llllllllllHllllllllllllllIllIHllllllllllllllllllllllllllllll 31293018017354 LIBRARY Michigan State University This is to certify that the dissertation entitled The Experiential Functions of Attitudes presented by Betty Helen La France has been accepted towards fulfillment of the requirements for Ph.D. Communication degree in jet/M [$3751 mm professor Date ”/6/9 3 MSU i: an Affirmative Action/Equal Opportunity Institution 0-12771 PLACE IN REI‘URN BOX to remove this checkout from your record. To AVOID FINE return on or before date due. MAY BE RECAUJ-ZD with earlier due date if requested. DATE DUE DATE DUE DATE DUE 1M Macs-p.14 THE EXPERIENTIAL FUNCTIONS OF ATTITUDES By Betty Helen La France A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Communication 1998 ABSTRACT THE EXPERIENTIAL FUNCTIONS OF ATTITUDES By Betty Helen La France This study examined the experiential attitude functions. When people hold an attitude based on past experience(s), that attitude is said to serve an experiential function. It was anticipated that messages that contain function-relevant information would be more persuasive than messages that contain function-irrelevant information. In addition. the more discrepant is one’s position from the position advocated in the message, the greater the conformity to message recommendations is expected, and this relationship is expected to be more substantial for strong arguments than for weak arguments. Each of 139 undergraduates were assigned randomly to one of four conditions in which a fictitious Executive Committee for Academic Integrity (ECAI) report, written for a university president, supplied strong/weak and pro-tenure/anti-tenure messages. Results indicated that these data were not consistent with the matched hypothesis. Conformity to message recommendations was contingent upon the discrepancy between the subject’s initial attitude and the position advocated in the message. Interestingly, the mean attitude change in every experimental condition was negative, indicating that regardless of the position advocated in the message, subjects became increasingly anti-tenure. Dedicated to my dad iii ACKNOWLEDGMENTS I would like to begin by thanking my advisor, Dr. Franklin J. Boster, to whom I am forever indebted for teaching me how to become a scholar. In addition, I also must thank my committee, Drs. Charles Atkin, David Imig, and Sandi Smith. for their patience and timely review of this work and its previous drafts. Thank you to Sean for being the brother I never had with all his help and support and to Marge (mom), Victoria, Craig. Jen, Cynthia, Jenn, Mo, and Chad for their patience and assistance. Melissa, it was because you loved me that I was able to finish this chapter of my life. To Bill, this work would have not been possible if it were not for your love and support; you never doubted me. Thank you for being my greatest fan. iv TABLE OF CONTENTS LIST OF TABLES ............................................................................. vii LIST OF FIGURES ........................................................................... viii CHAPTER 1 INTRODUCTION .......................................................................... 1 LITERATURE REVIEW ................................................................... 2 The Functional Perspective ........................................................ 2 The Neofunctional Perspective ................................................... 2 The Processes ....................................................................... 7 Valence of the Experiential Functions ........................................... 12 i Message Design ..................................................................... l3 Functional Target of Message ............................................ l3 : Argument Strength ......................................................... l6 2 Message Advocacy ......................................................... 17 CHAPTER 2 METHOD ..................................................................................... 2O Pretesting ............................................................................ 20 Argument Strength ........................................................ 20 Pro-tenure/Anti-tenure Message ......................................... 21 Subjects ............................................................................. 22 Design .............................................................................. 22 Procedure .......................................................................... 23 Instrumentation .................................................................... 24 Attitude ..................................................................... 24 Function ................................................................... 24 Argument Strength ...................................................... 24 Pro-tenure/Anti-tenure Message .......................................... 24 Need for Cognition ...................................................... 25 Message Discrepancy ................................................... 25 Attitude Change ......................................................... 25 Function Change ......................................................... 25 CHAPTER 3 RESULTS .................................................................................... 27 Measurement Model ................................................................ 29 Evaluating the Inductions ......................................................... 34 Evaluating the Hypotheses ........................................................ 39 Change in the Basis of the Attitude .............................................. 45 v CHAPTER 4 DISSCUSSION .............................................................................. 47 To Match or Mismatch ............................................................ 47 Discrepancy ........................................................................ 48 Positive Attitude, Negative Conformity to Message Recommendations 49 What? Functions Change? ....................................................... 50 Limitations ........................................................................... 52 APPENDIX ............................................................................... 55 REFERENCES .............................................................................. 67 FOOTNOTES ............................................................................... 70 vi LIST OF TABLES TABLE 1 The Prediction for the Relationship between Discrepancy and Attitude Change ............. 2 8 TABLE 2 Descriptive Statistics for the Constructs ....................................................... 30 TABLE 3 Confirmatory Factor Analysis Results for Pretest and Posttest Attitude Measures ........... 3 1 TABLE 4 Confinnatory Factor Analysis Results for Pretest and Posttest Function Measures .......... 3 2 TABLE 5 Confirmatory Factor Analysis Results for the Perceived Argument Strength Measure ...... 33 TABLE 6 Confirmatory Factor Analysis Results for the Perceived Message Advocacy Measure ...... 32 TABLE 7 Perceived Argument Strength Index Broken Down by the Advocacy and Argument Strength Inductions ............................................................................ 3 5 TABLE 8 Perceived Message Advocacy Index Broken Down by the Advocacy and Argument Strength Inductions ............................................................................ 3 7 TABLE 9 Pretest Attitude Index Broken Down by the Advocacy and Argument Strength Inductions ...................................................................................... 38 TABLE 10 Posttest Attitude Index Broken Down by the Advocacy and Argument Strength Inductions ...................................................................................... 40 TABLE 1 1 Mean Attitude Change Broken Down by the Advocacy and Argument Strength Inductions ...................................................................................... 42 Table 12 Unstandardized Slopes of Posttest Attitude on Message Discrepancy ....................... 44 vii LIST OF FIGURES FIGURE 1 Typologies Attitude Functions ............................................................. 3 FIGURE 2 Herek’s (1986) Neofunctional Theory .................................................... 5 FIGURE 3 The Experiential Functions Continuum ................................................. l 1 FIGURE 4 Study Predictions ............................................................................ 19 viii CHAPTER 1 INTRODUCTION The primary purpose of the current project is to test the hypothesis that messages that correspond with an attitude function (match) will be perceived as more persuasive than messages that do not correspond with that function (mismatch). The functions that will be studied here are two generated by Herek (1986): the experiential-schematic and experiential-specific functions. In the next sections the functional and neofunctional perspectives will be described, and the experiential functional process will be detailed. Next, three features of message content will be discussed: the functional target of the message (i.e., schematic or specific), the strength of the argument in the message (i.e., strong or weak), and the position advocated in the message (i.e., proattitudinal or counterattitudinal). Next, the relationships among the experiential functions, functional target of message, argument strength, and message advocacy will be detailed and specific testable hypotheses offered. The method for testing these hypotheses is then examined and the results from these analyses are discussed. LITERATURE REVIEW The Functional Perspective The fundamental tenet of the functional approach is that attitudes are held because they serve a purpose or purposes. This approach began in the late 19505 and early 19605 with two independent research programs, that of Smith, Bruner, and White (1956) and of Katz (1960). Although several scholars have identified various functions (Herek, 1986; Katz, 1960; Smith, Brunet, & White, 1956), many of these functions are consistent across typologies, sharing the same content if not the same label (see Figure 1 for a comparison). An attitude can elicit a single or multiple functions. In addition, various attitude objects may elicit different functions. A focus on the former, termed the personal characteristics (Herek, 1986) or person variations approach (Shavitt, 1989), describes under what circumstances functions should develop for a person with specific predispositions. For instance, ego-defensive attitudes should be held by those who extemalize their intrapsychic conflicts or project their fears onto others. A focus on the latter, termed domain characteristics (Herek, 1986) or object variations (Shavitt, 1989), describes how multiple attitude functions toward an attitude object exist. For example. attitudes concerning war might evoke experiential, utilitarian, and value-expressive functions. The Neofunctional Perspective Herek’s (1986) neofunctional theory views all attitudes as inherently utilitarian in the sense that they, “are strategies for satisfying psychological needs” as well as gaining Function Smith, Bruner, Katz (1960) Herek (1986) Definition & White (1956) Attitudes that help Obj cot-appraisal Knowledge/ Experiential organize reality; to Utilitarian Specific/ Schematic make sense of the environment in accord with our ongoing interests; based on personal experience (s). Attitudes that are strategies for coping with intrapsychic conflict by denying, repressing, or projecting concerns onto others in defense of one 's self-concept. Attitudes that help facilitate interpersonal relationships via gaining social acceptance or aligning oneself with important referent groups. Figure 1. Typologies of Attitude Functions. Extemalization Social adjustment Ego-defensive Value- expressive Ego-defensive Value/ Social expressive some type of benefit from either possessing or expressing the attitude (p. 99). In addition, he suggests that people hold specific attitudes for their symbolic or instrumental value. Symbolic attitudes meet personal needs that are satisfied by the expression of such attitudes. Instrumental attitudes, Herek (1986) argues, are based on whether the attitude object is detrimental or beneficial to the self. He arranges these two different types of attitudes as a taxonomy where the need for each is considered and used subsequently to determine the type of benefit derived from each (see Figure 2). To the extent that there is a high need to express a particular attitude, then the attitude will be served by different expressive functions. The expressive function category provides a means to an end where social support and internal harmony (as opposed to anxiety) are the needs to be met. The first function identified under this categorization includes the social-expressive function where the need to be accepted and gain acceptance within one’s environment is dominant. The second function, termed value-expressive, is based on the need to construct oneself by expressing significant values and by being positioned with important reference groups. The third and final expressive function is the ego-dc ensive function where attitudes are coping strategies for intrapsychic conflict. To the degree that there is a high need to have an attitude because the perceived valence of the attitude object is based on the rewards and punishments associated with that object, then the attitude will be served by different evaluative functions. An attitude formed via the evaluative functions, in contrast to expressive functions, is an end in and of itself. Herek (1986) identifies the anticipatory function where attitudes are established based on an expected future utility anticipated from holding such an attitude. Instrumental Attitudes: those that derive benefit from the attitude object itself. Low need High need Symbolic Attitudes: 5 those that High need EXPRESSIVE: BOTH derive benefit Social i from the Value expression of Ego-defensive attitudes. ............................................................... Low need NON FUNCTIONAL EVALUATIVE: i Experiential- Specific/Schematic Anticipatory Figure 2. Herek’s (1986) Neofunctional Theory. The second and third functions, labeled experiential-specific and experiential- schematic, are the foci of the current project. Figure 2 classifies expressive and evaluative attitude functions. Herek (1986) distinguishes between the two experiential functions because, although each is based upon an encounter with an attitude object, the attitude formed from that interaction differs. Specifically, a function that is experiential and specific is one in which “... after interacting with a particular instance of the attitude object category. it is treated as a unique entity, differentiated from its membership in the larger category, and evaluated in terms of its individual utility for the person” (Herek, 1986, p. 105). This type of attitude function can be expressed in the following example: Professor Armstrong is quite different from professors I’ve had for previous classes. Unlike other professors at the university, he makes me think about the possibility that there is life on other planets and in other galaxies. He stresses how I should grow spiritually. I really enjoy going to class because he makes learning fun. This explanation and example can be contrasted with a function that is experiential and schematic where “... treating the attitude object as representative of a larger category perceived as either beneficial or detrimental to oneself. In this case, past experiences with representatives of the category have lead to the development of a cognitive schema that guides subsequent interactions with members of the category” (Herek, 1986, p. 105). In the essays that Herek (1987) gathered from undergraduate students regarding attitudes toward homosexuality, the following is an excerpt explaining a subject’s positive attitude toward all homosexuals based upon her experience with a gay friend: “. .. I have come to know some of these people and find them no different from any other people. Little did I know that the guy in the next room [in college] was gay. We became good friends and did things together all the time. Eventually he told me [that he was gay] and it was then that I realized that homosexuals only differ in sexual preference” (p.288). Whereas Herek (1986) describes each of the experiential functions, he does not describe the process by which people develop either an experiential-specific or experiential-schematic attitude function. These processes are described below. The Processes Although Herek (1986) does not outline the conditions under which people‘s attitudes are classified as either experiential-schematic or experiential-specific. his definitions conceptually parallel the category-based and attribute-based process of impression formation. The former emphasizes top-down cognitive processes where impression formation occurs “... according to the categories into which [individuals] place other people,” (F iske, Neuberg, Beattie, & Milberg, 1987, p. 400). The latter. attribute-oriented perspective “... holds that people form impressions through an attribute- by-attribute consideration of isolated pieces of information” (Fiske et al.. 1987, p. 401). Fiske et al. (1987) and Verplanken, Jetten, and van Knippenberg (1996) provide evidence that suggests people use both categorical and attribute processes. For example, Fiske et a1. (1987) induced category-attribute consistency and inconsistency in which the consistent condition category of loan shark included the attributes opportunistic, shady, greedy, shrewd, and heartless; whereas, the inconsistent condition consisted of the label doctor, and included the attributes bored, obedient, unenterprising, uneducated. and efficient. Fiske et al. (1987) found that categories were used to form impressions of others rather than attributes when categorization was easy; however, when categorization was difficult, people relied on attributes to make judgments. They went on to state that: It is important to reiterate that subjects used both types of information in all conditions, but the relative emphasis depended on the information configuration. The current hypotheses seem useful especially as part ofa heuristic continuum that explains the current results and that integrates previous lines of research (Fiske et al.. 1987, p.422). The process for the experiential attitude functions may mirror this categorical- attribute continuum theory of impression formation offered by Fiske et al. (1987). Conceptualizing the experiential functions in a similar manner, as anchors of a continuum rather than two distinct dimensions, might be more useful than thinking of them as two distinct, independent functions. Similar to the continuum theory of impression formation, the degree to which people are willing to generalize their attitudes may vary depending on the attitude object being considered. According to this experiential function continuum, contact with a particular attitude object first would lead subjects to try to place that attitude object in a category. As cognitive misers “... people’s preferred mode is to categorize [others] whenever possible” (Fiske et a1. 1987, p.401). Then, a determination would be made as to whether or not characteristics of the attitude object were consistent with the category into which the object was assigned. To the extent that there was category-attribute (characteristic) fit, or consistency, people would form attitudes based on the category. Thus, the attitude toward the object would be included in that category. An example of this process is exemplified in one respondent’s essay about her dislike of male homosexuals. Herek (1987) reports the respondent’s sentiments: Personally, I don’t like most male homosexuals. I once worked under one and worked with some and they were everything homosexuals are stereotyped to be-- someone once said ‘male homosexuals have all the bad qualities of women‘ (shrewishness, pettiness, etc.)-- and unfortunately, for the men I worked with this statement applied... (p. 288) Alternatively, to the extent that there is poor fit, or category-attribute inconsistency, people would consider individual characteristics in forming an attitude. In this situation, people would form their attitude independent of any representative category. One example is expressed in the following statements: We must look at the Vietnam War as being different from all other wars that preceded and followed it. Its uniqueness distinguishes it from other wars because of several characteristics. For example, participation in the armed services to fight the enemy was involuntary. These descriptions are examples of the anchors of the experiential continuum, with the former an example of experiential-schematic thinking and the latter as an example of experiential-specific thinking (see Figure 3). It is not difficult to think of instances in which the function may be more or less experientially-schematic or experientially-specific. For example, using the excerpt from Herek (1986) above. although it is possible that male homosexuals could be grouped under a single broad category, people also may have subcategories in which they place homosexuals (e. g.. masculine, feminine). This subcategorization would lead to a function that is experientialIy-schematic; however, the schema is less general. As can be seen from these examples, the degree to which an attitude is experientially-specific or experientially-schematic based, depends upon the degree to which stimulus generalization or stimulus discrimination occurs. Miller and Steinberg (1975) describe stimulus generalization as a process of abstraction in which commonalities between and among a group of objects are recognized and differences between these same objects are ignored. Conversely, to the extent that distinguishable characteristics between attitude objects are sought, stimulus discrimination occurs. 10 4 ’ Schematic Specific Example: Example: Personally, I don 'I like We must look at the most male Vietnam War as being homosexuals. I once worked under one and worked with some and they were everything homosexuals are stereotyped to be-- someone once said ‘male homosexuals have all the bad qualities of women ' (shrewishness, pettiness, etc. )-- and unfortunately, for the men I worked with this statement applied Figure 3. The Experiential Functions Continuum. ll different/mm all other wars that preceded and followed it. Its uniqueness distinguishes it. from other wars because of several characteristics. For example, participation in the armed services to fight the enemy was involuntary Valence ofthe Experiential Functions To develop his Attitude Function Inventory, Herek (1987) coded essays for their functional content as well as the positive or negative valence attached to each of the functions. Interestingly, the experiential-schematic function, but not the experiential- specific function, was assigned a valence depending on the positive or negative attitude toward the object’s representative category expressed in the essay. Conceptualizing the experiential functions along a vertical continuum, with the valence of an attitude object aligned perpendicular to that continuum, offers an explanation as to why a valence is not attached to the experientially-specific function. Consider an attitude object that serves an experientially-schematic function. Whatever attitude, positive or negative, that is associated with an object will be generalized to the object’s representative category. For example, if a student believed a professor was lazy, did not like students, and was unfair, and this attitude served an experientially-schematic function, then the attitude toward all other professors would be negative also. Conversely, an attitude object that was the foundation for an experientially- specific function will not be generalized and therefore any valence, positive or negative. attached to the object will not be generalized either. Using the above example. a student who viewed a professor as lazy, did not like students, and was unfair, and whose attitude served an experientially-specific function, would not positively or negatively evaluate all professors. As such, in this case, evaluation of the attitude object is limited to that specific object. 12 In these cases the attitude valence of the category is at issue. This consideration has an important implication for the effects of messages on conformity to message recommendations. According to the explanation given above, a positive or negative evaluation of an object’s representative category can occur only with people whose attitude functions experientially and schematically. Thus, the valence assigned to the object’s representative category should be neutral for subjects whose attitude functions experiential-specifically. The main purpose of the current study is to identify which messages are more persuasive, given that an attitude is more experiential-specifically or experiential- schematically based. The issue of message design, including the functional target of message, argument strength. and message advocacy is discussed subsequently. Message Design Functional Target of Message. One implication of the functional approach to attitudes is that messages whose content appeals to a particular function served by that attitude will be more persuasive than messages that do not address the function (Herek. 1987; Katz, 1960). For example, someone who has joined The 700 Club because of the need to demonstrate a certain level of religiosity (e.g., value-expressive) will be persuaded more by a message whose content addresses this need for identification with religious groups than a message targeting psychodynamic processes (e.g.. ego-defensive). Message-function matching produces more persuasion because the arguments in the message are processed in a biased fashion. Specifically, messages that target a person’s attitude function are scrutinized more closely than messages that do not target the function (Lavine & Snyder, 1996; Petty & Wegener, 1998). Therefore, subjects 13 become more susceptible and responsive to messages that are functionally relevant (Snyder & De Bono, 1985). Research investigating this matched versus mismatched hypothesis has focused largely on the expressive or defensive functions. For instance, Snyder and De Bono (1985) found that high self-monitors, hypothesized to have attitudes based on the social- adj ustive/expressive function, were persuaded by messages that appealed to an object's image. Alternatively, low self-monitors, hypothesized to have attitudes based on the value-expressive function, were found to be persuaded by messages containing information about the quality of a product. In short, messages that targeted the object‘s image persuaded high self monitors, whereas messages that targeted the object’s quality persuaded low self monitors. Moreover, Petty and Wegener (1998) found that strong messages were persuasive if they matched respondents’ underlying social-expressive (high self-monitors) or value-expressive functions (low self-monitors). Conversely. when weak messages were used, matched messages were less persuasive than mismatched messages due to greater message scrutiny. Katz, McClintock, and Sarnoff (1957) investigated the impact of messages on people who were classified as high, moderate, or low ego-defensives. The content of these messages described the cognitive, psychodynamic process of defense mechanisms as they related to prejudice. Using several different measures of ego-defensiveness (including TAT cards, a subscale of the MMPI, the F Scale, the Negro Stereotype Scale. and the Bogardus Social Distance Scale), Katz et a1. (1957) provided some evidence that suggested subjects who were classified as moderately ego-defensive did demonstrate some attitude shifts about black people after receiving the message (message-function 14 match). Katz et al. (1957) argued that low ego-defensives would not be persuaded because the ego-defensive function was not the basis of their attitudes. Alternatively, high ego-defensives would be too stimulated by any material regarding prejudice resulting in a heightened level of defensiveness. As predicted, low and high ego- defensives showed little or no attitude change. Arguing that certain attitude objects tend to serve a single function, Shavitt (1990) demonstrated that attitudes about air conditioners and coffee were primarily utilitarian I based; whereas attitudes regarding greeting cards served more of a social-expressive function. A fourth product, perfume, was based equally from utilitarian and social- expressive functions. Furthermore, Shavitt (1990) found that advertisements for these ] products that contained function-relevant information were perceived by subjects as being more persuasive than advertisements that included function-irrelevant information. ' Message effectiveness of the experiential functions has not enjoyed widespread empirical investigation. Therefore, the current study attempts to fill this void by investigating message persuasiveness via the experiential-specific and experiential- schematic functions. Based on previous research, it is hypothesized that messages that address a person’s attitude function (match) will be perceived as more persuasive than messages that do not address that attitude function (mismatch) (Katz, McClintock, & Sarnoff, 1957; Snyder & De Bono, 1989). This hypothesis is consistent with Petty and Wegener’s (1998) findings. They noted that the effect of message strength can either enhance or detract from the functional match or mismatch of a particular message. Therefore, in addition to the functional target of the message, argument strength is 15 hypothesized to have an impact on effectiveness. Thus. argument strength is discussed next Argument Strength. Another element of a persuasive message is the strength of the arguments presented in the message. Not only is it important for a message to target the function of an attitude, but also it is critical that the message be cogent. Although the impact of argument quality on attitude functions has been examined (Petty & Wegener, 1998), what constitutes a strong argument has varied. Allen and Burrell (1992) distinguish between evaluating the structure of an argument and the persuasiveness of an argument. The former refers to formal logic in which a conclusion follows or does not follow from premises. A strong argument is an argument in which a conclusion follows from premises, and the premises are corroborated with evidence. The latter criterion, the persuasiveness of an argument, Allen and Burrell (1992) state as the concern with whether or not the receiver decides to believe the conclusion reached in the message. The persuasiveness of the message is defined as the degree to which subjects conform to message recommendations. Alternatively. Cacioppo, Petty. and Morris (1983) have measured argument quality, asserting that strong arguments evoke more favorable than unfavorable statements and weak arguments elicit more unfavorable than favorable statements on thought listing tasks. In previous work, however, Petty and Cacioppo (1979) constructed strong messages by incorporating evidence such as statistics and relevant studies and constructed weak messages using quotations and opinions. 16 Both argument structure and content comprise argument strength. Specifically, strong arguments are those that are sound. That is, they are logically valid if deductive or logically strong if inductive, and their content includes compelling evidence for the premises. Accordingly, the degree to which a message is perceived as being persuasive is hypothesized to be impacted by the functional target of the message. argument strength. and the position advocated (proattitudinal or counterattitudinal). The current study hypothesizes that strong messages will enhance the persuasiveness ofa matched message g (Petty & Wegener, 1998). Alternatively, a weak and matched message will produce less ' conformity to message recommendations than a strong and matched message. . Mismatched messages are expected to produce no change. The third and last element of message design to be discussed is the position advocated in the message. and it is described next. Message Advocacy. A third factor hypothesized to affect the suasory effect of messages is the position advocated by the message. A proattitudinal message contains a statement or statements with which the receiver is generally in accord; conversely. a counterattitudinal message contains information that challenges the receiver’s position on a topic. French (1956) suggests a linear discrepancy model of attitude change where the greater the distance between an initial position and the position advocated in the message. the greater the attitude change toward the message’s position. For example, ifa student was moderately in favor of universities retaining tenure for its faculty members, and the student was presented with a counterattitudinal message (e.g., a message that advocated 17 the abolishment of tenure), that person’s attitude would change toward the message so that the student would be less in favor of tenure for university faculty. Figure 4 shows the predictions for this study. Overall, the greatest conformity to message recommendations is expected only when the function and message is matched. discrepancy is large, and messages are strong. Matched weak messages, in which there is little discrepancy from subjects’ initial positions, will produce little conformity to message recommendations. If the data are consistent with these hypotheses, then those subjects in the strong, counterattitudinal message condition should experience the greatest conformity to message recommendations (Petty & Wegener, 1998). Subjects in the weak, counterattitudinal message condition should only display little conformity (French, 1956). There should be no conformity to message recommendations in the mismatched conditions. To this point, the functional and neofunctional approaches to attitudes have been described. The process underlying two important attitude functions. the experiential- specific and experiential-schematic functions, has been elucidated. The current study tests the proposition of the functional approach that messages tailored to the function of an attitude (match) will be more persuasive than messages that do not address the attitude function (mismatch). In addition, the strength of the argument and message advocacy are expected to affect conformity to message recommendations. No persuasion is hypothesized for subjects with experiential-specific functions. The following section describes the method used to test these anticipated relationships. 18 Experiential-Schematic Message Weak Message Strong Message Function Message Discrepancy Message Discrepancy Low High Low High Experiential- No change Some conformity No change Conformity to Schematic to message message recommendations recommendations Experiential- No change No change No change No change Specific Figure 4. Study Predictions. 19 CHAPTER 2 METHOD The functional target of the message was controlled in the study. Specifically, messages were designed so that all subjects would respond to a message that contained statements reflecting an experientially-schematic function. These statements advocated a particular position for all attitude objects that comprised a given category. One suggestion for functional research is that an induction will be most successful if it focuses on an attitude object with which subjects are unfamiliar (Shavitt. 1989). The challenge of the current study was that having experience with an attitude object was a necessary prior condition to test the hypotheses regarding experiential functions. Because the attitude object in the current study had to be something with which all possible subjects had some prior interaction, professors were particularly salient. The position advocated in the message was that universities should either continue to grant or that they should abolish tenure for all faculty members. Pretesting All the instruments were pretested (see the Appendix for entire survey). In addition, four messages were pretested to determine the extent to which they contained either strong or weak arguments, and the degree to which the messages were perceived as being either pro-tenure or anti-tenure. Argument Strength. Unlike the instrumentation used by Cacioppo. Petty. and Morris (1983), strong arguments were created so that conclusions followed probabilistically from realistic premises that were corroborated by evidence. Conversely. 20 weak arguments were created so that conclusions did not follow from premises. Moreover, no cogent evidence was offered to corroborate the premises. Subjects rated the messages using seven semantic-differential items (compelling/not compelling, convincing/unconvincing, logical/illogical, reasonable/unreasonable. sound/unsound, believable/not believable, and plausible/not plausible) measured on seven-point scales. Results from this pretest demonstrated that the weak message was perceived as weak (m=3.55, sd=1.44) and the strong message was perceived as fairly strong (m=4.98. sd=l.18), [t(33)=-3.23, r=.49, p=.003]. In addition to argument strength, arguments were generated so that they were either proattitudinal or counterattitudinal. Therefore, message advocacy is discussed next. Pro-tenure/Anti—tenure Message. To create messages that were either proattitudinal or counterattitudinal, an attitude regarding a specific aspect of professors was used. A topic that was anticipated to elicit highly variable attitudes from the targeted respondents of the study was the abolishment or continuation of tenure for professors. Accordingly, the messages detail a fictitious Executive Committee for Academic Integrity (ECAI) report that had been created for the president of the university to either continue or abolish tenure for faculty members. Students were asked to judge the degree to which the author of the message was either pro-tenure or anti-tenure using six, seven- point Likert-type items. The pretest demonstrated that subjects did perceive that messages designed to advocate the continuation of tenure of faculty were pro-tenure (m=5.87. sd=.84. where a value of seven indicated the most pro-tenure attitude) and the messages constructed to 21 .. mud-W ' L .4 in :A'aal‘ ' espouse the abolishment of tenure were perceived as being anti-tenure (m=l .36, sd=.47). [t(26)=-4.51, r=.96, p=.000]. Subjects A total of 139 subjects took part in the main data collection. Subjects volunteered to participate in the study, and all were students enrolled in undergraduate courses at a large, Midwestern university. Subjects were assigned randomly to receive either a strong or weak message, and a message that was either pro-tenure or anti-tenure for all university faculty members. The mean age in years for the sample was 22.50 and the modal year was senior. In addition, subjects reported that they had had a mean of 23.49 professors for various t» university courses in the past. Fifty-nine (42 %) subjects were male and 79 (57 %) were ‘ female, the remaining 1 % of subjects did not respond to this item. Of the total number of subjects, .7 % identified themselves as Pacific Islanders or Hispanic. 1.4 0 0 were Native American, 4.3 % indicated “Other,” 7 percent identified themselves as Asian. 17 % were African-American, and 67 % were Caucasian. The remaining 2.2 "/0 ofsubjects did not respond to this item. Design Although conceptualizing the current study as a 2 (argument strength: strong. weak) x 2 (message discrepancy: high, low) x 2 (functional target of message: match. mismatch) was useful for illustrative purposes. it was necessary for regression anal yscs to dichotomize the experiential function measure. How these and other relevant variables were assessed is discussed next. 22 Procedure Subjects were given a questionnaire containing seven measures (see the Appendix). After reading the directions for completing the questionnaire. the following definition of tenure was provided: Tenure is a procedure through which a faculty member’s job performance is reviewed against departmental guidelines approximately six years after the person has been hired by the university. These guidelines include assessment of various "Pl'lnw duties that a professor is expected to perform, including teaching. research, and service. Once tenure is granted by the university, a professor’s dismissal from that university only occurs under one of three conditions, moral perversion, insubordination, or gross incompetence (e.g., Alzheimer’s Disease). Subjects were first presented with the pretest attitude and function measures. Then, each subject was presented with one of four experientially-schematic messages that included a strong or weak argument that either advocated the continuation or abolishment of tenure for university faculty. Subjects then completed the argument strength measure. indicated the degree to which the message supported the continuation or abolishment of tenure for faculty, and were assessed on their level of need for cognition. Measurements assessing subjects’ posttest attitudes and posttest functions were taken last. 23 Instrumentation Except where noted, all measures were comprised of seven-point, Likert-type items. All items were submitted to confirmatory factor analysis and were found to pass both internal consistency and parallelism tests. Attitude. The pretest attitude and posttest attitude measures were comprised of four items where 1 indicated the most anti-tenure and 7 indicated the most pro-tenure attitudes. The reliability for these scales were .86 (M=4.70, _S_D=1.53) and .82 (M =4.01, a S]; =1.55), respectively (see the Appendix for specific items retained). ’ Function. In accordance with the conceptualizations of the experiential-schematic and experiential-specific continuum, the pretest function and posttest function instruments were comprised of five items where 1 demonstrated the most experientially- "" specific function and 7 indicated the most experientially-schematic function. Additionally, despite some methodological problems with the instrument, an adaptation of Herek’s (1987) Experiential-Schematic subscale of the Attitude Function Inventory also was included.2 The pretest function scale achieved a reliability of .45 (M =2.24, S_‘[_) flaw) and the posttest function measure had a reliability of .55 (M =2.43. SD=.93). Argument Strength. Argument strength was calculated with five semantic- differential items where 1 indicated that subjects perceived the message to be extremely weak and 7 indicated that the message was very strong. The standardized item alpha for this measure was .85 (M =4.20, §_[_)_=l .27). Pro-tenure/Anti-tenure Message. Six items were used to determine subjects’ perceptions of the degree to which the messages were either pro-tenure or anti-tenure. The reliability of this scale was .94 (M =3.78, $=2 26). 24 Need for Cognition. The Need for cognition scale (Cacioppo & Petty. 1982) was included to provide filler items. It too had an acceptable reliability of .78 (M =4.85, SQ SD=.98). This instrument was comprised of eight items where 1 indicated a low need for cognition and 7 indicated a high need for cognition. Message Discrepancy. Message discrepancy was not manipulated: students could perceive varying levels of discrepancy regardless of which message they received. In order to determine whether the messages were either proattitudinal or counterattitudinal. subjects’ existing attitudes toward tenure for professors were subtracted from their perception of the attitude of the author of the messages. Absolute values were not used. Instead, a negative value of this index indicated the degree to which the position r—uIr-i - signature“ . advocated in the message deviated from the subjects’ attitudes such that the messages position was more anti-tenure than the subject’s attitude. Alternatively, a positive value indicated that the position in the message was perceived as more pro-tenure than the respondent’s own attitude. This variable had a mean of -.92 (SQ=3.03). Attitude Change. Attitude change was calculated by subtracting subjects’ initial attitude score from their posttest attitude score. This index was then reverse scored so that a negative value indicated the degree to which subjects’ attitudes changed and became more anti-tenure for professors; whereas, a positive value indicated the degree to which subjects’ attitudes became more pro-tenure. The mean of this index was -.68 (flfl .07). The test-retest correlation was g=.76, with standardized item g=.59. Function Change. To determine if there was any function change. an index was created so that a negative value indicated that a subject’s function became more specific; where, a positive value suggested that a respondent’s function changed toward the 25 experientially-schematic end of the continuum. The distribution of this variable was skewed slightly distributed with M=.19 (S_D_=.69) and had a test-retest correlation of .69. with standardized item g=.26. 26 CHAPTER 3 RESULTS To the extent that the matching hypothesis is correct. the relationship between discrepancy and conformity to message recommendations. as measured by posttest attitude, will have a negative slope when the message and the functional base of the attitude matches. i.e., the basis of the participants’ attitude is schematic. This prediction is illustrated in Table 1. On the top one-halfof the table the discrepancy-posttest attitude relationship is illustrated for the pro-tenure message. These calculations assume that the pro-tenure message is perceived as advocating a position of 6 on a 7-point scale, that those holding an anti-tenure view have a pretest attitude of 2, those holding a neutral attitude have a pretest attitude of 4. and those holding a pro-tenure attitude have a pretest attitude of 6. Discrepancy is calculated by subtracting pretest attitude from the position advocated in the message (6). Attitude change is then computed by assuming that persons change one- half of the way toward the message. Finally, posttest attitude is the sum of pretest attitude and attitude change. The bottom one-half of the table illustrates the relationship for the anti-tenure message. The only difference in this portion of the table is that the anti-tenure message is assumed to advocate a position of 2 on a 7-point scale. Comparing the first and third rows of both the first one-half shows the negative relationship between discrepancy and posttest attitude. The first and third rows of the second one-half illustrate the same relationship. 27 Table l. The Prediction for the Relationship between Discrepancy and Attitude Change Discrepancy Attitude Change Posttest Attitude Discrepancy Attitude Change Posttest Attitude Pro-tenure Message Initial Attitude Anti-tenure Neutral Pro-tenure (2) (4) (6) 4 2 0 2 1 0 4 5 6 Anti-tenure Message Initial Attitude Anti-tenure Neutral Pro-tenure (2) (4) (6) 0 -2 -4 0 - l -2 2 3 4 28 _..- : HF-'1‘.’I-. ..- Measurement Model The items measuring attitude (pretest and posttest). function (pretest and posttest). perceptions of argument strength and message advocacy (to abolish or grant tenure) were analyzed. Confirmatory factor analysis, which consists of internal consistency and parallelism, was employed to test the measurement model (Hunter, 1993). A six-factor solution was obtained with a reduced set of indicators; the appendix shows the items that were eliminated based on this analysis. Table 2 shows the descriptive statistics for these variables. Tables 3-6 Show the factor loadings, item means, standard deviations. and the number of pairwise observations for each of the items that were retained for analysis. lntemal consistency tests showed that the errors calculated between items measuring the same construct were generally within sampling error. The parallelism test demonstrated similar results; the errors calculated between items measuring different constructs also were within sampling error. 29 Table 2. Descriptive Statistics for the Constructs. Construct M SQ M Pretest Attitude 4.70 1.53 139 Posttest Attitude 4.01 1.55 137 Pretest Function 2.24 .79 139 Posttest Function 2.43 .93 138 Perceptions of Argument Strength 4.20 1.23 139 Perceptions of Message Advocacy 3.78 2.26 139 30 Table 3. Confirmatory Factor Analysis Results for Pretest and Posttest Attitude Measures. Pretest Item I. I Oppose professors being granted tenure. 2. Tenuring deserving professors is something of which I am in favor. 3. I believe that abolishing tenure would be a good thing for the university. 4. I think that tenure should be granted to deserving faculty. Posttest 1. I oppose professors being granted tenure. 2. Tenuring deserving professors is something of which I am in favor. 3. I believe that abolishing tenure would be a good thing for the university. 4. I think that tenure should be granted to deserving faculty. 31 IT} .69 .91 .81 .72 .80 .81 .91 .44 IE 4.57 4.81 4.42 5.01 2.97 1.81 1.88 1.76 1.95 1.83 1.94 1.98 Table 4. Conlirmatory Factor Analysis Results for Pretest and Posttest Function Measures. Pretest Item 1'11 13 m U 12 1. One is likely to make errors when thinking that all professors are the same. .31 1.78 1 .26 139 to My opinions about professors mainly are based on whether or not someone I care about is a professor. .38 1.67 1.26 138 3. It is useful for students to make generalizations about professors based on professors they have had for classes in the past. .36 3.47 1.80 139 4. Student are generally right when they assume that all professors are the same. .55 1.83 1.30 139 5. It is useful for students to form attitudes about professors based on specific incidences rather than overall generalizations. .27 2.44 1.47 139 Posttest 1. One is likely to make errors when thinking that all professors are the same. .47 2.45 1.58 138 2. My opinions about professors mainly are based on whether or not someone I care about is a professor. .48 1.86 1.40 138 3. It is useful for students to make generalizations about professors based on professors they have had for classes in the past. .40 3.38 1.92 138 4. Student are generally right when they assume that all professors are the same. .62 2.03 1 .3 7 l3 7 5. It is useful for students to form attitudes about professors based on specific incidences rather than overall generalizations. IQ U1 [‘0 4:. UI LII Id 9) 3° 32 I ”.4 ‘4) Table 5. Confirmatory Factor Analysis Results for the Perceived Argument Strength Measure. Item I believe the Committee 's report regarding tenure for professors is: 1. 2 3 4. 5 C onvincing/ unconvincing Reasonable/unreasonable Unsound/ Sound Believable/not believable Not plausible/plausible 3.97 4.28 4.13 4.29 4.34 1.76 1.64 1.62 1.60 Table 6. Confirmatory Factor Analysis Results for the Perceived Message Advocacy Measure. Item The author... 1. of the message opposes professors being granted tenure. of the message is in favor of abolishing tenure for all faculty. of the message thinks that professors should be granted tenure if they merit it. of the message is in favor of tenuring professors. believes that abolishing tenure would be a good thing for the university. thinks that tenure should be granted to all faculty. 33 1‘11 .94 .91 .72 .97 .93 .62 12 4.00 4.02 3.71 4.01 4.13 2.87 Ix) .14 138 11‘ Evaluating the Inductions The mean of the responses to the five items presented in Table 5 were averaged to form the perceived argument strength index. This index was distributed normally with a mean of 4.20 and a standard deviation of 1.27 (_l\_l=139). The reliability of this measure was estimated by Cronbach’s alpha, and was found to be g=.85. The perceived argument strength index was broken down by the advocacy induction and the argument strength induction, and the resulting descriptive statistics are presented in Table 7. Homogeneity of variance tests indicated that the hypothesis of homogeneity of variance could not be rejected (EL(3,135)=1.61. p>.05). A two-way analysis of variance performed on these data indicated that there was a statistically significant and substantial main effect for the argument strength induction (E(1,135)=24.49, p<.05, r=.39, g’=.42), such that the strong arguments were perceived as stronger (M=4.77) than the weak arguments (M=3.77). Neither the advocacy induction main effect nor the argument strength X advocacy interaction effect were statistically significant. Moreover, neither effect size was substantial. Thus, the data are consistent with the hypothesis that the argument strength induction, and only the argument strength induction, was effective in inducing levels of perceived argument strength that varied in the anticipated direction. The mean of the responses to the six items presented in Table 6 were averaged to form the perceived message advocacy index. This index was distributed bimodally with a mean of 3.78 (M_d=4.00, $21-00) and a standard deviation of2.26 (_N_=139). The reliability of this measure was found to be g_=.94. 34 Table 7. Perceived Argument Strength Index Broken Down by the Advocacy and Argument Strength Inductions. Message Advocacy Argument Strength Pro-tenure Anti-tenure Strong M=4.51 5.00 SQ=1.29 1 . 1 8 _N_=28 32 Weak 3 .65 3 .89 1 .25 .94 40 3‘) 35 The perceived message advocacy index was broken down by the advocacy induction and the argument strength induction, and the resulting descriptive statistics are presented in Table 8. Homogeneity of variance tests indicated that the hypothesis of homogeneity of variance could not be rejected (£L(3,135)=2.30. p>.05). A two-way analysis of variance performed on these data indicated that there was a statistically significant and substantial main effect for the advocacy induction (£(1.135)=446.87. p<.05. _if=.87, _r_’=.90), such that the pro-tenure argument was perceived as more pro- tenure (M=5.84) than the anti-tenure arguments (M=1.85). Neither the argument strength induction main effect nor the argument strength X advocacy interaction effect were statistically significant. Furthermore, they were not substantial. Thus. the data are consistent with the hypothesis that the advocacy induction. and only the advocacy - induction, was effective in inducing levels of perceived advocacy that varied in the anticipated direction. The mean of the responses to the four items presented in Table 3 were averaged to form the pretest attitude index. This distribution of this index approximated closely the normal distribution. The mean of the index was 4.70 and a standard deviation of 1.53 (N=139). The reliability of this measure was found to be g=.86. The pretest attitude index was broken down by the advocacy induction and the argument strength induction, and the resulting descriptive statistics are presented in Table 9. Homogeneity of variance tests indicated that the hypothesis of homogeneity of variance could not be rejected (EL(3,135)=2.17. p>.05). A two-way analysis of variance performed on these data indicated that there was a statistically significant main effect for the advocacy induction (E(1,135)=5.24. p<.05,). such that those in the pro-tenure 36 Table 8. Perceived Message Advocacy Index Broken Down by the Advocacy and Argument Strength Inductions. Message Advocacy Argument Strength Pro-tenure Anti-tenure Strong M=6.04 l .69 S_D=.71 1.20 fl=28 32 Weak 5.63 2.00 1.08 1.26 40 39 37 Table 9. Pretest Attitude Index Broken Down by the Advocacy and Argument Strength Inductions. Message Advocacy Argument Strength Pro-tenure Anti-tenure Strong M=4.04 4. 8 1 SQ=1 .79 1 .66 N=28 32 Weak 4.69 5.10 1.22 1 .40 40 39 38 message condition were less pro-tenure initially (M=4.3 7) than those in the anti-tenure message condition (M=4.96). Although statistically significant, the effect size for this variance component was small (_r_'=-.18, [’=-.19). Neither the argument strength induction main effect nor the argument strength X advocacy interaction effect were statistically significant. Furthermore, they were not substantial. Therefore, the data indicate that pretest attitudes were not distributed uniformly across conditions as one would expect given random assignment. Consequently, it was necessary to control for pretest attitude in subsequent analyses. This goal was accomplished either by treating pretest attitude as a covariate or examining attitude change scores or both. Evaluating the Hypotheses The mean of the responses to the four posttest attitude items presented in Table 3 were averaged to form the posttest attitude index. This distribution of this index approximated closely the normal distribution. The mean of the index was 4.01 and a standard deviation of 1.55 (Er-137). The reliability of this measure was found to be g=.82. The posttest attitude index was broken down by the advocacy induction and the argument strength induction, and the resulting descriptive statistics are presented in Table 10. Homogeneity of variance tests indicated that the hypothesis of homogeneity of variance could not be rejected (EL(3,132)=.82, p>.05). A two-way analysis of covariance, with pretest attitude as the covariate, performed on these data indicated that there was a statistically significant main effect for the covariate (E(1,132)=184.72, p<.05). Thus, if persons had pro-tenure pretest attitudes they tended to have relatively 39 Table 10. Posttest Attitude Index Broken Down by the Advocacy and Argument Strength Inductions. Message Advocacy Argument Strength Pro-tenure Anti-tenure Strong M=3 .61 3 .59 S_I_)_=1 .74 1 .50 N__=26 32 Weak 4. l 5 4.47 1.49 1 .43 40 39 40 yam: “3:411“: 1.‘_" «IQ Cw LC”. pro-tenure posttest attitudes, and if they had anti-tenure pretest attitudes they tended to have relatively anti-tenure posttest attitudes (§=.76). Moreover. there was a statistically significant main effect for the advocacy induction (E(1,132)=4.20. p<.05). This main effect was overridden. however, by the advocacy X argument strength interaction (£(1,132)=4.45, p<.05). These results are paradoxical. given the pattern of the means displayed in Table 9. They reflect, however. the impact of controlling for pretest attitude scores that are not distributed uniformly across experimental conditions. The pattern ofthis interaction can. therefore, be viewed more clearly by observing change scores. These data are presented in Table 11. Levine’s test indicated that variances were homogeneous (E(3,133 =. 13. p>.05). A two-way analysis of variance performed on these data showed that there was a statistically significant main effect for the advocacy induction (£(1.133)=7.78, p<.05, 1_'=.21, r’=.27), indicating that those exposed to the anti-tenure message changed more in a negative direction (M=-.93) than did those who were exposed to the pro-tenure message (M=-.43). This main effect was overridden by the statistically significant advocacy X argument strength interaction effect (E(1,133)=5.38, p<.05). Observing the means in Table 10 suggests that for weak messages there is little difference in conformity to message recommendations between the pro-tenure and anti-tenure conditions. but that for strong messages there was substantially more conformity in the anti-tenure direction in the anti-tenure condition than in the pro-tenure condition. C orrelational analyses are consistent with this interpretation. The correlation between the advocacy induction and posttest attitudes is only [:04 (p>.05) in the weak message condition, but is [=40 (p<.05, one-tailed test) in the strong message condition. 41 [H41 -——-—. 3—1 Table 1 1. Mean Attitude Change Broken Down by the Advocacy and Argument Strength Inductions. Message Advocacy Argument Strength Pro-tenure Anti-tenure Strong M=-.31 -l .23 _SQ=.95 1 . 1 3 N=26 32 : Weak -.54 -153 ’ 1.03 1 .03 40 39 42 An additional striking feature of Table 1 1 is that all of the mean change scores are negative. Hence, on the average, attitudes changed in an anti-tenure direction. Within cell one-sample t-tests demonstrated that, statistically, change scores were significantly lower than zero with the exception of the strong message/pro-tenure message, which did not differ significantly from zero (t(39)=-3.27, p<.05 in the weak message/pro-tcnure condition; t(25)=-1.66. p>.05 in the strong message/pro-tenure condition; t(38)=-3.77. p<.05 in the weak message/anti-tenure condition; and t(31)=-6.06, p<.05 in the strong message/anti-tenure condition). The reason for these outcomes could not be predicted from the other variables measured in this experiment. To test the hypothesis advanced in the first chapter the functional type measure was dichotomized by dividing it at the median (Md_=1 1). excluding those participants who scored at the median. It is notable that the median was relatively low. 1 1 on a scale that ranged potentially from 5-35, indicating that the functional basis of the attitude for the vast majority of participants was specific, rather than schematic. The regression of posttest attitude on discrepancy was then calculated for each of the four combinations of functional type (specific/schematic) and argument strength (weak/strong). The unstandardized slopes, correlation coefficients, number of participants, and tests of the significance of the unstandardized slopes are presented in Table 12. From this table it can be seen that, in the main, the data are inconsistent with the matching hypothesis. Contrary to this prediction, discrepancy did exert a substantial effect on posttest attitudes in both specific conditions. Moreover, inconsistent with this hypothesis the effect of discrepancy on posttest attitudes was marginal in the schematic/weak argument condition.’ 43 n.‘ F_(' Table 12. Unstandardized Slopes of Posttest Attitude on Message Discrepancy. Argument Strength Strong Weak Function Specific Schematic _B=-.94 -.8 1 _r=-.41 -.41 [1:24 30 ;=-2.1 1 (p<.05) -2.36 (p<.05) -1.03 -.51 -.51 -29 38 .3.) -352 (p<.05) -1.67 (p>.05) 44 IT'— Change in the Basis of the Attitude Although functional attitude theorists emphasize that attitudes develop for various reasons, and others examine the manner in which matched or mismatched messages change attitudes, little consideration is given to the manner in which messages might affect the reasons. or functions, for which persons hold their attitudes. Because the functional bases of attitudes toward tenure were measured twice in this experiment, it was possible to examine if there was any change in the functional basis of the attitude as a . 1..) " '7! function of receiving schematic messages regarding tenure. The distribution of the change in the functional basis ofthe attitude exhibited a slight positive skew. and was leptokurtic. The mean of this distribution was .19 (_M_q=0, 26.8% changing negatively/specifically and 44.2% changing positive]y/schematical1y). a ‘- standard deviation of .69, and a reliability of g_=.55. A one-sample t-test on these data indicate that the mean differs significantly from zero (1( 1 37)=3.22. p<.05). Thus. overall. participants became more schematic from the pretest to the posttest. Neither argument strength nor position advocated nor discrepancy affected directly the change in the functional basis measure. Nevertheless. perceptions of the position advocated in the message did correlate substantially with this variable. depending upon whether the participants received a pro-tenure or anti-tenure message. For those in the pro-tenure condition the correlation between perceived message advocated and change in the functional basis of the attitude toward tenure was -.30 (p<.05, two-tailed test); whereas, for those in the anti-tenure condition this correlation was .14 (p>.05, two-tailed test). The difference between these two correlations was both substantial and Statistically significant (;=2.58. p<.05, two-tailed test). Ilence. for those 45 who were exposed to the pro-tenure message, the more pro-tenure they perceived this message, i.e., the more accurate was their perception of the message, the more specific became the basis of their attitude. And, for those exposed to the anti-tenure message, there was a trend in the opposite direction. That is, the more accurately one perceived the message the more schematic became the basis of one’s attitude. Moreover, attitude change correlated differentially with the functional basis of the attitude, contingent upon whether participants were exposed to the weak or strong .. message. For those participants in the weak message condition there was a slight tendency for an increase in attitude change to be accompanied by becoming more schematic (r=.20, p>.05, one-tailed test); whereas, for participants in the strong message condition there was a slight tendency for an increase in attitude change to be .2 accompanied by becoming less schematic (r=-. 14, p>.05, one-tailed test). In sum, becoming more schematic was associated with perceiving the pro-tenure message inaccurately or the anti-tenure message accurately. Moreover. it was associated with greater attitude change when one read the weak message, but less attitude change when one read the strong message. Becoming more specific was associated with perceiving the pro-tenure message accurately or the anti-message inaccurately. Furthermore, it was associated with less attitude change when exposed to the weak message, but greater attitude change when exposed to the strong message. Notably the causal direction of these data is unclear. That is, function change might be antecedent to the accuracy of the perception of message advocacy, or consequent to it. Similarly. attitude change could be either antecedent or consequent to function change. 46 CHAPTER 4 DISCUSSION To Match or Mismatch According to the functional approach to attitudes, messages that target an attitude function should be more persuasive than messages that contain function-irrelevant information. This study was conducted to test whether this matched hypothesis could be demonstrated with the experiential functions as it had in the past with other functions (Petty & Wegener. 1998; Snyder & De Bono, 1985; Katz, McClintock, & Sarnoff, 1957). These data were not consistent with this relationship; attitude function did not contribute to conformity to message recommendations. Shavitt (1989) asserts that some attitude objects are based on multiple functions. although one function may take precedence over the others. The attitude object here, tenure, was conceptualized more as a personal characteristic (Herek. 1986). assuming that students’ attitudes about their professors would be based largely on their experiences with professors. The inaccuracy of this assumption may explain the lack ofeffect for the functional target of the message. For instance, it is reasonable to expect that some students would have a utilitarian function with regard to their attitudes about tenure. asking themselves, “What benefits do I receive from professors being granted tenure?” A social-expressive function, where students are pro-tenure or anti-tenure in order to align themselves with important referent groups, seems equally plausible as well. One suggestion for future research would be to have subjects rate the extent to which they perceive that their attitudes toward a particular object are based on different functions. 47 Discrepancy Not only was the functional target of the message expected to motivate subjects to comply with message recommendations, but the discrepancy between students” initial positions and the position advocated in the message also was predicted to stimulate conformity to message recommendations. Table 1 demonstrated that ifa linear discrepancy model of attitude change existed (French, 1956), the data would produce a negative slope when posttest attitude scores were regressed onto message discrepancy. . Although there was conformity to message recommendations when students experienced discrepancy between their own attitudes and the perceived attitudes of the author of the message, this relationship existed for all experimental conditions (contrary to the matched hypothesis). This indicates that neither argument strength nor advocacy contributed " systematically to conformity to message recommendations. In other words. it did not matter if students received a weak/strong or pro—tenure/anti-tenure message; message discrepancy, and message discrepancy alone, led to conformity to message recommendations. Analysis of variance results indicated that respondents did perceive both the strength and advocacy of the message accurately. Why, then, did argument strength not affect posttest attitudes? Perhaps one answer to this question lies within the students themselves. It may have been the case that this was the first opportunity students had to actually think about their own positions regarding tenure. Although a definition of tenure was provided to the participants of this study, it cannot be determined whether or not they knew what tenure was prior to reading that definition. Therefore, regardless of argument strength. the topic alone stimulated students to evaluate the message in relation to the 48 position advocated in the message. In other words, merely exposing students to a message regarding tenure may have provided enough stimulation, argument strength notwithstanding, to experience conformity to message recommendations. Positive Attitude, Negative Conformity to Message Recommendations Results indicated that students’ initial attitudes toward tenuring professors generally was positive (M=4.70), although those in the anti-tenure condition were more pro-tenure initially. This can be compared to how attitudes changed after receiving the _....J message. The data demonstrated that, in general, after receiving a message regarding the tenuring of university faculty, attitudes became more anti-tenure, regardless of argument strength or message advocacy. Both these findings were interesting. Considering both r'n 3...;Jdm20'9v- L‘JI'LL ) . simultaneously, however, was perplexing. How do initial positive attitudes change and become more negative, despite the actual position advocated in the message? One explanation for these results can be explained partially by the boomerang or contrast effect (Hovland, Harvey, & Sherif, 1957). This effect is grounded in a latitude of rejection. Hovland and his colleagues argue that when respondents are given a message, the position advocated in the message is compared to their own position. If the position advocated falls within the subject’s latitude of acceptance, then the position is assimilated and little conformity to message recommendations takes place. Alternatively, if the subject perceived the position advocated was markedly different than his or her own position, then that message falls into the latitude of rejection. where there is no attitude change but reinforcement of an attitude already held. In the present study, however, there was a change in attitude. Extending the findings of Hovland, Harvey, and Sherif (1957), what if advocating a perceived extreme 49 position motivated subjects to generate internal messages? These internal messages. in turn, would move the subject’s attitude further away from the extreme position advocated. For example, a woman who initially identifies herself as pro-choice and receives a message advocating for banning abortions nationwide would generate internal messages against that recommended position and become more pro-abortion. Although this explains how the subjects in the anti-tenure conditions reacted. it does not address the change experienced by students in the pro-tenure conditions. What - stimulated the people in this latter condition to become more anti-tenure? None of the variables measured in the current study accounted for this change. '13? One explanation, however, for these data is what might be called a rebellion of FIJI n- 3.... absolutes in which students were opposed radically to continue granting tenure to faculty because what their perceptions of tenure were, rather than what was the actual definition of tenure. After receiving a message advocating that tenure should be granted to all faculty (schematic message), students accessed other beliefs associated with tenure. For example, one belief regarding tenure is that once a faculty member receives tenure, that member ceases to care about students, becoming rather lazy and uninvolved with university activities. Because the vast generalization of the pro-tenure message. it stimulated students access their own beliefs (stereotypical perhaps) to react negatively to the message and become anti-tenure. In short, the pro-tenure message enacted another schema about professors and tenure, and this schema was negative. What? Functions Change? One of the most fundamental tenets of the functional approach is the assumption of function stability—perhaps more stable than attitudes themselves. These data indicate 50 that functions do change, a finding which calls into question the stability of functions. Specifically, it was found that positive attitudes were associated with the experientially- specific function and negative attitudes were associated with the experiential1y-schematic function. Hence, students who were relatively pro-tenure based this attitude on individual interactions and did not generalize this attitude to include all professors. Alternatively, students who had more anti-tenure attitudes were more likely to generalize to other professors. [T Based on this evidence, a reconceptualization of the functional approach to attitudes and their change is in order. This result is exciting particularly because it indicates that targeting conformity to message recommendations might be premature. instead focusing on why the initial attitude is held (its function) would be more useful. Knowing how to change functions could then, in turn, help change attitudes. For example, ego-defensives often are prejudiced toward specific samples of the population including homosexuals (Herek, 1987) and African-Americans (Katz. McClintock. & Sarnoff, 1957). To change an attitude that serves an ego-defensive function is a difficult task because mere exposure to any message regarding this type of attitude object will serve to increase the subject’s defensiveness. Instead, a fruitful option would be to attempt to change the function for which the attitude is held. Thus. generating messages that stimulate change from an ego-defensive function to a value-expressive function would facilitate conformity to message recommendations in the long run because then one could target the value-expressive function in an attempt to change the attitude. 51 Limitations One limitation of the current study to be reiterated is the restriction in range for the attitude function measure. The highest score on this measure was 4.20 (where 1 indicated more of an experientially-specific function and 7 indicated more of an experientially-schematic function). This finding was surprising particularly in that it suggests that students do not generalize across professors. rather their experientially- based attitudes are formed from independent interactions with different professors. Another limitation that should be discussed is the attitude object of the study. The attitude measure contained items regarding tenuring professors; whereas the function measure consisted of statements regarding professors in general. Initially. this was done assuming that if functions were the basis of an attitude, than that attitude would be contingent upon the function it served. Specifically, if one had a experientially-schematic function in which professors were perceived positively, than attitudes toward tenuring professors should also be positive, based on the conceptualization of the functional approach. This assumption was premature. The attitudes students have toward professors refer to a more generalized attitude; whereas. the attitudes students have toward the different aspects or dimensions of a professor. such as tenure, are more specific. Therefore, subjects’ attitudes toward the tenure system for faculty may be negative; however, they simultaneously may have positive attitudes regarding professors in general. Future functional research in this area would benefit from elucidating the relationship between an attitude function regarding a general versus specific aspect of an attitude object. This suggestion might be difficult to pursue pragmatically depending 52 upon the attitude object of interest. If two constructs are inextricably linked, the specific versus general distinction would be tenuous. For example, assessing attitude functions regarding HIV/AIDS with certain samples of the population would be so strongly associated with homosexuals that respondents would be unable or unwilling to make the distinction between their general or specific attitudes about those two constructs. This illustration can be contrasted with another example in which subjects easily distinguish between general attitudes toward an attitude object and specific attitudes toward that same object. For instance, a person who has a generally positive attitude about drinking alcoholic beverages could have an equally negative attitude about driving while drinking. Thus, research examining attitudes and their functions must examine the relationship between general versus specific attitudes and their functions. Selecting the tenuring of professors as the attitude object for this study might have been problematic for another reason. When students think of the label professor. a myriad of characteristics are accessed. Intelligent, fair, old. nerd, hard working. research, outdated, uncaring, and egocentric are all characteristics that students may associate with being a professor. Tenure, however, may not be a readily accessible characteristic that students associate with being a professor because in their perception being a “tenured” professor has little or nothing to do with them personally. Future work would benefit by conducting focus groups to gain knowledge of the most salient characteristics identified with professors and using those traits to examine the functional basis of the attitudes regarding professors. The last limitation to be addressed is the finding that initial attitudes were not uniformly distributed across experimental conditions despite randomization. This result 53 was surprising but statistically controlled for using the pretest attitude measure as a covariate when evaluating the hypotheses. In summary, the data were not consistent with the matched hypothesis predicted from past research regarding other attitude functions (Katz. McClintock. & Sarnoff, 1957; Snyder & De Bono, 1989). Attitude function was not associated with conformity to message recommendations. Moreover. argument strength did not predict whether or not subjects conformed to message recommendations; both weak and strong messages r‘1 stimulated some conformity. Conformity to message recommendations was contingent upon the discrepancy between the subject’s initial attitude and the position advocated in i the message. Interestingly. the mean attitude change in every experimental condition was ; negative, indicating that regardless of the position advocated in the message. subjects L became increasingly anti-tenure. Last, despite the conceptual stability associated with functions. these data demonstrated that functions can and do change. Although the causal direction ofchange is ambiguous, it may be the case that a longer causal string exists with message advocacy and function change as mediator variables and argument strength and conformity to message recommendations as independent and dependent variables. respectively. This latter finding offers fertile ground for future empirical investigations. 54 APPENDIX 55 APPENDIX PRETEST AND DATA COLLECTION INSTRUMENT The following survey asks you questions regarding your opinions about professors you have had in the past. You will not be asked to identify these professors. rather the questionnaire asks you about your opinions about professors in general. In addition. you will be asked about your preferred methods of thinking. The survey will take you approximately 15 minutes to complete. after which your participation in this project is over. Your participation in this survey is completely voluntary. You may choose not to participate at all or refuse to answer any question or questions without penalty or loss of benefits. You may cease your participation at any time. Your name will not appear on any page on this survey: therefore. your anonymity is assured. If you have any questions about this survey or are interested in the results of this research project. ' 1‘ please contact: Betty H. La France at the Department ofCommunication. Michigan State University. 13. " Lansing, MI 48824, (517) 353-4466, email: lafranc2=apilotmsucdu. Please sign and date this sheet below. Fr: :1. «1.4-0 o'. " 1 (tear and separate from survey when completed) 1 have consented to being a voluntary participant in this research project. Name Date: 56 Directions: You are being asked your opinions and beliefs regarding tenure for university professors. What is tenure? Tenure is a procedure through which a faculty member’s job performance is reviewed against departmental guidelines approximately six years after the individual has been hired by the university. These guidelines include assessment of various duties that a professor is expected to perform, including teaching, research, and community service. Once tenure is granted by the university, a professor’s dismissal from that university only occurs under one of three conditions, moral perversion, insubordination, or gross incompetence (e.g., Alzheimer’s Disease). You are being asked to respond to each item by indicating the degree to which you agree or disagree with each statement. Use the following scale and place the number that best represents your opinion on the line before each statement. There are no right or wrong answers; please respond to each item as honestly as possible. Strongly 1 2 3 4 5 6 7 Strongly Disagree Agree 1. I oppose professors being granted tenure.* Ex) 1 am in favor of abolishing the tenure system for all faculty. __ 3. 1 think that professors should be granted tenure if they merit it. __ 4. Tenuring deserving professors is something of which 1 am in favor.* __ 5. I believe that abolishing tenure would be a good thing for the university.‘ 6. I think that tenure should be granted to deserving faculty.* 57 Directions: Think about the professors you have interacted with in the past. Using the following scale, please indicate the degree to which you agree or disagree, IN GENERAL, with the following statements. There are no right or wrong answers; please respond as honestly as possible. Strongly 1 2 3 4 5 6 7 Strongly Disagree Agree 1. It is beneficial for students to make assumptions regarding professors based on interactions they have had with professors in the past. [schematic] 2. It’s difficult for students to generalize from one professor to another. [specific] 3. My opinions about professors mainly are based on my personal experiences with people whose family members or friends are professors. [schematic] 4. One is likely to make errors when thinking that all professors are the same.* [specific] My opinions about professors mainly are based on whether or not someone I care about is a professor.* [schematic] 6. It is useful for students to make generalizations about professors based on professors they have had for classes in the past."‘ [schematic] 7. Students are generally right when they assume that all professors are the same.* [schematic] 8. My opinions about professors mainly are based on my judgment of how likely it is that I will interact with professors in any significant way. [schematic] 9. It’s useful for students to judge professors based on individual interactions rather than generalizations. [specific] 10. My opinions about professors mainly are based on my personal experiences with specific professors. [schematic] 1 1. It is useful for students to form attitudes about professors based on specific incidences rather than overall generalizations.‘ [specific] 58 Background and Instructions: In the fall of 1997, a Blue Ribbon Committee was formed to review the university policy regarding faculty tenure. This committee, named the Executive Committee for Academic Integrity (ECAI), was composed of members from the administration, faculty, staff, and community. The ECAI issued a report summarizing their findings. Following is the executive summary of this report. Please read it carefully. After you have read it, there will be some questions concerning your opinion of the executive summary. Executive Summary “We, the members of the Executive Committee for Academic Integrity, support the present tenure system at Michigan State. We have reached this conclusion based on the following four critical criteria our analysis identified. First, the faculty members at Michigan State are some of the best professors in the state. This fact is clearly demonstrated each semester by the computerized evaluations (SIRS forms) completed by students who rate their professors favorably. A second factor we have identified is the predominant research role of our faculty. Last year. our approximately 2,000-member faculty published 5,634 articles in prestigious academic journals, helping to make Michigan State one of the top-ranked universities in the Big Ten. A third criterion to be discussed is the faculty’s continued active role in our community. Over 90 percent of professors on campus have donated their time and expertise to help community organizations in Michigan, focusing on issues that alleviate important technological, agricultural, and social problems. Fourth, tenure provides faculty members the security and freedom to express unpopular points of view. This freedom is critical to professors' and students' academic pursuits. If the university continues to grant tenure, students are assured of being exposed to different views, not only to socially acceptable opinions, from all types of faculty. It is for these reasons that the Executive Committee for Academic Integrity strongly supports and advocates that Michigan State University retain its present tenure policy.“ (Strong, pro-tenure message) 59 Background and Instructions: In the fall of I997, a Blue Ribbon Committee was formed to review the university policy regarding faculty tenure. This committee, named the Executive Committee for Academic Integrity (ECAI), was composed of members from the administration, faculty, staff, and community. The ECAI issued a report summarizing their findings. Following is the executive summary of this report. Please read it carefully. After you have read it, there will be some questions concerning your opinion of the executive summary. Executive Summary “We. the members ofthe Executive Committee for Academic Integrity, support the present tenure system at Michigan State. We have reached this conclusion based on the following four critical criteria our analysis identified. First, the faculty members at Michigan State consistently state how much they enjoy the atmosphere on campus, noting the size and pleasant color of most of the offices in their departments. In addition, most ofthe offices have relatively large windows that provide them with pleasing views of campus. A second factor we have identified is the role our faculty take keeping up with research. Last year, our faculty reported they enjoyed watching research documentaries on television. In addition. the subject that many of these documentaries focused on included historical events. biological and physical marvels, and people’s psychological health. A third criterion is the faculty’s attitude toward our community. Approximately one-third of professors at Michigan State live in the community in which they teach. Many drive from areas such as East Lansing and Okemos. As such, they can relate better to one another as well as have shorter, more time-saving commutes to campus. Fourth, tenure provides professors with security so that they feel less pressured and stressed. A majority of professors want tenure because it makes them feel good about themselves and raises their self- esteem. Tenure will promote self-esteem and professors will be happy as a result. It is for these reasons that the Executive Committee for Academic Integrity supports that Michigan State University retain its present tenure policy.” (Weak, pro-tenure message) 60 Background and Instructions: In the fall of 1997, a Blue Ribbon Committee was formed to review the university policy regarding faculty tenure. This committee, named the Executive Committee for Academic Integrity (ECAI), was composed of members from the administration, faculty, staff, and community. The ECAI issued a report summarizing their findings. Following is the executive summary ofthis report. Please read it carefully. After you have read it, there will be some questions concerning your opinion of the executive summary. Executive Summary “We, the members of the Executive Committee for Academic Integrity, believe that the present tenure system at Michigan State University should be abolished. We have reached this conclusion based on the following four critical criteria our analysis identified. First, the faculty members at Michigan State are some of the worst professors in the state. This a fact is clearly demonstrated each semester by the computerized evaluations (SIRS forms) completed by 1 students who rate their professors poorly. L; A second factor we have identified is the subordinate research role of our faculty. Last year. our ’ approximately 2,000-member faculty only published 402 articles. In addition, these articles appeared in i popular magazines rather than in prestigious academic journals, failing to make Michigan State one of the top-ranked universities in the Big Ten. A third criterion to be discussed is the faculty’s continued inactive role in our community. Less than IO percent of professors on campus have donated their time and expertise to help community organizations in Michigan, directing research toward esoteric theoretical issues and not toward important technological, agricultural, and social problems. Fourth, tenure provides professors with the security not to work as hard as they once did. Studies show that 79 percent of professors who were granted tenure from their universities tended to decrease their workload by not teaching as many classes nor conducting as much research after they had received tenure. It is for these reasons that the Executive Committee for Academic Integrity strongly supports and 9 advocates that Michigan State University abolish its present tenure policy. ’ (Strong, anti-tenure message) 61 Background and Instructions: In the fall of I997, a Blue Ribbon Committee was formed to review the university policy regarding faculty tenure. This committee, named the Executive Committee for Academic Integrity (ECAI), was composed of members from the administration, faculty. staff, and community. The ECAI issued a report summarizing their findings. Following is the executive summary of this report. Please read it carefully. After you have read it, there will be some questions concerning your opinion of the executive summary. Executive Summary “We, the members of the Executive Committee for Academic Integrity, believe that the present tenure system at Michigan State University should be abolished. We have reached this conclusion based on the following four critical criteria our analysis identified. First, the faculty members at Michigan State consistently state how much they dislike the atmosphere on campus, noting the small size and plain color of most ofthe offices in their departments. In addition, most ofthe offices have relatively small windows that only provide them with views of other buildings on campus. A second factor we have identified is the role our faculty take in scarcely keeping up with research. Last year, our faculty reported they did not enjoy watching research documentaries on television. The subject that many of these documentaries focused on included historical events. biological and physical marvels, and people’s psychological health. A third criterion is the faculty’s attitude toward our community. Less than one-third of professors at Michigan State live in the community in which they teach. Many faculty commute from areas such as Ann Arbor, Battle Creek, and Grand Rapids. As such, they fail to relate to one another as well as have longer, more time-consuming commutes to campus. Fourth, tenure only provides professors with a false sense of security, which has little to do with their actual work teaching or conducting research. This, in turn, may lead to feelings of helplessness in that faculty will no longer feel that they can control their own destiny. This would result in professors feeling more stressful and pressured. It is for these reasons that the Executive Committee for Academic Integrity supports that Michigan State University abolish its present tenure policy.” (Weak, anti-tenure message) 62 Directions: Please judge the Committee’s report along the following dimensions. Circle the number that corresponds closest to your opinion. I. I believe the Committee’s report regarding tenure for professors is: a. Compelling I 2 3 4 5 6 7 Not Compelling b. * Convincing I 2 3 4 5 6 7 Unconvincing c. lllogical I 2 3 4 5 6 7 Logical d. “ Reasonable I 2 3 4 5 6 7 Unreasonable e. * Unsound I 2 3 4 5 6 7 Sound f. * Believable l 2 3 4 5 6 7 Not Believable g. * Not Plausible I 2 3 4 5 6 7 Plausible Directions: Use the following scale to respond to each statement. Place the number that best reflects your response to each statement on the line preceding that statement. Disagree I 2 3 4 5 6 7 Agree I. The author of the message opposes professors being granted tenure.* The author of the message is in favor of abolishing tenure for all faculty.* The author of the message thinks that professors should be granted tenure if they merit it.* 2 3 4. The author of the message is in favor of tenuring professors.* 5 The author believes that abolishing tenure would be a good thing for the university” 6 The author thinks that tenure should be granted to all faculty.* 63 Directions: Now you will be asked some questions about things you like and dislike. Please respond to the statements below using the following scale. There are no right or wrong answers; please respond as honestly as possible. Strongly 1 2 3 4 5 6 7 Strongly Disagree Agree I. I really enjoy a task that involves coming up with new solutions to problems. 2. I would prefer a task that is intellectual, difficult, and important to one that is somewhat important but does not require much thought. __ 3. I tend to set goals that can be accomplished only by expending considerable mental effort.‘ __ 4. I am usually tempted to put more thought into a task than the job minimally requires. __ 5. Learning new ways to think doesn’t excite me very much.* __ 6. I am hesitant about making important decisions after thinking about them. __ 7. I usually end up deliberating about issues even when they do not affect me personally. _ 8. I preferjust to let things happen rather than try to understand why they turned out that way. __ 9. I have difficulty thinking in new and unfamiliar situations. __ IO. The idea of relying on thought to make my way to the top does not appeal to me. __ I I. The notion ofthinking abstractly is not appealing to me. __ I2. I am an intellectual. __ I3. I only think as hard as I have to.‘ __ I4. I don’t reason well under pressure. __ IS. I like tasks that require little thought once I’ve Ieamed them. __ I6. I prefer to think about small, daily projects to long-term ones. __ I7. I would rather do something that requires little thought than something that is sure to challenge my thinking abilities. _ I8. I find little satisfaction in deliberating hard and for long hours.* __ I9. I more often talk with other people about the reasons for and possible solutions to international problems than about gossip or tidbits of what famous people are doing. 20. These days, I see little chance for performing well, even in “intellectual“jobs. unless one knows the right people. 2]. More often than not, more thinking just leads to more errors. 64 Strongly 1 2 3 4 5 6 7 Strongly Disagree Agree 23. I don’t like to have the responsibility of handling a situation that requires a lot of thinking. 24. I appreciate opportunities to discover the strengths and weaknesses of my own reasoning. 25. I feel relief rather than satisfaction after completing a task that required a lot of mental effort.‘ 26. Thinking is not my idea of fun.* __ 27. I try to anticipate and avoid situations where there is a likely chance I will have to think in depth about something.* __ 28. I prefer watching educational to entertainment programs. __ 29. I think best when those around me are very intelligent. __ 30. I prefer my life to be filled with puzzles that I must solve. __ 31. I would prefer complex to simple problems!“ __ 32. Simply knowing the answer rather than understanding the reasons for the answer to a problem is fine with me. __ 33. It’s enough for me that something gets the job done, I don’t care how or why it works. __ 34. Ignorance is bliss. __ 35. I enjoy thinking about an issue even when the results of my thought will have no effect on the outcome ofthe issue. Use the following scale and place the number that best represents your opinions regarding professors on the line before each statement. There are no right or wrong answers; please respond to each item as honestly as possible. Strongly I 2 3 4 5 6 7 Strongly Disagree Agree 1. I oppose professors being granted tenure.* 2 I am in favor of abolishing tenure for all faculty. 3. I think that professors should be granted tenure if they merit it. 4 Tenuring professors is something of which I am in favor.* 5. I believe that abolishing tenure would be a good thing for the university.* 6. I think that tenure should be granted to all faculty.* 65 Strongly I 2 3 4 5 6 7 Strongly Disagree Agree 1. It is beneficial for students to make assumptions regarding professors based on interactions they have had with professors in the past. [schematic] 2. It’s difficult for students to generalize from one professor to another. [specific] 3. My opinions about professors mainly are based on my personal experiences with people whose family members or friends are professors. [schematic] 4. One is likely to make errors when thinking that all professors are the same.* [specific] 5. My opinions about professors mainly are based on whether or not someone I care about is a professor.‘ [schematic] 6. It is useful for students to make generalizations about professors based on professors they have had for classes in the past.* [schematic] 7. Students are generally right when they assume that all professors are the same.* [schematic] 8. My opinions about professors mainly are based on my judgment of how likely it is that I will interact with professors in any significant way. [schematic] 9. It’s useful for students tojudge professors based on individual interactions rather than generalizations. [specific] I0. My opinions about professors mainly are based on my personal experiences with specific professors. [schematic] I I. It is useful for students to form attitudes about professors based on specific incidences rather than overall generalizations.* [specific] Please respond to the following questions. I. What year are you in college? D Freshman D Sophomore Cl Junior El Senior 2. Approximately how many professors have you had for classes in the past? 3. What is your age? years 4. Are you (please check one)? D Female B Male 5. Are you? DAsian DAfrican-American DNative American DCaucasian DPacific-Islander DHispanic DOther: Please specify * Indicates those items retained for analysis. 66 REFERENCES 67 REFERENCES Allen, M., & Burrell, N. (1992). Evaluating the believability of sequential arguments. Argumentation and advocacy. E, 135-144. Cacioppo, J. T., & Petty, R. E. (1982). The need for cognition. Journal of Personality and Social Psychology, 12,. 116-131. Cacioppo, J. T., Petty, R. E., & Morris, K. J. (1983). Effects of need for cognition on message evaluation, recall, and persuasion. Journal of Personality and Social Psychology, 45, 805-818. Edwards, K. (1990). The interplay of affect and cognition in attitude formation and change. Journal of Personality and Social Psychology, _5_9, 202-216. Fiske, S. T., Neuberg, S. L., Beattie, A. E., & Milberg. S. J. (1987). Category- based and attribute bases reactions to others: Some information conditions of stereotyping and individuating processes. Journal of Experimental Social Psychology, 2, 399-427. Fiske, S. T., & Taylor, S. E. (1984). Social cognition. New York: McGraw-Hill. French, J. R. (1956). A formal theory of social power. Psychological review, 63. 181-194. Herek, G. M. (1987). Can functions be measured? A new perspective on the functional approach to attitudes. Social Psychology Quarterly, 59, 285-303. Herek, G. M. (1986). The instrumentality of attitudes: Toward a neofunctional theory. Journal of Social Issues, Q, 99-1 14. Hovland, C. 1., Harvey, O. J ., & Sherif, M. (1957). Assimilation and contrast effects in reactions to communication and attitude change. Journal of Abnormal and Social Psychology, §_5_, 242-252. Hunter, J. E. (1993). Confirmatory factor analysis: Statistical analysis overview. Unpublished lecture notes. E. Lansing, MI.: Department of Psychology. Katz, D. (I 960). The functional approach to the study of attitudes. Public Opinion Quarterly, E, 163-204. Katz, D., McClintock, C., & Sarnoff, (1957). The measurement ofego-defense as related to attitude change. Journal of Personality, 2_5_, 465-474. 68 Lavine, H., & Snyder, M. (1996). Cognitive processing and the functional matching effect in persuasion: The mediating role of subjective perceptions of message quality. Journal of Experimental Social Psychology, 32, 580-604. Millar, M. G., & Millar, K. U. (1990). Attitude change as a function of attitude type and argument type. Journal of Personalityand Social Psychology, _5_9, 217-228. Miller, G. R., & Steinberg, M. (1975). Between people: A new analysis of interpersonal communication. Chicago: Science Research Associates, Inc. Petty, R. E., & Cacioppo, J. T. (1979). Issue involvement can increase or decrease persuasion by enhancing message-relevant cognitive responses. Journal of Personality and Social Psychology, _3_7, 1915-1926. Petty, R. E., & Wegener, D. T. (1998). Matching versus mismatching attitude functions: Implications for scrutiny of persuasive messages. Personality and Social Pchhology Bulletin, _2_4, 227-240. Shavitt, S. (1990). The role of attitude objects in attitude functions. Journal of Experimental Social Psychology, 26, 124-148. Shavitt, S. (1989). Operationalizing functional theories of attitude. In A. R. Pratkanis, S. J. Breckler, & A. G. Greenwald (Eds), Attitude Structure and Function (pp. 311-338). Hillsdale, NJ: Erlbaum. Snyder, M. & De Bono, KG. (1985). Appeals to image and claims about quality: Understanding the psychology of advertising. Journal of Personality and Social Psychology, fl, 586-597. Snyder, M., & De Bono, K. G. (1989). Understanding the functions of attitudes: Lessons from personality and social behavior. In A. R. Pratkanis, S. J. Breckler, & A. G. Greenwald (Eds), Attitude structure and function (pp. 361-381). Hillsdale. NJ: Lawrence Erlbaum. Smith, M. B., Bruner, J. S., & White, R. W. (1956). Opinions and personality. New York: Wiley. 69 FOOTNOTES ’ Other scholars have examined the suasory impact of matched and mismatched messages (Edwards, 1990 regarding cognition-based attitudes; Millar & Millar. 1990); however, these studies have investigated the relationship between cognitive-based versus affective-based attitudes and cognitive versus affective messages. not function-message match or mismatch. 2 Herek (1987) created an Attitude Function Inventory based on the categorizations of open-ended responses to the statement, “I have generally positive [or negative] attitudes toward lesbians and male homosexuals because..." (p. 287). An example of an item generated from this procedure is: My opinions about gay men and lesbians mainly are based on whether or not someone I care about is gay. I-Ierck (1987) asserted that this instrument was tapping the experiential-schematic function of attitudes. On close inspection, these items seem to tap other functions as well (e.g., anticipatory), as conceptualized by Herek (1986). Other limitations regarding these items also must be considered. It would be difficult to adapt these items to tap any other attitude object beside homosexuality (Shavitt, 1989). Another concern that can be raised with these items is the ambiguity regarding what an individual’s response indicates. For instance. recall the item noted above. If subjects answered “disagree” to this statement, does that indicate that they do not know someone who is homosexual, or does that indicate that their opinions are based on something beside their experience with a particular person? One last limitation considered here is the inherent assumption Herek makes: that people can tap their own 70 cognitive processes to explain why they hold certain attitudes: unfortunately. this expectation may be too high. 3 Because the range of the measure of the functional basis of the attitude was truncated, another method of testing the hypothesis is to dichotomize the discrepancy measure and examine the regression of posttest attitude on the attitude’s functional basis. These analyses were performed, and produced results which require one to draw the same conclusions as presented in the text. These additional analyses are not presented here because they are redundant. 71 "iiiiiiiiiiii