis a joy to work on. My committee members also deserve a world of thanks: Bob LaRose, who kept me employed and taught me statistics; and Sandi Smith, who connected me with both Mandy and Bob and influenced me to choose the thesis track. I am also very grateful to my amazing Telecommunication, Information Studies, and Media pals for providing a place to laugh, a place to gripe, knowledge of the system, proofreading, printing, and acting as pre- pretesters. Without all of you, I would not be where I am today. iii TABLE OF CONTENTS LIST OF TABLES ............................................................................................................... v Capitalization ....................................................................................................................... 1 Capitalization Responses ................................................................................................. 2 Celebratory Ego Support .................................................................................................. 6 Attribution Theory ........................................................................................................ 7 Capitalization and Self-Esteem ......................................................................................... 16 Self-Esteem and Likelihood of Capitalization ................................................................... 16 Self-Verification Theory ................................................................................................ l7 Self-Enhancement Theory ............................................................................................. l8 Self-Esteem and Perceived Capitalization Skillfulness ..................................................... l9 Self-Verification Theory and Perceived Capitalization Skillfulness ............................. 20 Pre-Test .............................................................................................................................. 22 Participants ..................................................................................................................... 23 Procedure ....................................................................................................................... 23 Study .................................................................................................................................. 35 Method ........................................................................................................................... 35 Participants ..................................................................................................................... 36 Procedure ....................................................................................................................... 36 Results ............................................................................................................................ 41 Discussion .......................................................................................................................... 50 Limitations ..................................................................................................................... 50 Conclusion ..................................................................................................................... 60 Appendix A ........................................................................................................................ 65 Appendix B ........................................................................................................................ 67 Appendix C ........................................................................................................................ 68 Appendix D ........................................................................................................................ 69 Appendix E ........................................................................................................................ 70 Appendix F ........................................................................................................................ 71 Appendix G ........................................................................................................................ 72 Appendix H ...................................................................................................... ~. ................. 73 References ...................................................................... 74 iv LIST OF TABLES Table 1 Summary of Capitalization Dimensions Shared by Gable and Rusbult ................. 5 Table 2 Samter’s Coding Scheme of Celebratory Events ................................................. 13 Table 3 Capitalization Message Attribution Examples ..................................................... 17 Table 4 Stability Dimension Attribution Based on Message Rating Comparison ........... 27 Table 5 Locus Dimension Attribution Based on Message Rating Comparison ................ 28 Table 6 Globality Dimension Attribution Based on Message Rating Comparison .......... 29 Table 7 Emotion Reliabilities ........................................................................................... 33 Table 8 Pre-Test 2 Attribution Based Message Rating Comparison ................................. 35 Table 9 Means, Standard Deviation, and Reliabilities by Dimension ............................... 40 Table 10 Means and Standard Deviations for Rating of Capitalization Message ............ 43 Table 11 Means and Standard Deviations for Positive and Negative Emotion for H6 ..... 50 Attribution based capitalization response messages: Assessments of their effectiveness and relationship to self-esteem When something good happens in life, we commonly turn to friends, family, and significant others to share the good news. The initial responses and appraisals we receive from others greatly affect our self-concepts and the feelings we have about positive events (Gable, Reis, Impett & Asher, 2004; Gable, Gonzaga & Strachman, 2006; Langston, 1994; Greca, 1982). Thus, it may be important for individuals who have experienced a positive event to engage in capitalization, or the act of sharing a positive event with another person (Gable et al., 2004; Gable et al., 2006; Langston, 1994). This thesis examines the capitalization process using attribution theory, self-verification theory, and self-enhancement theory. In particular, the proposed study will (a) identify the characteristics of skillful capitalization messages, (b) examine how self-esteem moderates the likelihood of capitalization, and (c) examine how self-esteem moderates preferences for different types of capitalization messages. Capitalization Capitalization refers to the act of seeking out another person to share a positive event. Research indicates that capitalization has multiple positive personal and relational effects for the recipient. One of the positive effects of capitalization is temporarily making a person feel better about the event (Langston, 1994). In addition, capitalization has implications for relationships. For example, Gable and colleagues (2006) found that individuals felt more committed, satisfied, and loved in their relationships when their capitalization attempts were met with a sincere, enthusiastic response. Other capitalization attempts (e.g., those that were unenthusiastic or pointed out the drawbacks of the event) were correlated with relational dissatisfaction. Gable et al. (2004; 2006) also found that capitalization is associated with greater overall life satisfaction. In addition to living happier lives, people retain and recall information as well as derive greater satisfaction about the positive experience after the act of verbally discussing the positive event (Gable etal., 2004). Though the effects of capitalization on recipients’ mental and physical health have not yet been researched, extensive psychological research points to the beneficial health effects of expressing positive emotions. Multiple studies have indicated that being aware Of and thoroughly experiencing positive emotions positively affects a person’s physical (e.g., mortality) and mental health (e.g., mood) (Brown & McGill, 1989; Burton & King, 2003; Danner, Snowdon & F riesen, 2001; Fredrickson, Tugade, Waugh & Larkin, 2003; Lyubomirsky, Sousa & Dickerhoof, 2006). Although the literature on experiencing positive emotion centers on cognitive processes, capitalization may have similar effects on mental and physical health. Capitalization Responses The capitalization process was initially studied by Langston (1994) in a series of diary studies. Langston had participants keep a daily diary of positive and negative events and their emotional responses to those events. The study found that when positive responses from others were received for good events, the recipient experienced greater positive affect in general. Capitalization responses are the feedback one receives afier sharing a positive event. Gable and her colleagues (2004) replicated Langston’s diary study and also found that when positive responses from others were received for good events, the recipient experienced greater positive affect in general. Gable extended Langston’s work by identifying two dimensions of capitalization responses, active/passive and constructive/destructive. The active/passive dimension describes how engaged the receiver is in the individual’s positive event. Active responses tend to be more elaborate (e.g., “That’s awesome! I knew you would do great. When do you start? Tell me all about the interview”), whereas passive responses minimally address or even dismiss the event (e.g., “That’s nice”). The constructive/destructive dimension describes how encouraging or positive the response is to the recipient. Constructive messages are uplifting in tone and content (e.g., “You’re so smart. You deserve the A.”), whereas destructive messages highlight the potential drawbacks of the event (e.g., “How are you going to pay for school? Are you sure you can handle the work?”). These dimensions may be crossed to produce four categories of capitalization messages: active/constructive, active/destructive, passive/constructive, and passive/destructive. Active/constructive messages demonstrate enthusiastic support for the event whereas active/passive message demonstrate quiet, understated support (Gable et al., 2004, p. 232). Active/destructive messages quash the significance of the event and passive/destructive messages ignore the event (Gable, et al., 2004, p. 232). In their 2006 article, Gable and her colleagues give the example of a woman, Maria, coming home from her job as a lawyer and telling her husband that she has just been promoted to lead lawyer on a very important case. According to Gable, an active/constructive response to Maria’s situation would be “wow, this is great news! Your skills and hard work are definitely paying off; I am certain that your goal to make partner will happen in no time. What is the case about?” A passive/constructive response would be, “that’s nice, dear.” An active/destructive response would be, “wow, I bet the case will be complicated; are you sure you can handle it? It sounds like it might be a lot of work; maybe no one else wanted the case. You will probably have to work even longer hours this month,” and a passive/destructive response would be, “what do you want to do for dinner?” (p. 905). Gable theorized that active/constructive responses are more skillful than passive/destructive messages because they help the recipient internalize the positive event; thus, the event becomes more salient in the recipient’s mind (Gable et al., 2004; Gable et al., 2006). Gable’s dimensions are derived from previous research (Rusbult, Zembrodt, & Gunn, 1982) that examines reactions to dissatisfaction in relationships. Gable and her colleagues use the dimensions identified by Rusbult et al. (active/passive and constructive/destructive); however, Gable et al. apply them to positive events as Opposed to reactions to relational dissatisfaction (see Table 1 for a comparison of Rusbult et al. and Gable et al.’s application of the dimensions). The active/constructive dimension for Gable reflects enthusiastic support for the positive event, whereas Rusbult’s active/constructive dimension, “voice,” refers to active discussion of the problems within the relationship. Gable’s passive/constructive category demonstrates an understated support of the positive event, whereas Rusbult’s passive/constructive dimension, “loyalty,” demonstrates the act of staying in the relationship and hoping for improvement. Gable’s active/destructive dimension demonstrates the act of calling out potentially negative consequences of the positive event, whereas Rusbult’s active/destructive dimension, “exit,” refers to the act of formally separating and ending the relationship. Finally, Gable’s passive/destructive dimension refers to the act of completely ignoring the positive event, whereas Rusbult’s passive/destructive dimension, “neglect,” refers to the act of ignoring the partner as well as the problems within the relationship (Rusbult et al., 1982). Theoretically, it seems problematic to apply Rusbult et al.’s scheme to responses to positive event disclosures. Not only do Rusbult et al.’s dimensions refer to responses to negative, as opposed to positive, events they also refer to reactions to prolonged relational dissatisfaction as opposed to acute events (like capitalization). Additionally, Rusbult et al.’s dimensions refer to situations that do not necessarily involve communication, whereas capitalization is a distinctly communicative phenomenon. Table 1 Summary of Capitalization Dimensions Shared by Gable and Rusbult Shared Gable’s Gable’s Examples Rusbult’s Rusbult’s Dimensions Definition Dimension Examples Active/ Enthusiastic “I sometimes get the Voice “He/she talks Constructive Support sense that my partner about what is is even more happy going on” and excited than I am” Passive / Quiet, “My partner tries not Loyalty “He / she gives Constructive understated to make a big deal out me the benefit of support of it, but is happy for the doubt and me” forgets about it” Active / Quashing the “My partner ofien Exit “He / she Destructive event finds a problem with considers it” breaking up with me” Passive / Ignoring the “My partner often Neglect “He / she avoids Destructive event seems disinterested” me for a while” The specific ways in which these dimensions are applied by Gable et al. appear to lack face validity. For example, Gable et al.’s passive/constructive capitalization measure, “my partner tries not to make a big deal out of it, but is happy for me,” does not seem analogous to the corresponding example for passive/constructive responses to relational capitalization, “he/she gives me the benefit of the doubt and forgets about it.” Nor is the active/destructive response, finding fault with the accomplishment, analogous to Rusbult et al.’s active/destructive response, leaving the relationship. The distinction between the content and form of active, passive, constructive, and destructive messages is also ambiguous. Gable et al. (2004, 2006) do not clearly state what the specific messages within the categories look like. Although she does code for her dimensions over the course of her studies (Gable, et al., 2006), coders focus on the valence of verbal and nonverbal activities “devoid of content” (p. 908). Although this approach provides information with regard to immediacy and nonverbal behavior, messages are not examined in terms of verbal Skillfulness. As psychologists, Gable et al.’s primary concern is how people feel about the responses they receive, and thus there is little focus on the content and form of capitalization messages. However, this thesis emphasizes the importance of message quality in capitalization attempts. To that end, Samter’s (1989) conceptualization of celebratory ego support is next discussed. Celebratory Ego Support A division of social support coined “celebratory ego support” may be conceptualized as a form of capitalization. Samter (1989) defines celebratory ego support as “messages aimed at making someone feel even better about an accomplishment” (p. 381). In her dissertation, Samter developed a scheme for coding celebratory support messages based on attribution theory. Attribution Theory According to attribution theory, causal attributions, or inferences about the perceived cause of events (Smith, Haynes, Lazarus, & Pope, 1993), are strongly related to the emotions associated with the event (Weiner, 1985). Attributions are classically characterized by three dimensions, locus, stability, and controllability (Weiner, 1985), and have been expanded to include the dimension of globality (Katz, Anderson, & Beach, 1997; Abramson, Seligman, & Teasdale, 1978). With regard to locus, events may be attributed to internal causes, indicating that the event was due to a trait or characteristics of the person (e. g., intelligence, effort), or external causes, indicating that the event was due to a force outside of the person, such as fate, luck, or God. Events may be attributed to stable causes, or those that are constant over time (e.g., intelligence), or to unstable causes, which may vary (e. g., effort). Finally, events may be attributed to controllable causes, or those that may be changed by the actor (e.g., effort), or uncontrollable causes, which are not subject to personal control (e. g., intelligence). Attribution theory is commonly used in the social support literature to examine attributions for upsetting events (Burleson & Goldsmith, 1998; Holmstrom, 2008; Samter, 1989). This research indicates that the attributions helpers make for events influence recipients’ emotional reactions to those events. For example, Burleson and Goldsmith’s (1998) theory of conversationally~induced reappraisal suggests that skillfiil emotional support messages have their effects by helping the target make sense of upsetting events. Holmstrom (2008) used attribution theory to explain why particular messages are more helpful in enhancing self-esteem after an esteem-threatening event; she found that messages that attributed negative events to internal, stable, uncontrollable causes were viewed as less helpful in enhancing self-esteem than messages that attributed esteem-threatening events to external, unstable, and controllable causes. Instead of focusing on attributions for negative events, Samter’s (1989) coding scheme for celebratory ego support examines attributions for positive events with a primary focus on the locus dimension. Skillfulness of celebratory ego support messages is rated based on “the extent to which the Speaker explicitly asserts that attributions of the target are responsible for his/her success” (Samter, 1989, p. 95). Higher level messages are more skillful because they “reinforce the reality of an achievement” (Samter, 1989, p. 387) and help the target to realize that they are responsible for their achievement. For example, the speaker may state, “you did great on the test because you are so smart,” indicating that the recipient is responsible for their good fortune. However, the Speaker could also say “the test must have been very easy,” thus indicating the ease of the test is responsible for the recipient’s good fortune, not the recipient themselves. This makes it acceptable for a person to celebrate their accomplishment with the speaker because then they are not breaching cultural norms by flaunting success (Samter, 1989). Additionally, skillful messages also are assumed to help reaffirm the relationship between the speaker and the target because they acknowledge the mutual understanding of the significance of the positive event (Samter, 1989). The coding scheme developed by Samter consists of three major divisions with six sublevels: lower divisions represent less skillful messages than higher-level divisions (Table 2 provides definitions and examples for each division and sublevel). In the first division, the speaker denies that the target is responsible for his/her success — that is, success is outside of target’s control (e.g., it is attributed to fate, luck, chance, or God). Associated with division one is sublevel one: the speaker makes an external attribution for the target’s success. An example of this would be, “Wow, you’re really lucky to get that internship.” In division two, the speaker implicitly acknowledges that target’s attributes may be responsible for his/her success, but no explicit mention is made of specific characteristics, abilities and/or behaviors. There are three subdivisions associated with level two (subdivision two, three and four of the coding scheme). Subdivision two acknowledges success and offers to celebrate in some public way but does not directly attribute success to internal attributes of the target. An example of this would be, “You got the internship! We should go out to celebrate!” In subdivision three, the speaker asks questions about the event and expresses his/her own feelings for the target’s success. For example, “I’m so excited for you! When do you start?” In subdivision four, the speaker implies that some global characteristic, ability and/or behavior of the target may be responsible for the event but specific attributes are not mentioned. For example, “You must have really stood out in the interview!” In division three, the speaker explicitly asserts that specific attributions of the target are directly responsible for success by citing characteristics, abilities and/or behaviors contributed to the achievement. Division three contains subdivisions five and six of the coding scheme. In subdivision five, the speaker draws attention to the person by crediting the achievement to a specific attribute of the target. An example of this would be, “That’s awesome! All of your hard work really paid off.” Finally, in subdivision six, the speaker provides an elaborate description of how specific attributes contributed to the target’s success. For example, “That’s awesome! You’re so smart and organized - I just knew all of your hard work would pay off. You really deserve it.” This thesis will expand Samter’s work by developing a modified version of her coding scheme to examine the Skillfulness of capitalization messages. Like Samter, this study will examine capitalization in the context of success as opposed to other types of positive events, such as luck or good fortune. Weiner’s attribution theory (Weiner, 1985) and attribution research by Abramson et al. (1978) are used to elaborate on Samter’s coding scheme by including not only the locus dimension of attributions, but also the additional dimensions of stability and globality in order to assess the Skillfulness of capitalization messages. Emotion. According to Weiner (1985), it is not an event itself that leads to emotion, but rather the attributions made about that event. Capitalization messages affect the attributions made in regards to the event, thus the attribution change leads to a change in emotion (Dillard & Meijnders, 2002; Nabi, 2002). Therefore, the same event (e.g., getting a job), may lead to happiness or sadness depending on the attributions made about the event. Emotions are commonly characterized as short lived, acute, intense, and in response to an external stimulus such as an event (N abi, 2002). Planalp (1999; 2003) addresses emotion theory as a process with specific components: a precipitating event, an appraisal process, physiological changes, action tendencies, expressions of emotions, and a regulation process. Although all of the components may be a part of evoking emotion, they do not all have to be present in order for the emotion to occur. Some of the key emotions, as identified by Lazarus (1991), include: anger guilt/shame, sadness, happiness/joy, and pride. Two perspectives have emerged in communication research as 10 the dominant way of examining emotion: the dimensional perspective and discrete perspective (Nabi, 2010). The dimensional perspective is a more generalized look at arousal and valance (Nabi, 2010). The bi-polar valance model is one way to dimensionally look at emotion. In the bi-polar valance model, emotion is measured along a continuum in pairs such as good-bad and happy-sad (Dillard & Meijnders, 2002). The pleasure-arousal theory, another dimensional model of emotion, posits that affect can be defined in terms of valance (i.e. good or bad) and the extent to which a person experience the emotion (Dillard & Meijnders, 2002). Although this is commonly used and serves and a good starting point for examining emotion (N abi, 2010), dimensional models are generally characterized as requiring a less cognitive response from the participant (Dillard & Meijnders, 2002). The discrete perspective posits that emotions can be seen in distinct, unique states and have specific appraisal patterns (Nabi, 2002; Dillard & Meijnders, 2002). The discrete emotion perspective allows us to look at both emotional valence and intensity measured categorically (i.e. happy, sad, joyful, mad), which allows for variation within the emotion to be observed and recorded (Nabi, 2010; Dillard & Peck, 2001). These discrete emotions result in action tendencies that later can affect perception, cognitions and behaviors (N abi, 2010). Because the discrete perspective allows for looking at multiple emotions at one time and intricately tease out which emotions are occurring and which are not, it is especially important in communication research (N abi, 2010). Therefore, this thesis approaches emotion from the discrete perspective. Discrete emotions such as sadness, guilt, joy and happiness, can then be loaded 11 onto two dimensions: positive and negative (Nabi, 2002). Skillful messages are expected to increase positive affect (Gable et al., 2004; Gable et al., 2006; Langston, 1994) where as research on low quality messages shows that unskillful messages decrease positive affect (Vangelisti & Young, 2000). In fact, the entire body of literature on cold comfort, (or miscarried) support documents than unskillful support messages, although well intended, may cause a person to experience negative affect (Brachman & Guerrero, 2006; Burleson, 2003; Holmstrom, Burleson & Jones, 2005; Samter, 1989; Vangelisti & Young, 2000). Additionally, Langston (1994) asserts that it is possible that less skillful capitalization messages may evoke negative emotion. Therefore, it was important to measure positive affect in response to capitalization messages. Because particular messages (i.e. external/stable/ global and external/unstable/global) were predicted to be low quality messages, it was necessary to also measure negative affect in addition to positive affect. Locus. The locus dimension is especially important in terms of how people feel. Internal attributions for success lead to feelings of happiness or joy whereas external attributions lead to feelings of sorrow or sadness (Peterson & Seligman, 1984; Weiner, 1985) Previous research has not addressed the distinction between attributing a success to an external/stable cause versus an external/unstable cause. If a success is attributed to an external/stable cause, there is the possibility the event will happen again, which may lead to positive emotion. However, the external attribution will cause people to still feel the event is outside of their doing. When a person attributes a success to an external/unstable cause, they may be even less happy than when the event is attributed to 12 an external/stable cause, because they cannot count on succeeding again. Therefore, it is possible that external/stable attributions for success will lead to greater positive emotion than external/unstable attributions. However, it may be that an external/unstable attribution opens the door for future successes to be attributed to internal causes. Table 2 Samter ’s Coding Scheme of Celebratory Events Division Sublevel Example I. Denial of Responsibi lity Il. Implicitly recognizes responsibil ity/no attribution lII. Explicitly recognizes responsibil "Y Explicit external attributes for success Offers to celebrate success Ask questions surrounding the success and express personal feelings Global characteristic responsible for the event but the attribute is not mentions Terse, but explicitly recognizes specific attributions responsible for success Elaborately recognizes specific attributes responsible for success 13 “You got so lucky getting that internship.” “That’s great! Let’s go out tonight!” “I’m so excited for you! Do you know where you’re going to live yet?” “You really deserved the internship.” “Being so disciplined in your studies really paid off.” “You really deserve this because you worked so hard. You’re going to do so well because you’re really organizes and smart. You’ve given up a lot lately, so I’m really glad it turned out well.” Due to the lack of research examining differences in preferences for external/stable and external/unstable attributions for success, the following research question is posed: RQl: Will participants differ in their ratings of messages that attribute success to external/stable causes versus those that attribute success to external/unstable causes? Globality. Closely related to the stability dimension is globality. How a person feels about a stable/unstable attribution is dependent on whether or not the success is global or specific (Peterson & Seligman, 1984) Global attributions indicate that a certain characteristic or trait of the actor affects a broad range of situations (e.g., “I’m really smart”), whereas specific attributions indicate that a certain characteristic or trait affect a narrow range of Situations (e.g. “I’m really smart at math”) (Abramson et al., 1978). Abramson and colleagues (1978) theorize that when people attribute a failure to global causes, they feel completely helpless even if the situation changes because the trait or characteristic is seen as constant. When the global failure attribution is paired with a stable attribution, feelings of hopelessness are amplified (Peterson & Seligman, 1984). However, if a failure is attributed to specific causes, the individual may have other abilities to help compensate for the shortcoming; thus they are more likely to feel they have the ability to be successful in the future. In parallel fashion, if a person attributes a positive event to a specific characteristic, they will feel better about having a strong, definite characteristic (e.g., “You are so intelligent in math”) than if the situation is defined in terms of a vague, global characteristic (e.g., “You are good”) (Peterson & Seligman, 1984). Additionally, 14 communication research notes that attributing successes to global characteristics may be less skillful than attributing successes to specific characteristics because explicit, concrete (specific) attribution messages (e.g., “You’re really good at biology”) are more elaborate and tailored to the individual than (global) messages (e.g., “You’re so smart”) (Samter, 1989) Similar to the locus dimension, Samter discusses globality, but only in the context of internal attributions. Her coding scheme dictates that more skillful messages elaborate on specific attributes rather than global attributes of the target. Thus, the following hypothesis is forwarded: H3: Messages that make internal/specific attributions for success will be rated more positively than messages that make internal/global attributions for success. According to Samter (1989), internal/specific attributions for successes are more skillful than internal/global attributions. Other studies on attribution have cited that attributing failures to global characteristics (e.g., “I can’t learn”) leads to depression and harms self-esteem (Peterson & Seligman, 1984; Abramson et al., 1978). However, the extant literature fails to distinguish between extemal/specific and external/ global attributions for success. To that end, the second research question is posed to examine ratings of capitalization messages that make external/specific and external/global attributions for success: RQZ: Will participants differ in their ratings of messages that attribute success to external/specific causes versus external/global causes? Summary of Attribution- Based Capitalization Messages 15 The previous sections theorize that attribution theory may be used to define what constitutes skillful and unskillful capitalization messages. Previous research shows that the locus dimension (i.e. internal/external) of attributions are the most important to self- esteem, followed by stability and globality respectively (Samter, 1989; Peterson & Seligman, 1984). Thus, the most skillful attributions for success are internal, stable, specific attributions (Peterson & Seligman, 1984). The next most skillful messages then would be those that attribute success to an intemal/stable global cause, followed in order by: intemal/unstable/specific, intemal/unstable/global, external/stable/specific, external/stable/global, external/unstable/specific, and external/unstable/global attributions. For a description of the categories with example messages, see table 3, which provides examples of messages for a situation in which a student succeeds in a class. Capitalization and Self-Esteem Gable and her colleagues (2004, 2006) speculated that self-esteem plays a considerable role in the act of capitalization as well as how a person perceives capitalization responses. Thus, this study will also examine how recipient self-esteem influences propensity to capitalize, as well as how it moderates perceptions of more and less Skillful capitalization messages. Self-Esteem and Likelihood of Capitalization Gable and her colleagues (2004, 2006) have noted that self-esteem may affect the capitalization process. More specifically, self-esteem may play an important role in whether or not a person is inclined to capitalize and it may also be a moderating factor in the capitalization process itself (Gable et al., 2004; Gable et al., 2006; Langston, 1994). 16 Table 3 Capitalization Message Attribution Examples Internal External Dimension Stable Unstable Stable Unstable Global “You always do “That’s great! “You are “Wow- you Student great because You were really always so were having a Success you’re good!” on top of things lucky.” really lucky that day.” day.” Specific Student “That’s why it’s “You worked “Tests in “I’m surprised — Success your major! You really hard on that class tests in that always do so great that exam.” are always class aren’t in required pretty easy.” usually so classes!” easy.” Note: Inspired by Abramson & Seligman, 1978 The impact of self-esteem on the capitalization process may be examined by using two rival self-esteem theories. Swann’s self-verification theory (SVT) is based on dissonance and balance theories in the sense that it assumes people strive for “cognitive symmetry” (Swarm, 1990, p. 414). In opposition to SVT, Self-Enhancement Theory (SET) posits that people “strive to maintain high levels of self-esteem” through their interactions with others (Swarm, 1990, p.410). In the next section, discuss rival predictions derived from these theories are discussed. Self-Verification Theory According to Self-Verification Theory (SVT), most people do not like to have their beliefs challenged, especially when they are beliefs about themselves (Swann, 1983). SVT asserts that “stable self-views provide people with a crucial source of 17 coherence, an invaluable means of defining their experience, predicting future events, and guiding social interaction” (Swarm et al., 2003, p. 369). By self-verifying, people are able to predict what is going to happen to them; thus, they can feel a sense of control over their environment (Swarm et al., 1992). Regardless of self-esteem level, people actively work to sustain their self- concepts because it bothers them to question who they are (Swarm, 1983). As a consequence, SVT predicts that people tend to choose interaction partners based on who will support their self-views (Swarm, Pelham & Krull, 1989; Swann, 1983; Swarm & Read, 1981). When people with LSE self-verify, they feel the need to seek information that will confirm who they are, “even if it means bringing others to recognize their flaws and limitations” (Swarm, 1990, p. 415). When Views are not self-verified, cognitive dissonance results. Therefore, self-verification for people with LSE may involve negative information about the self. In contrast, self-verification for individuals with high self- esteem supports their positive self-concepts and the discomfort of dissonance is not experienced. As such, SVT predicts that people with LSE will be less likely to capitalize, or seek feedback regarding the event, than people with HSE because doing so would create a social context for their negative self-concept to be challenged and cognitive dissonance to arise. H4: Self-esteem will be positively associated with the likelihood of capitalization. Self-Enhancement Theory In contrast to SVT, previous researchers have speculated that capitalization may be especially important for people with low self-esteem, because they need to receive the 18 message that their personal strengths are acknowledged and validated (Gable et al., 2006). As a consequence, SET asserts that people want to receive “feedback that is more positive than an individual’s self view” (Morling & Epstein, 1997, p. 1269). Additionally, SET predicts that people prefer to spend time with others who think of them more positively than they think of themselves (Brown et al., 1988) and that everyone attempts to create social interactions to present themselves in a positive light (Swarm, in press). Previous research on self-esteem and self-enhancing behaviors has found that people with LSE prefer self-enhancing feedback opposed to self-verifying feedback (Swann, Pelham & Krull, 1989). In order to achieve this goal, SET would predict that self-esteem will not moderate one’s likelihood of capitalizing; that is, people with high self-esteem will capitalize as frequently as people with low self-esteem. However, since SET suggests no difference (i.e., a null hypothesis), we do not forward a formal hypothesis here. Self-Esteem and Perceived Capitalization Skillfulness In addition to affecting the frequency of capitalization, self-esteem may also affect the types of capitalization messages people tend to prefer in response to sharing the event. Although it seems intuitive to believe that everyone would want to hear positive messages about themselves, self-verification theory states that this is not always the case. SVT asserts that experiencing a success may in fact be a very stressful event for a person with LSE because it counteracts how the person feels about themselves (Swarm & Predmore, 1985). However, self-enhancement theory asserts that people will enjoy the success and want to receive a positive response to it, despite their level of self-esteem (Swarm, 1987; Betancourt & Weiner, 1982). Self-verification theory and self- 19 enhancement theory are next used to predict how self-esteem may moderate the types of capitalization messages that a person will prefer to receive. Self-Verification Theory and Perceived Capitalization Skillfulness Using SVT, Swarm and Read (1981) determined that how we think about ourselves is extremely stable; thus, we want to process social information in a way that is congruent with our core thoughts and beliefs. That is, people consciously, as well as subconsciously, strive to solicit feedback that is congruent with their self-perceptions. Receiving information that is counter to self identity is a stressor (Swarm & Predmore, 1985). Whether or not this information is positive or negative does not matter; the feedback may be equally stressful (Swarm & Predmore, 1985). Therefore, there may be differences in the ways people perceive responses to capitalization messages based on self-esteem. A person with high self-esteem (HSE) may rate an unskillfirl message that makes an external attribution for success (e.g., “you got really lucky in that class”) as negative because this statement goes against what they believe about themselves (i.e., that they are responsible for their success). However, a person with LSE may rate this same message more positively than the person with HSE because their own thoughts are congruent with the statement (i.e., that they could not be responsible for their success and that it must be luck) (Swarm & Predmore, 1985). Therefore, people with LSE may perceive less skillful messages as higher quality messages more than HSE individuals because they reduce the stress of receiving divergent feedback. H5: People with low self-esteem will perceive less skillful messages more positively than people with high self-esteem. 20 SET proposes that everyone, despite their level of self-esteem, attempts to create uplifting social interactions during the self-enhancement process (Swarm, 1990). A study conducted by Brown and colleagues (1988) using SET found that people with low self- esteem “distort and bias personal information in a self-enhancing direction” (p. 445). Multiple SET studies show that neither people with HSE nor people with LSE enjoy hearing negative feedback (Swarm, 1987; Swarm et al., 1989), and the goal of everyone, regardless of their level of self-esteem, is to enhance how they feel about themselves as well as how others view them. Therefore SET suggests that everyone, regardless of their level of self-esteem, will generally prefer to receive skillful messages and that both people with low self-esteem and high self-esteem will rate messages that make internal attributions for success more positively than messages that make external attributions for success. Again, we do not forward a formal hypothesis, as SET suggests the null. Although everyone self-enhances, research has shown that people with HSE verses people with LSE do so differently (Swarm, 1990; Brown et al., 1988). People with HSE may self-enhance by giving more detail about the positive aspects of the success than people with LSE. Additionally, people with HSE may be looking for more elaborate messages from the helper because they want to openly celebrate the success whereas people with low self-esteem may not provide an initial message that invites an elaborate, skillful response from the helper even though they do capitalize. This is not to say that people with LSE will not self-enhance; they may simply do so differently than people with HSE. People with HSE may desire more skillful, elaborate self-enhancing message to feel good, because they already think positively about themselves (Swarm, 1990; Brown et al., 1988). On the other hand, people with LSE may 21 not require as skillful a self-enhancing message to feel good about themselves, because a person with LSE may only be enhanced to the starting point of the person with HSE. A message may not have to elicit as strong of a positive feeling for people with LSE to self- enhance. However, this is still enhancing their affect even though it is not to the extent that it is for people with HSE (Swarm, 1990). Because of the differences in how people with different levels of self-esteem self- enhance, SVT predicts that people with HSE will distinguish more clearly between messages that make internal attributions for success whereas SET predicts there will be no difference. In order to test these propositions, hypothesis six was formulated: H6: Self-esteem will moderate preference for messages that make internal attributions for success, such that LSE individuals will not differentiate between the four levels, but HSE individuals will rate internal/stable/specific and internal/stable/global levels more positively than the internal/unstable/specific and internal/unstable/global levels. Pre—Test 1 Method Before formally evaluating the hypotheses and research questions, a pre-test was conducted to ensure that the hypothetical scenarios designed for the study were seen as (a) a success and (b) important. It was also conducted to ensure that (c) the messages created for the study were seen by the participants as embodying the intended causal attributions. Participants were randomly assigned to rate nine attribution-based capitalization messages for one of three hypothetical scenarios: an academic success, an 22 internship success, or a romantic success. The pre-test was conducted as a paper and pencil survey. Participants Participants in this pre-test were 119 college students: 62 (52%) male, 51 (42%) female, and 6 (5%) who did not provide their sex. They were recruited from undergraduate communication classes at Michigan State University. Extra credit was awarded to the participants for their participation in the study. Procedure Participants were contacted during class meetings and given the option to complete the extra credit survey on their own time. Participants were presented with an informed consent form and a questionnaire consisting of (a) one of three hypothetical scenarios, (2) nine messages, (3) nine assessments of each message, three each assessing causal locus, stability and globality, (4) eleven items assessing emotional reactions to each message, (5) one item measuring event importance, (6) four items measuring likelihood of capitalization, and (7) three items measuring perceptions of whether or not the event was a success. All measures are included in Appendix B. The instrument contained a total of 189 items. Hypothetical scenarios. Participants were randomly assigned to read one of three hypothetical situations depicting themselves experiencing a success. Hypothetical success situations included: an academic success (N = 39), a romantic success (N = 41), and an internship success (N = 39). Situations are attached in Appendix A; sample directions are included in Appendix C. Event manipulation checks. A manipulation check was included in pretest one to ensure participants rated the scenarios as both important and as a success. One 7-point 23 Likert-type item assessed the perceived degree of importance for the event (e.g., “How important do you think it is to have an internship?”) and three 7-point items anchored by 1, “not at all a success” and 7, “completely a success,” assessed the degree to which participants view the event as a success (e.g. “DO you feel the event is a success?”). A complete list of the items can be found in Appendix B. All three scenarios were rated as important by the participants, as means were above the midpoint on the 7—point scale (academic M = 5.58, SD = 1.97; romantic M = 6.07, SD = 1.99; internship M = 5.51, SD = 1.99). All three of the situations were rated above the midpoint on the 7-point scale for success as well, indicating that the events were viewed as successes: academic, or =.79, M = 6.50, SD = .92; romantic, or =.85, M = 5.00, SD = 1.44; internship, or = .68, M = 5.71, SD = .90. The mean difference between the midpoint and participant’s scores for event success were significant in all three situations: for the academic situation, t (39) = 16.93, p < .001, for the romantic situation, t (40) = 4.44, p < .001, and for the internship situation, t (38) = 11.83, p < .001. The mean difference between the midpoint and participant’s scores for event importance were significant in all three Situations: for the academic situation, t (39) = 5.04, p < .001, for the romantic situation, I (40) = 11.14, p < .001, and for the internship situation, t (38) = 4.72, p < .001. Likelihood of capitalization. Participants responded to four 7-point Likert-type questions about how likely they are to share the event with others. Two of the items developed by Langston (1994) were used; the others were created for this study. Examples of these questions include, “Would you brag to others about this event?” and “Would you tell someone about this success?” A complete list of the items can be found in Appendix B. On average, ratings were above the midpoint on the scale, indicating that 24 participants were likely to capitalize for all three events: academic, or = .65, M = 4.86, SD = 1.25; internship, or =.82, M = 5.58, SD = 1.21; relationship, or =.57, M = 6.04, SD = .78. The mean difference between the midpoint and participant’s scores for likelihood of capitalization were significant in all three situations: for the academic situation, I (39) = 4.33, p < .001 , for the romantic situation, I (40) = 16.75, p < .001, and for the internship situation, t (38) = 5.58,p < .001. Messages. Participants rated eight messages derived from attribution theory and one control message that made no attributions. In line with the hypotheses for this study, messages were created to represent specific causal attributions varying on three dimensions: causal locus (internal or external to actor), stability (stable or unstable), and globality (global or specific). Planned comparisons were performed in pretest one to ensure that participants accurately identified the attributions made by each of the eight capitalization messages. Three 7-point Likert items (Appendix D) were designed to assess each of the three dimensions: locus, stability, and globality. The eight capitalization messages represent the following causal attributions for the success: intemal/stable/specific, internal/stable/global, intemal/unstable/specific, intemal/unstable/global, external/stable/specific, external/stable/global, extemal/unstable/specific, and external/unstable/global. The messages can be found in Appendix E. Tables 5 — 7 Show the means, standard deviations, and results of the paired samples t-tests comparing messages within the three dimensions. The following comparisons were made: with respect to stability, internal/stable/specific was compared to internal/unstable/specific, external/stable/specific was compared to 25 external/unstable/specific, internal/stable/global was compared to intemal/unstable/ global, and extemal/stable/ global was compared to external/unstable/global (Table 5). With respect to locus, internal/stable/specific was compared to extemal/stable/specifrc, internal/unstable/specific was compared to external/unstable/specific, intemal/unstable/global was compared to external/unstable/global, and internal/stable/ global was compared to external/stable/global (Table 6). With respect to globality, internal/stable/specific was compared to intemal/stable/ global, extemal/stable/specific was compared to external/stable/global, internal/unstable/specific was compared to intemal/unstable/global, and extemal/unstable/specific was compared to extemal/unstable/global (Table 7). The results indicate that participants accurately perceived the attributions made for all of the messages except seven. In the romantic scenario, the intemal/stable/global and extemal/stable/global messages were not seen as different. In the internship scenario, the stability dimension had two pairs of messages that were not significantly different (intemal/stable/global compared to intemal/unstable/global and extemal/stable/global compared to extemal/unstable/global). Also within the internship scenario, the globality dimension had one pair of messages that were not rated as significantly different (internal/unstable/specific compared to intemal/unstable/global). 26 owemmofi 3:3 52.32. 80:03....6 “gamewa 38.2.. matomcoasm fiesta $82 .320 A. A. 338?: .3. mud Sm. mmé 3.... ....m \EESxm a a .220. v .5... mm mm... «m. mogm cod ov Sam M... afim cod mm case. me. mad DESmZmEme LEA. .320 A. a \03825 3.. mad 3.. Ram co. 3mm 22.5.5 a m .mno_0\ m cm... mm cod .N. 2.6 cc... ow owe mm. mm.m cod mm on. mm. mud 038m2m585 ...mA. 058% n A. n \ozfimcD :1 EN 8.. 28 no. RH 258%.. .... a a 2.60%. N cod mm owe ow. .mm cod ow $6 3.. .06 mod mm omN mm. and oESmEeEme :5. 050on A. n A. \ozfimcb 3.. Sam 8.. mod 22 0.3 2.3.8:. a e a oEooam\ . 8.0 mm 8.». .9. one cod 9. .ws 3.. end cod mm 3...... on. mm.m Basile—:8... can. 1% S a a 2 me E e on x. me E a S. E 9562:. c.2853. c.8033. zomtemfiob Mztem swans}. Epsom «5:3..th 29538.5 REESE v 05m... 27 .owmmmoE 3.82. 503.2. meanest... .58..in 88.3.. Eatofloqsm anoba .632 S. E a . .. sum 8.. a .320 seam. .EEexm 3a a: 8... ..m .3... mm. a .m... .... 8.. ...... m...“ 8... ..m 2% .N. a 30.32.393525 .. 8A B; can. 5.. E 3... B .... E .320 .m... w . .... mam . 05925 . .535. 8... ..m 2... on. a 8... .... 9% 3.. a 8... ..m .....m ..2 m2... .965 m $6 86 \ 0.8.3.5 \ .mEoE. ...». 2.. E 8.. E e... E 2.6on Nam 8.... Sm . 0392: . .aEaxm 8... mm a... .... a 8... .... 8a a... a 8... ..m cm... on. a 8:6on N a... 5... man 3332:. .26.... .5. an. on. E 2.. Exam ocean. mm... .....m . 839m . .aesxm 8... mm 8... E. a 8... .... Nam .m. a 8... ..m New 8.. “2% unsafensfaeez. . 8a 8... ...... A; B . cm x. m: S . a x. w: B . a. x. BEES... 3.583. 0.888... eatemfiob M582 swayed: .335 2339......V =c.....§§.5 2.3% m 2%... 28 .mowemmoe 02.3 503.02 30:20....0 Enocfiwfi 0.8.0:. Sarasota Baoba $.02 A. a n .320 we... ....m 2... N3. 3.. a: .2222: . ....Esxm a a a 2...“.on v 8... ..m mam 2.... mm... ...... 2. SN 2.. ....m 8... ..m 8a 3.. .N... 2222335220 22. a n 2 320.0 a... N. .m e... .3 t... 2.... 3.2% . .aeexm om... a a a 2.28% m 2.... ..m - 2.. .3 2.... 2. .3 .2. 3m 2.... ..m o... 3.. S... 33233505 22. B .220 2... .3 ...... s as a... a}. .2222: . see... ...... E ... 00.8% m S... ..m - ..m. 2% 2.... .... E... 9... 8... 8... .3 8a .... 2... 2228225522 ha... 2 2 n .0920 mm... :H 3.. an... a... mam 3.2% . see... a a . 2.63m . 8... ..m ...... an. S... 8... .... 3... E. a; 8... ..m .....m an. “”2 22923622 ...: 1m: B . Q... s. w: B . Q... 2 w... B . a. 2 @2285. 0:52:51 0.80.03... temtemfiob m33- swans: 330m =0.§2...SV 20.30:..5 5:305 0 0.2m... 29 Dependent variables. Based on previous research by Langston (1994) and Gable et a1. (2004; 2006), receiving a skillful capitalization message should increase positive affect following a successful event. Therefore, hypotheses and research questions assessed how the person anticipated they would feel after receiving capitalization messages exhibiting varying degrees of Skillfulness. Affect was assessed with a modified version of the Positive Affect and Negative Affect Schedule (PANAS) (Watson, Clark & Tellegen, 1988). The scale consists of 20 items; 10 measure positive affect and 10 measure negative affect. In Langston (1994) and Gable et al.’s (2004) studies, participants were asked to rate how they feel about the 20 adjectives on a daily basis from “not at all” to “extremely. ” Given the nature of the current study, a modified version of the PANAS was used. On a 7-point semantic differential scale, participants rated select polar adjectives from the PAN AS (e. g., very happy-not at all happy; very distressed-not at all distressed). Instead of indicating how they would feel on a daily basis, participants indicated how they would feel after receiving each capitalization message. Because particular messages (i.e. extemal/stable/global and extemal/unstable/ global) were predicted to be low quality messages, it was necessary to also measure negative affect in addition to positive affect. Items were chosen based on which adjectives had the greatest face validity for assessing affect following a capitalization message. Affect was assessed with the following eleven emotions: guilty, ashamed, happy, cheerful, sad, proud, depressed, confident, distressed, upset, and accomplished. The items happy, cheerful, proud, confident and accomplished comprised the positive affect measure. The items guilty, 30 ashamed, sad, depressed, distressed and upset comprised the negative affect measure. The items used can be found in Appendix F. Averaged across all nine messages in each of the scenarios, the positive and negative affect scales exhibited acceptable levels of reliability (academic scenario: positive affect, a = .84, negative affect, a = .83; romantic scenario: positive affect, a = .73, negative affect, a = .77; internship scenario: positive affect, a = .81, negative affect, a = .84). See Table 8 for descriptive statistics and reliability information for the scales within individual messages. Discussion Pre-test one was conducted to ensure that the hypothetical scenarios designed for the study were seen as (a) a success and (b) important. It was also conducted to ensure that (c) the messages created for the study were seen by the participants as embodying the intended causal attributions. The results indicate that the events were generally seen as a success and important. Overall, the majority of the messages were rated as intended, however, there were a few that were not. All of the messages from the academic situation were rated as intended. Six of the eight messages from the romantic situations were rated as intended, thus there were two messages that needed to be reconstructed. These messages were the on the locus dimension of the intemal/stable/global and extemal/stable/global messages. Unfortunately, five of the eight messages from the internship situation were not rated as intended. Four were on the stability dimension: the extemal/stable/global, extemal/unstable/global and intemal/stable/global and intemal/unstable/ global messages. Two were on the globality dimension: the internal/unstable/specific and intemal/unstable/global messages. In order to ensure that 31 these messages were viewed by participants as they were intended, a second pre-test with reworked messages was necessary. 32 ...... ...... 3... no... 8... 8... a... 2.... a... 228.. ...... ...... a... ..o... No... .5... 2.... ...... 8... 22.82 0:52:3— 3... ...... mo... 8... ...... 3... 2.... 8... No... 928.. ...... 5.... 8... R... a... a... ...... 2.... ...... 3.332 952.8:— .5... .8... 2.... No... ...... ...... 2.... mo... 2.... 26.8.. 2.... ...... ...... 2.... 2.... ...... Q... R... 8... 2.2.02 2:533. .82.. .....uomm .320 2.28% .2502 028.... .820 .320 .28.... .2222: 2.225 22% 22% 2.225 .2222: 22% 22% \_aE2xm \EEme >9...me E8335 >mEBE \EEBE $.58:- \EESE 3.......2.§ .235. 5 22¢ 33 Pre-Test Part 2 Pre-test 1 revealed there were seven messages not seen by participants as embodying the intended causal attributions. Additional participants (N = 57) were recruited from undergraduate communication classes at Michigan State University to assess seven new messages created to better instantiate the intended causal attributions. Participants in the first group (N = 27) were asked to read the internship scenario and rate three messages. Participants in the second group (N = 30) were asked to read both the internship and romantic scenarios and rate two messages based on each scenario. The following internship messages were changed for the pre-test: intemal/stable/specific, intemal/stable/global, intemal/unstable/global, extemal/stable/global, and extemal/unstable/global. The following romantic messages were changed for the pre—test: internal/stable/global and external/stable/ global. Paired samples t-tests were performed to determine that the messages were rated as significantly different on the appropriate dimension. For the romantic scenario, intemal/ stable/ global was compared with extemal/ stable/ global. For the internship scenario, two pairs of messages were compared on the stability dimension (intemal/stable/ global was compared to intemal/unstable/ global and extemal/stable/global was compared to extemal/unstable/global) and one pair of messages was compared on the globality dimension (internal/unstable/global was compared to intemal/unstable/specific). Complete message alterations are listed in Appendix B. Table 9 shows the means, standard deviations, and results of the paired samples t—tests comparing attributions within the reworked messages. As may be seen, all messages differed significantly on the intended attribution dimensions. 34 Table 8 Pre-T est 2 Attribution Based Message Rating Comparison Locus Stability Stability Globality Note: Different superscripts indicate significant differences between paired messages. The purpose of the second pre—test was to ensure that participants saw the new messages as embodying the intended causal attributions. Results indicate that these messages were indeed viewed as intended. Thus, the final set of messages was obtained lntemal / Stable / Global External / Stable / Global lntemal / Stable / Global lntemal / Unstable / Global External / Stable / Global External / Unstable / Global lntemal / Unstable / Specific lntemal / Unstable / Global for the study. This study was conducted as an empirical test of hypotheses 1-6 and research questions 1 and 2. Power to detect medium sized effects (f=.3) was .95. Power was lntemship M SD 0.84 1.41 1.20 2.76 1.15 2.50 1.03 1.85 1.04 t df sig 7.59 28 0.00 4.28 26 0.00 3.09 29 0.00 Discussion Study Method 35 Romantic M SD 5.87 t df sig_ 0.78 5.56 26 0.00 calculated using G*Power 3.1 (F aul, Erdfelder, Buchner & Lang, 2009). Participants were given a questionnaire consisting of (a) a self-esteem measure, (b) demographic items, (0) likelihood of capitalization, (d) one of three hypothetical situations, (e) nine capitalization messages, (f) 11 items assessing positive and negative emotion for each message, and (g) other individual difference measures not pertinent to the current study. Participants Three hundred and twelve participants were recruited from undergraduate communication classes at Michigan State University using the communication department’s Experimetrix Participant Pool website (https://experimetrix2.com/msucom/) and from an undergraduate communication class at San Diego State University. One hundred twenty-five (40%) were male, 187 (60%) were female, and 3 (<1%) did not indicate their sex. Student’s ages ranged from 18 to 49 (M = 22.18, SD = 3.24). Forty- nine (15.7%) were freshmen, 42 (13.4%) were sophomores, 89 (28.4 %) were juniors, 131 (41.9%) were seniors, and 2 (0.6%) were graduate students. Most participants were Caucasian (N = 203, 64.9%); 41 (13%) identified themselves as “other”; 27 (8.6%) were Hispanic American; 24 (7.7%) were African American; and 18 (5.8%) were Asian American / Pacific Rim. Participants received extra credit or research credit for their participation. Procedure Participants clicked on a link that directed them to an online survey website (www.qualtrics.com), where they were randomly assigned to complete one of the three versions of the questionnaire. Following completion of the questionnaire, students were linked to a second website where they entered their personal information in order to 36 receive credit for participation. The information on the second website was not linked to the first to maintain the anonymity of participant questionnaire responses. Demographic Items. The demographic items (attached in Appendix G) assessed participant age, sex, year in school, major, and ethnicity. Hypothetical scenarios. Participants were randomly assigned to read one of three hypothetical situations depicting themselves experiencing a success. Situations are located in Appendix A; sample directions are included in Appendix C. Hypothetical success situations, validated in the pre-test, include: academic success (N = 103), romantic success (N = 104), and internship success (N = 108). Event success manipulation check. Three 7—point Likert-type items assessed the degree to which participants viewed the event as a success (e. g., “Do you feel this event was an accomplishment?”), with higher numbers equating a greater success. A list of these items can be found in Appendix B. The results show that each of the scenarios was viewed as a success, as means were above the mid-point on the 7-point scale: academic, M= 6.39, SD =1.09, a = .84; romantic, M= 5.39, SD = 1.57, a = .95; and internship, M= 6.04, SD = .1.29, a = .89. The mean difference between the midpoint and participant’s scores for event success were significant in all three situations: for the academic situation, I (102) = l6.32,p < .001, for the romantic situation, t (103) = 2.33, p < .02, and for the internship situation, I (106) = 7.71, p < .001. Likelihood of capitalization. Participants responded to four 7-point Likert-type questions on how likely they are to share the event with others, with higher ratings indicating a greater likelihood to capitalize. Results indicate that participants were generally likely to capitalize for the event, as means were above the mid-point on the 7- 37 point scale: academic, M = 4.86, SD = 1.25, a = .71; romantic, M = 6.04, SD = 0.78, a = .78; and internship, M = 5.58, SD = 1.21, a = .68. These items can be found in Appendix B. The mean difference between the midpoint and participant’s scores for likelihood of capitalization were significant in all three situations: for the academic situation, t (102) = 17.39,p < .001, for the romantic situation, t (103) = 13.91,p < .001, and for the internship situation, I (106) = 22.83, p < .001. Messages. Participants were randomly given one of the three scenarios. Participants were then asked to rate eight messages derived from attribution theory that were developed in pretests 1 and 2, as well as a control (no attribution) message that corresponded with the scenario they were given. A complete list of messages for each hypothetical situation is included in Appendix A. Dependent variables. Positive affect was measured using five 7-point semantic differential items including happy, cheerful, proud, confident, and accomplished. Negative affect was measured on the same scale using guilty, ashamed, sad, distressed, depressed, and upset. A breakdown of the scale reliability, means, and standard deviations by situation and message is located in Table 10. Self-esteem. The Rosenberg Self-Esteem Scale (1965) was used to measure general, trait self-esteem. Participants rate ten statements about the self, five positive and five negative (reverse scored). Participants indicate their agreement with each item on a 7-point scale bounded by strongly disagree and strongly agree, with higher scores indicating greater self-esteem. Sample items include “On the whole, I am satisfied with myself” and “I feel I have a number of good qualities.” The items measuring self-esteem displayed good reliability (a = .88). A median split was created at 5.7 to separate 38 participants with high self-esteem versus those with low self-esteem. The items from the Rosenberg Self-Esteem Scale can be found in Appendix H. 39 2.28% d m N d d db 0 n a d N m Nd \ 0380:: ad v. ..m ad ..N 36. wad mm. .m dad 0.. a... dd m. m6 ad dd .0 \ 380...”.— .0220 d m d d d 00 a a d w _ m ad \ 03825 ad .... m.m ad .2 Swm and 3..— .v mad 0.. m.m dd m. ..m ad a... .m 20:00....”— .0220 _ w w h w mm d v N _ 0 m mm \ 0305 ad m. m.m ad d.m am.m wad em. .m mad .... dd ad .... Wm ad _d.m .v \ BEBE 00.8.... d . m m 0 vm a d w a ... V cm \ 0305 ad m. 06 ad d.~ mm... vad dd. .0 mad 2 ad dd 2 m6 ad .dN .m \ _0E0.xm v m m d . 3. m d m . v d 0. ad 2 dd ad d.m m0... dad S... m mad .... dd ad .1 Wm ad 3.. .m .8502 0500.5 d 0 n d a co m _ d m d _ we \ 03825 ad _._ a.m ad 5.. ..dd .ad 3... .m bad x. N... ad .... ..m ad wm. .m 20:00.5 .0220 a m _ m d wm m N a N m ... m. \ 0380:: dd m. Wm ad m. 2% add mm. .m dad w. dd dd 2 m6 ad 00.. .v 22:0...— .320 m m n v m m. d d a d m d E \ 030% ad ad v.0 ad d._ 9.0 .ad ad; .0 wad m. dd dd 2 dd ad am... .m \ RES... 0.0.00.5 a a m m a mm a N. m d a a 3 \ 030% ad ad vd ad NJ 3.6 wad N: 6 _ad ad vd ad 2 dd ad 5; .m 22:25 a QM 3 d QM 3 d QM V< .o QM E .0 9m. 3. .c QM E .00t< 02.032 000.c< 0338a 0000“: 0>3mw02 000...: 03:00.. 000.c< 02.0»02 000...}. 030.09— cedficou 95:00:: 03:083— 0.8033. 20....20ENQ Ad w0.....~.56..~0- 350 .QQNBNSQ 3.335%. .330: a 030,—. 4O Results Tests of Hypotheses and Research Questions For each hypothesis, results are reported separately for each of the three situations in this study: academic, romantic, and internship. Hypothesis 1. Hypothesis I predicted that messages that make internal attributions for success would be rated more positively than messages that make external attributions for success. Paired sample t-tests were conducted comparing the four messages that make internal attributions (intemal/stable/specific, internal/stable/global, intemal/unstable/specific, intemal/unstable/global) with the four messages that make external attributions (extemal/stable/specific, external/stable/global, external/unstable/specific, external/unstable/ global) on the two dependent variables: positive emotion and negative emotion. Descriptive statistics related to H1 can be found in Table 11. Negative Emotion The mean difference between the internal attribution and external attribution messages was significant in all three situations: for the academic situation, t (102) = 4.31, p < .001, for the romantic situation, I (103) = 1.96, p < .05, and for the internship situation, t (106) = 8.14, p < .00]. Inspection of the means shows that, as expected, negative emotion was greater for the external attribution messages than the internal attribution messages. Positive Emotion The difference between the internal attribution and external attribution messages was significant in all three situations: for the academic situation, I (102) = 6.63, p < .001, 41 for the romantic situation, I (103) = 6.66, p < .001, and for the internship situation, I (106) = 13.14, p < .001. Inspection of the means shows that, as expected, positive emotion was greater for the internal attribution messages than the external attribution messages. Therefore, hypothesis one is supported. Hypothesis 2. Hypothesis two predicted that messages that make internal/stable attributions for success will be rated more positively than messages that make intemal/unstable attributions for success. Paired sample t-tests were conducting comparing intemal/stable/specific and intemal/stable/global messages with intemal/unstable/specific and intemal/unstable/global messages on negative and positive emotion. Means, standard deviations and effect sizes can be found in Table 11. The mean difference between the internal/stable and intemal/unstable messages was significant in all three situations: for the academic situation, t (102) = 7.91, p < .001, for the romantic situation, t (103) = 8.45, p < .001, and for the internship situation, t (106) = 6.40, p < .001. In each situation, negative emotion was rated higher for the intemal/unstable message. These results support H2. Positive Emotion The mean difference between the internal/stable and internal lunstable messages was significant in all three situations: for the academic situation, t (102) = 13.12, p < .001, for the romantic situation, t (103) = 10.87, p < .001, and for the internship situation, t(106) = 9.57, p < .001. In each situation, positive emotion was rated higher for the internal/stable message. These results support H2. 42 Table 10 Means and Standard Deviations for Ratings of Capitalization Messages: H I -RQ 2 Academic Romantic lntemship M SD r2 M SD r2 M SD r2 Hypothesis lntemal Negative 3 a 1'8; 1 Emotion 2.37 1.04 0.02 2.21 0.96 0.00 0.84 0.12 External Negative b b 2‘5?) Emotion 2.73 1.2 2.37 1.03 1.17 Hypothesis lntemal Positive a a 5'5: 1 Emotion 4.64 1.19 0.08 5.15 1.02 0.06 1.10 0.3] External Positive b b 3‘83) Emotion 3.87 1.41 4.56 1.15 1.50 Hypothesis lntemal/Stable a a 6'02 2 Positive Emotion 5.40 1.32 0.24 5.91 0.98 0.28 1.09 0.15 lntemal/Unstable b b 5'0: Positive Emotion 3.87 1.35 4.38 1.47 1.37 Hypothesis lntemal/Stable a a 1‘5: 2 Negative Emotion 1.94 1.08 0.1 1 1.73 0.99 0.15 0.79 0.07 lntemal/Unstable b b 2'03 Negative Emotion 2.79 1.27 2.68 1.23 1.08 Research Extemal/Stable a a 4'02 Question 1 Positive Emotion 3.97 1.48 0.00 5.11 1.35 0.15 1.68 0.02 Extemal/Unstable b b 3 '517) Positive Emotion 3.76 1.51 4.01 1.25 1.57 Research Extemal/Stable a a 2'3 Z Question 1 Negative Emotion 2.59 1.26 0.00 1.94 1.05 0.12 1.22 0.02 Extemal/Unstable b b 2'7?) Negative Emotion 2.82 1.24 2.79 1.22 1.32 Hypothesis lntemal/Specific a a 5'42 3 Positive Emotion 4.50 1.29 0.01 5.05 1.05 0.00 1.20 0.00 lntemal/Global b b 5 '6: Positive Emotion 4.78 1.29 5.25 1.12 1.12 Hypothesis lntemal/Specific a 3 Negative Emotion 2.40 1.10 0.00 2.31 1.02 0.00 1.78 0.84 0.00 lntemal/Global Negative Emotion 2.33 1.12 2.1 1 1.06 1.84 0.94 Research Extemal/Specific a a 25% Question 2 Negative Emotion 2.84 1.26 0.01 2.54 1.01 0.02 1.21 0.00 Extemal/Global b 2’5 3 Negative Emotion 2.58 b 1.24 2.18 1.17 1.18 Research External/Specific a a Question 2 Positive Emotion 3.70 1.46 0.01 4.29 1.20 0.04 3.85 1.56 0.00 Extemal/Global b b Positive Emotion 4.04 1.50 4.83 1.31 3.75 1.60 Note: Different superscripts indicate significant differences between paired messages. 43 Research Question 1. The first research question asked if participants would differ in their ratings of messages that attribute success to external/stable (external/stable/specific and extemal/stable/ global) causes versus those that attribute success to external/unstable (extemal/unstable/specific and extemal/unstable/global) causes. Paired-samples t-tests were used to compare external/stable to external/unstable messages on negative and positive emotion for all three situations. Means, standard deviations and effect sizes can be found in Table 11. Negative Emotion In all three situations, the mean difference between the external/stable and external/unstable messages was significant: for the academic situation, t (102) = 2.29, p = .02, for the romantic situation, t (1 03) = 8.67, p < .001, and for the internship situation, I (105) = 3.79, p < .001. In all three situations, the external/stable message was rated more highly on negative emotion than the external/unstable message. Positive Emotion In the three situations, the mean difference between the external/stable and external/unstable messages was significant. For the academic situation, I (102) = 2.06, p = .04, for the romantic situation, t (103) = 9.28, p < .001, and for the internship situation, t(105) = 3.78, p < .001. In all three situations, the external/stable message was rated more highly on positive emotion than the external/unstable message. Hypothesis 3. Hypothesis three predicted that messages that make internal/specific attributions for success would be rated more positively than messages that make internal/global attributions for success. Paired-samples t-tests were used to compare intemal/stable/specific and internal/unstable/specific messages to the 44 internal/stable/global and intemal/unstable/global messages on negative and positive emotion. Means, standard deviations and effect sizes can be found in Table 11. Negative Emotion For the academic situation, the mean difference between the internal/specific and internal/global message was not significant, t (102) = -0.93, p = .35. For the romantic situation, the mean difference between the internal/specific and the intemal/ global messages was significant, t (103) = -2.58, p = .01. For the internship situation, the mean difference between the internal/specific and internal/global messages was not significant, t (106) = 1.19, p = .23. Inspection of the means revealed that internal/global messages were rated as better in both the academic and romantic situations, though the difference was only significant in the romantic scenario. However, the internship scenario was different. The means from the internship scenario show that the internal/specific messages were liked better than the internal/global messages, however, this was not a significant difference. Thus, the hypothesis was not supported. Positive Emotion The mean difference between the internal/specific and internal/global messages was significant for all three situations: for the academic situation, t (102) = -2.79, p < .001, for the romantic situation, t (103) = -2.74, p < .001, and for the internship situation, t (106) = 2.22, p = .03. Although significant, the means indicated that internal/global messages evoked greater positive affect among the participants than internal/specific messages, thus hypothesis three is not supported. Research Question 2. Research question two was concerned with whether participants would differ in their ratings of messages that attribute success to 45 extemal/specific causes versus external/global causes. Research question two was assessed with paired samples t-tests comparing extemal/stable/specific and extemal/unstable/specific messages to the extemal/stable/ global and extemal/unstable/global messages on negative and positive emotion. Means, standard deviations and effect sizes can be found in Table 11. Negative Emotion For the academic situation, the mean difference between the extemal/specific and external/global messages was significant, t (102) = -2.69, p < .001. For the romantic situation, the mean difference between the extemal/specific and extemal/ global messages was significant, t (103) = -4.75, p < .001. For the internship situation, the mean difference between the external/specific and external/global messages was not significant, t (105) = 1.43, p = .15. Inspection of the means indicates that the external/global messages evoked less negative emotion than extemal/specific messages, though the difference was not significant in the internship situation. Positive Emotion For the academic situation, the mean difference between the extemal/specific and external/global messages was significant, t (102) = -3.71, p < .001. For the romantic situation, the mean difference between the external/specific and extemal/ global messages was significant, t (103) = -5.55, p < .001. For the internship situation, the mean difference between the external/specific and extemal/ global messages was not significant, t (105) = 1.03, p = .31. Inspection of the means indicates that the extemal/ global messages evoked more positive emotion than extemal/specific messages, though the difference was not significant in the internship situation. 46 Hypothesis 4. Hypothesis 4 predicted that self-esteem would be positively associated with the likelihood of capitalization. The correlations for the academic situation, r(101)= .34, p < .001, and the romantic situation r(102) = .35, p < .001, were significant, however, the correlation for the internship situation was not significant r(106) = .17, p < .05 (one-tailed test) . Thus, the hypothesis was supported in all situations. Hypothesis 5. Hypothesis 5 predicted that people with low self-esteem would perceive less skillful messages more positively than people with high self-esteem. Hypothesis 5 was assessed with independent samples t-tests comparing high and low self- esteem participants on ratings of positive and negative emotion for messages that make external attributions (extemal/stable/global, extemal/unstable/global, extemal/stable/specific, and extemal/unstable/specific). Negative Emotion There was no difference between high (M = 5.49, SD = 1.34) and low self-esteem (M = 5.10, SD = 1.05) participants on ratings of messages that make external attributions in the academic situation, t(101) = -1.65, p = .10, r2 = .03; nor were there differences between high (M = 5.59, SD = 1.11) and low (M = 5.28, SD = 1.22) self-esteem participants in the internship situation, t(105) = -1.34, p = .18, r2 = .02. However, there was a significant difference between high (M = 6.06, SD = .89) and low (M = 5.28, SD = 1.01) self-esteem participants on message ratings in the romantic situation, t(102) = -4.19, p < .001, r2= .15. Therefore, people with LSE had less negative affect about the low quality messages than people with high self-esteem in the romantic situation, however, this difference was not supported in the academic or the internship situation. Thus, there was only support for H5 in the romantic situation. 47 Positive Emotion There was no difference between high (M = 3.88, SD = 1.68) and low self-esteem (M = 3.85, SD = 1.16) participants on ratings of messages that make external attributions in the academic situation, t(101) = -.12, p = .91, r2 = 00; nor were there differences between high (M = 3.92 = 1.54) and low (M = 3.69, SD = 1.46) self-esteem participants in the internship situation, t(105) = -.83, p = .41, r2 = .01. However, there was a significant difference between high (M = 4.91, SD = 1.11) and low (M = 4.28, SD = 1.12) self-esteem participants on message ratings in the romantic situation, t(102) = -2.89, p < .01 , r2 = .08. Therefore, people LSE had significantly less affect toward low quality messages in the romantic scenario than people with HSE. This finding is in contradiction to H5. Therefore, the hypothesis was not supported. Hypothesis 6: Hypothesis 6 predicted that self-esteem would moderate preferences for messages that make internal attributions for success, such that LSE individuals would not differentiate between messages that make internal/stable attributions and those that make intemal/unstable attributions, but HSE individuals would rate internal/stable messages more positively than intemal/unstable messages. Hypothesis 6 was assessed with 2 x 2 repeated measures ANOVA, with self-esteem as a between- groups factor. The repeated measure was capitalization message level, and intemal/stable/global and intemal/stable/specific messages were contrasted with internal/unstable/ global and intemal/unstable/specific messages on both negative and positive emotion. Means and standard deviations can be found in Table 12. Negative Emotion 48 There was not a significant main effect for self-esteem in the academic situation (M = 5.42, SD = 0.93 for low self-esteem, M = 5.90, SD = 1.13 for high self-esteem), F (1,101) = 5.59, p =.02, partial n2 = .05. However, there was a significant main effect for self—esteem in the romantic (M =5.44, SD = 1.01 for low self-esteem, M = 6.21, SD = 0.71 for high self-esteem), F(1,102) = 19.46, p <.001, partial n2 = .16, and internship situations, (M = 5.98, SD =0.99 for low self-esteem, M = 6.37, SD = 0.61 for high self- esteem), F (1,102) = 6.12, p =.01, partial n2 = .05. People with high self-esteem rated these messages as evoking more negative emotion than people with low self-esteem. The interaction between self-esteem and message level was not significant for any situation: For the academic situation, F (1 ,101) = 1.21, p = .27, partial n2 = .01, for the romantic situation, F (1,102) = 1.34, p = .25, partial n2 = .01, and for the internship situation, F ( 1,105) = .08, p = .77, partial n2 = .00. Thus, the hypothesis was not supported. Positive Emotion There was not a significant main effect for self-esteem in the academic situation (M =4.65, SD = 0.93 for low self-esteem, M = 4.61, SD = 1.47 for high self-esteem), F (1,101) = .03, p =.85, partial n2 = .00. However, there was a significant main effect for the self-esteem in the romantic (M =4.88, SD = .94 for low self-esteem, M = 5.47, SD = 1.03 for high self-esteem), F (1,102) = 9.30, p <.003, partial n2 = .08 and internship situations, (M =5.33, SD = 1.30 for low self-esteem, M = 5.79, SD = 1.01) for high self- esteem, F (1,102) = 4.98, p =.02, partial n2 = .04. Means indicate that overall, high self- esteem participants rated messages with internal attributions for success better than did low self-esteem participants. 49 The interaction between self-esteem and message level was not significant for any situation: For the academic situation, F(1,101) = .35, p = .55, partial n2 = .00, for the romantic situation, F (1,102) = .05, p = .83, partial n2 = .00, and for the internship situation, F (1 ,105) = 1.93, p = .17, partial n2 = .02. Thus, hypothesis 6 was not supported. Table 11 Means and Stand Deviations for Positive and Negative Emotion for H6 Negative Positive Emotion Emotion Low Self- High Self- Low Self- High Self- Esteem Esteem Esteem Esteem M SD M SD M SD M SD Academic lntemal/Stable Composite 5.90 1.06 6.26 1.08 5.45 1.14 5.34 1.52 lntemal/Unstable Composite 4.95 1.07 5.54 1.43 3.86 1.06 3.89 1.67 Internship lntemal/Stable Composite 5.98 1.1 1 6.62 0.70 5.63 1.01 6.25 0.82 lntemal/Unstable Composite 4.91 1.26 4.13 1.29 5.82 1.00 4.69 1.62 Romantic lntemal/Stable Composite 6.27 0.97 6.63 0.50 5.78 1.36 6.39 0.58 lntemal/Unstable Composite 5.70 1.18 6.12 0.94 4.88 1.42 5.20 1.32 Discussion Gable (2004; 2006) asserts that the quality of responses to capitalization people receive influences the quality of their relationships. However, Gable et al. (2004; 2006) failed to examine the features of high and low quality responses to capitalization— that is, what are the components of a skillful capitalization response message, and what are the components of a less skillful capitalization response message? This thesis expanded on Gable et al’s research (2004; 2006) by focusing on content and forms of the capitalization message. This thesis also expanded on Samter’s (1989) coding scheme of celebratory ego 50 support messages, by testing capitalization messages derived from attribution theory. Finally, this thesis examined if self-esteem influenced the likelihood of capitalization or moderated liking for capitalization messages differing in levels of Skillfulness, as suggested by previous research (Gable et al., 2004; Gable, et al., 2006). Many predictions were supported by the findings of the study, though a few received little or no support are discussed. In the following sections, the results of the study in terms of their implications for capitalization research, attribution theory, self-enhancement theory, and self- verification theory. Attribution Theory As predicted, messages that made internal attributions for success were rated significantly better than messages that made external attributions for success. This is consistent with previous research that elucidates the benefits of making internal rather than external attributions in celebratory scenarios (Samter, 1989; Peterson & Seligman; 1984, Weiner, 1985). Samter’s (1989) coding scheme for celebratory ego support messages, which was developed post hoc by examining messages generated by participants, places the greatest emphasis on the locus dimension of capitalization messages, because it indicates whether or not the target is responsible for the positive event (Abramson et al., 1978). Research indicates that there are benefits to making internal rather than external attributions for success, which include helping a person to feel more happiness or joy because of an event (Weiner, 1985). This thesis had congruent findings. Participants indicated that messages that made internal attributions for success led to greater positive affect and less negative affect than messages that made external attributions for success. 51 The results of hypothesis two were also consistent with previous research that asserts that messages that make internal/stable attributions for success will be rated more positively than messages that make internal/unstable attributions for success (Samter, 1989; Peterson & Seligman; 1984; Weiner, 1985). Peterson and Seligman (1984) emphasize stability as the second most important component of a skillful attribution people make about their own successes, after locus, because attributing the success to a stable cause leads the target to believe the event can occur again. Thus, that people rate messages which make internal/stable attributions for success as more helpful than those that make internal/unstable attributions may be because such messages suggest not only that the success is of their own doing (internal attribution), but also that they also can expect the success to happen again (Weiner, 1985). The results of this thesis indicate that the attributions people make for others’ successes can have similar results for targets as those attributions targets make for their own successes. Although previous research had not examined variation within messages that make external attributions for success, this study found that messages that attribute success to external/stable causes were rated as better than those that attribute success to external/unstable causes. Extemal/stable messages and external/unstable messages were ranked in a manner similar to internal/stable and internal/unstable messages. That is, participants liked messages that made stable attributions better than those that made unstable attributions, despite the locus of the message. Thus, regardless of whether the attribution suggested that the participant caused the event, he or she still liked messages that indicated the event may happen again. However, it is important to note that the locus dimension appears to be more important than the stability dimension, since messages that 52 made internal attributions for the success were rated better than messages that made external attributions. This thesis supports other previous research that found the locus dimension to be the most influential to how a person feels. This research (Peterson & Seligman, 1984; Weiner, 1985) has shown that internal attributions for negative events lead to negative emotion; however, this research shows that internal attributions for positive events leads to positive emotion. Therefore, in line with Peterson and Seligman (1984), when successes are seen as the result of internal, stable characteristics, a person has a better outlook on life and is happy because they can expect the result again. Hypothesis 3 predicted that participants would like internal/specific messages better than intemal/ global messages; however, this hypothesis was not supported. This hypothesis was derived from Samter’s (1989) coding scheme for celebratory ego support messages. Based on her coding, Samter (1989) asserted that global messages are less tailored to the recipient than specific messages; therefore, they are lower in quality. For example, the message may allude to a global characteristic (e.g. “you really deserved the internship”) or the message may elaborate on very specific characteristics of the target (e.g. “you really deserve this because you worked so hard. . .you’re organized and smart”) However, the participants in the current study reported that global messages produced greater positive affect than specific messages, and that specific messages increased negative affect. This may be because when the attribution is global, the positive event can be seen as constant across situations (Abramson etal., 1978), indicating that the participant may expect success in other areas of his or her life as well. Specific 53 messages may imply that the target is not good at other things outside of the specific situation, which may indicate a shortcoming in other areas. It is also possible that this hypothesis was not supported because Samter’s coding scheme was created post hoc, using messages generated by participants in response to a hypothetical scenario. This scheme was not empirically tested after development. Thus, Sarnter’s assumptions about the quality of messages that made specific attributions may have been based on the nature of her specific data. For example, Samter had participants generate messages as if they were speaking with a close, same-sex friend. Although the current study specified that the messages were coming from a friend, it is possible that Samter’s participants generated messages for people they are particularly close with. Specific messages may indicate that the sender and the receiver have a particularly close relationship that allows the sender to develop a message highlighting specific attributes of the target. If the participants from the current study did not consider the participants to be in a particularly close relationship, the importance of making a specific attribution may have been diminished. The same pattern was found in research question two: participants rated external/global messages to be better than extemal/specific messages. Even though the external messages attribute the success to a force outside of the target, global messages indicate the target will experience a broad range of success rather than a narrow range of success. Again, it is important to note that the locus dimension appears to be more important than the globality dimension, as participants rated messages that made internal attributions better than messages that made external attributions, regardless of globality. Self-Enhancement vs. Self- Verification Theory: Likelihood of Capitalization 54 Hypothesis four predicted that people with higher self-esteem would be significantly more likely to share a positive event with others than those with lower self- esteem. This hypothesis was supported: self-esteem was significantly, positively associated with capitalization in the romantic, academic, and internship situations. Theoretically, the findings in these scenarios support the notion that people like to receive messages that are congruent to their self-esteem (Swarm, 1990). There are a variety of reasons that people may capitalize differently based on self-esteem. Swann (1990) noted that one of the reasons a person with LSE may be more reluctant to share about a positive event is out of fear for the event’s longevity. Because individuals with LSE may be more worried about the event not lasting than individuals with HSE, LSE individuals may be less likely to capitalize, as well as become more uncomfortable with messages that attribute the success to a stable cause. Another reason low self-esteem individuals are less likely to capitalize is that in doing so, they experience dissonance between what they think about themselves and what others think about them. Sharing a success would help others form a positive opinion about the individual that the individual would not agree with, an opinion that would be disjointed with his or her self-views. In order to avoid this dissonance, a person with LSE should be less likely to capitalize about the event. However, this is not to indicate that people with LSE never capitalize. In fact, the average rating for likelihood of capitalization across situations for LSE individuals (M = 5.91, SD = 1.09) was well above the midpoint on the 7-point scale, and not much below that for HSE individuals (M = 6.32, SD = .77). Thus, LSE people are still likely to capitalize, just not to the extent that HSE individuals do. 55 Throughout this study, the internship situation consistently arose as being different than the romantic and the academic scenarios. There are a variety of reasons that can explain the difference with the internship situation. Students do not commonly like to share their grades with others, even when one does well. Those who do openly share their academic performance are even looked down upon. Therefore, participants may feel they are breaching social norms of bragging if they capitalize on an academic situation. However, the internship is an individual success and it does not breach etiquette to tell fiiends and family about a new position. Because HSE individuals are more confident in themselves and their success, they may be more comfortable capitalizing on receiving a good grade; thus, there is a significant difference. Furthermore, regardless of their own self-esteem, people may need to inform their friends and family about having received an internship, because an internship will change their schedule for a period of time. Though the same may be said for the romantic scenario, the romantic relationship is inherently dyadic. Therefore, the low self-esteem individual may rely on his or her relational partner to perform the capitalization act (i.e., telling friends and family, having an engagement party). The internship scenario is different; applying for and receiving an internship is an independent act, and thus the target may be more likely to capitalize for themselves rather than having a third party perform the capitalization acts. Another reason for the unique findings in the internship scenario is associated with the participant pool that was used in the study. Because the study was conducted at the end of the winter semester, many of the participants may have been actively looking for internships at the time of completing the survey. Due to the economy, participants 56 may have been having trouble finding internships of their own; therefore, this scenario was seen as particularly exciting. LSE people may be more likely to share because everyone is having a difficult time finding an internship therefore it is not necessarily a reflect of their abilities. Additionally, the current economy has made receiving an internship an event that is no longer strictly driven by internal causes; that is, even well- qualified candidates are having trouble finding paid work. There are fewer organizations currently willing to take on interns, especially paid interns, during tough economic times (Meglio, 2010). Students may see their peers receiving internships because of networking or through the efforts of parents and family members, not because of their unique talents. Therefore, receiving an internship may truly be seen as a function of being in the right place at the right time, not because the recipient is a superior candidate. Although self-esteem was associated with likelihood of capitalization in the academic and romantic scenarios, ratings of capitalization messages generally did not vary much based on self-esteem. The data do show that although there were no interactions between ratings of messages varying in quality and self-esteem, there was a main effect for self-esteem on message ratings in several cases. Overall, people with HSE rated both internal/stable and internal/unstable messages better than people with LSE. This may partly have to do with the LSE person’s self-confidence. People with HSE, being confident with themselves, are excited to share the event with others, and may see any type of message more positively than an LSE individual, even when that message is lower in quality (i.e., internal/unstable). On the other hand, people with LSE may be more wary of any capitalization message, because they are not confident they will live up to the standards the new success has set. 57 Previous research (Swarm, 1990; Brown, et al., 1988) contends that people with HSE will prefer to receive more skillful messages because they already think better about themselves. Thus, hypothesis five predicted that people with LSE would perceive less skillful messages more positively than people with HSE. Hypothesis six predicted that LSE individuals would not differentiate between messages varying in quality, whereas HSE individuals would. Neither hypothesis was supported. Thus, although people with LSE did not capitalize as frequently, when capitalization does occur, they prefer high quality messages in the same way people with HSE do. Consistent with these findings, and in support of SET, Swarm (in press) found that everyone, despite self-esteem, wants to be seen in a positive light. Although people with LSE do not seek others out to share the event as frequently, once the event is shared, LSE individuals prefer messages that enhance their self-view, and do not perceive less skillful messages more positively than people with HSE. In fact, it may be more important for an individual with LSE to receive a high quality capitalization response than people with HSE, because self-enhancement is more valuable to them. People with HSE already feel good about themselves; however, quality capitalization messages may be necessary for people with LSE to self—enhance and even reinforce the benefits of sharing positive news with others (Brown, et al., 1988). With respect to Hypothesis 5, interesting findings emerged in the romantic situation. Although not significant in the academic or internship situations, the means indicate that participants with HSE associate greater negative emotion with low quality messages (i.e., messages that make external attributions) than participants with LSE. Though not significant, the means also indicate that in the academic and internship 58 situations, HSE participants associate greater negative emotion with low quality messages than do participants with LSE. Interestingly, people with HSE also reported greater positive emotion in regards to the same messages. This may indicate a pattern of people with HSE rating messages as evoking stronger affect in general - affect which is not necessarily valence specific. The romantic situation required participants to place themselves in a situation where they have recently become engaged. Although each of the hypothetical situations is a segment of a dyadic relationship (i.e., a recruiter gave the candidate an internship, the professor awarded the student with a good grade), a romantic relationship is inherently different. Romantic relationships require a greater sense of co-dependency within the dyad than do the internship or academic scenarios. Therefore, attributing the engagement to an external cause may be seen as particularly upsetting because the message may appear to be insulting to the recipient as well as the person they love. Therefore, people with both HSE and LSE report negative emotion associated with the low quality messages, though HSE individuals report less negative emotion. People with LSE may view the low quality messages as reinforcing of thoughts and attitudes they already possess. However, people with HSE also report more positive emotion is associated with these messages than people with LSE. People with HSE may be reading into the positive elements of the message and feeling good about specific aspects. For example, the extemal/unstable/specific message (e.g. “Your fiancé must be in a loving mood lately. Do you think it will last?”) evoked both positive and negative emotions from people with HSE. The positive emotions can possibly be traced back to the part about their loved one 59 being in a loving mood. However, people with HSE also see the instability insinuated by the comment (e.g. “Do you think it will last?”) as negative. Additionally, the external messages in the romantic situation did not carry as negative a connotation as the external messages in the other situations; therefore, the low quality messages may have been seen as better than the other scenarios. For example, the romantic extemal/stable/specific message was, “It’s just fate for you guys to get married” is less negative than the other scenarios, “Tests in that class are always easy and always will be.” Although fate is an external attribution, many people believe that fate is heavily involved in a successful romance, whereas fate may not determine the grade an individual receives on a test or whether he or she receives a summer internship. Limitations One limitation of this study is that it included only successful events that were portrayed as having internal causes. There are many other positive events people experience every day that are not due to personal attributes, such as finding money on the sidewalk. Additional research is needed to understand how people respond to these positive events. Also, little is known about how culture may moderate capitalization processes. For example, it is possible that positive life events are responded to differently in collectivistic cultures than in individualistic cultures. Personal expression is generally considered to be specific to individualistic cultures (Rime, 2007). Collectivist cultures do not generally encourage people to talk about themselves and their successes and doing so may be seen as irrelevant and inappropriate (Singh-Manoux & F inkenauer, 2001). Therefore, capitalization may be less likely and may have fewer positive effects for recipients. 60 Another limitation of this study is that it involved hypothetical situations. Messages may have a larger impact when the participant is an actual recipient of the message. However, that significant effects were found for hypothetical situations suggests that the effects in real life may be even larger. One additional limitation to the study’s design was the order participants received messages. Each participant rated messages in the same order, starting with intemal/stable/specific, which was predicted to be the highest quality message, to extemal/unstable/global, which was predicted to be the lowest quality message. Therefore, the results could show some ordering effect as well as fatigue from the participants. Finally, it may be argued that a limitation of the study is the number of t-tests that were conducted. Convention dictates that as more significance tests are conducted, the likelihood of Type I error occurring becomes inflated (O’Keefe, 2003). However, this belief has been challenged. O’Keefe (2003) argues that by adjusting the alpha level to accommodate for significance testing, researchers are needlessly reducing statistical power. Additionally, researchers are not required to adjust the alpha level for other statistical methods, such as ANOVA and regression; therefore, it is unnecessary to do so for t-tests. Thus, we do not view this as a limitation to the current study. Future Directions All messages were constructed by the researcher for this study. Therefore, there is likely a range of other message types that were not included here, and that may be rated differently by participants varying in self-esteem. Another option for exploring various types of capitalization messages is through a retrospective study. Participants could report 61 on sharing a positive event in the past and the messages they received in response to that disclosure. It is also necessary to study capitalization messages in a variety of contexts. As was found in this study, in some cases the internship scenario did not produce the same results as did the academic and romantic scenario, and in others the romantic situation produced different results. Therefore, it may be fruitfirl to test messages that make similar attributions with additional success scenarios such as sports, landing a dream job, or buying a car. Another way to expand on this research is by examining capitalization messages in situ. For example, participants could come to the lab and respond to capitalization attempts. A study could be done with pairs of friends or significant others, or with participants and confederates. Upon arrival, one member of the team could be given good news such as “you just won our $100 prize!” Following the delivery of the good news, the capitalization process could be studied. Conclusion Understanding good responses to capitalization messages has a variety of benefits. Skillful capitalization response messages help people to experience greater positive affect surrounding the positive event and reduce possible negative affect. It is important to help people experience positive emotion because it can positively affect a person’s physical health as well as their mental health (Brown & McGill, 1989; Burton & King, 2003; Danner, Snowdon & F riesen, 2001; Fredrickson, Tugade, Waugh & Larkin, 2003; Lyubomirsky, Sousa & Dickerhoof, 2006). This study’s findings were similar to 62 Langston (1994) and Gable et al’s (2004; 2006) in that skillful capitalization response messages helped people experience greater positive affect and less negative affect. This study provides information on what more and less skillful capitalization response messages look like, allowing us to make recommendations to people who would like to provide such messages. This study shows that intemal/stable/global messages are viewed as being the highest quality capitalization response messages. As predicted, messages that make internal attributions are rated as better than those that make external attributions, and messages that make stable attributions are rated as better than those that make unstable attributions. Contrary to prediction, messages that make global attributions are rated as being more positive than messages that make specific attributions. Although everyone (regardless of level of self-esteem) is able to differentiate between high and low quality capitalization messages, it is important to be considerate of someone else’s self-esteem. A friend or acquaintance may not capitalize as often as a person with HSE, but they still experience more positive affect and less negative affect after receiving skillful messages. Thus, it is important for people with LSE to receive high quality capitalization response messages, even if they do not solicit them. Because capitalization has been relatively unexplored, future research on the sharing of positive life events is necessary. This study helps to fill a significant void in the literature on capitalization research; specifically, it looks at the character of capitalization messages. Previous research has demonstrated that self-esteem may play a significant role in capitalization processes, but prior to this study there was no research on the interaction between self-esteem and ratings of capitalization messages. Although there is still a great deal of research that needs to be done in order to fully understand the 63 capitalization process, this thesis provides a solid launching point for research examining the features of capitalization messages and addressing the relationship between self- esteem and capitalization. 64 Appendix A Appendix: Capitalization Situations and Messages Academic Success Situation: Imagine that you are very excited about next semester because you will finally be able to officially enter your chosen major. This major is very important to you, and you have dreamed about getting a degree in this field for a long time. All that is required of you now is one more B in an important pre-major course, and you have worked to the best of your abilities in that course. After the final, you check your grade for the semester and find out that you got an A in the class. You feel great about yourself. In fact, you feel very smart at this point. Messages: Internal/stable/ specific: “That’s why it’s your major! You’re always so smart in your major!” lntemal / stable/ global: “You always do great because you are so good at everything!” Intemal/ unstable / global: “You are doing great at everything lately! I hope it lasts because not everything can go right all the time like it has.” lntemal / unstable/ specific: “You worked really hard on that one exam. It seems too much to hope for your other exams to go that well.” Neutral: “Do you want to go out tonight?” External / stable/ specific: “Tests in that class are always easy and always will be.” External / stable/ global: “Luck is always with you with everything that you do.” Extemal/ unstable / specific: “I’m surprised — tests in that class are usually tough but you must have gotten an easy one. I bet the next one will be hard.” External / unstable/ global: “Wow- luck has really been with you all week with everything you have been doing. Wish this happened more often for you!” Romantic Success Situation: Imagine that you have been dating someone seriously for three years. Your relationship is going well and you are sure this person is “the one.” One night, the two of you get engaged. You feel great about yourself and feel very excited about the future. Messages: Intemal/stable/ specific: “Wow! You have always been a really great girlfriend/boyfiiend to him/her.” lntemal / stable/ global: “That's great! You are always good at everything - even relationships” Intemal/ unstable / global: “You've been great at everything you do lately. I hope it lasts!” 65 lntemal / unstable/ specific: “That's great! I bet it's the way you've been acting because you have been really sweet to him/her the past few weeks. I hope you can keep it up!” Neutral: “Do you want to go out tonight?” External / stable/ specific: “It's just fate for you guys to get married” External / stable/ global: “Luck has really been with you and everything always works out. Plus, it stays good, too.” Extemal/ unstable / specific: “Wow! Your Fiance’ must be in a loving mood lately. Do you think it will last?” External / unstable/global: “That's great. Didn't she/he just get a new job, too? She/he has been generally great with everything lately! I hope all this success lasts!” Internship Success Situation: Imagine that you are applying for a summer internship. The internship is a perfect fit for your career goals, and you hope to land a job with the company after graduation. You complete your first interview with the company and you receive a call to come in for a second interview. You learn that the company is extremely pleased with you and they would like you to work for them the next summer. You feel very excited about the opportunity to work in the field. Messages: Intemal/stable/ specific: “You always do great in interviews because you're always so smart and win over the interviewers!” lntemal / stable/ global: “You always get what you try for because you are so good at everything you do!” Internal/ unstable / global: “You've worked hard lately and it's paying off in every area of your life. I hope it lasts because not everything can go right all the time.” lntemal / unstable/ specific: “You really found the right interviewing strategy this time. Do you think you can do it again?” Neutral: “Did you call your parents?” External / stable/ specific: “Interviews have been easy everywhere for the past few years. I bet it's going to stay like this for a long time.” External / stable/ global: “Every company has been having easy interviews for a long time now; not just the place you got the internship. I bet it will stay like this for a long time, too.” Extemal/ unstable / specific: “I wonder if there weren't many people who applied to that particular job. I bet that will change once graduation happens. ” External / unstable/ global: “Everything you do has been easy for you lately. You've been getting everything you go for. Too bad luck like this doesn't usually last.” 66 Appendix B Appendix: Manipulation Check Items Importance How important do you think it is to do well in your classes/ become engaged and get married / find an internship? Likelihood of Capitalization How likely are you to tell someone about your grade / becoming engaged / getting the internship? Would you share your grade / becoming engaged / getting the internship with family and fiiends? Would you brag about your grade / becoming engaged / getting the internship? Would you celebrate getting an A / becoming engaged / getting the internship? Success Reliabilitv Check How upsetting would it be if you had failed? How disappointed would you be if you had not succeeded? Do you feel receiving an A was a success/ becoming engaged / getting the internship? Stabilitv Reliability Check This comment would lead me to think my success is due to something that will last. This comment would lead me to think my success is due to something temporary. This comment would lead me to think my success is due to a consistent factor. S eci zci Reliabili Check This comment would lead me to think I am successful in other situations. This comment would lead me to think my success is due to something that helps me in other situations. This comment would lead me to think my success is specific to academics/relationships/internships. Locus Reliabilitv Check This comment would lead me to think my success is because of something I did. This comment would lead me to think my success is due to something other than me. This comment would lead me to think my success is due to me. 67 Appendix C Appendix: Sample Pre-test Directions Support Messages Survey This questionnaire asks you to provide your opinion about several matters related to yourself and communication, especially supportive communication you have received when experiencing a successful event in your life. There are no right or wrong answers to any of the questions. We are simply trying to learn what you think about these matters. Please answer every question as honestly as possible. It should take no more than 20 minutes to complete the questionnaire. Please remember than your participation is completely voluntary and you may choose to stop participating at any time. Instructions: The next section of this questionnaire is designed to help us learn about how people feel in response to receiving particular messages after sharing a positive event. Please imagine yourself in the following hypothetical situation. Then, read through this questionnaire carefully and answer questions as honestly as possible. Please indicate your answers on the provided scantron. Rate these items on the following scale by marking the appropriate circle on your scantron. Please begin with question scantronl, form A. After question 150, you will proceed to form B, the second scantron. PARTICIPANTS COMPLETE MANIPULATION CHECKS REGARDING HYPOTHETICAL SCENARIO Imagine that after you found out about getting your grade, you run into a same-sex friend. You tell your friend about your engagement, and he/she responds with a message. Egg rate all of the messages on the following pages based on the scen_ario. You will respond to_a total of we messages. Rate these items on the following scale by marking the appropriate circle on your scantron. PARTICIPANTS COMPLETE MANIPULATION CHECKS REGARDING THE ATTRIBUTION DIMENSIONS 68 Appendix D Appendix: Items assessing Dimensions Locus Reliability Check Items '1. This comment would lead me to think my success is because of something I did. 2. This comment would lead me to think my success is due to something other than me. 3. This comment would lead me to think that my success is due to me. Stability Reliability Check Items 1. This comment would lead me to think I am successful in situations other than academics, too. 2. This comment would lead me to think my success is due to something that helps me in situations other than academics. 3. This comment would lead me to think my success is specific to academics. Globality Reliability Check Items 4. This comment would lead me to think my success is due to something that will last. 5. This comment would lead me to think my success is due to something temporary. 6. This comment would lead me to think my success is due to a consistent factor. 69 Appendix E Appendix: Messages Altered in Pretest 2 Internship Scenario Internal/ Stable / Specific Original: “You always do great in interviews because you always really research the company and it shows.” New: “You always do great in interviews because you’re always so smart and win over the interviewers!” Internal / Stable / Global Original: “You always do great in interviews because you’re good at everything!” New: “You always get what you try for because you are so good at everything you do!” Internal/ Unstable / Global New: “You’ve worked hard lately and it’s paying off in every area of your life. I hope it lasts because not everything can go right all the time. External / Stable / Global New: “Every company has been having easy interviews for a long time now; not just the place you got the internship. I bet it will stay like this for a long time, too.” External / Unstable / Global New: “Everything you do has been easy for you lately. You’ve been getting everything you go for which is totally different from last year. Too bad luck like this doesn’t usually last!” Romantic Scenario Internal / Stable / Global Original: “That’s great! You’re good at everything -— even relationships!” New: “That’s great! You are always good at everything — even relationships!” External / Stable / Global Original: “Everything always works out for you and stays good, too.” New: “Luck has really been with you and everything always works. Plus, it stays good, too.” 70 Appendix F Appendix: PANAS This response would lead me to feel.... Very Guilty 1 2 3 4 5 6 7 Not At All Guilty Very Ashamed 1 2 3 4 5 6 7 Not At All Ashamed Very Happy 1 2 3 4 5 6 7 Not At All Happy Very Cheerful 1 2 3 4 5 6 7 Not At All Cheerful Very Sad 1 2 3 4 5 6 7 Not At All Sad Very Proud 1 2 3 4 5 6 7 Not At All Proud Very Depressed 1 2 3 4 5 6 7 Not At All Depressed Very Confident 1 2 3 4 5 6 7 Not At All Confident Very Distressed 1 2 3 4 5 6 7 Not At All Distressed 71 Appendix G Appendix: Demographic Items General Instructions: Please answer the following questions about yourself. 1. What is your grade level? A = Freshman B = Sophomore C = Junior D = Senior E = Graduate 2. What is your sex? A = Male B - Female 3. What is your major? 4. What is your ethnicity? A = African American B = Asian American / Pacific Rim C = Hispanic American D = European American (white/Caucasian E = Other 72 Appendix H Appendix: Rosenberg Self-Esteem Scale 1.0n the whole, I am satisfied with myself. 2. At times I think I am no good at all.* 3.1 feel that I have a number of good qualities. 4.1 am able to do things as well as most other people. 5.1 feel I do not have much to be proud of.* 6. I certainly feel useless at times.* 7. I feel that I‘m a person of worth, at least on an equal plane with others. 8. I wish I could have more respect for myself.* 9. All in all, I am inclined to feel that I am a failure.* 10. I take a positive attitude toward myself. *Items are reverse scored. 73 References Abramson, L. V., Seligman, M. E. P., & Teasdale, J. D. (1978). Learned helplessness in humans: Critique and reformulation. Journal of Abnormal Psychology, 87, 49-74. Brown, J. D., Collins, R. L., & Schmidt, G. W. (1988). Self-esteem and direct versus indirect forms of self-enhancement. Journal of Personality and Social Psychology, 55, 445 — 453. Brown, J. D., & McGill, K. L. (1989). The cost of good fortune: When positive life events produce negative health consequences. Journal of Personality and Social Psychology, 5 7, 1 103-1110. 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: 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. (2003). Producing emotional(al) messages. Communication Theory, 10, 221-250. Burton, C. M., & King, L. A. (2004). The health benefits of writing about intensely positive experiences. Journal of Research in Personality, 38, 150-163. Campbell, J. D. (1986). Similarity and uniqueness: The effects of attribute type, relevance, and individual difference in self-esteem and depression. Journal of Personality and Social Psychology, 5 0, 281-294. Danner, D. D., Snowdon, D. A., & Friesen, W. V. (2001). Positive emotions in early life and longevity: Findings from the nun study. Journal of Personality and Social Psychology, 80, 804-813. 74 Dillard, J. P. & Peck, E. (2001). Persuasion and the structure of affect: Dual systems and discrete emotions as complementary models. Human Communication Research, 27, 38-68. Dillard, J. P., & Meijnders, A. (2002). Persuasion and the structure of affect. In J .P. Dillard & M.W. Pfau (Eds.), The persuasion handbook: Developments in theory and practice (pp. 309 — 328). Thousand Oaks, CA: Sage. F aul, F., Erdfelder, E., Buchner, A., & Lang, A.-G. (2009). Statistical power analyses using G*Power 3.1: Tests for correlation and regression analyses. Behavior Research Methods, 41, 1149-1160. Fredrickson, B. L., Tugade, M. M., Waugh, C. E., & Larkin, G. R. (2003). What good are positive emotions in a crises? A prospective study of resilience and emotions following the terrorist attacks on the United States on September 11'“, 2001. Journal of Personality and Social Psychology, 84, 365-3 76. Gable, S. L., Reis, H.T., Impett, E.A., & Asher, ER. (2004). What do you do when things go right? The intrapersonal and interpersonal benefits of sharing positive events. Journal of Personality and Social Psychology, 87, 228-245. Gable, S. L., Gonzaga, G. C., & Strachman, A. (2006). Will you be there for me when things go right? Supportive responses to positive event disclosures. Journal of Personality and Social Psychology, 91, 904-917. Gecas, V. (1982). The self-concept. Annual Review of Sociology, 8, 1-33. Lazarus, R. S. (1991). Emotion and adaptation. New York: Oxford University Press. Holmstrom, A. J. (2008). The development and assessment of a theory of esteem support messages. Unpublished doctoral dissertation. Purdue University, West Lafayette. Katz, J ., Anderson, R, & Beach, S. R. H. (1997). Dating relationship quality: Effects of global self-verification and self-enhancement. Journal of Personality and Social Psychology, 14, 829-842. 75 Langston, C. A. (1994). Capitalizing on and coping with daily-life events: Expressive responses to positive events. Journal of Personality and Social Psychology, 6 7, 1112-1 125. Lyubomirsky, S., Sousa, L., & Dickerhoof, R. (2006). The costs and benefits of writing, talking, and thinking about life’s triumphs and defeats. Journal of Personality and Social Psychology, 90, 692-708. Meglio, F. D. (2010, July 22). Houston: A Mecca for college graduates. Bloomberg Businessweek. http://businessweek.com Morling, B., & Epstein, S. (1997). Compromises produced by the dialectic between self- verification and self-enhancement. Journal of Personality and Social Psychology, 73, 1268 - 1283. Nabi, R. L. (2010). The case for emphasizing discrete emotion in communication research. Communication Research, 77, 153-159. Nabi, R. (2002). Discrete emotions and persuasion. In J. Dillard & M. W. Pfau (Eds.), The Persuasion Handbook: Developments in Theory and Practice (pp. 289-308). New York, NY; Sage. O’Keefe, D. J. (2003). Searching for a defensible application of alpha-adjustrnent tools. Human Communication Research, 29, 464-468. Planalp, S. (1999). Communicating emotion: social, moral, and cultural processes. New York: Cambridge University Press. Planalp, S. (2003). The unacknowledged role of emotion in theories of close relationships: How do theories feel? Communication Theory, 13, 78-99. Peterson, C. & Seligman, M. E. (1984). Causal explanations as a risk factor for depression: Theory and evidence. Psychological Review, 91, 347-374. 76 Rammstedt, B. & John, O. P. (2007). Measuring personality in one minute or less: A 10- item short version of the Big Five Inventory in English and German. Journal of Research in Personality, 41, 203-212. Rime, B. (2007). The social sharing of emotion as an interface between individual and collective processes in the construction of emotional climates. Journal of Social Issues, 63, 307-322. Rosenberg, M. (1965). Society and the adolescent self-image. Princeton, NJ: Princeton University Press. Rusbult, C. E., Zembrodt, I. M., & Gunn, L. K. (1982). Exit, voice, loyalty, and neglect: Responses to dissatisfaction in romantic involvements. Journal of Personality and Social Psychology, 43, 1230- 1242. Samter, W. (1989). Communication skills predictive of interpersonal acceptance in a group living situation. Unpublished doctoral dissertation. Purdue University. Singh-Manoux, A., & F inkenauer, C. (2001). Cultural variations in social sharing of emotions: An intercultural perspective on a universal phenomenon. Journal of Cross-Cultural Psychology, 32, 647-661. Smith, C. A., Haynes, K. N., Lazarus, R. S., & Pope, L. K. (1993). In search of the “hot” cognitions: Attributions, appraisals, and their relation to emotion. Journal of Personality and Social Psychology, 65, 916-929. Swann, W. B. (in press). Self-verification theory. In P. Van Lang, A. Kruglanski, & E. T. Higgins (Eds.), Handbook of theories of social psychology. London: Sage. Swann Jr. W. B. (1987). Identity negotiation: Where two roads meet. Journal of Personality and Social Psychology, 53, 1038-1051. Swarm Jr, W. B., Pelham, B. W., & Krull, D. S. (1989). Agreeable fancy or disagreeable truth? Reconciling self-enhancement and self-verification. Journal of Personality and Social Psychology, 5 7, 782-791. 77 Swarm Jr, W. B., Rentfrow, P. J ., & Guinn, J. S. (2003). Self-verification: The search for coherence. 1n Leary, M. R. & Tangney, J. P. (Eds.), Handbook of Self and Identity (pp. 367-383). New York: Guilford Press. Swarm Jr, W. B., Wenzlaff, R. M., Krull, D. S., & Pelham, B. W. (1992). Allure of negative feedback: Self-verification strivings among depressed persons. Journal of Abnormal Psychology, 101, 293-306. Tomaka, J ., Thompson, S., & Palacios, R. (2006). The relation of social isolation, loneliness and social support to disease outcomes among the elderly. Journal of Aging and Health, 18, 359—384. Watson, D., Clark, L. A., & Tellegen, A. (1988). Development and validation of brief measures of positive and negative affect: The PAN AS scales. Journal of Personality and Social Psychology, 54, 1063-1070. 78 "7111111111111 1113111111 llllflllllfllllllll“