WWW $ i \ illHlWHIWWWW“WWII __THS LIBRA l Illlllllllll’llll will f H 31293 01554 45 LIBRARY Mlchlgan State Unlverslty This is to certify that the thesis entitled The Role of Argumentativeness and Verbal ressiveness in Compliance-Gaining: Predicting Persis_ence, Obstacle Focus, and Strategy Use in Telephone Requests for Blood Donation presented by Laurel Irene Humphreys has been accepted towards fulfillment of the requirements for Master of Atrsiegree in Communi cati on 5:24 i V/m Major professor Dategfl/é’b 0-7539 MS U is an Affirmative Action/Equal Opportunity Institution PLACE N RETURN BOX to romovo this chockout from your rocord. TO AVOID FINES rotum on or boforo doto duo. DATE DUE DATE DUE DATE DUE MSU Io An Afflrmotlvo Action/Equal Opportunity lnctltutlon WM! THE ROLE OF ARGUMENTATIVENESS AND VERBAL AGGRESSIVENESS IN COMPLIANCE-GAINING: PREDICTIN G PERSISTENCE, OBSTACLE FOCUS, AND STRATEGY USE IN TELEPHONE REQUESTS FOR BLOOD DONATION By Laurel Humphreys A THESIS Submitted to Michigan State Univeristy in paritial fulfillment of the requirements for the degree of MASTER OF ARTS Department of Communication 1996 ABSTRACT THE ROLE OF ARGUMENTATIVENESS AND VERBAL AGGRESSIVENESS IN COMPLIANCE-GAMING: PREDICT ING PERSISTENCE, OBSTACLE FOCUS, AND STRATEGY USE IN TELEPHONE REQUESTS FOR BLOOD DONATION By Laurel Humphreys The present study examined the relationship between the individual difference traits of argumentativeness and verbal aggressiveness and compliance-gaining behavior. High argumentative, low verbal aggressive sources were predicted to be more persistent, obstacle focused, and more willing to use confrontive strategies than other trait combinations during telephone requests for blood donations from prior donors of the American Red Cross. The predicted relationships did not hold for the high argumentative, low verbal aggressive group. Instead, the results revealed that the low argumentative, high verbal aggressive group was significantly more persistent and confrontive than the other three combinations of traits. No significant findings emerged for obstacle focus. The results obtained were significant and directly opposite of the predictions. Potential explanations of the unexpected findings for all three of the hypotheses center around Politeness Theory, and the speculated relationships are elaborated in the final chapter. ACKNOWLEDGEMENTS It would seem prudent to acknowledge the influence of significant individuals in my education in a somewhat temporal order of events. In so doing I must first acknowledge the influence of my parents, Clarence and Dorothy Humphreys. As educators they both instilled in me a reverence for knowledge and a belief that my achievements would be limitless given will and desire. I am particularly grateful for the enduring support of my mother who has made it possible financially and spiritually for me to fulfill my dreams. Second I wish to acknowledge the influence of Steven McCornack whose passion for Communication infected me with a similar passion. He was the first to suggest the existence of a career in academe and articulated the path one might take to pursue it. Throughout my Master's program he continued to share his passions in the classroom and guide my learning of communication. His inspiring pep talks still echo in my head when I am filled with self doubt. I am additionally thankful for his assistance and comments as a member of my committee in writing this thesis. Rather by accident or Fate I met Steven Wilson who has served in the most influential role of academic advisor and committee chair. He was my stable refuge in a particularly unstable environment for graduate work. His encouragement and guidance both as advisor and as professor have been instrumental in my achievements. In particular he provided me with the opportunity to collect the data for this study. While pursuing his own research objectives he offered me the opportunity to collaborate on his project and simultaneously extract this study of my own. Specifically, the project was structured with the purpose of collecting data that could be utilized for both our objectives: his interest in examining natural interactions that might be used for the benefit of the American Red Cross; the testing of the hypotheses posited in this thesis. His comments on the numerous drafts of this document have aided me immensely. Much of what I have learned is attributable to him. I also wish to thank Frank Boster for his influence on this thesis. Indirectly, through his own research in compliance-gaining and individual differences, his extant work has guided my project. His direct influence as a committee member has taken the form of critique and suggestions for improvement for which I am much appreciative. James Stalker deserves recognition for his role as a committee member and for his comments and suggestions throughout the editing process. . Due to the unorthodox circumstance of completion of this thesis after beginning doctoral study, I am indebted to Brant Burleson for his immeasurable help in analyzing the results of this study and suggestions for discussion. Finally, I would like to thank the graduate students whose help I could not have survived without. Ken Levine devoted many hours to the project in the form of transcription and coding. Maria Lapinski served numerous roles including friend, sanity saver, and coder. Adrianne Kunkel served a significant support role in data entry and analysis. Michael Dennis acted as a coach and cheerleader in the late stages when I was mentally spent. His comments on the final chapter were particularly insightful. Scott Caplan provided essential support in the formatting stage. To all whom I have mentioned I am eternally grateful. TABLE OF CONTENTS LIST. OF TABLES ..................................................................................... vi INTRODUCTION ........................................................................................ 7 CHAPTER 1 THEORETICAL RATIONALE .................................................................... 9 Resistance, Obstacles, and Persistance ............................................... 9 Argumentativeness and Verbal Aggressiveness ................................. 10 Hypotheses ....................................................................................... 12 CHAPTER 2 METHODS ................................................................................................. 15 Participants ...................................................................................... 15 Procedures ....................................................................................... 16 Dependent Variables ......................................................................... 17 Independent Variables ...................................................................... 20 Unit of Analysis ................................................................................ 21 CHAPTER 3 RESULTS ................................................................................................... 22 Hypothesis 1 .................................................................................... 22 Hypothesis 2 .................................................................................... 26 Hypothesis 3 .................................................................................... 28 CHAPTER 4 DISCUSSION ............................................................................................. 32 Review of the Current Study ............................................................ 32 Review of Boster et a1. (1993) .......................................................... 33 Differences Between the Two Studies ............................................... 34 Speculations ..................................................................................... 36 Limitations ....................................................................................... 41 Future Directions .............................................................................. 42 Conclusions ...................................................................................... 44 APPENDD( A CATEGORIES OF OBSTACLES RAISED BY DONORS ......................... 45 APPENDD( B NINE COMPLIANCE GAINING STRATEGIES ....................................... 46 LIST OF REFERENCES ............................................................................. 47 LIST OF TABLES Table 1 - Means and Standard Deviations For Independent and Dependent Variables in the study ................................................... 22 Table 2 - Means and Standard Deviations For Persistance of Compliance-Gaining Strategy Use as a Function of Argumentativeness and Verbal Aggressiveness ................................ 24 Table 3 - Means and Standard Deviations For Percentage of Obstacle Focus as a Function of Argumentativeness and Verbal Aggressiveness .............................................................. 27 Table 4 - Means and Standard Deviations For Use of Confrontive Strategies as a Function of Argumentaiveness and Verbal Aggressiveness .................................................................... 29 vi INTRODUCTION The study of compliance-gaining has been prominent in communication research since the late 1970's. Exemplary reviews of the domain of this literature recently have been conducted by Kellerman and Cole (1994) and Seibold, Cantrill, and Meyers (1994). Compliance-gaining involves goal-oriented communication generated by a source who wants a target to perform a desired action. This study examines the manner in which the individual difference variables of argumentativeness and verbal aggressiveness influence features of compliance-gaining behavior, namely, persistence, obstacle focus, and the use of confrontive strategies in telephone requests for blood donation. Explication of these variables and their relationships will appear in the sections to come. The methodology of the present paper follows the approach of a recent study examining compliance-gaining as an interactive process. An integration of the conversation analysis (CA) method with quantitative interaction analysis was recently undertaken by Wilson, Levine, Humphreys, and Peters (1994) to investigate the sequential nature of the compliance-gaining process in genuine interactions. Specifically, the interactions were audio-taped conversations of research participants' telephone calls to former donors of the American Red Cross requesting a repeat blood donation. The data collection in the Wilson et al. (1994) study was intentionally structured to allow for the testing of hypotheses posited here. Hence, the current data set includes data used in that study as well as individual difference measures and additional data from audiotapes that were not transcribed or analyzed at the time of the Wilson et a1. (1994) report. To provide a research rationale, chapter one begins with a review of compliance-gaining literature emphasizing resistance, obstacles, and persistence, and an explication of the individual difference variables of argumentativeness, and verbal aggressiveness. Chapter one concludes with the hypotheses to be tested. Information regarding the sample, data collection, transcription, and coding of compliance-gaining strategies is provided in chapter two. Chapter three reports the results of the primary tests of the hypotheses as well as supplemental analyses of the findings. Chapter four is a speculative discussion of the findings and possible reasons for the results. The chapter concludes with implications for future research. Chapter 1 THEORETICAL RATION ALE Resistance Obstacles And Persistence Relationships among several variables will be suggested in the following paragraphs, and to aid comprehension, definitions are offered first. Crucial variables in any interactive study of compliance-gaining are resistance and obstacles. Resistance will be conceived of as the general failure of the target to comply with an initial request from a source. Wilson et al. (1994) classified the initial response of the target in four ways: unconditional yes, conditional yes, maybe, and no. The unconditional yes consisted of a response of "yes" or "sure". Such a response can be considered a grant of the initial request for a repeat donation, although it often required further communicative effort by the source to secure an appointment. To best understand the other three response types found in the Wilson et al. (1994) data, an explanation of adjacency pairs is provided. Conversational analysts Schegloff and Sacks (1973) describe adjacency pairs as two sequential speech acts, such that the occurrence of the first speech act, or first pair part (FPP) necessitates a second pair part (SPP). In compliance-gaining a request functions as a FPP which necessitates a relevant SPP of a grant or refusal (Wilson et al., 1994). The linguistic formula of SPP's are somewhat constrained by the concept of preference (Blimes, 1988). SPP's may be preferred or dispreferred, and when following a request, the preferred SPP is a grant of the request, as it affirms the FPP. Refusals are dispreferred, as they reject the action of the prior turn, and should be "marked" for this rejection. The unconditional yes is best classified as a grant, and the other three response types are best classified as refusals. The conditional yes is a response of "willingness to comply prim that certain conditions (obstacles) could be overcome" (p. 13). Both the conditional yes and maybe responses are classified as refusals, and hence resistance, because they are dispreferred SPP's which fail to promptly complete the adjacency pair initiated by the request. In the maybe category targets did not say yes or no and instead either "asked 10 questions" or "mentioned external circumstances" (p.13). Obviously a target who explicitly says "no" also offers resistance. Obstacles, as subcategories of resistance responses, are conceived of as specific instances or reasons, either stated or inferred, for non-compliance. The obstacles raised by targets in the Wilson et al. (1994) study can be found in Appendix A. The most common obstacles were medical ineligibility, such as pregnancy or a recent donation, and general schedule constraints. Inherently tied to resistance is the concept of persistence, defined as secondary attempts by the source to gain compliance after resistance is encountered. That is, persistent sources do not give up their goal of gaining compliance after a target resists, and their turns subsequent to an obstacle raised by a target contain additional and diverse strategies to obtain compliance. Generally this variable is measured in terms of the total number of compliance-gaining strategies used, as well as the number of different strategies used by the message source in a compliance attempt. Occasionally, the length of the interaction is used as an additional measure of persistence (Boster, Levine, & Kazoleas, 1993; Wilson, Cruz, Marshall, & Rao, 1993). A highly persistent source uses both a large variety of strategies and numerous strategies while one who is not persistent uses few strategies with little diversity. Argumentativeness And Verbal Aggmssiveness Past research has suggested that individual differences play a role in compliance-gaining strategy choice (Boster & Levine, 1988; deTurck 1985; Boster et al., 1993; Wilson et al., 1993). Two variables that seem particularly relevant to the study of compliance-gaining are argumentativeness and verbal aggressiveness. Infante and Rancer (1982) define argumentativeness as "a generally stable trait which predisposes the individual in communication situations to advocate positions on controversial issues and to attack verbally the positions which other people take on theses issues" (p.72). Infante and ll Wigley (1986) define verbal aggressiveness as a "personality trait that predisposes persons to attack the self-concepts of other people instead of, or in addition to, their positions on topics" (p.61). Argumentativeness is viewed as constructive while verbal aggressiveness is generally seen as a destructive trait. (Infante & Wigley, 1986). The two traits are conceived to be opposing, yet conceptually distinct (Sabourin, Infante, & Rudd, 1993). The nature of their opposition is evidenced in findings of verbal aggressiveness moderating the effects of argumentativeness, such that when both argumentativeness and verbal aggressiveness are high then constructive outcomes are diminished (Infante & Rancer, in press). Alternatively, constructive outcomes result when argumentativeness is high but verbal aggressiveness is low (Boster et al., 1993; Rudd, Burant, & Beatty, 1994; Bayer & Cegala, 1992; Sabourin et al., 1993). Both argumentativeness and verbal aggressiveness are measured in this study to enhance the chance of finding the constructive outcomes of argumentativeness. However, few verbally aggressive messages are expected given the pro-social nature of the task. That is, our participants who telephone on behalf of the American Red Cross are not expected to become verbally aggressive toward the donors they call given the social constraints against such behavior, regardless of a participants' trait verbal aggressiveness. Given this situational constraint, individuals who are high in both argumentativeness and verbal aggressiveness may temper their normally confrontational style in order to avoid appearing rude or impolite. Specific instructions were given to volunteers requesting that they recognize that they were representing the American Red Cross and to be polite to the donors they called. In other words, individuals scoring high in verbal aggressiveness are not expected to argue when they are instructed to be polite because they are conceived as not being able to argue without attacking at the same time. Emergence of the verbal aggressiveness trait is restricted by the highly constrained nature of the task. Thus, no predictions are made for the construct beyond the interactions assumed to exist between argumentativeness and verbal aggressiveness. 12 As defined, the construct of argumentativeness suggests that individuals who score high may be more likely to pay close attention to the positions of others than individuals who are low in argumentativeness. Hence, it is not too distant to claim that they may also be more likely to focus on specific obstacles raised by a target in a compliance-gaining episode. Further, the individual high in argumentativeness is conceived as being skilled in argument, or as possessing a better repertoire of arguments than one who is not argumentative (Infante, 1988; Infante et al., 1984), and thus may be cognitively more adept at responding to resistance raised by targets. This reasoning is supported by the argumentative skill deficiency model (Infante, Chandler, & Rudd, 1989), which suggests that individuals high in argumentativeness are more skilled at argument than those who score low on the trait. Related findings suggest that individuals who are high in verbal aggressiveness lack verbal skills for dealing with frustrations, and possess a limited repertoire of compliance-gaining messages (Infante, 1988; Sabourin et al., 1993; Infante, Riddle, Horvath, & Tumlin, 1992; Infante et al., 1984; Infante & Wigley, 1986; Boster et al., 1993). Therefore, based on the preceding review of resistance, persistence, argumentativeness, and verbal aggressiveness, the following hypothesis is posited: Hymtheses H1 Sources who are high in argumentativeness but low in verbal aggressiveness will be more persistent than sources who possess any of the other three combinations of argumentativeness and verbal aggressiveness (high/high, low/high, low/low). In the study by Wilson et a1. (1993) it was noted that sources differed in their orientation to the specific obstacles raised by the target. Namely, some sources focused on the specific obstacles raised while others used general persuasion strategies. Wilson et al. (1993) suggest individual differences might be related to the focus on specific obstacles. To further explore the influence of individual differences on obstacle focus selection of relevant individual difference variables is necessary. The argumentativeness and verbal l3 aggressiveness traits studied here seem appealing for this task as they are inherently tied to the behavioral characteristics of interest. It was stated in a previous section that Infante and Rancer's (1982) definition of argumentativeness implies a relationship between the trait and the amount of attention paid to the positions taken by others and more specifically the obstacles raised by a target. For example, messages focused on a specific obstacle (see Appendix B) either seek to clarify it, by asking for more information about it, deny or refute the obstacle raised, or suggest plans for overcoming the obstacle (Wilson et al., 1993; Wilson et al., 1994). General persuasion strategies, on the other hand, are not focused on the reason offered for non-compliance, and often consist of arguments for why the target should comply, using for example resource arguments or normative arguments (Wilson etal., 1993; Wilson et al., 1994). Based on the observation of distinctive source behavior with regard to obstacles and the conceptual definition of the argumentativeness trait a second hypothesis is posited. H2 Sources who are high in argumentativeness but low in verbal aggressiveness will focus a larger percentage of their talk on the specific obstacle( s ) raised than sources who are high/high, low/high, or low/low. The propensity to use particular strategies should be influenced by the argumentativeness and verbal aggressiveness traits as well. In examining the compliance-gaining strategies found by Wilson et al. (1994), (Appendix B) it seems that two strategies, the Refute Obstacles and Request Specific Commitment strategies, may be more likely to be used by individuals who score high in argumentativeness and low in verbal aggressiveness as they are inherently reflective of the argumentativeness trait characteristics as described by Infante and Rancer ( 1982). Specifically, argumentativeness is defined in part as the propensity to refute the positions of others, and thus the trait should predict the use of the Refute Obstacles strategy. This reasoning is further supported by findings suggesting high argumentatives are more skilled in argument than low argumentatives, and thus have a larger repertoire of arguments to draw from in refuting obstacles raised by a donor l4 (Infante, Chandler & Rudd, 1989). A recent article reviewing the literature on argumentativeness and verbal aggressiveness (Infante & Rancer, in press) describes argumentative communication as a "forceful" form of communication. It is "forceful" in the sense that the communication is used as a force to attain dominance and interpersonal goals (Infante & Rancer, in press). It is argued here that use of the Request Specific Commitment strategy is an attempt to force the donor to commit to an appointment time, and as the literature suggests is more likely to be used by participants who score high on the trait. For the purposes of distinction the Request Specific Commitment and Refute Obstacle strategies are described as confrontive in comparison to the other strategies types (see Appendix B). H3 Sources who are high in argumentativeness but low in verbal aggressiveness will be more likely to use the confrontive strategies of Request Specific Commitment and Refute Obstacles than sources who are low/low, high/high, and low/high. Chapter 2 METHODS Overview The present study examines the data obtained by Wilson et al. (1994). Their study examined compliance-gaining interactions between volunteer telephone solicitors and prior blood donors of the American Red Cross, the method made possible through the collaborative assistance of the American Red Cross Great Lakes Regional Blood Center. The procedure required two meetings of the volunteers on different evenings. On the first evening the volunteers received cursory training from representatives of the American Red Cross and responded to questionnaires measuring individual differences. At the second meeting, the volunteers received a list of names and telephone numbers of prior blood donors whom they were to telephone. Upon reaching a donor, the volunteers obtained permission to audiotape the telephone call and then made a request for a repeat donation. Following the phone calls the volunteers revealed the extent of their experience in telemarketing and persuasive sales. Finally, the volunteers listened to each of the calls where contact was made with a donor and responded to questions regarding their perceptions of the calls. (The volunteers’ perceptions are not analyzed here). The audiotaped conversations were transcribed and coded for the compliance-gaining strategies used by the volunteers and the obstacles raised by the donors. The specific analyses of the transcribed conversation will be discussed elsewhere. Participants: Volunteers and Prior Donors The volunteer telephone solicitors were undergraduate students enrolled in communication courses at Michigan State University who received extra credit for their participation. Thirty-two participants volunteered for the study. Thirteen cases in the 15 16 original sample were discarded for the following reasons: seven of the audio tapes were not audible; one tape was lost; individual difference measures were missing for five participants. Analyses of the individual difference measures and compliance-gaining strategies were based on the remaining participants (N = 19 ). The American Red Cross Great Lakes Regional Blood Center provided lists of approximately 1000 prior donors in the Greater Lansing area with A+ blood type who had donated blood within the last four years. The names of 20 prior donors were randomly selected for each volunteers' call list. The former donors had no prior knowledge of the project until the volunteer called and informed them, and their responses to the requests are genuine. Procedures The volunteers reported the first night for a one-hour training session conducted by John Dobias, Director of Donor Resources for the Great Lakes Regional Blood Center. Dobias informed the volunteers of the blood donation processes, uses of blood products, maintenance of donor confidentiality, medical eligibility criteria, and tips for dealing with commonly raised obstacles to donation. Volunteers were given the opportunity to ask questions. Volunteers completed the Argumentativeness and Verbal Aggressiveness scales at the end of the session, after which they signed up for a second night to call donors. Two to four weeks later the volunteers returned to call the prior donors. They received a list of 20 names with background information (addresses, telephone numbers, date of last donation), notes on the medical eligibility criteria, a schedule of the donor room hours at the Great Lakes Regional Blood Center, the phone number for the center, and a procedural guide to be followed in each call. The procedures required volunteers to follow three steps in each call: obtaining permission to audiotape the call; verifying background information; and requesting the donor to make an appointment for a repeat l7 donation. Upon contacting a donor the volunteers were to ask for permission to audiotape by following as closely as possible the script below: This is and I'm calling for the American Red Cross and Michigan State University. The Red Cross is conducting research into their blood donation program, and it would greatly assist this effort ifI might audiotape this telephone conversation. Please be assured that all information will be held confidential. Do I have your permission to record this call? Upon obtaining permission to audiotape, the volunteers were to verify the donor's background information (e.g., address, date of last donation) and then to ask if the prior donor was willing to donate blood again. The following phrase was suggested, " I'm calling to see if we could schedule you to donate blood again. Can we count on you again?" Once having made the initial request, it was up to the volunteer to decide how to proceed. Thus the calls are initially highly scripted for all participants, but generally unique dialogue emerged throughout the rest of the call. Demndent Variables Seven dependent variables were examined in this study. Persistence was operationalized as a composite of diversity, or the average number of different strategies used by a source per call, and frequency or the average total number of strategies used by a source per call. Diversity and frequency were examined individually and combined to form an index of persistence. The persistence index was created by standardizing scores for diversity and frequency, and summing those scores for a composite. The fourth dependent variable, obstacle focus, was calculated by determining the ratio of use of obstacle strategies to the total number of strategies used. Finally, two confrontive strategies, Request Specific Commitment and Refute Obstacles were examined both individually and as a combined index called confrontive strategies. Compliance-gaining strategies. After transcribing the audiotapes of the telephone calls, the transcripts were coded for the compliance gaining strategies used by the volunteers. 18 The coding scheme developed by Wilson et al. (1994) was used in the present analysis as it was developed for initial analysis of the same data analyzed here. The scheme was developed by drawing the four general categories from existing typologies of compliance- gaining strategies (Wilson et al., 1993; Kellerman & Cole, 1994), and adapting the specific strategy types within the general categories to fit the data and context (see Appendix B). Two independent coders first unitized the compliance-gaining strategies. The percentage of disagreement as calculated by Guetzkow's U was three percent. The strategies were then coded for strategy type by two independent coders achieving satisfactory agreement. Cohen's kappa was .90. The first general category is labeled Request Strategies, which consist of "statements/questions in which volunteers asked a prior donor to do something s/he otherwise would not have done" (Wilson et al., 1994, p. 11). Two distinct strategy types comprise this category. The first, general requests, "include direct requests for blood donation or indirect queries about the donor's ability or willingness to donate again" (p. 11). The second type, requests for a specific commitment, were requests which concentrated on securing a specific appointment time as opposed to securing consent to donate again in general. Resource Strategies, the second general category, are accounts of the donors' resource outcomes given a donation. Just one strategy of this type, downgrading costs, appeared in our data. Statements of this type suggested that donating would take less time than the donor might have imagined. In the third category are Obstacle Strategies which are oriented to stated or inferred reasons for why the donor may not agree to donate again. Strategy types in this category included: "clarifying specific obstacles (asking for more information about why the donor can not donate again), providing information about the range of options for complying (e.g., asking the prior donor to call the Regional Blood Center for further information 19 about their ability to donate), and refuting obstacles (explicitly/implicitly asserting that a state of affairs should not prevent the prior donor from donating again)" (p. 11-12). The fourth category, Interpersonal Strategies, stated the social benefits of donating. The other benefit strategy suggested that the prior donor could help by giving blood to save lives. The expressing gratitude strategies were statements of thanks and appreciation for past blood donations. Source mrsistence, orientation to obstacles, and confrontive strategies. In addition to coding volunteers' use of the above strategies, five more global variables were calculated for each telephone call. First, a measure of source persistence was calculated to test hypothesis one. The persistence variable was operationalized by computing both the average number of total strategies used by a source per call (frequency), and the average number of difi‘erent strategies used by the source per call (diversity). Computation of the averages was necessary as the number of calls made by each participant varied from one to nine (i.e., the number of calls in which each participant actually reached and spoke to a prior donor). Frequency and diversity scores were highly correlated, r = .70, p < .001. To test the second hypothesis, the percentage of talk focused on obstacles was operationalized as the number of obstacle strategies used by a source (see Appendix B) relative to the total number of strategies used by a source (i.e. the proportion of obstacle strategies used). For example, if a volunteer used a total of five compliance-gaining strategies across their phone calls, and two of these strategies were clarifying obstacles and refuting obstacles, then that volunteer received a score of 2/5 = 40% obstacle focused talk. The average frequency of source use per call of the Refute Obstacles and Request Specific Commitment strategies were examined to test hypothesis three. A composite index (i.e., the summed average frequency for these two strategies also was analyzed). 20 e ndent Variables Following the training session on the first evening, the volunteers answered two questionnaires measuring individual differences: Infante and Rancer's (1982) measure of trait argumentativeness; and Infante and Wigley's (1986) measure of verbal aggressiveness. The argumentativeness scale is a 20 item self-response measure made up of 10 items measuring the general tendency to approach arguments (ARGap) and 10 items measuring the general tendency to avoid arguments (ARGav). The scores on the both the ARGav items and the ARGap items (ranging from 1 to 5) are summed separately. Then the sum of the ARGav items are subtracted from the sum of the ARGap items to arrive at the measure of argumentativeness. The reliability of the scale as assessed by test-retest procedures and Cronbach‘s alpha is well established as is the validity (criterion, content, convergent, and divergent) of the scale (Boster & Levine, 1988; Infante & Rancer, 1982; Rancer, Baukus, & Infante, 1985). In the current study Cronbach's Ma for the 20 item Argumentativeness scale was .87. The verbal aggressiveness scale contains 20 items, 10 positively worded and 10 negatively worded. The negative items are reverse scored and then all items are summed for a verbal aggressiveness score. The reliability and validity of the scale is supported by prior research (Boster & Levine, 1988; Infante & Wigley, 1986). Cronbach's Ma for the Verbal Aggressiveness scale in the current study was .87. Median splits divided the sample into high and low groups for the argumentativeness and verbal aggressiveness traits. Four groups of subjects were created based on their argumentativeness and verbal aggressiveness scores. In one group (N = 6), subjects scored below the median on both argumentativeness and verbal aggressiveness. In the second group (N = 3), subjects below the median on argumentativeness and above the median on verbal aggressiveness. A third group (N = 3) consisted of subjects who scored above the median on argumentativeness and below the median on verbal aggressiveness. Finally, subjects in the fourth group (N = 7) scored above the median on both argumentativeness and verbal 21 aggressiveness. The correlation between individuals' scores on argumentativeness and verbal aggressiveness was .58. Unit pf Analysis The hypothesized outcomes involve individual differences in seeking compliance, hence, "participant" was the unit of analysis. Each participant telephoned 20 prior donors, but individual participants actually reached and talked with different numbers of donors (range = one to nine calls per participant). This made it necessary to compute averages for each of the dependent variables over the number of calls completed by the participant. This was done by averaging each participants' scores on each dependent variable across all of their calls where a donor was reached. For example, if participant one reached three donors, then frequencies of the number of strategies used was recorded for each of the three calls. The frequencies of strategies used were then summed across the number of calls completed and that sum was divided by the number of calls completed to generate the index of average total number of strategies employed. A similar procedure of averaging scores across calls was used to calculate each participants' score on all but one of the other dependent variables. The exception was for the dependent variable percentage of obstacle focus which was calculated by dividing the total number of obstacle strategies used by the total number of strategies used. Chapter 3 RESULTS This draft reports the results of both direct tests of the hypotheses as well as supplemental analyses of the findings. The results for each hypothesis are reported in turn, reviewing the direct tests first followed by any additional findings of the supplemental analyses. Means and standard devi ations for all independent and dependent variables are presented in Table 1. Descriptive Statistics Table 1 Means and Standard Deviations for Independent and Dependent Variables in the Study Mean Standard Deviation Verbal Aggressiveness 42.63 12.17 Argumentativeness 1 .63 1 1.90 Total Strategies 3.88 1.83 Diversity 1 .23 .68 Obstacle Focus .33 .12 Request Specific Commitment .75 .88 Refute Obstacles .10 .17 Confrontive Composite .84 .89 Note: N= 19. Dependent variable means computed by summed standard scores. Hypothesis 1 Direct Test of Hmthesis 1 Hypothesis I predicted that sources who are high in argumentativeness but low in verbal aggressiveness would be more persistent than sources who possessed any of the other three combinations of argumentativeness and verbal aggressiveness. To test Hypothesis 1 22 23 apfipri contrasts were carried out for each of the two persistence indices, frequency of strategy use and diversity of strategies, as well as for the composite index combining frequency and diversity. Specifically, for each dependent measure those participants who scored high in argumentativeness and low in verbal aggressiveness were contrasted with those participants having other combinations of argumentativeness and verbal aggressiveness scores (i.e., low/low, low/high, high/high). Means and standard deviations for each of the combinations of argumentativeness and verbal aggressiveness are presented in Table 2 for the three indices of persistence (frequency, diversity, composite). Visual inspection of these means shows that the sources high in argumentativeness and low in verbal aggressiveness were only irregularly more persistent than sources having the low/low, low/high, or high/high trait combinations. The mgcontrast for frequency (the total number of strategies used) found that sources high in argumentativeness and low in verbal aggressiveness (M = 4.62 ) did not employ a significantly larger number of strategies than the other participants (M = 3.74), t (14) = -0.47, p > .60. Similarly the am contrast for the diversity of strategies found that sources high in argumentativeness and low in verbal aggressiveness (M = 1.43) did not significantly differ from the sources who were low/low, low/high, high/high (M = 1.19), t (14) = -0. 17, p > .85. Finally, the am contrast for the composite index of persistence found that sources high in argumentativeness and low in verbal aggressiveness (M = .38) were not significantly more persistent than sources who were low/low, low/high, high/high (M = -.07 ), t (14) = -0.37, p > .70. Thus Hypothesis 1 was not supported. 24 Table 2 Means and Standard Deviations for Persistence of Compliance-gaining Strategy Use as a Function of Argumentativeness and Verbal Aggressiveness Persistence Argumentativeness Index Low High Frequency Low Verbal Aggressiveness 3.57ab 4.62ab (1.42) (2.20) High Verbal Aggressiveness 5.83b 2.99a (2.33) (1.32) Diversity Low Verbal Aggressiveness 1.05a 1.43ab (0.29) (0.29) High Verbal Aggressiveness 2.13b 0.92a (1.23) (0.45) Composite Low Verbal Aggressiveness -.24a .38ab (0.56) (.85) High Verbal Aggressiveness 1.30b -.52a (1.67) (.54) Note: Figures in parentheses are standard deviations. Means with any superscript letter in common do not differ significantly (p < .05) by the Least Significant Difference test. 25 Supplemental analyses on mrsistence indices To further explore the effects of argumentativeness and verbal aggressiveness on persistence in a compliance-gaining context, a 2 (low and high argumentativeness) x 2 (low and high verbal aggressiveness) AN OVA was conducted on each of the persistence indices (frequency, diversity, composite). For the two-way AN OVA on the number of strategies used there was no significant main effect for either argumentativeness [If (1, 14) = 1.32, p > .25] or verbal aggressiveness [E (1, 14) = 0.11, p > .75]. There was however, a significant two-way interaction between the argumentativeness and verbal aggressiveness factors [If ( 1,14) = 5.63, p < .03]. Decomposition of this significant interaction by means of the Least Significant Difference (LSD) test revealed that those low in argumentativeness and high in verbal aggressiveness (M = 5.83) employed significantly (p < .05) more strategies than did those high in both argumentativeness and verbal aggressiveness (M = 2.99). As is apparent in Table 2 participants in the other two groups (high/low and low/low argumentativeness/verbal aggressiveness) fell in between the prior two groups and did not differ significantly from either. The 2 x 2 AN OVA on the diversity of strategies employed obtained results similar to those observed for the total number of strategies employed. Specifically, the main effect for argumentativeness was not significant [F (1,14) = 2.42, p > .10] nor was the main effect for verbal aggressiveness [13 (1,14) = 0.87, p > .35]. However, there was a significant interaction between the factors of argumentativeness and verbal aggressiveness, [E (1 ,14) = 7.97, p < .01]. Decomposition of this interaction assessed by the LSD test revealed that those high in verbal aggressiveness and low in argumentativeness (M = 2.13) used significantly more diverse strategies than did those who were high in both argumentativeness and verbal aggressiveness (M = 0.92) and those who were low in both argumentativeness and verbal aggressiveness (M = 1.05). See Table 2. The 2 x 2 AN OVA on the composite index of persistence obtained results consistent with those of the individual indices. The main effect for argumentativeness was not 26 significant, [E (1, 14) = 2.35, p >. 10] nor was the main effect for verbal aggressiveness, [E (l, 14) = .50, p> .45]. However, there was significant two-way interaction between the argumentativeness and verbal aggressiveness factors [E = (1, 14) = 8.72, p> .01]. Decomposition of this interaction by means of the LSD test revealed that individuals high in verbal aggressiveness and low in argumentativeness were more persistent (M = 1.30) than individuals low in both traits (M = -.24) or individuals high in both traits (M = -.52). See Table 2. In sum, Hypothesis 1 received no support. However, supplemental analyses revealed that persons low in argumentativeness and high in verbal aggressiveness were most persistent in their compliance-gaining efforts, employing both a greater number of and more diverse strategies with their targets. Mia; Direct Test of Hymthesis 2 Hypothesis 2 predicted that sources who are high in argumentativeness but low in verbal aggressiveness would focus a larger percentage of their talk on the specific obstacle(s) raised by the target than sources who are low/low, low/high, high/high. To test this hypothesis aLrio_ri contrasts were carried out for the measure of obstacle focus, the ratio of obstacle strategies to total strategies. The participants high in argumentativeness and low in verbal aggressiveness were contrasted with the average of the other three trait combinations (i.e., low/low, low/high, high/high). Means and standard deviations for the use of obstacle strategies for the four combinations of argumentativeness and verbal aggressiveness are presented in Table 3. Visual inspection of these means revealed that the sources high in argumentativeness and low in verbal aggressiveness did not use a greater percentage of obstacle strategies than the other three groups. The arm contrasts found that sources high in argumentativeness and low in verbal aggressiveness (M = 0.27) did not focus a 27 proportionately greater percentage of their talk on obstacles raised by the targets than did participants in the other three groups (M = 0.34), t (14) = 1.03, p > .30. Thus, Hypothesis 2 was not supported. Table 3 Means and Standard Deviations for Percentage of Obstacle Focus as a Function of Argumentativeness and Verbal Aggressiveness Argumentativeness Percentage of Obstacle Focus Low High Low Verbal Aggressiveness 0.31 0.27 (.07) (.13) High Verbal Aggressiveness 0.40 0.34 (.10) (.16) Supplemental Analyses on Obstacle Focus A 2 (high and low argumentativeness) x 2 (high and low verbal aggressiveness) ANOVA was used to assess the effects of argumentativeness and verbal aggressiveness on the dependent variable obstacle focus. For the two-way AN OVA on the percentage of talk focused on obstacles there was no significant main effect for either argumentativeness [E (1, 14) = 0.64, p > .40] or verbal aggressiveness [F (1,14) = 1.63. p > .20]. No interaction between argumentativeness and verbal aggressiveness was found either [E ( l , 14) = 0.01, p > .90]. Hence, neither the primary nor supplemental analyses provided support for Hypothesis 2. 28 Hypothesis 3 Direct Test of Hymthesis 3 ' The third hypothesis predicted that sources who were high in argumentativeness but low in verbal aggressiveness would be more likely to use the confrontational strategies of Request Specific Commitment and Refute Obstacles than would sources who were low/low, high/high, and low/high. To test Hypothesis 3 a_p_ri_qr_i contrasts were conducted for each confrontive strategy individually and the two strategies combined. Sources high in argumentativeness and low in verbal aggressiveness were compared to the other three trait combinations for use of the confrontive strategies. Table 4 presents means and standard deviations for the use of the confrontive strategies by each of the four trait combinations of argumentativeness and verbal aggressiveness (high/low, low/low, low/high, high/high). Visual inspection of the means suggests no consistent differences between the four groups in the use of summed confrontive strategies. The apri_or_i contrast for the combined index of Refute Obstacles and Request Specific Commitment strategies found that sources high in argumentativeness and low in verbal aggressiveness (M = 1.11) were not significantly more likely to use the confrontive strategies than the other groups (M = 0.79), t (14) = -0.21, p > .80. Source use of the Refute Obstacle strategy as assessed by the am contrasts did not reach significance. Sources high in argumentativeness and low in verbal aggressiveness (M = 0.15) were not more likely to use the Refute Obstacles strategy than the other participants (M = 0.09), t (14) = -0.23, p > .80. Similarly, no difference in source use of the Request Specific Commitment category was detected by the ami_ori contrasts. Sources high in argumentativeness and low in verbal aggressiveness (M = 0.96) were not more likely to use the Request Specific Commitment strategy than the other sources (M = 0.70), t (14) = -0.15, p < .85. The a priori contrasts thus failed to support Hypothesis 3. 29 Table 4 Means and Standard Deviations for Use of Confrontive Strategies as a Function of Argumentativeness and Verbal Aggressiveness Argumentativeness Confrontive Index Low High Request Specific Comrrritrnent Low Verbal Aggressiveness 0.49ab 0.96% (.34) (.72) High Verbal Aggressiveness 1.70b 0.46“l (2.0) (.29) Refute Obstacles Low Verbal Aggressiveness 0.05a 0.15ab (.08) (.13) High Verbal Aggressiveness 0.32b 0.02a (.33) (.05) Composite Low Verbal Aggressiveness 0.54a 1.11“ (.36) (.84) High Verbal Aggressiveness 2.02b 0.49‘5l (1.75) (.27) Note: Figures in parentheses are standard deviations. Means with any superscript letter in common do not differ significantly (p < .05) by the Least Significant Difference test. 30 Supplemental Analyses on Confrontive Strategies Further analyses of the influence of argumentativeness and verbal aggressiveness on the use of confrontive strategies was assessed through a series of 2 (high and low argumentativeness) x 2 (high and low verbal aggressiveness) AN OVAs. For the two-way ANOVA on the summed use of the two confrontive strategies (Refute Obstacles and Request Specific Commitment) no significant main effects were found for argumentativeness [E (l, 14): 1.81, p > .15] or verbal aggressiveness [E (1, 14) = 1.15, p > .30]. However, a significant interaction between argumentativeness and verbal aggressiveness was found [E ( 1, 14) = 7.93, p < .01]. Decomposition of this interaction by means of the LSD test revealed that sources low in argumentativeness and high in verbal aggressiveness (M = 2.02) were significantly more likely ( p < .05) to use the confrontive strategies than the low/low (M = 0.54) or high/high (M = 0.49) combinations of argumentativeness and verbal aggressiveness. A two-way AN OVA was also conducted on the use of the Refute Obstacles strategy. No significant main effects were found for either argumentativeness [E (1, 14) = 2.12, p > .15] or verbal aggressiveness [E (1,14) = 0.85, p > .35] Again, however, a significant interaction effect was found between argumentativeness and verbal aggressiveness [E (1, 14) = 7.99, p < .01]. Decomposition of this interaction by means of the LSD test revealed that sources low in argumentativeness and high in verbal aggressiveness (M = 0.32) were significantly (p < .05) more likely to use the Refute Obstacles strategy than sources who were low/low (M = 0.05) or sources who were high/high (M = 0.02). The two-way AN OVA on the Request Specific Commitment strategy obtained similar results to those of the Refute Obstacles strategy. Specifically, the main effect for argumentativeness was not significant [E (1, 14) = 0.98, p > .30] nor was there a significant main effect for verbal aggressiveness [E (1, 14) = 0.69, p > .40]. The interaction between argumentativeness and verbal aggressiveness was significant [E (l, 14) = 4.43, p < .05]. Decomposition of this interaction by means of the LSD test found that 31 sources low in argumentativeness and high in verbal aggressiveness (M = 1.70) were significantly more (p < .05) likely to use the Request Specific Commitment strategy than sources high in argumentativeness and high in verbal aggressiveness (M =0.46). To summarize, Hypothesis 3 was not supported by the data. However, the supplemental analyses revealed significant effects opposite of the predictions. Sources low in argumentativeness and high in verbal aggressiveness were most likely to use the Refute Obstacle strategy, the Request Specific Commitment strategy, and both confrontive strategies combined. Chapter 4 DISCUSSION The purpose of this paper was to explore the influence of the individual difference variables of argumentativeness and verbal aggressiveness in compliance-gaining interactions by examining telephone requests for blood donations. Although the findings did not support the hypothesized relationships between the individual difference traits of argumentativeness and verbal aggressiveness and the dependent variables in the study, significant results were obtained that allow for interesting speculations to be made. This chapter will offer post hoc reasoning for the findings that were significant as well as those that were not. The findings of the present study will be compared to existing findings of a closely related study with the intent of providing an explanatory account of inconsistencies. The comparison will exclusively focus on the Boster et al. (1993) study providing both a review and a speculative explanation of the inconsistencies. Exclusive focus on the one study is necessitated by the difficulties of offering an interpretive account of available studies on argumentativeness and verbal aggressiveness, as there is little overlap in the existing literature in terms of situation and method. The chapter will conclude with a discussion of the study's limitations and provide directions for future research. Review of findings of the current study It was hypothesized that sources high in argumentativeness and low in verbal aggressiveness would differ in compliance-gaining behavior from sources classed into the other three combinations of these traits. Specifically, the high argumentative low verbal aggressive combination was predicted to be associated with more persistence, more focus on obstacles, and greater willingness to use confrontive strategies. The findings failed to support these predictions for the high argumentativeness low verbal aggressiveness group, 32 33 and instead revealed significant relationships directly opposite of the predictions. That is, two of the three hypothesized relationships were found to exist for the low argumentative, high verbal aggressive group. Specifically, sources who were low in argumentativeness and high in verbal aggressiveness were found to be significantly more persistent ( i.e., used more strategies overall and used a greater number of different strategies) than the high/high and low/low groups. Secondly, the low argumentative, high verbal aggressive group was found to be significantly more willing to use the confrontive strategies (i.e., Request Specific Commitment and Refute Obstacles) than the high/high and low/low groups. N 0 significant differences were detected between the low argumentative, high verbal aggressive group and the high argumentative, low verbal aggressive group on any of the dependent measures. Qurrent findings compared to Boster et a1. (1923) The findings of the present study are perhaps best compared to Boster et al. (1993) for the high degree of overlap between the two studies in terms of method and independent and dependent variables examined in the studies. Specifically, argumentativeness and verbal aggressiveness were independent variables in both studies. Additionally, both studies examined natural messages for the number of strategies used and diversity of strategies. (The present study analyzed these two measures individually and as a composite index called persistence). However, despite the similarity in the two studies, remarkable differences emerged in the findings. Boster et al. (1993) found an increase in persistence (number of strategies) for high argumentative, low verbal aggressive sources. Similarly, the authors found a proportionate decrease in persistence for sources high in both traits. For diversity (number of different strategies) an effect for argumentativeness was detected. N 0 effects were found for diversity and verbal aggressiveness, nor were there any interactions between the two traits affecting diversity. These findings differ from the present study in 34 that an increase in persistence (frequency and diversity) was found for the low argumentative, high verbal aggressive group, and no main effects were found for either group. Situational differences between tlLtwo studies Given the identical overlap of independent variables and similar operationalization of the dependent variables persistence and diversity, the inconsistent findings most likely are a function of the differences between the two studies in the compliance-gaining situation designed for examination. Specifically, the Boster et a1. (1993) study involved a simulated negotiation between interactants over the price of a car. Participants enacted the role of either buyer or seller where points were earned for buying low or selling high. For each turn, written messages were produced by the source and transmitted via an experimenter. The points earned were translated into extra credit for the undergraduate participants. The present situation involved undergraduates who participated for extra credit earned for performance of the task of making requests of prior blood donors for a repeat blood donation. Messages were produced and transmitted in natural conversation. The crucial differences in the two studies originate in the genuine versus simulated nature of the tasks, the. magnitude of the requests, the relationship between the interactants, the range of potential final outcomes, the implications of success and failure, and the identity of the source as represented in the interaction. First, genuine and simulated interactions can be reasoned to differ substantially. Although points here are merely speculations as to the nature of those differences, one important difference stems from the absence of one to one transmission of vocal utterances in the Boster et al. (1993) simulation. Interactants are separated by time, space and medium in a way that does not exist in the present study. Secondly, the role play nature of the simulation might be a source of important differences, possibly a tendency by the source to envision and enact behaviors thought to be correct for the role being played 35 out that might differ from one's own behavior. In enacting a simulated role the source may feel less responsibility for the impact of one's messages, and subsequently be less careful in constructing and editing messages. Regardless of the true differences between a simulation and a genuine conversation, it is sufficient to note that differences surely exist and are grand enough to prohibit expectation of identical results. The magnitude of the request is conceived to be much larger in the present study than in the Boster et al. (1993) study and constitutes a second major difference in situations. Specifically, participants in the current study make requests of strangers for a blood donation. Compliance would require the donor to invest a substantial amount of time, create time in one's schedule, travel to a donation site, incur fatigue, and most importantly, subject oneself to intrusive procedures designed to remove an essential bodily fluid. In contrast, compliance with a request in the car sale negotiation simulation may result in a loss of extra credit points in a college class. Clearly, the larger request magnitude of the current study ought to be associated with differences in request strategies. The third important difference lies in the relationship between the interactants. In the present study the participants call strangers who are not expecting the call. Based on a review of the tapes of the calls it is the author's assumption that the participants were generally younger in age than the donors they called, and thus reflect the presence of status differences. In the Boster et al. (1993) study participants made requests of strangers who were also fellow college student participants with advance knowledge of the request, and who can be expected to be similar in age and status. The fourth situational difference derives from differences in potential fr_n_al_ outcomes. The final outcome in the Boster et al. (1993) was a negotiated price which could fall along a range of prices, blue book to retail. In the current study just two bipolar opposite @ outcomes were possible: the donor made an appointment to donate or the donor refused to do so. Obviously a range of responses would be transmitted throughout the course of any given call, thus the emphasis here is on the final response of the donor. 36 The fifth source of influential difference exists in the implications of success and failure in gaining compliance in the two studies. Success for the participant making a request for a blood donation brings about a prosocial and altruistic act that benefits someone other than the source. The participant's role in the act is indirect in that the participant is not the one donating blood, however, the success of the participant initiates a chain of events that benefit those in need of blood. In contrast, success for the participant in the Boster et al. (1993) study involves self benefit of extra credit. Failure in the simulated situation is by degrees, where less than optimal extra credit points may be earned. Failure of the source in the current study carries no significant implications for self-benefit, and instead carries influence for third parties in need of blood. Finally, the two studies differ in that participants in the car negotiation represented themselves (an identity that may be freed of some responsibility by role play) while participants in the request for a blood donation represented both themselves and the American Red Cross (an identity that may potentially add responsibility). Smpplative acpount pf reasons for rp’ cpnsistencies in light of situatipnal differences Having presented the inconsistent findings between Boster et al. (1993) and the present study, as well as a delineation of distinct features of the two compliance-gaining situations, it now possible to posit some interpretations to account for the foregoing discussion. Many of the differences argued to exist between the two studies can be coherently explained by politeness theory (Brown & Levinson, 1978). According to the theory persons are said to want both a positive face, which is the positive self-image claimed by interactants, and a negative face which is the claim to "freedom of action and freedom from imposition" (Brown & Levinson, 1978, p. 66). Brown & Levinson (1978) propose that some speech acts are inherently face threatening, for example, criticizing and requesting. These speech acts are labeled "face—threatening acts" (FI‘As). Most important to the discussion here is the assumption that the amount of face threat created by an FT A 37 is a function of three factors: the social distance between speaker and bearer, each parties‘ relative power or status, and the culturally constituted degree of imposition of a given act. According to the theory the act of request in compliance-gaining situations contains an inherent threat to the negative face of a target of the request. Thus, both studies under review should contain a threat to negative face. Where the two studies differ is in the amount of face threat that exists in the situations. One key situational difference referred to the magnitude of the requests, or the amount of imposition on the target, and it was argued that the size of the request in the present study is much larger than in the Boster et al. (1993) study. The amount of imposition present in a request is directly addressed by politeness theory as a functional source of face threat. Thus, the situation of the present study is argued to be more face threatening than the Boster et a1. (1993) situation due to the larger magnitude of the request. The relationship between the interactants comprised another important source of situational difference between the two studies, and is a functional source of face threat according to politeness theory. Specifically, the differences argued to exist stemmed from the donor as a stranger with no prior knowledge versus a co-participant as a stranger with prior knowledge. Hence, it is argued that the social distance between the interactants in the present study is greater than the social distance in the Boster et al. (1993) study. This increased distance is assumed to be associated with a greater face threat. Another aspect of the relationship between the interactants addressed by politeness theory is the relative status of the interactants. It was posited earlier that status differences were likely greater in the present study than in the Boster et al. (1993) study due to age. Hence, this status difference amounts to greater face threat in the present study. A final important source of difference between the two studies related to politeness theory is the range of potential final outcomes. Specifically, the differences were bipolar opposite outcomes for the present study versus a range of outcomes for the Boster et al. study. Negative face threats are defined as those that impede freedom of action (Brown & 38 Levinson, 197 8). It seems plausible that the current situation carries greater face threat as there is less freedom of choice between two options than a choice made from a range of options. Again it is argued that the face threat is greater in the present study. The foregoing argument posits that the situation of the current study carries inherent threats to negative face that exceed such threats in the Boster et al. (1993) study. The implication of this is that the greater threat to negative face in the current study is the source of inconsistent findings. Specifically, the account offered here is that the high argumentative, low verbal aggressive group was particularly sensitive to the threats and did not act as predicted or as consistent with Boster et al. (1993) because they were responding to the amount of face threat. The upcoming section provides a framework for the claim that the high argumentative, low verbal aggressive sources may be more sensitive to negative face threats and alter their compliance-gaining behavior. To provide a framework for interpretation of these findings it is helpful to look away from the argumentativeness and verbal aggressiveness literature and instead focus on findings from compliance-gaining literature in general. Research focused on compliance- gaining strategy use has posited that compliance-gaining strategies can be arranged along a single continuum with positive, prosocial strategies at one end, and negative, antisocial strategies at the other end (Hunter & Boster, 1978; Hunter & Boster, 1987). Hunter and Boster ( 1987) further posit that an individual's ethical threshold constitutes the midpoint of this continuum, and the ethical threshold is presumed to vary by individual and situation. The ethical threshold is that point at which available strategies are bisected such that strategies believed by the source to be acceptable fall above the ethical threshold and strategies believed by the source to be unacceptable fall below the ethical threshold. An individual who has a high ethical threshold in a given situation will consider fewer strategies to be acceptable in a given situation than an individual who has a has a low ethical threshold. 39 The ethical threshold model provides a possible explanation for the findings of both studies under review here. Hunter and Boster (1978) suggest individual difference factors and situational differences should be associated with an individual's ethical threshold. The individual difference variables of argumentativeness and verbal aggressiveness and situation factors of increased threats to negative face may constitute a source of varying ethical thresholds, and if so may account for the findings of both studies. It is possible to imagine the findings in terms of ethical thresholds. High argumentative, low verbal aggressive sources in the Boster et al. study revealed greater persistence, hence lower thresholds according to the model. This is as should be expected for the high argumentatives, who are defined as having a large repertoire of arguments. Hence, they possess a greater number of available arguments in the prosocial domain above the ethical threshold that they are willing to use given the minimal amount of face threat in the simulated car situation. Effects in the Boster et al. (1993) study were also found for the high argumentative, high verbal aggressive group that showed decreased persistence, hence a higher threshold. It is posited that sound arguments for this group are confounded with rudeness which they recognize to be antisocial, regardless of the amount of face threat limiting the number of prosocial strategies available to them. This finding was replicated in the present study. Similar visualization for the inconsistent findings of the present study in terms of the model yields consistency when paired with perceptions of negative threat. Specifically, greater sensitivity to the amount of negative face threat in the situation by the high argumentative, low verbal aggressive group should raise the ethical threshold attenuating persistence. This speculation is supported by previous findings of high argumentativeness attenuating willingness to use compliance-gaining strategies that create negative feelings in receivers (Infante & Rancer, in press). In this situation, given the magnitude of face threat to the donor, antisocial behavior may be seen by these sources as not only evident in any given strategy, but also as a function of using multiple strategies. That is, strategies that may qualify as prosocial by themselves become more antisocial in combination with one another. Moreover, once the donor resists compliance with a request to do a prosocial and altruistic deed it is necessarily more face threatening to persist and expose the donors unwillingness to be prosocial and altruistic. Conversely, the greater persistence found for the low argumentative, high verbal aggressive group is reflective of this group's failure to recognize the inherent face threats and that multiple and different strategies would aggravate the face threats despite a given strategy's social value on the continuum. Consequently a lower ethical threshold results. It would also be consistent with the findings to presume that the ethical threshold for this group may be constant across situations, and the differences between groups appeared due to high variation across situations for the high argumentative, low verbal aggressive group. The ethical threshold model also supports the findings of greater use of the two confrontive strategies (Request Specific Commitment and Refute Obstacles) by the low argumentative, high verbal aggressive group compared to the high/high and low/low groups, the assumption being that these strategies fall above the ethical threshold for the low argumentative, high verbal aggressive group and thus in the realm of acceptable strategies for use, but are considered unacceptable for the other groups falling below their ethical thresholds. This reasoning is also consistent with the possibility that the ethical threshold of the high argumentative, low verbal aggressive group is more variable such that in situations containing high threats to negative face, this group will drastically raise their ethical threshold while other groups may remain fairly constant and be more willing to be confrontive across situations. The second hypothesis predicted that the high argumentativeness low verbal aggressiveness group would be more focused on obstacles than the other trait combinations. This hypothesis found no support in any direction. It is possibly the case that the high threat to negative face resulted in a higher ethical threshold of this group 41 prohibiting their use of obstacle oriented strategies to avoid aggravating the face threat. It is probable that the ethical threshold of the more persistent, more confrontive group (low argumentativeness, high verbal aggressiveness) was high for this group as well, (they were not more obstacle focused) because they lacked the skill required to focus on obstacles, and not because they decided obstacle focused strategies were antisocial. Litni—tatm An obvious limitation of this study is the small sample size. The small sample size is particularly problematic when the sample is divided into the four groups of trait combinations, yielding as few as three participants in two of the groups. This limitation is partially offset by enhanced reliability gained from using multiple measures of each of the dependent variables. Confidence in the findings would be enhanced by increased sample size, as the possibility of outliers could be ruled out. An additional limitation of the study is its lack of attention to conversational sequencing. The point here is that although interactive data was obtained, the analyses of the data did not capitalize on the potential information to be gleaned. Specifically with regard to the operationalization of obstacle focus, the specific obstacles raised by the donors are not matched one to one to the obstacle strategies used by the donors. Further problems with the operationalization of this variable are discussed next as an additional limitation of the study. It is likely that the compliance-gaining task and the operationalization of obstacle focus did not allow for differences among groups to emerge. In looking at the strategies that comprised a portion of the measure it became apparent that two of the obstacle strategies would be used by virtually all participants regardless of their individual difference traits. Specifically, they are the Range of Options and Third Party strategies. Wilson et al. (1994) found high frequency of use for both of these strategies compared to the other strategies. Recall that the Wilson et al. ( 1994) study examined much of the same data analyzed here. Additionally, the Wilson et al. (1994) study found that the Range of 42 Options strategy was frequently used as both a second and third strategy in consecutive attempts to secure an appointment. The authors also found that the Third Party strategy was often used as the last strategy before termination of the interaction. Findings of frequent use of these two strategies combined with the realization that in order to schedule an appointment virtually all sources would necessarily have to provide information about the range of options (e.g., donor room hours) helps to explain the lack of findings for this hypothesis. It might be better to operationalize obstacle focus as a ratio of the use of the Clarify Obstacles and Refute Obstacles strategies to all strategies used. A source would not necessarily have to use these strategies in order to carry out the task. This operationalization might also better tap the obstacle focus observed by Wilson et al. ( 1993) which was that some sources sought to clarify and refute obstacles. An additional point on this hypothesis is that in the Wilson et al. (1994) study the largest percentage of obstacles raised by donors involved medical ineligibility (e.g., recent donation, pregnancy) thus placing a ceiling on opportunities to refute or clarify the obstacles raised. These possibilities need to be considered with discretion however, as the additional transcripts included in the study were not analyzed for the obstacles raised or the order of strategy use by the volunteers. Future Directions It was posited in this chapter that the amount of face threat present in situations may alter an individual's ethical threshold depending on the individual difference traits of the individual. Much of the reasoning is highly speculative at this point, and therefore the goal of future research should be to test the hypothesized relationships between argumentativeness and verbal aggressiveness with regard to perceptions of face threats in various situations. Specifically, the research design should include a high threat and low threat situation with predictions for persistence, obstacle focus, and use of confrontive strategies in each condition by individual difference variables. It would be advantageous 43 to know whether the ethical threshold is subject to more variation for some levels of and interactions among individual difference traits than for others. Questionnaires might be designed to gauge perceptual differences of the high and low face threat conditions of individuals with differing levels of trait combinations. To explore the possibility of constancy of ethical thresholds versus variability associated with individual differences, predictions could also be made for specific traits across a variety of communication situations. The implications of such future work might be to determine appropriate forms of communication skill enhancement training and for who they might be most beneficial to, for example, the high argumentative, high verbal aggressive source who has been speculated to have particular difficulty producing socially appropriate messages. Additional comments are made next regarding an interesting result found while performing "snooping" analyses. Post hoc analyses searching for any significant differences between argumentativeness and verbal aggressiveness and strategy use revealed significant negative correlations between trait verbal aggressiveness and the use of the Express Gratitude strategy and use of the Other Benefit strategy. Note that this effect was for trait verbal aggressiveness by itself and was not examined as an interaction with argumentativeness. Use of the Express Gratitude strategy seems to be a sort of reward strategy, perhaps an instantiation of either the positive affect (use of charm) or pregiving (doing positive and nice things in advance of request) strategies (Kellerman & Cole, 1994). Use of the Other Benefit strategy seems to involve a realization of the significance of compliance in the broader realm of human need, going beyond the importance of simply securing compliance via an appointment. Use of both of these strategies tended to co—occur with a General Request strategy (Wilson et al., 1994) suggesting that those who use them might be more able to handle multiple goals in communication. The corollary of this is that those who fail to use the Other Benefit and Express Gratitude strategies might be less able to handle multiple goals, take broad alternative perspectives, and produce reward strategies. Related findings suggest that individuals high in verbal aggressiveness are less sensitive to the hurtfulness of a message to the target than are low verbal aggressives (Infante et al., 1992). It is possible that they are also less likely to recognize when and how to attenuate the hurtfulness or face threats of messages via prosocial strategies. The majority of the research on verbal aggressiveness addresses the propensity to use negative, antisocial strategies as opposed to the failure to generate positive, prosocial strategies. One study that does examine this direction found a significant negative relationship between verbal aggressiveness and laudativeness (predisposition to praise verbally) (W igley, Pohl & Watt, 1989). These findings suggest that one area future research should further explore is the verbal aggressiveness trait as a deficit in producing prosocial communication as well as a tendency to produce antisocial communication. Such a deficit might be explored in situations where the production of prosocial messages is expected, such as a comforting situation, where differences among individuals with differing trait levels could be examined. Conclusions The present chapter has sought to sort out perplexing and inconsistent findings in a comprehensible manner. To do this similarities and differences with prior findings were presented, followed by speculations regarding perceptions of face threats and altering of ethical thresholds that might explain inconsistent findings. The study’s limitations were acknowledged, providing suggested areas and means for improvement. Suggestions for future research endeavors concerning the argumentativeness and verbal aggressiveness traits in compliance-gaining behavior were made to provide direction for motivated persons willing to begin testing the speculated explanations of the results. In sum, the present endeavor has failed to replicate prior work in the manner desired, but it has illuminated potential fruits of the communication research field. Appendix A Categories of Obstacles Raised by Donors 1 No, Not Willing 2 Health Obstacles a Prior Bad Experience b Medically Ineligible 3 Inconvenience Obstacles a) Work Prevents b) General Schedule Prevents c) Location Prevents 4 Targeted Donation Obstacles a) Only Give At Alternative Cite b) Only Give For Specific Individuals 45 Appendix B Nine Compliance-Gaining Strategies 1 Request Strategies a) General Requests b) Request Commitment 2 Resource Strategies a) Downgrade costs 3 Obstacle Strategies a) Clarify Obstacles b) Information about Options c) Third Party (1) Refute Obstacles 4 Normative Strategies a) Other Benefit b) Express Gratitude 46 LIST OF REFERENCES Bayer, C.L., & Cegala, DJ. (1992). Trait verbal aggressiveness and argumentativeness: Relations with parenting style. Western Journal of Communication, M, 301-310. Blimes, J. (1988). The concept of preference in conversation analysis. Language in Society, M, 161-181. Boster, El. (1990). An examination of the state of compliance-gaining message behavior research. In J. P. Dillard (Ed.), Seeking compliance: The production of interpprsonal influence messages (pp. 7-20). Scottsdale, AZ: Gorsuch Scarisbrick. Boster, F.J., & Levine, TR. (1988). Individual differences and compliance-gaining message selection: The effects of verbal aggressiveness, argumentativeness, dogmatism, and negativism. Communication Research Remrts, _5_, 114-119. Boster, F.J., Levine, T., & Kazoleas, D. (1993). The impact of argumentativeness and verbal aggressiveness on strategic diversity and persistence in compliance-gaining behavior. Communication Quarterly, 4_l_, 405-414. Burleson, B.R., Wilson, S.R., Waltrnan, M.S., Goering, E.M., Ely, T.K., & Whaley, BB. (1988). Item desirability effects in compliance-gaining research: Seven studies documenting artifacts in the strategy selection procedure. Human Communication Research, M, 429-486. deTurck, MA. (1985). A transactional analysis of compliance-gaining behavior: Effects of non-compliance, relational contexts, and actors' gender. Human Communication Research, _1_2, 54-78. Goodwin, M. H. (1990). He-said-she-said: Talk as social organization among Black ghildren. Bloomington, IN: Indian University Press. Goffman, E. (1967). Interaction ritual: Essays in face to face behavior. Garden City, NY: Doubleday Anchor. Grice, H. P. (1975). Logic and conversation. In P. Cole & J. Morgan (Eds), Syntax and semantics, (vol. 3. PP.41-58). New York: Academic Press. Hunter, J .E., & Boster, El. (1978). An empathy model of compliance-gaining strategy selection. Paper presented at the annual meeting of the Speech Communication Association, Minneapolis, MN. 47 48 Hunter, J .E., & Boster, F.J. (1987). A model of compliance-gaining message selection. Communication Monographs, 3, 63-84. Infante, DA. (1988). Argping constructively. Prospect Heights, IL: Waveland Press. Infante, D. A., Chandler, T. A., & Rudd, J .E. (1989). Test of an argumentative skill deficiency model of interspousal violence. Communication Monographs, E. 163- 177. Infante, D.A., & Rancer, AS. (1982). A conceptualization and measure of argumentativeness. Journal of Personality Assessment, 46, 72-80. Infante, D.A., & Rancer, AS. (in press). Argumentativeness and verbal aggressiveness: A review of recent theory and research. Communication YearboolMQ. Infante, D.A., Riddle, B.L., Horvath, C.L., & Tumlin, SA. (1992). Verbal aggressiveness: Messages and reasons. Communication Ouarterle, 49, 116-126. Infante, D.A., Trebing, J .D., Shepherd, P.E., & Seeds, DE. (1984). The relationship of argumentativeness to verbal aggression. The Southern Smch Communication Jopmal, 50, 67-77. Infante, D.A., & Wigley, CH. (1986). Verbal aggressiveness: An interpersonal model and measure. Communication Monographs, 13, 61-69. Kellerman, K.K., & Cole, TC. (1994). Classifying compliance-gaining messages: Taxonomic disorder and strategic confusion. Communication Theogé, 3-60. Lim, TS. (1990). The influences of receivers' resistance on persuaders' verbal aggressiveness. Communication Quarterly, 3, 170-188. Manusov, V. ( 1989). How can you resist me? Compliance resistance among friends. Paper presented at the meeting of the International Communication Association, San Francisco, California. O'Keefe, B. (1988). The logic of message design: Individual differences in reasoning about communication. Communication Monographs, fi, 80-103. Rancer, A.S., Baukus, R.A., & Infante, DA. (1985). Relations between argumentativeness and belief structures about arguing. Communication Education, 3_4, 37-47. Sabourin, T.C., Infante, D.A., & Rudd, J.E. (1993). Verbal aggression in marriages: A comparison of violent, distressed but nonviolent, and nondistressed couples. Human Communication Research, _2_Q, 245-267. 49 Schegloff, E., & Sacks, H. (1973). Opening up closings. Semiotica, 1, 289-327. Seibold, D.R., Cantrill, J.G., & Meyers, R. (1994). Communication and interpersonal influence. In M.L. Knapp & G.R. Miller (Eds.), flandbookpf interpersonal communication (2nd ed., pp. 542-588). Thousand Oaks, CA: Sage. Weiner, B. (1985). An attribution theory of achievement motivation and emotion. Psychological Bulletin, 92, 548-573. Wigley, C. J., Pohl, G.H., & Watt, M.G.S. (1989). Conversational sensitivity as a correlate of trait verbal aggressiveness and the predispositionto verbally praise others. Communication Research Repprts, 2, 92-95. Wilson, S.R., Cruz, M.G., Marshall, L.J., & Rao, N. (1993). An attributional analysis of compliance-gaining interactions. Communication Monographs, M), 352-372. Wilson, S.R., Levine, K., Humphreys, L., & Peters, H. (1994). Seeking and resisting compliance: Strategies and sequences in telephone solicitations for blood donation. Paper presented at the annual meeting of the Speech Communication Association, New Orleans, Louisiana. ICHIGQN STRT E UNIV. LIBRARIES l“ H 1WlllWlWlHlll"lllllUHll 9 544574 13112 13(31!5