, . ‘ fl — _, - fl--- - a“ _, x - .. .'-..-:..P:1 s . d C .. .. -._— “f'm-‘-:‘vwv ——.— ‘3 .1} ,. ammm.4n~p-<~ ' . .- . - . ‘n _ ‘ it? ”5“.\..”? ‘ .‘o. 2.“;le I - . - . .' .. V -- do, P‘. . Q .430“ ”‘1! “.4 I . -‘ . h. .' J." V . ‘ " r .«I‘- _ o (‘6‘ 7_ :.‘.:.§:’ . . A. _ fl) .‘3‘ “9“" _ "‘l‘ '.:. 3'.” ‘ -3\ .‘£ -M l a 3.- a. .0.» ‘ a -v . '1. »o-.-: *‘fit;" 3‘: 1‘. y. 4 . .‘t- ‘~ ~’- ’a‘ x- - Won-J49?" ‘ "" '3'.” J1 ‘ ”‘ ‘.. .i";‘-.‘: ;. '93 .‘ N ' ‘j‘ !. . o‘ . .. _ _ . n . .- .-., o - . z. D _ _ :51 ‘_ €“&~.1~.‘\.9_- gl- ~-‘ 0 3;“ ‘7' '6 I'.'.' f. . 33.,»fi1'o, .5. .t I. 3 w. '9 Or ' '_' ‘1‘" ‘3‘ o , ' - > . . .. \- s- 4-. -‘\.. } ‘9 ,. “a .,.p<. .04 o .1 -_ . .u u . . ‘ - 5“‘~‘ . s \. ‘3‘.“rv ' -d‘ - -\ _, '5‘ 5- a . ' ’ 'u‘ "‘ ”.5” ..'. 9-3". ‘..'_‘ 3nta.k_-- .b‘& I ’ 0' ‘ ." c I <. I - o ‘ A“ 7“ J n f “I ' “C ‘1'. ~5491.”. a". ‘1- . ; .-\ 6.\V.’!~- 0 . ' __. ac - ‘0 . .0100 ‘\ 5,.6' .u J's . (”,3 ..,...¢“Q. .(“-.-. 51- ~¢> ~-.. . tn.-. ' -V\ D ."o- — ‘ I . - .0 0-1.. "9 ., . < -\V ' ' 'fi' o’wl‘v‘ " " “"' [7' 't l“: ' 0' .5 ’ ”t l:- v a ‘f' 00‘ ‘0 \ N - ‘ '- . ‘ ‘ ' °“'£ ‘ . ... “v , ‘“,,‘}, ,. .— o . .‘ V. ' us In. 1 g . ." t ‘ ’zUy-‘ "‘ 4' L ‘h‘:\ V :.nfl,“ 3;.‘|;O "L‘. : .. - . . _ . . _ . , - 2,2»; - w; a “:vvw— ': =,“,".:. .1: z. . f, ‘ '"* " "' A... o w, ‘. .Q'. 1.-.! bu? 0 1.,“ ‘ 4' 3;. U .4. ._ '4 -d ‘ ~ :5} ‘W., , ._,, ,, _. go ., -.n~ ,_ ‘ ‘9 a (‘ . ~ ‘ - . y p 0“. Cl » J. . 4- 4:1 ( . . _1 .-_ q- _. ”4.in ..~ 3. A. -- , - . - _, - - u .r n:- -. o v; . I .~ v =-" ‘ '~ ‘ ~ -.......’r- '3, 2:332 ..- T‘- A- . . , .- . 0 5 . - « 4. ~- -~ ':‘.'1~',:'~~., '3': we #...'.‘ r, 8-: , ., ' - 'c - -. g -. .u- c "-" . r.. ,.‘ o. t ‘0 .. . a; - »" 3&6 ,. ’4 .._. f.- c .‘ z,- t‘ .0 I d. .b‘ .9 p -g‘ 5 I'. Q . J .A 3': ,‘ O C , U . . J . ‘Q 0 f ‘. . ..v. X. ‘ CI"I. ' . r} -=. Q in {v " ‘. .4! '\ r 4 ..;.‘ - , - . . . _ ,. ‘ fl .-- , . . V, c .g.:.. m. . . . .. . m 5 “Va? T’s' .. - _ ‘qtn . v“--_~ . ‘ . ‘ ,. . .fi. _ ._.. .V. -.-.ul1 . . . ‘5 1. - o " ' .r‘ ‘ ' . I .p- 1‘ L .3 ;{ ’.‘ Aer; .3: .217" ' ’.. a'_€£a‘r-:f : 3". 0 q. m}: 3?;“3 55: r I t?! :4- l & ‘3 I " 11".. \x 4‘. 7 I 1" I . X ‘ 3-‘1 C '.I . .. ’35“ ‘3 ‘ o :1 . -0 ‘F L r U '6 I .- ..I .. c I 1 I .I ‘I . .. 4 1 . pl 0.:00 O . ‘?i; mm AC l' I 3. , 2 t,--- ":.- Q.‘ J ‘:‘-Iflz'.-. 0‘: 1"" .. ,v v ..o '0'." 3'4“. I: }~,«.‘I‘Ti"-' .1. . ' 5a ‘ 5'} t. 9 $1.“: .. - 1' 1‘7: 'p“. 1-) ~ ' '14P- l‘?"‘|‘:r- . . .Q .. 4 .3- 013.“ ,1 19‘ .‘NA L-‘c -' -_.“ a”?! 3' . . . :u . .' '3;.. -.'.e 'Q'Ak- -‘ '.; \hv.‘ . ..T_|"..';‘L.;L." 7.43:}: . A ..‘{: ‘ n . 940.2. 9.5:.» 4““ .7 rd p Q C 0" " " "I‘ a .51— . : ‘Iv ‘ ‘6- a “34“ - C's ' fl 3 .v ’0'. .,o . "o - ‘ 0-1“! ,. ... o ~‘4 ‘-‘. . :f (3:7‘! _ ( flying-«c434. :.l -‘l- ‘ ...,,1‘. ,, «..--9 . VJ. ‘..'... L_ 7‘“ i’”~.‘~u ’1.- I... . :& z 5 .?.v ‘L N . 3 ‘~. ..._ _. _ ,. ‘ .,., (I and 3m": nut-v, . f _ ‘3. r 3.0. uh). v. .9". . - “‘3‘ if“ Jirb190-r:.} .1 “A"; :1 .. ‘ ‘_-'D'l ‘. - - ' . I, .Q.‘. ‘ . ~ .. ‘4 "_1.'- ,1...“ <5! I ~ " . . . . . ... . ‘ . -.Q‘- ~ . . , t}. v,.n.-’3')o‘ 3Q? ‘. I“ J . 09’- ‘. '. ..‘ 4‘ I.- ,-.".‘.t u.‘ .0 ‘fi‘ . fi . ‘ 01' — "" Ho ‘ J"- -“ .6" 0‘. J, a ' ’0 ‘ 531‘.“ ’31; {1:3 .P .I I ' n ‘4 0 I .46.» I, ‘ ‘t \- ' léo‘.‘ ‘ y . s “lg" ..h4::ui:‘ o o‘ 3'. o. ' 4 ‘Qvfiu . ~'Q f.‘ “ "" I ‘ ‘ .. 3 ‘ - ..I0.ln1'¢v - I ‘4 ‘V P“ 1‘» 3-“, h." ; .':'-' -\‘D‘ u. r .< .‘ .5 0. its}. ‘ .5. . .5?» .3: . ..g":‘a (My! 3;". “Asa-J")- i‘ofi‘f. A ’\ . ,- .. {I .v a J {‘0‘ .0“. 'I”!J‘;B?£Iiu‘5 . -. '. . v .o’ , o. x ,. ‘ ”~‘~".i.-- «‘c ° - v- '-"‘ ""“§:::Eer-1m.,~ 7. ..~ . A, . ._- n m. r3: 'N".Q'. N”-‘ g ‘ ‘ . ._ 1‘ {'22. ‘ . ~ .. ’3‘ "I o ('7 .’:'«‘1.'A!i? ‘o‘ 1"" '17:: $‘::‘ ’5 . (ltfic'g .L “a .y‘ 4' 7 ritr: .‘t c?“ 1%,“, Li: . ‘ ~ 3.: _ ‘3’. ‘ .‘ ."l‘n ‘1 - I (3.: - Q, '-"‘-‘~ JY. |_‘.‘1;fu. - fm9 :1 .4: - I. . ‘ f1 .0 I. {:‘Y". i 0 1",}:fi 1 f. .y gw-i- . 1.“.‘3'1A(-,Ct “33‘ 3. ”2 '0 ”3.43; z. ' . . o ‘ . ' ‘a Q - Q'- a - . ‘. ~E;:§'l‘: -“_‘2.’ 3;: ’. t, '. ‘ V ‘ ' 3 ‘39; a": ‘0' g ’ ' ’ . _- «n V ' ‘ ‘| {7%,- T’tl‘.'k,"‘0.f“.:gi. ‘zg’:10!r;$::‘: """.~o' #r' :, .‘._' " .5 1'. » m1 , ' ’ 53”,? ‘ ‘ 91rd"... fit .' ’ Yflz-‘zd‘b‘gk ."“‘ 1... "n‘ 1.- I 3,3. " - ”‘5‘ I‘ ‘ .‘f-n . g a. '7'.‘.l 1.). ... . . ‘ S~ ‘ "may run-3’ ‘5. - tm . - '- ' r '. o p re: - - an?» - =. ~s»§=~-'?='3’ ve‘ . ‘u” t." ‘ "’1‘“! ."t' “ '1 ;.‘ w‘imfie: m ~ m < . V “3.3%“ pr; .I;.% .0 31"er 311;.(1 ? (3." "3"- ~-.-. .3153, ”7-" .31 ~. 1“ .r Q"A "wc- .,. ' 'u -- .W . .1 . 05' Um 01". 1 . Fl Q- f’WS-n- I“ ‘J ‘ £0 ‘lr'lw"‘ ’ ‘ _"‘ g a! 1‘“ 9K- .n. .4: L ..I x 3:: , ¢ - .31 ‘9 :4 3:3 .5. 3V3 - .4 . '*.ro'v ‘- 'I s 211;: r: :;;;.;."e . i , 2.... .0 O‘ ‘ 'I‘ ‘O ‘ 0 $ . . ' a"?! a"; 5‘ 3-2. I ‘3'}: 7' H ' 9-1595- "Ava- ‘ " 0’ A 0 2‘4""?! 33‘91‘5; 33331,. . gum. ‘:‘ ' ”14‘. «3‘ _ 43d; 71:.1’, t’,‘ . x ‘1: ".553 m: -=;g~:s.§axE§?%-.I§%m, «.5: .. fiwgz'taj' 'P'o‘d' '4“ ‘ ‘ 50:43. _ 'L’Q‘J‘.‘ _ ,,-.'f 5?. “'"’w ' “’9': 9"“‘3‘ ”.9 f! k tr chi/f." ' ‘ij; m;fi.fi cffi‘c 4. “ '“rrwrf. . 3"}23? . 5E”;JX‘",~,‘9;,§337 11“:ch '15..."- 9 v a: :1." «5‘43” ‘3'?! - "£2=*aa.-,v'--'-:¢W.a:..sa‘érgvzsi'aniv‘ g5" aw; W51“ if” ,4? '22:“ cigasf' “£9? flwmrww§.vfi 5:4,~ _.. "01...; ‘43:? k l:-"... ‘0-'5: .- .p'l- g-écvgtw? n; 421:1: mfg-E a," 3 m We}: 375-3, 79"?33’.’ Q5 ffffi'j‘ 7:19” ‘ = < $12” 35".” f 3“ 11.. 1 up; '; ‘-Qa Trig! ."'0‘ ;;%‘(:':'f Q0)! f‘v I" D - b — F' .l . .3- fl“.1L:i-§vxa ; 1. Jim - “"1 :.i" "'3" I“ 5"" 3‘"; ”.;Iu~'.:.':;' 3:15. . {~ "v 'x n "H m ‘ «our . ..'2":'-fl fa. '¢>:§.Ll_-;-~ ’ ER. 9 I. u. ~04 l: .0. '_. '_ g ‘f‘: C‘. '1' «' 32.113; - . ‘,” v . .A‘u u a}: £04. $43?" ‘3 “W‘- Y... (3'31: 3' er . o 4",: - 1. ' - - « ”-1? LP ' 0 h 1 I u . H1 69* .N ‘ ‘ :‘VVzteflfiif"" ' .:'£=& " i‘ ‘1.“ ‘gtuq’d rig-’3‘ 0 “1:5! {35:3 fa‘ _ _. , ' “I". :, 9.)}? -'W‘ .V s...”‘.;t‘.‘ 91: r1 "$"t§‘,:z 3;" g‘i%.m o 2'2“" 3‘3“ . ‘ r!" | lay ‘ cw ' . 335% ' s...‘ :-:.: a.” 7n“ \ .‘X‘Hgg" -: "fiw‘i' .‘0 a ‘ .v ‘0 .‘ .. I 0Q «In 2" , .. 3 0-9» 1" :- ug. THE-33‘. This is to certify that the thesis entitled The impacts of increased goal complexity on message production: A message design logic approach presented by Dedra M. Campbell has been accepted towards fulfillment of the requirements for Master ' 5 degree in Communication jor professor 2 9/94 Date / 0-7639 MS U is an Ajfirmative Action/Equal Opportunity Institution 1293 01025 895 .1Milli!iiiiillll’lilliiiiliii LIBRARY Michigan State University PLACE IN RETURN BOX to roman this checkout Irom your record. To AVOID FINES «turn on or baton dd. duo. DATE DUE DATE DUE DATE DUE MSU In An Afflmutivo Action/Em Opportunity Institution Wanna-n THE IMPACTS OF INCREASED GOAL COMPLEXITY ON MESSAGE PRODUCTION: A Message Design Logic Approach BY Dedra M. Campbell A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF ARTS Department of Communication 1994 ABSTRACT THE IMPACTS OF INCREASED GOAL COMPLEXITY ON MESSAGE PRODUCTION: A Message Design Logic Approach BY Dedra M. Campbell Message design logics is a model of communication which maintains that individuals construct mental representations of situations which lead them to systematically select different elements for expression. The purpose of this research is to test the assertion that as goal complexity of a regulative communication situation increases, message variation between individuals who adhere to different messages design logics will increase. This research employed two levels of goal complexity: interethnic and intraethnic contexts. The interethnic context was argued to be more complex. When the ethnicity of the message target was held constant (i.e., the two levels of goal complexity were collapsed into one category) the results of these analyses indicate that there were significant differences observed between expressives, conventionals, and rhetoricals that were consistent with the hypotheses of this study. When the ethnicity of the message target was taken into account, there was a different pattern in how individuals who adhere to different message design logics address the features of a complex situation. The implications of these findings are discussed and suggestions for future research proposed. ACKNOWLEDGMENTS I would like to thank Dr. Steven Wilson for his comments on earlier versions of this manuscript and for encouraging me to pursue this research, Dr. Steven McCornack for taking hours of his time to brain storm, code hundreds of messages (several times) and assist me in developing the major frameworks of this research, and Dr. William Donohue and Dr. Jacob Climo for their valuable comments and words of encouragement during a long and often difficult process toward completion of this manuscript. iii TABLE OF CONTENTS Introduction Message Design Logics Current Research Directions The increased complexity of a situation Situational Theme Theme Theme Method Context of communication task: interethnic versus intraethnic Message source role Inconsistent information about the message target Multiple objectives of the task Themes 1 2 3 Independent Variables Dependent Variables Analyses Results Discussion Appendices Appendix A Appendix B 'Appendix C References iv 13 14 17 25 26 27 29 29 35 39 41 42 45 51 77 84 84 87 92 Table Table Table Table Table Table Table Table Table Table Table Table Table Table Table 10 11 12 13 14 15 LIST OF TABLES 46 47 48 49 49 53 58 59 62 63 66 68 69 74 7S Introduction Message design logics is a model of communication which purports that in producing and comprehending messages, individuals can employ any one of three different ways of reasoning about the fit between messages and tasks (O’Keefe, 1988; O'Keefe & Lambert, 1989,- O’Keefe, 1991). The three message design logics are: (1) expressive, in which self- expression is the chief message function and the principles connecting messages to the context are affective and idiosyncratic; (2) conventional, in which the point of messages is to secure some specific response and in which messages are connected to their context by institutionalized norms of appropriateness; and (3) rhetorical, in which messages are connected to their context by models of character, motivation, and social coordination and in which the point of messages is to create a consensus between message sender and receiver. In her article discussing message design logics and individual differences in reasoning, O'Keefe (1988) asserts \that as the complexity of a situation increases, message variation between individuals who adhere to different message design logics should also increase. However, there has been 2 no systematic examination of this relationship. Hence, the purpose of this research is to investigate the impacts of task complexity on how individuals, who adhere to different message design logics, address the communication situation. The first factor used in this research to increase goal complexity was presenting ethnicity of message target in the communication task. Because research in the area of interethnic studies suggests that situations are more complex when the interaction is interethnic versus intra-ethnic (e.g., Bodenhausen & Lichtenstein, 1987; Hecht, Ribeau, &Alberts, 1989; McConahay, Hardee, Batts, 1981; Feagin, 1991) the present research will explore how these issues play out across individuals who adhere to different message design logics. The difficulty of the message source role was also increased, contradictory information about the message target was imparted, and the objective of the task was increased. It is hypothesized that interethnic encounters are more complex than intraethnic encounters, hence, there should be more variation between individuals who adhere to different message design logics in the interethnic context versus the intraethnic. To investigate these issues, individuals who adhere to one of the three message design logics will be asked to respond to a complex, regulative communication situation in which the message target will be either African American or Euroamerican. The participants messages will be coded based on three themes: (1) "how'do individuals address the objective 3 task at hand," (2) "how do individuals address the message target’s current excuse regarding a family emergency, " and (3) "how'do individuals address consistent.patterns of the message target’s missing work." This paper is organized into eight main sections. The first section reviews prior message design logic research. The second and third sections discuss the current research directions, present the coding scheme of the research and propose hypotheses. The forth and fifth sections present the methods and analyses. The final section presents the results of these findings and discusses the implications. Message Design Logics Considerable attention has recently been given to message design logics (e.g. , O’Keefe & Shepherd, 1987; O'Keefe & McCornack, 1987; O'Keefe, 1988; O’Keefe 8: Lambert, 1989; O’Keefe, 1990; O’Keefe, 1991). O’Keefe and Lambert (1989) purport that in. producing and. comprehending' messages, a communicator relies on an implicit model of communication that embodies an individual’s knowledge about how'to relate message forms and functions (O’Keefe and Lambert, 1989) . These models of communication are referred to as "message design logics." O’Keefe (1988) maintains that a person can employ any of three different message design logics or ways of reasoning about the fit between messages and tasks. The three message design logics are: (l) expressive, in which self-expression is the chief message function and the principles connecting messages to the context are affective and idiosyncratic; (2) conventional, in.which.the point of messages is to secure some specific response and in which messages are connected to their context by institutionalized norms of appropriateness; and (3) rhetorical, in which messages are connected to their context by models of character, motivation, and social coordination and the point of messages is to create a consensus between 5 message sender and receivemn Put another way, O'Keefe argues that individuals can have distinctly different conceptions of what is relevant to a communication situation. Differences in how'one views communication can cause individuals toiconstruct the context for communication differently which lead individuals to different decisions about what ideas are appropriate for expression (O'Keefe & Lambert, 1988). Because a message design logic is the body of working knowledge an individual possesses about a communication process, one way O'Keefe (1988) argues that message design logics can be observed is by its manifestation in messages that individuals produce. Thus, individuals can not simply articulate their beliefs or theories about communication - it must be observed in how they actually design messages (O'Keefe & Lambert, 1989). However, variation in message design logics will be suppressed in simple communication situations and revealed as situations become complex (O'Keefe, 1988). The complexity of a situation refers to the number of competing situational goals and obstacles to goal achievement. As the number of relevant situational goals increases, especially if they are competing goals, variation in goals adopted by individuals is encouraged (O'Keefe, 1988). If the number of potentially relevant goals is small, there are less possible subsets of goals that persons might select. Hence, there will be little individual variation.inugoals selected for pursuit. In short, 6 expressives, conventionals, and rhetoricals generate similar messages in simple communication, but as the complexity of the situation increases, message variation as a function of message design logics increases (O’Keefe, 1988). A task used frequently in this research to elicit an individuals' message design logic is the "Group Leader Scenario" (see O’Keefe, 1990). In this scenario, subjects are given a regulative task in which the message producer must deal with. a procrastinating' group member' who failed to complete work on time. IParticipants are instructed to provide the actual words they would use to deal with the situation. O’Keefe defines this task as a complex communication situation (i.e. , a communication encounter with multiple goals) in which one person is faced with the need to regulate the erring behavior of another person (O’Keefe, 1988). O’Keefe states that the circumstances that make a situation regulative can provoke a good deal of negative thoughts about the message target that are not relevant to the attainment of goals of the situation (O’Keefe, 1990). Similarly, the inability' or unwillingness of the message target to meet existing situational requirements combined with the intrinsically face- threatening character of attempts at regulation of conduct, make redefining the communicative context desirable (O’Keefe, 1990) . The group leader scenario has two primary features which make it effective in eliciting message variation from 7 individuals whose ‘message design logics differ (O’Keefe, 1990). First, Ron (the message target in the situation) is represented as having repeatedly failed to conform to the group’s decisions and procedures. This leads the participant to have a negative impression of Ron and a number of salient beliefs about his past conduct that are not particularly relevant for dealing with the present situation. Second, the group is portrayed.as having a particular organization and.set of commitments that may or may not be subject to renegotiation but which are relevant to dealing with Ron's failure to perform. In addition to this, the ways in which the group's arrangements might be altered to secure a good outcome are fairly easy to see and plainly advantageous (O’Keefe, 1990). O’Keefe summarizes this point by stating "the presence of salient but irrelevant and negatively valenced beliefs about Ron’s past behavior and the desirability of reorganizing the situation make this scenario especially useful for detecting differences in message design logic (O'Keefe, 1990, p.92) Thus, an individual’s message design.logic is elicitediby using a system described by O'Keefe (1988; 1991) which specify the properties of messages indicative of each of the three message design logics. Messages are then classified as either expressive, conventional, or rhetorical. In the Group leader scenario, messages which contain properties of an expressive message design logic will produce:a particular type of message when addressing this task” There may be a lack of editing (so 8 that socially unacceptable content is included in the message) and a failure to engage the immediate task to be accomplished in the situation (e.g., there may' be a focus on past transgressions instead of attempting to get the task accomplished) (O’Keefe, 1990). The principle premise of expressive messages is expression of affective feelings and emotions. The following are two examples of an expressive message. (1) I will do the research myself and I'm telling our teacher all of the chances we gave you and that you should not get credit in helping us on our project. (2) Ron, I can’t believe you haven’t finished your research. ‘You have been inconsiderate to the group all along. Several members even suggested that you be taken out of the group but we decided to give you a chance. Now what are we supposed to do? It was your responsibility and you backed out. I'm afraid that I’m going to tell the T.A. that you haven’t done your share. I will be so mad at you if we get a bad grade on this - I need an A in this course. Messages which contain properties of a conventional message design logic will also produce a particular type of message when addressing this task. These messages will focus a great deal on directing a future-goal related action of the message target. Rights and obligations of the situation will also be given as reasons why the message target should perform as required. In other words, the point of messages is to secure some specific response using institutionalized norms of appropriateness according to the context. The following are two examples of a conventional message design logic. 9 (1) Ron you knew that we were all counting on you. You just have to get the project done for tomorrow because there is no more time. If you don’t come up with it I am going to have to give you a failing grade. Now I don't want to have to do that, so don’t let me down or the group down. (2) Well, Ron, I’m sorry you don’t have your part of the project done. We have given you several breaks thus far and I don’t see how we can give you any more. The whole group is depending on you so I would suggest to you to get it done or at the most bring in what you have got done. If you don't get this done I’m going to have to give you a F for the project. If you can’t hold up your responsibility with this group even under these adverse conditions (family problems) how are you going to make it in life. Messages which contain properties of a rhetorical message design logic lead to a third type of message. This type of message is derived from the view that situations are created in the process of communication. Thus, there will be motivational appeals, rhetorical labelling, explicit redescriptions of the context, and so forth. Moreover, the sole purpose of these messages is to create a consensus between message sender and receiver. The following are two examples of a rhetorical message design logic. (1) When can you get done by? Would two days be enough (and if he agrees and sets a day . . .)? O.K., please make sure you get it completed because we really need to get the final draft done soon, O.K.? And. if you need some help, please tell me or someone else from the group. IHow are things at home? (2) Ron, I’m sorry you won't have everything ready by tonight. Can you give me some good reasons I can tell the group? I know you’ve been having problems all along and that’s obvious to the rest of the group too. You need to be able to see how your personal problems have interfered in the group’s completion of the project. I’ll be the first to sympathize with you but now we'll have to come up with some concrete solutions. You can't expect someone else to take over your research workload. 10 O'Keefe argues that these three message design logics reflect three different ways of reasoning from goals to messages rather than three alternative message forms derived from the same way of reasoning about communication (O’Keefe, 1988). Thus, there is reason to believe that differences in messages produced (e.g., differences in how individuals address relevant goals in a complex communicative situation) result from individual differences in more fundamental processes guiding message construction and interpretation. To move away from using communication as the only way of measuring an individual's fundamental processes guiding message construction and interpretation, O’Keefe and Lambert (1989) developed a method of eliciting an individual's situationally relevant beliefs about communication. These authors argued that different message design logics lead individuals to have different substantive beliefs about the situation (O’Keefe & Lambert, 1989). Moreover, individuals will also have different beliefs about what they should try to accomplish and how communication can be used to accomplish goals. O’Keefe and Lambert (1989) had participants complete the "Group leader problem" and the "Broken date problem" (another situation frequently used in this research) to elicit an individuals message design logic (see O’Keefe and Lambert, 1989). Individuals completed a thought checklist about the thoughts they had while completing the Broken date problem; 11 121 common thoughts were generated iJI the pretest. Participants employing different message design logics showed differential selection patterns for 30 of the 121 common thoughts. 'Thus, individuals in this research reported similar sets of thoughts about the Broken date scenario, irrespective of message design logic. However, based on the results of a discriminant analysis on. participants thoughts about the situation, those who employed expressive, conventional, and rhetorical message design logics disagreed about two functions (O’Keefe & Lambert, 1989). The first function reflects differences in focus of attention. Expressives and conventionals focus on what Terry (the message target in this situation) has done and its’ subjective meaning (i.e., how has his or her behavior affected me?). Thoughts associated with this focus are expressions of irritation with Terry, feelings that the message producer should stop seeing Terry, and believing that Terry needs to know of the message producers anger. On the other hand, rhetoricals tended to focus on their friendship with Terry and what could be done to maintain it. Thoughts associated with this focus stressed the importance of the relationship, the importance of understanding, the suppression of hurt feelings, and recognized alternative ways to maintain the friendship. The second function that those who employed expressive, conventional, and rhetorical message design logics disagree on 12 regards the thoughts about Terry’s character (e.g., whether Terry is being dishonest or truthful about his or her excuses). Expressives tended tx> have negatively valenced judgments about Terry’s character, mention grievances about Terry's conduct, express extreme annoyance, express wishes that Terry be punished and desire reparations. Conventionals and rhetoricals on the other hand tended to express confidence in.Terry’s good faith and acceptance of Terry's account of the situation. Although the effect sizes for the distinctions of thoughts between the different message design logics were not large, albeit significant, they'do suggest that these thoughts about the situation and message target are consistent with the message design logic model. Put another way, differences in message design logics may be understood as reflecting differences in situational representations. Moreover, how one perceives the situation leads to the production of a particular expression of these beliefs. Current Research Directions Message design logics is an important model of communication because it attempts to conceptualize message production as being a function of how individuals reason from situational goals to messages rather than messages reflecting three alternative message forms derived from the same thoughts about communication (O’Keefe, 1988). However, one of the primary limitations of this research is the lack of research which examines if expressives, conventionals, and rhetoricals systematically vary in how they design messages in communication tasks with increased complexity; This research will explore this issue by incorporating ethnicity of message target, by increasing the difficulty of the source role, by offering contradictory information about the message target, and.by increasing the objectives of the task; This complexity’ is compared to the situations used in prior message design logic research. Moreover; interethnic communication encounters are argued to be more complex than intraethnic communication encounters. The next section explores each of these issues. 13 14 The increased complexity of a situation One of the limitations of this theory is that a larger variety of situations are not employed to investigate how the different message design logics respond to them. As mentioned previously, as the goals of a situation increase, message variation as a function of message design logics should also increase (O’Keefe, 1989). However, no research has systematically examined if goal complexity actually increases message variation between individuals who adhere to different message design logics. Moreover, in what ways these messages vary is also yet to be explored (i.e., what features of the situation, if any, are consistently addressed.by expressives, conventionals, and rhetoricals when the goal complexity is increased). For example, the Group leader problem presents to the message source one primary objective to the task - to get the group member to complete his work“ The message source is also provided with consistent information about the message target’ 3 (i.e. , Ron) previous behavior as being unreliable and problematic to achieving the groups goals of completing a class group project. There are also a set of commitments that are relevant to dealing with the message target’s failure to perform, such as Ron's responsibility to complete a portion of the research to receive a grade in the course. Thus, the task 15 objective is straightforward enough that the three message design logics can be easily distinguished from each other by the features of the situation addressed in the message. Expressives generally do not focus on the objective task (i.e., getting the message target to complete his work). Conventionals generally focus on the objective task by reminding Ron of his obligations and responsibilities to complete his work. IRhetoricals tend to focus on the objective task of getting the message target to complete his work by negotiating some consensus between the message target and him/herself. Thus, the primary goal of this research is to test the assertion originally presented by O’Keefe (1988) that as the goal complexity of a situation increases, message variation increases. To investigate this issue further, a complex communication situation was created for this research. The situation employed is "The manager situation" and was modeled after the Group leader problem in terms of providing the participant with a regulative task and ambiguous pieces of information about the message target that can be construed negatively or positively by the message source. By creating two contexts, interethnic and intraethnic, based. on the ethnicity of message source and.message target, two levels of task complexity were created. Before reading the situation, message sources were provided with an "employee information sheet" which presented a picture of the message target, 16 "Donald." Donald was either an African American or an Euroamerican male. Thus, this situation differs from previous message design logic research in that the salience of the message target’s ethnicity is encouraged. Depending on the ethnicity of the message source, the context of the communication situation is either interethnic or intraethnic. In the current study’s communication task, message sources are asked to imagine that he or she is a manager of a catering service. The responsibilities of the manager are described in the scenario (e.g., in charge of the food service and preparing schedules for employees, keeping up—to—date reports on each employee, evaluating the employee, and hiring and firing these employees). The message source is then told about the message target, Donald, a good employee who has recently been late for work several times, has had several unexcused absences, and has previously given the excuse that his car broke down and that he's having family problems. After message sources are given this background information, they are presented with a current dilemma. The message source is preparing for a very important event, and at the last minute, Donald calls in to say that an emergency has come up with his family and he will not be able to work that day (see appendix I for the situation). Message sources are asked to write down exactly what they would say to Donald in this situation. Thus, in general, there are four primary issues that make 17 this situation complex compared to previous communication tasks used in message design logic research: (1) the context of the situation (i.e., interethnic or intraethnic) (2) the role of the message source, (3) the inconsistent information given about Donald, the message target, and (4) the multiple objectives that potentially can be addressed in this situation. Specifically, however, to jpresent. a. direct test of O'Keefe and Lambert’s (1989) assertion that as goal complexity’ increases, variation between individuals who adhere to different message design logics should also increase, two levels of goal complexity were created for this research. These levels are interethnic and intraethnic contexts. The interethnic contexts is argued to be more complex, hence, we should see more variation in this context versus the intraethnic context. The next section explores each of the four issues of complexity in more detail. Context of communication task: interethnic versus intraethnic An important issue relating to goal complexity in a regulative, communication situation is the ethnicity of the message target. The tasks frequently used.in this research do not specify any characteristics about the message target. Thus, one can assume that the message target is ethnically 18 nondescript (i.e., not an issue to be considered in the situation). In actual interactions, characteristics of one’s interactionel.partner'(e.gu, ethnicityy gendery age, perceived attractiveness) play an important role in how we construct representations of the situation which lead to our message design. v/ Moreover, research examining interethnic communication suggests that ethnicity is a relevant factor in interethnic situations which influences subsequent behaviors (e.g., Devine, 1989; Bodenhausen.& Lichtenstein, 1987). Thus, there is reason to believe that ethnicity of message target may be an influential factor in how individuals address particular features of a situation. Ethnicity of message target may also influence the three message design logics differently. To explore these issues further, in relation to message design logics, research examining interethnic interactions will be discussed. Relevant Interethnic Literature Researchers have attacked issues encompassing interethnic strife and communication problems by attributing them to differences in culture and styles of communicating. For example, Kochman (1982) challenges the assumption that African Americans and Euroamericans share identical speech and cultural conventions by claiming there are differing norms and social styles which affect communication. Kochman points to l9 divergent patterns of intonation, expressive intensity, spontaneity, aggressiveness, and argument. Hecht and Ribeau (1984) report differences in the kinds of intra-ethnic communication African Americans and Euroamericans find satisfying. Asante and Noor—Aldeen (1984) found.a pattern of indifference between African American and Euroamerican interactants when interracial dyads were observed as well as a general pattern of racial isolation. Such studies demonstrate potential as well as real tensions involved in interethnic interactions and ways in which these tensions can diminish prospects for effective and satisfactory communication. Other researchers have focused on the role that stereotypes play in subsequent behaviors produced by Euroamericans during interactions with African Americans. Due to the negative connotations embedded in stereotypes of African Americans (e.g., laziness, imitative behaviors, violence, hostility, and poverty). (Devine, 1989; Dividio, Evans, & Tyler, 1986; Molnar, 1989), researchers have associated ethnic stereotypes with racism (Allport, 1954; Hamilton, 1981; Tajfel, 1969). Although knowledge about a stereotype does not necessarily indicate prejudice toward members of the stereotyped group (both high- and low- prejudiced people have been found to possess similar information regarding stereotypes of African .Americans), research does suggest implications for subsequent behaviors. 20 Both high- and low-prejudiced individuals may unknowingly display' stereotype-congruent Ibehaviors (i.e., negative responses) when faced with complex-decisions regarding African Americans as well as when judging ambiguous behaviors under time constraints (Devine, 1989). Similarly, Duncan (1976) demonstrates how ambiguous behaviors are rated stereotypically by Euroamerican subjects. After observing a videotape of a simulated ongoing interaction occurring in another room, Euroamerican subjects evaluated ambiguous shoves as being more violent when initiated by an African American than when initiated by an Euroamerican (Duncan, 1976). ZBodenhausen.and Lichtenstein (1987) presented subjects with a hypothetical situation in which they role- played jurists for a criminal trial. Subjects gave more guilty judgments and perceived a Hispanic defendant as more aggressive than a nondescript defendant. Consequently, stereotypes of African Americans may interfere with effective interethnic communication because it influences the perceptions of individuals. Rich (1974) found that when African Americans regarded Euroamericans as "evasive" and "concealing" they ignored any communication coming from. Euroamericans which. may' have suggested "directness" or "honesty". On the other hand, Euroamericans who were predisposed to view African Americans as "hostile" and "aggressive," ignored any overtures by members of these groups to cooperatively share information. 21 Campbell (1993) claims that given the growing body of research, media attention, and awareness of behaviors associated with racism towards African Americans, Euroamericans are becoming increasingly anxious about the topics, choice of words, nonverbals, and so on, that they employ when interacting with African Americans. In other words, in their interactions with African Americans, Euroamericans may face a dilemma of not wanting to be perceived as "racists" by producing the "telltale signs" described by researchers and the mass media. Their concern with not being perceived as racists may deter their use of openly confrontational and/or negative interactions toward .African Americans even where they would use such behaviors in interactions with other Euroamericans. In a pilot study examining the extent of argumentative communication, Campbell (1993) found that Euroamericans are likely to be significantly more confrontational when interacting with other Euroamericans as opposed to African Americans. When Euroamerican subjects participated in simulated. negotiation. sessions 'with. the Euroamerican confederate, subjects used more aggressive and assertive communication such as blatant disagreements and/or rejecting the other’s proposal or offeru In contrast, when.Euroamerican subjects interacted with the African American confederate, they used more neutral communication such as asking questions or responding to the immediately preceding utterance without 22 providing additional information or introducing a new topic (Campbell, 1993). These findings suggest that Euroamericans feel more comfortable being argumentative, in.both.conflict or negotiation situations, when the interaction is intra-ethnic as opposed to interethnic. In difficult communication situations (e.g., conflict, regulative or compliance-gaining situations), adding an interethnic dimension intensifies the tensions involved. Communication interactions between.African.Americans and Euroamericans in. particular have had a long history' of conflicts on many different levels. Beginning with the historical impact of slavery (Asante and Noor—Aldeen, 1984; Litwack, 1979), African Americans and Euroamericans have had political, economic, and.social conflicts. IRecent examples of these conflicts include attempts to desegregate public facilities and school systems, fights for civil rights and riots triggered by racial discrimination (McConahay, Hardee, Batts, 1981). A current example of an on-going issue between Euroamericans and African Americans which.may'be the source of problematic interactions is the degree to which racism exists today. Public opinion polls examining perceptions of racial prejudice today indicate that Euroamericans believe nondominant groups experience very little racial prejudice and discrimination (Feagin, 1991; McConahay, 1981). In contrast, African Americans do not share this perception; both their 23 life experiences, and those close to them indicate far more prejudice and discrimination. Sui a study examining these issues, Feagin (1991) interviewed middle-class African Americans. Even though the interviewees had achieved a high- level of education, successful careers and financial stability, they still experienced many different forms of discrimination and racism. Thus, while an Euroamerican who has not experienced racism might say, "It is unfortunate that discrimination still exists, but at least. it is ‘not as extensive as it used. to be" an .African .American. would disagree. .African Americans may become very upset and.respond to this statement by saying something to the effect, "Racism is still as strong as it used to be and just as infuriating and painful. The only difference is that it is more difficult to detect." The list of potential communication sCenarios are infinite, yet one thing remains the same - cultural experiences influence individual perceptions about racism, what is an insensitive comment and so on. Different perceptions of such emotionally charged issues may lead to conflict between different ethnic groups. Thus, an exchange like the previous example may leave the interactants with negative attitudes toward the other person. African Americans may'be extremely' "sensitizedfl and.emotional regarding racism, discrimination and inequality. Because these feelings of anger and intense emotion are difficult for an Euroamerican to understand (probably because of the lack of shared L 24 experiences) he or she may feel African Americans are overreacting and hence, feel uncomfortable introducing topics which may potentially lead to the "bristle-like" reactions. Such issues are cultural “baggage" brought into interethnic interactions between African Americans and Euroamericans making these encounters different from intra— ethnic interactions. Campbell (1993) terms these issues the "walking-on eggs" syndrome. This syndrome experienced by Euroamericans while interacting with African Americans may influence features observed in the messages produced. This is particularly important when a conflict or regulative situation arises. In such situations Euroamericans may feel unsure of how to interact with.African Americans because they' perceive them as paranoid or overly sensitive. Thus, Euroamericans may avoid any direct confrontational or argumentative style of communication while interacting with African Americans. Moreover, they' will. be jparticularly' cautious approaching .African. Americans in. a regulative situation. due to the negative connotations associated with arguing and overly confrontational styles of communication (Infante et a1. , 1984) that could be perceived as ethnic- or race-related. Thus, since interethnic communication interactions tend to be more complex than intra-ethnic interactions, it is reasonable to assume that ethnicity of the message target will be an important factor in how individuals construct representations of a situation. However, because the message 25 design logic model proposes that expressives, conventionals, and.rhetoricals employ three different ways of reasoning about communication (i.e., the fit between messages and situational tasks), it follows that ethnicity'of the message target should influence individuals who adhere to one of the three message design logics differently. Message source role Regarding the issue of the message source’s role complexity, one can.assume that a position.with this magnitude of responsibilities came from hard work and the ability to deal with complex situations. Put another way, the message source as manager earned his/her position as manager. In the Group leader situation, the message source was appointed the role of leader by the professor. As manager for the catering service, there is also the issue of earning a living through the position, having an investment of time and energy in achieving the position, having one’s ego attached to the success or failure of doing a good job, and the threat of failing to perform services adequately, resulting in unhappy customers or lower profits. Even if the role of being a manager for a catering service is only a temporary job, there is an increased involvement to this position versus being a group leader for one group project. Moreover, a manager with this much responsibility also has to be concerned with 26 policies and procedures regarding firing and hiring which could potentially lead to serious legal ramifications if handled incorrectlyn These are issues which.the manager would have to be responsible for if problems arose. This is partially due to the expectation of being a competent manager who is able to handle problems in the best way possible. In the group leader situation, the leader has the option of dealing with.the message target by "telling the professor" and letting him or her handle it. This option is not readily available for the manager situation. In short, the manager situation presents the message source with increased role complexity. Inconsistent information about the message target The previous behavior of the message target also increases the complexity of this task versus other tasks used in this research (e.g., the group leader problem). Donald's previous behavior had earned him the respect of being a "good workeru" However, his recent behavior as an employee has been problematic because he has missed several days and has been late to work several times. The message target has given the excuse once of having car trouble and twice (including the current excuse) of having family problems. Because the message target has not consistently been a problem, the issue of whether or not Donald’s excuse of having family problems 27 is valid becomes problematic. Ihnni if the message source believes that Donald’s current. behavior' is due to irresponsible behavior, his past behavior as a good worker may be a salient issue in how one perceives the message target's behavior and character. Based on the message design logic model, expressives, conventionals, and rhetoricals should differ in terms of how they focus on the message targets behavior. Multiple objectives of the task There are multiple objectives, long— and short-term, related to this task which also make this situation complex. The most obvious short-term objective is to deal with Donald’s request of having the day off. And since the situation described suggests that Donald is needed to work, the main objective would be to persuade him to come into work. Related to this objective is how the message source assesses the credibility of Donald’s current excuse. If the message source believes that Donald did have a family emergency, then this excuse may be considered important enough to grant him the day off. This would most likely not be the case if the message source believed Donald to be making the excuse up. How the message source addresses the current excuse (if he or she does at all) is also related to the thoughts he or she has about Donald's previous excuses, absences and tardiness. The objective of addressing the message target's long ternrbehavior is an additional issue in this task. 'Unlike the group leader problem, the relationship between Vmanager- employee is more long term and entails more factors than the relationship between group leader-group member. In the group leader problem, the primary objective was to get the message target to complete his work. In the manager situation, the message source‘mag not only be concerned with the immediate , ** r” (ax/u wom objective of getting Donald to work that day, but with addressing and trying to resolve his problems (if the message source believes that there are legitimate problems) so that the message target can continue to be a productive employee, hence, long-term benefits for the manager. In this case, the manager may forsake the short-term goal of getting Donald to work that day so that he can take care of his family problems and become a good worker again. This option may benefit the manager's overall objective of being efficient by eliminating the costs of having to hire and train somebody new, therefore, taking a chance when the manager knows that Donald has been a good worker in the past, therefore, has the potential to be a good employee again. On the other hand, message sources may forsake any long-ternibenefits of allowing Donald.to deal with his family problems and to get himself on track again by focusing on the short-term goal of getting him to come in. If Donald fails to comply with this goal, he may be fired. In 29 some cases, the message source may perceive that Donald is not worth the trouble and may not address any objective tasks, short- or long-term. There is also the option of removing Donald permanently from worry, by choosing to fire him. In summary, this task provides the message source with a complex, regulative situation with multiple competing goals which can be addressed. The complexity of this situation stems from the message source's role as manager the inconsistent information given about the behavior of the message target, and the various long and short-term goals which can be addressed in this situation. Situational Themes Based on this situation, there. are three situational themes which message sources can address in the messages they produce for this task: (1) "how do individuals respond to the message target's request of having the day off," (2) "how do individuals address the message target’s current excuse regarding problems," and (3) "how do individuals address consistent patterns of the message target’s missing work. These themes were generated by examining the situation itself and. potential responses based. on. a sample of responses generated by subjects in this study. 30 Theme One The first theme concerns how message sources address the immediate objective task at hand .of confronting Donald’s request for the day off. This paper identifies a priori five potential ways this request could be handled: (1) to not address this request at all, (2) to not address the request by firing Donald, (3) to demand that Donald has no choice but to come into work or else (i.e., using authority over Donald to get him into work), (4) to allow Donald to get away with not coming in for this-occasion, but attaching a contingent threat to the concession, and (5) allowing Donald to have the day off. When the context of the situation is considered, the message design logic model suggests that. individuals who adhere to'different message design logics should.orient to the immediate task of addressing the message target’s request differently. In other words, individuals who adhere to different message design logics should produce messages which vary in how this objective is dealt with. Specifically, because expressives tend to produce messages which focus on past transgressions instead of attempting to get the task accomplished and their principle premise of communication is expression of affective feelings and emotions (O'Keefe, 1991) , one would expect that expressives will be more likely to not address the message target's request of having the day off or 31 choosing to fire him. Hypothesis la is posed: Hypothesis 1a: Expressives, versus conventionals or rhetoricals, will be more likely to not address the immediate objective or to fire the message source. Conventionals on the other hand tend to produce messages which focus a great deal on directing some action of the message target. Rights and obligations of the situation will often be given as reasons why the message target should perfonm as required (O’Keefe, 1991). Thus, conventionals, versus expressives or rhetoricals, will be more likely to demand that Don has no choice but to come into work based on his responsibilities and obligations. Hypothesis 1b is posed: Hypothesis 1b: Conventionals, versus expressives or rhetoricals, will be more likely to demand that the message source has no choice but to come into work or he will be punished (e.g., fired or placed on probation). Rhetoricals tend. to jproduce messages with the main purpose of creating a consensus between message sender and receiver (O’Keefe, 1991). Thus, rhetoricals will be more likely to allow Don to get away with not coming in for the occasion by attaching a contingent threat to the concession or indicating a complete willingness to allow'Don to have the day 32 off. Hypothesis 1c is posed: Hypothesis 1c: Rhetoricals will be more likely to allow Don to have the day off versus expressives or conventionals. Because the interethnic context is argued to be more complex, the variation between message design logics should be increased. Hypothesis 1d is proposed: Hypothesis 1d: The variation in the proposed hypotheses (i.e., hypotheses 1a—1c) for theme one between expressives, conventionals, and rhetoricals should be more amplified in the interethnic context than the intraethnic context. Theme 2 The second theme concerns how message sources address Donald’s current excuse regarding his family problems. This research identifies a priori four possibilities of how this issue could be addressed: (1) do not address current excuse; (2) accuse him of making it up; (3) acknowledge legitimacy of current problems; and (4) acknowledge and attempt to address/resolve problems. Although prior message design logic research has not assessed. how' individuals 'may' address excuses of 'message 33 targets, this model does suggests that individuals who adhere to different message design logics will address it differenthn Specifically; O'Keefe and. Lambert’s (1989) research focusing on the impact that message design logics has on the communication of intention suggests that expressives, conventionals and rhetoricals disagree on the thoughts about the message target's characterx 'These thoughts or perceptions are related to whether or not individuals believe the message target is being honest about particular accounts about a story (O’Keefe & Lambert, 1989). In the broken date scenario (see O’Keefe & Lambert, 1989), expressives tended to have negatively' valanced. judgments about Terry’s (i.e., the message target in this situation) conduct, express extreme annoyance, and express wishes that Terry be punished. This should influence how expressive message design logics address the message target’s excuse regarding family problems. Hypothesis 2a is proposed: Hypothesis 2a: Expressives will be more likely than conventionals or rhetoricals to not address the current excuse or to accuse the target of making it up. O’Keefe and Lambert (1989) found that conventionals and rhetoricals tended to express confidence in the message target's good faith and accept his or her account of the situation (see O'Keefe & Lambert, 1989). Thus, individuals 34 who adhere to these two message design logics will be more likely to acknowledge the legitimacy of the message target’s excuse in the current situation” IHowever, rhetoricals tend to stress the importance of the relationship between themselves and the message target and.possess thoughts which focus on the importance of understanding, the suppression of hurt feelings, and recognize alternative ways to maintain the relationship more than conventionals (O’Keefe & lembert, 1989). Hence, these two message design logics should address the excuse differently . Hypotheses 2b and 2c are proposed. Hypothesis 2b: Conventionals will be more likely, than expressives or rhetoricals, to acknowledge the legitimacy of the message source’s current problem. Hypothesis 2c: Rhetoricals will be more likely, than expressives or conventionals, to acknowledge and attempt to address/resolve current problems of the message target. Because interethnic contexts are argued. to Ibe more complex, this variation should also be more amplified in this theme between the three message design logics. The variation in the proposed hypothesis for theme 2 between expressives, conventionals, and rhetoricals should be more amplified in the interethnic context than the intraethnic context. Hypotheses 2d is proposed: 35 Hypothesis 2d: The variation in the proposed hypothesis for theme 2 (i.e., hypotheses 2a - 2c) between expressives, conventionals, and rhetoricals should be more amplified in the interethnic context than the intraethnic. Theme 3 Theme three focuses on how message sources address the message target’s pattern of missing work. This paper identifies a priori three potential issues which message sources can address through. messages: (1) sources could express subjective inconvenience and irritation caused by Don's behavior (e.g., casting absences and excuses as a problem for me), (2) sources could express objective inconvenience and irritation caused by Don’s behavior (i.e., how is this affecting the company's services, other workers, etc.,.) and, (3) messages attempting to generate solutions to work through or solve Don’s personal problems (proactive measurements) i.e., do the participants probe Don, suggest he work fewer hours until his problems are worked out or that he receive counseling? Consistent with the message design logic model and the current study's previous hypotheses, individuals who adhere to different message design logics should also produce varying messages which address this issue. Specifically, because the 36 premise of expressive messages is expression of affective feelings and emotions (O’Reefe, 1988) with a focus on how the message target’s behavior affects and annoys them (O’Keefe & Lambert, 1989), it follows that. messages produced by expressive message design logics should contain properties which reflect subjective inconvenience and irritation. Hypothesis 3a is proposed. Hypothesis 3a: Expressives, versus conventionals or rhetoricals, are more likely to produce messages which contain properties of subjective inconvenience and irritation caused by the message target's behavior. Conventionals tend to have thoughts which reflect a subjective preoccupation with how the message target’s behavior effects them (O’Keefe & Lambert, 1989). However, unlike expressives, a conventional message tends to contain message properties which stress the situational rights and obligations of the message target. Thus, instead of simply focusing on subjective issues, conventionals are more likely to focus on issues which are situationally relevant or task- related. Hypothesis 3b is proposed: 37 Hypothesis 3b: Conventionals, versus expressives or rhetoricals, are more likely'toqproduce messages which contain properties of objective inconvenience and irritation caused by the message target's behavior (i.e., how'is this affecting the company’s services, other workers, etc.,.). Rhetoricals tend to have thoughts which focus on the importance of relationships and.maintaining the relationships (O’Keefe & Lambert, 1989). Rhetoricals are also more inclined to produce messages which contain motivational appeals, and a flexible, negotiative stance with the sole purpose of creating a consensus between message sender and receiver. Thus, rhetoricals will focus on the issue of missing work by focusing less on the Tproblem" aspect of the circumstance, and more on generating solutions to the problem by enlisting the assistance of the message target. Hypothesis 3c is proposed. Hypothesis 3c: Rhetoricals, versus expressives or conventionals, are more likely to produce messages which attempt to generate solutions to work through or solve the message target's personal problem. Because interethnic communication contexts are more complex than intraethnic communication contexts, these relationships should vary more'when.the context is interethnic versus intraethnic. Hypothesis 3d is proposed. 38 The proposed hypotheses for theme 3 (i.e., hypotheses 3a - 3c) between expressives, conventionals, and rhetoricals should be more amplified in the interethnic context versus intraethnic. Method Research Participants Participants (n=197) were recruited from communication classes at Michigan State University and received course extra-credit for participating. The participants consisted of 113 women and 84 men. Participants ranged from 18-31 years old.(M=20.31, SD=5.4S). The participants were composed.of 160 Euroamericans, 27 African Americans, 2 Hispanics, 7 Asian Americans, and 1 participant who specified the "other" category. The research project was approved by the Human Subjects Committee, Michigan State University (1993). Research Design This study formed a 3 x.2 research design, with three levels of message design logics (expressive, conventional, rhetorical) crossed with two types of goal complexity (black or white message target). To elicit the participant’s message design logic, participants completed the standard form of the Group Leader problem in which their messages were classified as being expressive, conventional, or'rhetorical” The message target's ethnicity was manipulated by presenting respondents 39 40 with a photo of either a black employee or white employee in a.hypothetical communication.situation. ZUJ.other information was identical. Procedures Participants were told that the purpose of the study was to investigate "the types of messages people employ when presented with multiple amounts of information. " Participants were confronted with the standard group leader scenario, which is a regulative communication task. They were instructed to write the actual words they would use in addressing the message target and not just describe the general actions they would take. Messages elicited by this problem were classified as reflecting either expressive, conventional, or rhetorical message design logic. There were 71 (36%)expressives, 90 (45.7%) conventionals, and 36 (18.3%) rhetoricals (see O’Keefe, 1988, for a description of the coding system). Intercoder reliability was computed using Cohen’s (1960) _K_app_a_ (.86). This was calculated based on two coders, working independently, across 30 messages. The participants were then asked to complete the Manager Situation. This situation, developed by the author for this study, offered. the participant a ‘more complex task for participants to address. This task manipulated ethnicity by attaching an "Employee Information Sheet" with a picture of 41 either an African American male or an Euroamerican male. 'The two cover sheets were randomly given to participants with identical information on each. Thus, the Manager Situation had two message target conditions. Condition one presented an Euroamerican message target (n=99; 50.3%); condition two an African American message target (n=98; 49.7%). Participants were given the same instructions received in the Group Leader problem. .After completing both tasks, participants completed a brief demographic questionnaire and were debriefed (see appendix A for the task given to participants). Independent Variables Message Design Logics The message design logic of an individual was elicited using the standard Group leader scenario. By providing the actual words they would use to address this situation, the participants' messages were classified as reflecting an expressive, conventional, or rhetorical message design logic. Increased goal complexity The second independent variable was increased goal complexity. The problem used for this situation is the 42 "Manager problem." Compared to the standard Group leader problem, the manager problenihas increased.goal complexity for several reasons: (1) the ethnicity of the message target, (2) the increased.responsibility'of the message source’s role, (3) the multiple objections that could be potentially addressed in the task, and (4) the mixed messages regarding the message target’s prior behavior. There are two levels of goal complexity based on the ethnicity of the message target and message source: the interethnic context and the intraethnic context. Although, both of these contexts are argued to be more complex than the tasks used in previous message design logic research, the interethnic context is argued to be more complex than the intraethnic context. Dependent Variables Message Features The first goal of the current study was to assess how individuals who adhere to different message design logics address the features of a complex situation. The hypotheses in the current study suggest that expressives, conventionals, and rhetoricals should systematically vary in how they address the features of the manager situation. The second goal of this research was to investigate if ethnicity of message target influenced message production and if these differences 43 varied between expressives, conventionals, and rhetoricals. Messages were coded in terms of the "message themes" present in individual responses. The three message themes analyzed within messages generated by respondents were: (1) how do message sources address the objective task at hand of responding to the message target's (i.e., Donald) request of having the day off; (2) hOW’dO they address the current excuse of having a family emergency; and (3) how do sources address the message target's previous pattern of missed work. To code for the presence of these message themes, messages were read three times each. The first reading examined the first theme, and gave the message source one score, ranging from 0-5 (0=other, 1=do not address the objective task, 2=fire Don, 3=demand that Don has no choice but to come into work or else, 4=allowing Don to get away with not coming in for this occasion, but attaching a contingent threat to the concession, and 5=messages which indicate a willingness to allow Don to have the day off). Intercoder agreement for theme one was 97%, resulting in.a Cohen’s (1960) Kappa = .96. The messages were read a second time, focusing on theme two~ iMessages were given a second score ranging from 0- 4 (0=other, 1=don’t address current excuses, 2=accuse him of making it up, 3=acknowledge legitimacy of current problems, and 4=acknowledge and attempt to address/resolve problems). Intercoder agreement for theme two was 90%, resulting in a Cohen's (1960) Kappa = .85. Messages were read a third time 44 examining theme 3. Messages were coded for the presence of (1) expressions of subjective inconvenience and irritation caused. by' Don’s behavior (intercoder' agreement was 93%, resulting in a Cohen's (1960) Kappa = .75) (2) expressions of objective inconvenience and irritation caused by Don's behavior (Intercoder agreement was 97%, resulting in a Cohen's (1960) Kappa = .90), and (3) messages attempting to generate solutions to work through or solve Don’s personal problems (Intercoder agreement was 100%, resulting in a Cohen’s (1960) Kappa = 1.00). Participants were given a score of 0 for no presence of any of these message properties and 1 if properties of this issue were present in the message. Intercoder reliabilities for each category were computed by two coders working independently across 30 messages (See appendix three for examples of messages). Analyses To test this study’s hypotheses, a series of crosstabs tables were employed to analyze the message features produced by individuals who were classified as having different message design logics and to investigate if they were significantly different from each other. Additional analyses were also employed to investigate if there was more variation between message design logics in the interethnic versus intraethnic context . Descriptive analyses Table 1 presents the frequencies and percentages for themes 1, i.e., how do message sources respond to the message target’s request of having the day off. 46 Table 1 Frequencies and percentages for Theme 1 Category Frequency Percent 0 6 ‘ 3.0 1 20 10.2 2 32 16.2 3 73 37.1 4 45 22.8 5 21 10.7 Mean=2.985 SD=1.227 47 Table 2 presents the frequencies and percentages for theme 2, i.e., how do the message sources respond to the message target's excuse of family problems. Table 2 Frequencies and percentages for Theme 2 Category Frequency Percent O 5 2.5 1 81 41.1 2 28 14.2 3 66 33.5 4 17 8.6 Mean=2.046 SD=1.094 48 Table 3 presents the frequencies and percentages of the number of messages that expressed subjective inconvenience for the message target’s behavior. Table 3 Frequencies and percentages for subjective inconvenience Category Frequency Percent O 157 79.7 1 40 20.3 Mean =.203 SD =.403 *note, 0 = not-present, 1 = present 49 Table 4 presents the frequencies and percentages of the number of messages that expressed objective inconvenience for the message target’s behavior. Table 4 Frequencies and percentages for objective inconvenience Category Frequency Percent 0 167 84.8 1 30 15.2 Mean = .152 SD = .360 *note, 0 = not-present, 1 = present Table 5 presents the frequencies and percentages of the number of messages that generated solutions for the message target's problems. Table 5 Frequencies and percentages for generating solutions Category Frequency Percent 0 181 91.9 1 16 8.1 Mean = .081 SD = .274 *note, 0 = ton - present, 1 = present 50 In theme 3, the category "messages generating solutions" was dropped from the analysis because less than ten.percent of the respondents in the current study produced messages which addressed this issue. Results Theme 1: "How'do message sources addreSs the objective task at hand?" Overall, for theme one, there were significant differences in how individuals, who adhere to different message design logics, address the message target’s request of having the day off (chisq=24.94, df=10, p<.01). Hypothesis 1a Hypothesis 1a proposed that expressives wouLd be more likely not to address the message target's request to have the day off or to fire the message target. The data supported this hypothesis. In comparison to the other design logics, expressives were more likely not to address this task (12, 60%) versus conventionals (6, 30%) or rhetoricals (2, 10%). Expressives were also more likely to fire the message target (16, 50%) versus conventionals (9,28%) or rhetoricals (7, 21.9%). 51 52 Hypothesis 1b Hypothesis 1b proposed that conventionals would be more likely to demand that the message target come into work or else. The data supported this hypothesis. Conventionals were more likely to demand that the message target come into work by attaching a contingent threat (41, 56.2%) versus ). o\° expressives (24, 32.9%) or rhetoricals (8, 11 Hypothesis 1c Hypothesis 1c proposed that rhetoricals would be more likely to allow the message target to have the day off with or without a contingent threat” The data partially supported the hypothesis. Conventionals were actually more likely to allow the message target to have the day off with a contingent threat (24, 53.3%) versus rhetoricals (8, 17.8%) or expressives (13, 28.9%). In fact, rhetoricals ranked on the bottom for this particular category. However, as predicted, rhetoricals were more likely to allow the message target to have the day off without any threat to the message target (9, 42.9%) versus conventionals (8, 38%) or expressives (4, 19%). Table 6 summarizes the findings of hypothesis 1a-1c. 53 Table 6 Theme One Count Col Pct Row Pct Categories Expressive Conventional Rhetorical 1 Do not 12 6 2 address 16.9 6.7 5.6 60.0 30.0 10.0 2 Fire Don 16 9 7 22.5 10.0 19.4 50.0 28.1 21.9 3 Demand come 24 4O 8 in or else 33.8 45.6 22.2 32.9 56.2 11.0 4 give day off 13 24 8 w/threat 18.3 26.7 22.2 28.9 53.3 17.8 5 willingly give 4 8 9 day off 5.6 8.9 25.0 19.0 38.1 42.9 0 other 2 2 2 2.8 2.2 2.6 33 3 33.3 33 3 Chisq=24.94, df=10, p<.01 54 Hypothesis 1d Hypothesis 1d argued that the variation in the proposed hypotheses for theme one between expressives, conventionals, and rhetoricals should be more amplified in the interethnic context versus the intraethnic context. To examine this hypothesis, only responses from Euroamerican participants were used because they were the largest ethnic group. If the message target was Euroamerican, then the interaction was considered intra-ethnic. If the message target was African American, then the interaction was considered interethnic. The "select if" (SPSS, version 4.0) command was employed to create the intraethnic or interethnic context. In the intra-ethnic context (i.e., Euroamerican message source, Euroamerican message target), there were very significant differences between expressives, conventionals, and rhetoricals in how they responded to this issue (chisq=27.13, df=10, p<.01). In the interethnic context (Euroamerican message source, African American message target) there were no significant differences between individuals who adhere to different message design logics in how they responded to this issue (chisq= 5.14, df=10, p>.05). Overall, the data suggest significant differences in the messages that participants produced in the intra-ethnic 55 context across the different message design logics (chisq=27.13, df=10, p<.01). However, there were no significant differences in the messages that participants produced in the inter-ethnic context across the different message design logics (Chisq=5.4, df=10, p>.05). These results suggest that the ethnicity of the message target influenced the participants representations of this particular situation which in turn, lead to different types of messages produced. Specifically, in the intra-ethnic situation, expressives, conventionals, and rhetoricals responded in.ai manner that is consistent to the hypotheses. In the interethnic context, there is less of a clustering around the hypothesized categories, and more of a lax dispersion of the features addressed in the situation based on an individual’s message design logic. In the intra-ethnic context, expressives produced 80% of the messages "which do not address the task," and 55.6% of the messages which "fired Don." In the interethnic context, expressives produced only 50% of the messages which did not address the task and 46.7% of the messages which "fired Don." What is interesting about the messages expressives produced regards the category "of demanding Don to come in or else." While 40.7% of the total messages produced by expressives in the intra-ethnic context fell into this category of "demanding Don to come in or else, only 29% of these messages were produced 56 by' expressives in time interethnic context. Thus, there appears to be a consistent pattern of messages produced by expressives in.the intra-ethnic context versus the interethnic context. For example, while there 'is a large portion of expressive message sources firing Don in the interethnic context (22.6%) versus in the intra-ethnic context (18.5%), more expressives have willingly allowed Don to have the day off in the interethnic context (9.7%) than in the intra-ethnic context (0.0%). In short, there is more variance in the interethnic context than in the intra-ethnic context for expressive message design logics. For conventionals, the pattern is similar: There is more variance in the interethnic context than in the intra-ethnic context. In the intra-ethnic context, conventionals produced messages which are consistent with the current hypothesized relationships. In the intra-ethnic context, 60.5% of the total messages produced by conventionals "demanded that Don come into work or else," and 23.7% of the messages gave Don the day off with a threat. However, in the interethnic context, only 3.14% of the total messages produced by conventionals "demanded he come into work or else and 28.6% gave him the day off with a threat. However, there was more variation in the messages produced by conventionals in the interethnic context than in the intra-ethnic. For example a larger portion of conventionals fired Don (14.3%) than in the intra-ethnic 57 context (5.3%). On the other hand, a larger amount of conventionals willingly gave Don a day off (14.3%) in the interethnic context than in the intra-ethnic (7.9%). The context of the situation (i.e., interethnic versus intra-ethnic) appears to impact the way rhetoricals respond to the task of addressing Don’s request of having the day off - they tend to "not address the task" and "fire Don" more in the interethnic context than the intra-ethnic context. 2n: the interethnic context, 14.3% of the total messages produced by expressives did not address the task while 0% did not address the task in the intra—ethnic context. In the interethnic context, 21% of the total messages produced by rhetoricals fired Don in the interethnic context versus 13.3% in the intra-ethnic context. Thus, the implications of these findings is that the complexity of the situation brings in more individual variance and causes a disorder in the data instead of the clear distinctions that was hypothesized. Thus, in the interethnic context, there is an overall variance versus and increased variance between message design logics. See Table 7 and 8 for a summary of these results. Theme One; Count Col Pct Row Pct Categories 1 Do not address 2 Fire Don 3 Demand come in or else 4 give day off w/threat 5 willingly give day off 0 other Expressive 11 4 40. 28. 25. 36. GOO GOO CDKO 58 Table 7 Conventional 23 60.5 60.5 Intra-ethnic context Rhetorical 000 00 13.3 22.2 26.7 10.5 20.0 15.8 26.7 57.1 13.3 100.0 59 Table 8 Theme One; Interethnic context Count Col Pct Row Pct Categories Expressive Conventional 1 Do not 5 1 address 16.1 8.6 50.0 30.0 2 Fire Don 7 5 22.6 14.3 46.7 33 3 3 Demand come 9 11 in or else 29.0 31.4 39.1 47.8 4 give day off 5 10 w/threat 16.1 28.6 27.8 55.6 5 willingly give 3 5 day off 9.7 14.3 27 3 45.5 0 other 2 1 6 5 2.9 Rhetorical 2 14.3 20.0 21. 20. Orb 21.4 13.0 21.4 16.7 21.4 27.3 000 CO 60 Additional analyses: Hypothesis Id In analyzing theme one to examine how the context of the situation (i.e., intra-ethnic versus interethnic) there were concerns with the low sample sizes located in some of the cells. To analyze the differences between intra-ethnic contexts and interethnic contexts, the sample was split in half} Moreover, .Asian. Americans, Native .Americans, and Hispanics were not used in this analysis. Thus, there were a number of cells which were zero and several which contained three or less participants. Thus, even though the data suggest there were significant difference in the messages between message design logics produced in the intra-ethnic context across the different message design logics (chisq=27.13, df=10, p<.01) and no significant differences in the messages between message design logics produced in the interethnic context (Chisq=5.4, df=10, p>.05) these small cell sizes may have inflated the chi-squares. Thus, the data was reduced and re-analyzed. For theme 1, "do not address the task" and "fire Don" were collapsed into one category and "give day off with threat" and "willingly give day off" were collapsed.into one categoryu 'The "other" category was deleted from analyses because of the small percentages of individuals who produced messages that were not classified into categories 1, 2, 3, 4, or 5. 'With this data reduction, the analyses were 61 re-run and new chi-squares produced. The chi-square for the intra-ethnic context was chisq=10.17 (df=4, p<.05). For the interethnic context the chisq=3.267 (df=4, p>.05). These chi- squares are consistent with the firSt analysis in that the data suggest that there are significant differences in how individuals address the message target’s request for the day off in the intra-ethnic context but not in the interethnic context. Table 9 and 10 summarize these results. 62 Table 9 Additional Analyses Intra—ethnic context Count Col Pct Row Pct Categories Expressive Conventional Rhetorical 1 Do not 9 3 2 address 33.3 7.9 14.3 or fire 64.3 21.4 13.3 Don 2 Demand come 11 23 4 in or else 40.7 60.5 26.7 28.9 60.5 10.5 3 give day off 7 12 7 w/threat 25.9 31.6 46.7 or willingly 26.9 46.2 26.9 w/no threat Chisq=10.17, df=4, p<.05 63 Table 10 Additional Analyses Interethnic context Count Col Pct Row Pct Categories Expressive Conventional Rhetorical 1 Do not 12 3 2 address 38.7 7.9 14.3 or fire 48.0 32.0 20.0 Don 2 Demand come 9 11 3 in or else 31.0 32.4 21.4 39.1 47.8 13.0 3 give day off 8 15 5 w/threat 27.6 44.1 42.9 or willingly 27.6 51.7 20.7 w/no threat Chisq=3.267, df=4, p>.05 Theme two: "How do message sources address current excuses regarding family problems?" Overall, for theme 2, there were significant differences in how individuals, who adhere to different message design logics, address the current excuse of the message target (chisq=16.55, df=8, p<.05). 64 Hypothesis 2a Hypothesis 2a proposed that expressives were more likely than conventionals or rhetoricals to "not address the excuse" or to "accuse the message target of meking it tux" This hypothesis in not supported by the data. In comparison to the other message design logics, conventionals were more likely to not address the excuse (36, 44.4%) than expressives (32, 39.5%) or rhetoricals (13, 16%). Conventionals were also more likely to accuse the message target of making up the excuse (36, 44.4%) than expressives (12, 42.9%) or rhetoricals (3, 10.7%). Hypothesis 2b Hypothesis 2b proposed that conventionals were more likely to acknowledge the legitimacy of the message target’s current excuse versus rhetoricals and expressives. The data supported this hypothesis. In05). When the context is interethnic (i.e., Euroamerican message source, African American message target) there were significant differences between individuals who adhere to different message design logics in how they responded to this issue (chisq= 16.94, df= 8, p<.05). See Table 12 and 13 for a summary of these results. 68 Table 12 Theme Two:Intra-ethnic context Count Col Pct Row Pct Categories Expressive Conventional Rhetorical 1 Do not 14 20 5 address 51.9 52.6 33.3 35.9 51.3 12.8 2 Accuse of 6 5 1 making up 22.2 13.2 6.7 50.0 42.7 8.3 3 acknowledge 6 8 8 legitimacy 22.2 21.1 53.3 27.3 36.4 36.4 4 address/ 0 5 1 resolve 0.0 13.2 6 7 0.0 83 3 16 7 0 other 1 0 . 0 3.7 00.0 00 0 100.0 00.0 00.0 chisq= 12.73, df= 8, p>05 69 Table 13 Theme Two:Interethnic context Count Col Pct Row Pct Categories Expressive Conventional Rhetorical 1 Do not 14 9 5 address 45.2 25.7 35.7 50.0 32.1 17.9 2 Accuse of 5 5 2 making up 16.1 14.3 14.3 41.7 41.7 16.7 3 acknowledge 6 17 3 legitimacy 19.4 48.6 21.4 23.1 65.4 11.5 4 address/ 2. 4 4 resolve 6.5 11.4 28.6 20.0 40.0 40.0 0 other 4 0 0 12.9 00.0 00.0 100.0 00.0 00.0 chisq= 16.94, df= 8, p<.05 Overall, there appeared to be insignificant differences in how individuals, who adhere to different message design logics, addressed the message target’s excuse regarding family problems in ‘the intra-ethnic context (chisq=12.73, df=8, p>.05). However, there were significant differences in the interethnic context (chisq=16.94, df=8, p<.05). This is 7O interesting considering that for theme one (i.e., how do individuals address the message target's request for the day off), there were significant differences across the different message design logics in the intra-ethnic context, but not in the interethnic context. The data does suggests, however, that the difference between the intra-ethnic and interethnic contexts in theme 2 is not as drastic as the differences observed in theme 1. When examining the percentages across categories, there did.not appear to be any large deviations in how expressives addressed the excuse regarding family problems. The only difference is a small percentage of expressives in the interethnic context did attempt to "address/resolve the message target's problems" (6.5%) while none did so in the intra-ethnic context (0%). The data suggest that the differences are more pronounced in how conventionals address this issue between contexts. In, the intra-ethnic context, over half of the total amount of messages produced by conventionals did not address the excuse (52.6%). In the interethnic context, only 25.7% of the messages produced by conventionals did not address the task. This relationship is reversed when messages which “acknowledge the legitimacy of the excuse" are examined. While only 21.1% of the total messages produced by conventionals fall into this category in the intra-ethnic context, almost half of the total messages fall into this category in the interethnic context (48.6%). This suggests that conventionals may feel more 71 compelled to acknowledge the legitimacy of the message target's excuse when the context is interethnic versus intra— ethnic. The majority of messages produCed by rhetoricals in the intra-ethnic context acknowledged the legitimacy of the message target’s excuse (53.3%) compared to only 21.4% in the intra-ethnic context. However, a larger proportion of rhetoricals produced.messages in the interethnic context that attempted to "address/resolve the message target's problems (28.6%) versus 6.7% in the intra-ethnic. Additional Analyses: Theme Two Although the original analyses suggested that there were significant differences in how expressives, conventionals, and rhetoricals addressed the message target’s current excuse regarding family problems in the interethnic context, and not in the intra-ethnic context, the additional analysis suggests that there were no statistically significant differences in how expressives, conventionals, and rhetoricals addressed the current excuse regardless of the context. Thus, it appears as if the small cell sizes may have inflated the chi-square for the interethnic context. When the data was reduced and the zero-numbered cells were removed from. the analysis, the significance in the interethnic context no longer existed. 72 However, the differences in the first analysis in the distribution across contexts is still apparent. For example, conventionals are more likely to "not address the current excuse" if the interaction is intra-ethnic (20, 52.6%) versus interethnic (9, 25.7%) . However, conventionals are more likely to "acknowledge the legitimacy of the excuse" or "attempt to address or resolve the message target’s problems" in the interethnic context (21, 60%) versus in the intra—ethnic context (13, 34.2%) . Expressives and rhetoricals have similar distributions in how they address the excuse of family problems irrespective of context. Theme three: "How do message sources address consistent pattern of missing work" Hypothesis 3a Hypothesis 3a proposed that expressives would be more likely, than conventionals and rhetoricals, to produce messages which contain properties that reflect subjective inconvenience and irritation caused by the message target’s behavior (i.e., how is this affecting me). There were no significant differences in the number of messages expressives, conventionals, and rhetoricals produced which contain subjective properties related to this issue (chisq=4.95, df=2, p>.05). Thus, the data did not confirm hypothesis 3a. 73 However the data is patterned in the hypothesized direction, but because only 20% of the total sample produced messages which contained subjective properties regarding the message target’s pattern of missed work, the cell sizes may have been too small to produce a statistically powerful chi-square. However, expressives were more likely than conventionals to produce messages which contained these message properties (20, 50%) than conventionals (16, 40%) or rhetoricals (4, 10%). Table 14 summarizes these results. 74 Table 14 Expression of subjective inconvenience Count Col Pct Row Pct Categories Expressive Conventional Rhetorical 0 subjective 51 74 32 message 71.8 82.2 88.9 properties 32.5 47.1 20.4 not present 1 subjective 20 16 4 message 28.2 17.8 11.1 properties 50.0 40.0 10.0 present Chisq=4.95, df=2, p>.05 Hypothesis 3b Hypothesis 3b proposed that conventionals would be more likely to produce messages which contain properties of objective inconvenience and irritation caused by the message target’s behavior (i.e., how is this affecting the company’s services, other workers, etc.,.). There were no significant differences in. the :number' of 'messages expressives, conventionals, and rhetoricals produced which contain objective properties related to this issue (chisq=3.49, df=2, p>.05). This is not surprising considering that only 15% of the total sample produced messages with these message properties. Table 15 summarizes these results. 75 Table 15 Expressions of objective inconvenience and irritation Count Col Pct Row Pct Categories Expressive Conventional Rhetorical 0 objective 56 78 33 message 78.9 86.7 91.7 properties 33.5 46.7 19.8 not present 1 objective. 15 12 3 message 21.1 13.3 8.3 properties 33.5 46.7 19.8 present Chisq=3.49, df=2, p>.05 Hypothesis 3c Hypothesis 3c proposed that rhetoricals would be more likely to produce message properties which attempt to generate solutions to work through or solve the message target's personal problems. Because less than 10% of the total sample produced messages which contained this feature, this category was dropped from the analysis. Overall, there were no significant differences in how individuals, who adhere to different message design logics, addressed the consistent pattern of missing work. However, this issue was difficult to assess because there was such a small amount of messages produced in the total sample which 76 addressed these issues. Due to this small percentage, hypothesis 3d was not analyzed. Discussion The primary purpose of this research was to investigate the impacts of task complexity on how individuals, who adhere to different message design logics, address the communication situation” Compared to the tasks used.in prior message design logic research, message complexity was increased by presenting the ethnicity of the message target (i.e., either African American or Euroamerican), by increasing the difficulty'of the message source role, by giving contradictory information about the message target, and by increasing the objectives of the task. Based on O’Keefe's (1988) assertion that as goal complexity increases, variation between individuals who adhere to different message design logics should increase, we expected more variation in the interethnic communication context versus the intraethnic context. By given participants a complex situation to address, their messages were analyzed in terms of how they addressed the message target’s request of having the day off, how they addressed the message target’s current excuse of having a family' emergency, and. how sources addressed the message target's previous pattern of missed work. Thus, this paper argued that expressives, conventionals, and rhetoricals should address these issues differently and that their differences 77 78 should be amplified in the interethnic versus the intraethnic context. The findings of this research were intriguing. When examining the various themes, the intraethnic and interethnic context were collapsed to examine specifically if there were differences in how expressives, conventionals, and rhetoricals addressed the issues in the communication task. The results of this research. provided. partial support for' O’Keefe’s assertion (1988). When examining how individuals addressed the message target’s request for the day'off, the data suggest that there were significant differences in how the different message design logics. addressed the situation. Comparatively, expressives tended to not address the task or fire the message target, conventionals focused on getting the message target to work by employing a contingent threat or allowing the message target the day off with a contingent threat, and rhetoricals tended to produce messages which willingly gave the message target the day off. However, though individuals of different message design logics tended to cluster around their hypothesized categories, the variation in the messages produced by different message design logics was not as great as would be expected” .Although "increased variation" between message design logics is an ambiguous statement, these data suggest that the distinctions between the message design logics do overlap. For example, 45.6% of all messages produced by conventionals demanded the message target come in 79 or else. However, 33.8% of all messages produced by expressives fell into this category as for 22.2% of all message produced by rhetoricals. Thus, this research suggest that individuals who.adhere tOidifferent message design logics saw similar aspects of this situation similarly, as in the case of how individuals responded to the message target's request. Yet, when individuals adhering to different message design logics do diverge, they deviate in different directions. These results were similar in how individuals addressed the current excuse regarding family problems. While the data suggest that individuals who adhere to different message design logics view many of the situational features similarly, they tend to diverge differently. Expressives, conventionals, and rhetoricals all tended to acknowledge the legitimacy of the target's excuse. .A large portion.of the messages produced by expressives, conventionals, and rhetoricals also did not address the task at all. This finding could be influenced by several factors. One is that the excuse was taken by message sources as being truthful, therefore there was no need to 'mention it. It could also be the case that the excuse was perceived as irrelevant and unimportant to the task of getting the message target to come in. In future research, questionnaire data should be given exploring what thoughts influence the message targets responses (i.e. , do expressives, conventionals, and rhetoricals have systematically different 80 thoughts even when they produce similar messages?). Theme three was concerned. with how' message sources address the consistent pattern of missing work. No significant differences were found in how individuals, who adhere to different message design logics, addressed this issue. The small percentage of the total sample producing these message features is probably partially responsible for this finding. When the context of the communication taskwwas taken into consideration, the findings became complex; It was hypothesized that the interethnic context was more complex than the intraethnic context. Therefore, we should see more message variation in this context between individuals who adhere to different message design logics. The findings of this research suggested that this was not the case. For theme one, there was more variation.between message design logics in how individuals addressed.the message target’s request for the day off in the intraethnic context versus the interethnic. V’This finding suggests that other variables such as personal attitudes toward members of different ethnic backgrounds becomes more salient in the interethnic context versus the intraethnic context.»4Thus, instead of individuals responding in.a manner that is expected for the message design logic they 3 ?) adhere to, they responded individually based on how they arejfjjyu‘ C influenced by the ethnicity'of the message target. Looking at it MAI/(M \m VII/VI) 0f 47kt \SSWAQCICLi'Cj/J; the dispersion of numbers in the interethnic context versus 24(ka (.19 u (3" k 0x. .9 m, hDL 81 the intraethnic context, it is apparent that individuals overall are more varied in their responses, irrespective of. message design logic. Future research should examine the impact.of«other important characteristics that individuals use to construct messages such as age, sex, perceived status,:/‘"WC/ attitudes of liking, and so forth. These findings suggest that biases, attitudes, and stereotypes may have mo . influence on how individuals construct the representations of 1392y££ a situation which lead to message design and production than an overriding message design logic. Overall, the results of this study suggest that an individual’s message design logic has some influence on how individuals address features of a situation. The variation among individuals who adhere to the same message design logic of course suggests that there are other variables and factors that need to be incorporated into this model of understanding how thoughts about a communication situation lead to message design. All individuals in this world do not fit into one of three message design logics. Factors such as culture, social class, mental health, upbringing, level. of communication “Li apprehension, as well as the situation itself are examples ofxgvAB MA or’ 0N1i3;, 3 construct representations of situations which lead totxiuhi relevant and important factors which influence how we Jr ' 1' expression. Thus, research in this area could benefit gre tlyd‘ ) *2“ by incorporating these issues into the message design loFic “03“" to model. The current research has made an attempt “)1" 82 investigate and understand how particular factors influence individuals who have been classified as expressive, conventional, and rhetorical. Ebrincreasing the complexity of the situation, when compared to other tasks frequently used in this field of research, this study attempted to see how each of the different message design logics perceived the issues in this study by analyzing the messages produced. Also, because q3y0}features of our communication interactants are rarely unimportant for how individuals design messages, this study incorporated ethnicity and issues of intraethnic communication versus interethnic communication to the message design logic approach. These issues are becoming increasingly more important to examine in.our society due to an increased.multi— cultural environment as well as with the preponderance of ethnic conflict around the world. ..Efiu§ (NJJ‘104 gigzxjttuzvi 9"“) v M] , , ,~ mum! However, there are several 11m1tations to the current+,guuj research, as well as the general body of message design logicl\::i“”‘ research to-date, concerning the issue of methodology. This ¢¥N¢V‘ study used hypothetical situations to investigate, predict, and explain complex communication processes. This method gives researchers in this area an indication of how an individual’s message design logic may come into play in that particular situation. What it does not explore is how individuals who are classified as expressives, conventionals, or rhetoricals interact in face-to-face situations or how the interaction in difficult communication contexts evolves as 83 x ,1 Kg) talk and nonverbals are exchanged between interactants. What types of outcomes occur (i.e., perceived degree of success, satisfaction) when two expressives communicate, an expressive and a rhetorical, and so forth, is also an issue to investigate. -How these issues relate to conflict resolution and management, relational development, uncertainty reduction, and other theoretical perspectives is also an intriguing research endeavor. The message design logic model would suggests that individual's adhering' to «different ‘message ‘design logics should be influenced by these processes but to what magnitude is unknown. In summary, the results of this research suggest that message design logics is an important model of communication which would also make an interesting framework for understanding other issues and theoretical perspectives in communication. More research should. be conducted to investigate what situations or circumstances encourage message variation between different message design logics and which suppress it. This line of research would also benefit by incorporating more interactional and face-to-face simulations (i.e., moving away from relying primarily on paper and.pencil hypothetical situations), linking culture, social class, education, and other theoretical perspectives in communication which influence our representations of situations and.message design. Appendix A Manager Situation Imagine you are a manager in a catering service. Your job is to oversee the preparation of food, its transfer to the sight of the event, and it's service to the guests. An important part of your job is to make sure that the employees in your department are on time and show up for their shift. When you are preparing the schedule, you carefully calculate how much time you need in preparation for a party and the amount of people needed for each function. Your duties also include keeping an up-to-date report on each of the employees in your department, writing out evaluations, and hiring and firing employees. Donald, an employee in your department who had previously been a good worker, has been giving you problems recently (Donald is the employee pictured on the previous page). He’s been late several times, called in twice in.a month and.didn't show up one shift without calling in to tell you. The next day, Donald apologized for not showing up and mentioned something about his car breaking down. He volunteered to take over another person’s shift to make up for it. Today, your company is catering a very important function and scheduled more employees than usual to make sure this event runs smoothly, Donald calls and tells you that he’s not going to be able to work because an emergency with his family has come up. He apologizes and says that he’ll work a double shift next week. You really needed Donald to come in today. What would you say to Donald? 84 Appendix B Manager Situation Coding Scheme: Theme one: "How do they address the task of responding to the message target’s request of having the day off?" (i.e., getting Don into work today) Mutually exclusive categories 1. Do not address the objective task. 2. Fire Don. 3. Demand that Don has no choice but to come into work or else (i.e., is the focus on the present with little opportunity for negotiation, instead using authority' as manager' over‘ Don, the employee, to get him to do what you want?). 4. Allowing Don to get away with not coming in for this occasion, but attaching a contingent threat to the concession (e.g., vDon will be terminated if he screws up again). 5. Messages which indicate a willingness to allow Don to have the day off (taking a negotiative or flexible stance regarding Don’s situation, or specifically his request for the day off). Score for theme one 86 Theme two: "How do they address current excuses regarding family problems?" Mutually exclusive 1. Don't address current excuses. 2. Accuse him of making it up. 3. Acknowledge legitimacy of current problems. 4. Acknowledge and attempt to address/resolve problems. Score for theme two Theme three: "How do they address consistent pattern of missing work" Independent message features 1. Expression of subjective inconvenience and irritation caused by Don’s behavior (e.g., casting absences and excuses as a problem for me). 2. Expression. of objective inconvenience and irritation caused by Don's behavior (i.e., how is this affecting the company’s services, other workers, etc.,.) Are his absences casts as a violation-of rules and obligation? .3. Messages attempting to generate solutions to work through or solve Don's personal problems (proactive measurements) . Do the participants probe Don, suggest he work less hours until his problems are worked out or that he receive counseling? Appendix C Examples of messages for theme one and two (message produced by participants in this study) Theme one: "How do they address the task of responding to the message target’s request of having the day off?" (i.e., getting Don into work today) Mutually exclusive categories 1. Do not address the objective task. (1) Well Donald, you have been late and came up with every excuse possible. I’m sorry to say if the problem continues, I will have to let you go. (2) I don’t believe that I would fire him, but I would make sure he knows all he is doing is being documented if he ever wants another job. I would also knock his pay down maybe 5 cents an hour to give him incentives to be on time. 2. Fire Don. (1) It would be different if you never missed a day of work but with numerous missed days and tardiness it is hard to believe you at this time. I will have to let you go to find someone more dependable. (2) This seems to be a reoccurring situation. I can understand missing a few days but this is getting ridiculous. I really needed you today and you knew it was important. I am afraid I’m going to have to fire you. I'm sorry but it is too important for you to miss any more time. Possibly when you get things straightened out you can apply later on. I'm sorry. You can pick up your salary at the end of the week. 87 88 3. Demand that Don has no choice but to come into work or else (i.e., is the focus on the present with little opportunity for negotiation, instead using authority as manager over Don, the employee, to get him to do what you want?). (1) Look Donald, we’re really in a tight situation here. This is a big event and I need everyone I can get my hands on. I’ve noticed you’ve been late several times, and you’ve missed a few shifts. Are you having serious family problems? If so, and it's going to keep effecting your work, I suggest you resign. I know you're a good worker, but I really need someone who’s going to be here. So, you can either come in today and keep your job, or you can resign. I’m sorry Donald, but I have to do this. (2) I would say that if he couldn't make it in today that he would be fired. This is a business and we need people that can be depended on. The choice is up to him. 4. Allowing Don to get away with not coming in for this occasion, but attaching a contingent threat to the concession (e.g., Don will be terminated if he screws up again). (1) Donald, I really need for you to fill your shift today. In the last few months you have seemed to have a reoccurrence of family problems, I would suggest to you that you find a resolution to them. I realize you are generally a good employee, but you’ve recently fallen of track. I would like for ‘you to bring in some type of documentation stating the problem which is keeping you from fulfilling your obligations to this company. As our company policy must be upheld, I must inform you that if you are absent or late again I will have to let you go, but for now I recommend that you call your fellow employees and get someone to cover your shift. 5. 89 (2) Donald, you came into this job because you needed one. I expect you to devote yourself to this and be a committed person. If you can’t make it on time or not at all, then why did you ever bother applying for a job here? You can have the day off, but if you don’t Work the double shift next week, you will have to be terminated. Messages which indicate a willingness to allOW'Don to have the day off (taking a negotiative or flexible stance regarding Don’s situation, or specifically his request for the day off). (1) Don, I really need you to come in today, at least to help set up? Well Don, I want to see you tomorrow. Your absences are hurting the ability 6 this department to function. Now, I don’t want to know what your family problems are, unless you want to share that information with me. We need to discuss how we're going to get you to work. If that means finding you some counseling help, or other types of assistance, we’ll discuss it. But I can’t run this service without the participation of one of my better workers. (2) Donald I am willing to listen to your situation you know that, but put yourself in my position . I really needed you today and this isn’t the first time you've let me down. Donald f you want to be able to count on this job we nadtn be able to count on you as an employee. Think about that. Don’t worry about the extra hours, you'll have to just miss these hours you've given up. When you get back, we’ll talk. I hope everything works out. Theme two: "How do they address current excuses regarding family problems?" Mutually exclusive 1. Don't address current excuses. (1) I will give you one final chance but this is the last one for any reason. If you don’t shape you will be fired. P 90 (2) I don’t like people who do not take responsibility about this kind of thing. Therefore, I can’t trust you. I will fire you if you don't come into today and hire someone else. 2. Accuse him of making it up. (1) If you don’ t come in today you are fired . You can only use the family emergency excuse so many times . If you aren’ t willing to come to work, I can find someone else who wants the job and will show up all the time. (2) Donald I'm sick and tired of all of your s——-! I don't have time for all of your lies. Don’t bother coming in at all ever! 3. Acknowledge legitimacy of current problems. (1) I was really counting on your help today Donald, but I understand that your family needs you. Could you possibly find a sub before you leave? If not I will schedule you for a double shift for next week; no backing out. I hope everything turns out okay, thanks for calling to let me know! (2) I understand your having family problems, but I have to get this job done. This is the last time you can get away with this. Next time I will have no choice but to replace you with a more dependable worker. 4. Acknowledge and attempt to address/resolve problems. (1) You should realize the situation you put me in. If this happens again, you will be fired. If there is a problem with your family I can help you. We can sit down tonight and discuss it. I would like you to stay on but I will not be able 0 compensate for anymore absences. 91 (2) This is a very important event Donald, and I need you working tonight not next week. Have you checked to see if someone else can take your shift? (no) I have to be honest with you, you have been late repeatedly or have not shown up. That kind of behavior can not be tolerated. If you are having family problems, you may want to take some time off. We can’t rely on you to come to work. Let’s figure something out. References Allport, B. P. (1975). Social distance and admiration reactions of unprejudiced whites. Journal of Personality, 43, 709-726. Ashmore, R. D., & Del Boca, F. K. (1981). Conceptual approaches to stereotypes and stereotyping. In D.L. Hamilton (Eds.), Cognitive processes in stereotyping and intergroup behavior (pp.1-35). Hillsdale, NJ: Eribaum. Bargh, J. A., Bond, R. N., Lombardi, W. J., & Tota, M.E. (1986). The additive nature of chronic and temporary sources of construct accessibility. Journal of Personality and Social Psychology, 50, 869-878. Bargh, J. A., & Pietromonaco, P. (1982). Automatic information processing and social perception: the influence of trait information presented outside of conscious awareness on impression formation. Journal of Personality and Social Psychology, 43, 437—449. Bodenhausen, G. V. & Lichtenstein, M. (1987). Social stereotypes and information-processing strategies: the impact of task complexity. Journal of Personality and Social Psychology, 52, 871-880. Crosby, F., Bromley, S., & Saxe, L. (1980). Recent unobtrusive studies of black and white discrimination and prejudice: a literature review. Psychological Bulletin, 87, 546-563. Devine, P. G. (1989). Stereotypes and prejudice: their automatic and controlled components. Journal of Personality and Social Psychology, 56, 5-18. Dovidio, J. F., Evans, N., & Tyler, R. B. (1986). Racial Stereotypes: the contents of their cognitive representations. Journal of Experimental Social Psychology, 22, 22—37. 92 93 Donnerstein, E., & Donnerstein, M. (1972). White rewarding behavior as a function of potential for black retaliation. Journal of Personality and Social Psychology, 24, 327-333. Duncan, B. L. (1976). Differential social perception and attribution of intergroup violence: testing the lower limits of stereotyping of blacks. Journal of Personality and Social Psychology, 34, 590-598. Hamilton, D. L. (1981). Stereotyping and intergroup behavior. Some thoughts on the cognitive approach. In D. L. Hamilton (Ed.) Cognitive processes in stereotyping and intergroup behavior (pp. 333-353). Hillsdale, NJ: Eribaum. Hecht, M. L., & Ribeau, S. (1984). Ethnic communication: a comparative analysis of satisfying communication. International Journal of Intercultural Relations, 8, 135-151. Hecht, M. L., Ribeau, S., & Alberts, J. K. (1989). An Afro—American perspective on interethnic communication. Communication Monographs, 56, 386- 409. Karlins, M., Coffman, T. L., & Walters, G. (1969). On the fading of social stereotypes: studies in three generations of college students. Journal of Personality and Social Psychology, 13, 1-16. Katz, P. (1976). Racism and social science: towards a new commitment. In P. A. Katz (Ed.) Towards the elimination of racism. New York: Pirgamon Press. Kochman, T. (1982). Black and white: styles in conflict. Chicago, IL: University of Chicago Press. Linville, P. W., & Jones, E. E. (1980). Polarized appraisal of out-group members. Journal of Personality and Social Psychology, 42, 193-211. Litwack, L.F. (1979). Been in the storm so long: The aftermath of slavery. Vintage Books: New York. McConahay, J. B., Hardee, B. B., & Batts, V. (1981). Has racism declined in America? It depends on who is asking and what is asked. Journal of Conflict Resolution, 25, 563-579. 94 O’Keefe, B. J. (1988). The logic of message design: Individual differences in reasoning about communication. Communication monographs, 55, 80-103. O'Keefe, B. J. (1990). The logic of regulative communication: Understanding the rationality of message designs. In Dillard (Ed.), Seeking compliance: The production of interpersonal influence messages, 87-106. Scottsdale, AZ: Gorsuch- Scarisbrick. O'Keefe, B. J. (1991). Message design logic and the management of multiple goals. In K. Tracy (Ed.), Understanding face-to-face interaction: Issues linking goals and discourse, 101-117. Hillsdale, NJ: Erlbaum. O'Keefe, B. J., & Lambert, B. L. (1989). Effects of message design logic on the communication of intentions. Paper presented at the annual meeting of the Speech Communication Association. San Francisco, CA, November. O'Keefe, B. J., & McCornack, S. A., (1987). Message design logic and message goal structure: Effects on perceptions of message quality in regulative communication situations. Human communication research, 14, 68-92. ‘ O’Keefe, B. J., & Shepherd, G. J. (1987). The pursuit of multiple objectives in face-to-face persuasive interactions: Effects of construct differentiation on message organization. Communication Monographs, 54, 396-419. Putnam, L. L., & Roloff, M. E. (1992). Communication perspectives on negotiation. In L. L. Putnam & M. E. Roloff (Eds.), Communication and negotiation, 1- l7. Putnam, L. L. & Wilson, 8. (1987, June). Argumentation and bargaining strategies as discriminations of integrative and distributive outcomes. Paper presented at the First International Conference of the Conflict Management, Group, George Mason University. Rich, A. L. (1974). Some problems in interracial communication: an interracial group case study. Central States Speech Journal, 13, 228-235. 95 Sanbonmatsu, D. M. & Fazio, R. H. (1991). Construct Accessibility: Determinants, Consequences, and Implications for the Media. In J. Bryant and Zillman (Eds.), Responding to the Sceen: Reception and Reaction Processes (45-59). Sigall, H., & Page, R. (1971). Current stereotypes: A little fading, a little faking. Journal of Personality and Social Psychology, 18, 247-255. Taijfel, H. (1969). Social and cultural factors in perception. In G. Lindrey & E. Aronson (Eds.), Handbook of Social Psychology. Mass: Addison- Westley. vanDijk, T. A. (1987). Communicating Racism: Ethnic prejudice in thought and talk. Newbury Park, CA: Sage. Wilson, S. R., & Kang, K. H. (1991). Communication and unfulfilled obligations: Individual differences in causal judgements. Communication Research, 18, 799- MICHIGAN STATE UNIV LIBRARIES l l29