Published on January 17, 2008
Defining Attitude Concept: Defining Attitude Concept Eagly & Chaiken (1993): emphasize the tripartite (multicomponent) classification. •”tendencies to evaluate an entity with some degree of favor or disfavor, ordinarily expressed in cognitive, affective, and behavioral responses” and formed on the basis of cognitive, affective, and behavioral processes. Evaluating = refers to all classes of evaluative responding, whether overt (verbal) or covert (nonverbal), cognitive, affective, or behavioral. Defining Attitude Concept: Defining Attitude Concept Tripartite (“trilogy of mind”) originally linked to “Faculty Psychology” Tripartite view of attitudinal responding: do attitudes have all three aspects? Grounded in 18th C. Enlightenment view of attitude (Cognition, Affection, Conation – act of striving). Kant, Leibniz, Scottish School: interest in consciousness and introspection. Debates about how many innate “faculties” of mind existed. Defining Attitude Concept: Defining Attitude Concept Preceded development of experimental psy in 19th C., and faded with its rise of latter in early 1920’s. Wundt, late 19th C in Germany, associationism was anti-introspection and discredited Faculty Psychology. But trilogy of mind remained in Psychology’s vocabulary. William McDougall (1923), Outline of Psychology (wrote 1st social psy text in 1908) Defining Attitude Concept: Defining Attitude Concept McDougall (1923): “We often speak of an intellectual or cognitive activity; or of an act of willing or of resolving, choosing, striving, purposing; or again of a state of feeling. But it is generally admitted that all mental activity has these three aspects, cognitive, conative, and affective…” Defining Attitude Concept: Defining Attitude Concept Tripartite view in contrast to Thurstone (1931: “Attitude is the affect for or against a psychological object.”); & later, Fishbein & Ajzen. Influenced Allport (1935): “An attitude is a mental or neural state of readiness, organized through experience, exerting a directive or dynamic influence on the individual’s response to all objects and situations to which it is related.” Defining Attitude Concept: Defining Attitude Concept Tripartite view played central role in attitude theory and attitude change research in its heyday in ’50s and ’60s. Rosenberg & Hovland (1960) model: Attitude is an inferred property of the 3 response classes, and the consistency of responses (formed on the basis of 3 different types of processes). Defining Attitude Concept: Defining Attitude Concept Zanna & Rempel (1986) evaluative appraisal model. Do attitudes have to have all 3 aspects? Z&R: categorization of a stimulus object along an evaluative dimension based upon 3 general classes of information: cognitive, affective/emotional, past behaviors or behavioral intentions. Model suggest that attitudes are separate cognitive entities which may be accessed from memory independent of the affective, cognitive, or behavioral information on which they are based. Defining Attitude Concept: Defining Attitude Concept 6 implications of this view: That these classes of information can determine evaluations separately or in combination. When evaluations are based primarily on utilitarian beliefs about an attitude object, the model is belief based. When evaluations are based primarily on affect produced by the object, the model becomes single component (evaluative, preferences) Defining Attitude Concept: Defining Attitude Concept 4. When evaluations based on inferences from past behavior, model is like self-perception. 5. If attitudes are based on different sources of information, do equivalent evaluations based on different sources differentially predict and guide behavior? (Priming) 6. Are such attitudes differentially susceptible to different methods of persuasion? Defining Attitude Concept: Defining Attitude Concept Attitudes as “tendencies to evaluate” - -there is an implicit or explicit response to an entity based on the “evaluative residue” of past experience (or beliefs or feelings) that predisposes the person to a favorable or unfavorable response. Attitudes can have varied antecedents on the input side, and varied consequences on the output side. But the attitude is not the response per se. Attitude is the tendency or latent property of the person that gives rise to judgments and categorizations. Defining Attitude Concept: Defining Attitude Concept Attitudes as Enduring or temporary constructions. Some attitudes are relatively enduring (formed early in life and carry through life; others are formed then changed; some formed but fade) N. Schwartz: Attitudes-as-construction view. Most if not all attitudes are unstable, constantly emerging anew in specific situations. Equates variability in the expression of attitudes with variability in the evaluative tendency that constitutes attitudes. Not same as context effects – latent construct can be stable but sensitive to context. Implicit and Explicit Attitudes: Implicit and Explicit Attitudes Chen & Bargh (1999): categorize good vs. bad. Access attitude from memory. Nonconsciously predisposes behavior toward object. Attitudes of which the person is not conscious at the moment of action (implicit attitudes) are also strongly predictive of behavior. Implicit and Explicit Attitudes: Implicit and Explicit Attitudes D. Myers (1990): “our attitudes predict our actions…if, as we act, we are conscious of our attitudes” (p.90). Bias toward the conscious operation of attitudes, not automatic activation. Greenwald & Banaji (1995) on implicit attitudes: “Implicit attitudes are introspectively unidentified (or inaccurately identified) traces of past experience that mediate favorable or unfavorable feeling, thought, or action toward social objects.” Implicit and Explicit Attitudes: Implicit and Explicit Attitudes What about those times when people have more than one evaluation of the same attitude object, one of which is more accessible than the other? Dual Attitude Model (Wilson, Lindsey, & Schooler, 2000) Dual Attitude Model: Dual Attitude Model Working Example: A White American reared in a racist family who learned to be prejudiced against African Americans. As an adult, he adopts egalitarian views and abhors prejudice of all kinds. What is this person’s attitude toward African Americans? Dual Attitude Model: Dual Attitude Model Dual Attitude Model (Wilson, Lindsey, & Schooler, 2000) ●Model proposes that people can have “dual attitudes,” which are different evaluations of the same attitude object (one is automatic, implicit attitude; other is explicit attitude). ●Proposes that the attitude people endorse at any point in time depends on whether they have the capacity to retrieve the explicit attitude, and whether explicit overrides implicit. Dual Attitude Model: Dual Attitude Model Remember: Implicit attitudes are “evaluations that have an unknown origin (people are unaware of the basis of their evaluation), are activated automatically, and influence implicit responses (uncontrollable responses and ones that are not seen as an expression of attitude and therefore are not controlled)” Greenwald & Banaji, 1995. Dual Attitude Model: Dual Attitude Model 5 Basic Hypotheses: A(e) and A(i) toward same attitude object can coexist in memory. When dual attitudes exist, A(i) is automatically activated; A(e) requires more capacity and motivation to retrieve from memory. When able to retrieve A(e), it overrrides A(i) and A(e) is reported. Dual Attitude Model: Dual Attitude Model 3. Even when A(e) is retrieved, A(i) influences implicit responses (i.e., uncontrollable responses like nonverbal behaviors) or responses that they do not view as an expression of their attitude and do not attempt to control (e.g., neural imaging). 4. A(e)’s change relatively easily, whereas Ai, like old habits, change more slowly. Attitude change techniques target A(e) but not A(i). Dual Attitude Model: Dual Attitude Model 5. Dual attitudes not same as ambivalent attitudes or attitudes with discrepant affective and cognitive components. People with dual attitudes report the attitude that is most accessible; don’t experience a subjective state of conflict from holding dual attitudes. Define attitudinal ambivalence vs. dual attitudes. Segue to Measurement: Segue to Measurement Direct measures: rely on self-reported attitudes. Asked direct questions about their thoughts, feelings, or behaviors toward attitude objects. Indirect measures: do not alert respondents to the identity of the object of the attitude being measured. Indirect measures rely on more circuitous methods of obtaining info. Assume that self-reports are of questionable validity because people are frequently unaware of their attitudes or unwilling to disclose them publicly. Different Types of Evaluative Response: Different Types of Evaluative Response Implicit-Explicit Measures (Hofmann, et al. 2005): Implicit-Explicit Measures (Hofmann, et al. 2005) 5 accounts for low r’s between explicit and implicit: Motivational biases in explicit self-reports (e.g., prejudicial attitudes). Lack of introspective access to implicitly assessed attitudes (introspection may increase awareness). Factors influencing the retrieval of information from memory [dual attitudes model; A(e) that are spontaneous correlate more highly with A(i)] Method-related characteristics of the two measures (e.g., lack of “correspondence”). Complete independence of the underlying constructs. Implicit-Explicit Measures (Hofmann, et al. 2005): Implicit-Explicit Measures (Hofmann, et al. 2005) Quantitative meta-analysis (126 studies) •Used IAT as implicit measure; various explicit measures used. •Overall effect size = .24 (sd=.14) •Moderators (e.g., research topic involved; awareness of A(i); effortful retrieval? higher r’s with spontaneous self-report (less thought) Attitude Measurement: Attitude Measurement Attitude Measurement: Attitude Measurement Attitude Measurement: Attitude Measurement Behavioral Indicators : Behavioral Indicators Assumption: that proper measurement on the behavior side is equally important and that we do not have to abandon attitude construct as long as we use properly scaled behavioral criteria and a valid attitude measure. Fishbein & Ajzen’s behavioral criteria: Self-report validity problems can also be addressed by measuring behavior appropriately. Behavioral Indicators: Behavioral Indicators Specific act or single act criterion: Should include 4 elements (action, time, context, target). Measure can be dichotomous or continuous. Repeated observations of same single act: repeated observations of same behavior at different observation times (e.g., unobtrusive measure of popularity of an art exhibit). Observations combined into repeated observation criterion. Behavioral Indicators: Behavioral Indicators 3. Multiple act criterion: Observation of different behaviors. 4. Multiple act, repeated observation: Gold standard. Cell entries can be summed, averaged, scaled. Attitude Measurement: Attitude Measurement Return to Multimethod Approach: Return to Multimethod Approach Guglielmi (199): Multidimensional view of prejudicial attitudes that makes use of multimethod strategies . Argues for both implicit and explicit measures of cognitive, affective, behavioral components. Esp. focused on psychophysiological methods. Return to Multimethod Approach: Return to Multimethod Approach Long history, beginning with Bogus Pipeline. Sigall and colleagues (1971): trying to account for decline in anti-Black sentiment using the adjective checklist procedure. Was the change due to social desirability? Hooked up participants to fancy machine; attached electrodes; used info collected earlier to establish “accuracy” Return to Multimethod Approach: Return to Multimethod Approach The asked to rate on 7-point scale how characteristic each of 22 traits was of Blacks and Whites (half rated each group). In order to determine “to what extent people are in touch with their real feelings” E allegedly checked participant’s verbal response against machine’s reading. Control: same task, no machine. Return to Multimethod Approach: Return to Multimethod Approach Found that students were much more likely to assign negative traits to Blacks under the bogus pipeline condition than Control. Significant racial prejudice exposed. Same concept as polygraph and lie detection – suspects need to believe that the machine will unmask their deception; leads to them spilling their guts, and the polygraph industry claiming efficacy. Same with No Lie MRI, Inc. Return to Multimethod Approach: Return to Multimethod Approach Which approaches are best? How does one choose? Does it really matter which technique one uses? What general conclusions should be reached? Some general conclusions: Some general conclusions Caveat: Answers depend in part on the attitude “objects” under investigation. Various assessment techniques are not interchangeable. One’s choice of measurement strategy can affect the results obtained and conclusions drawn about focal attitude (esp. intergroup attitudes). Some general conclusions: Some general conclusions Use of more subtle self-report measures and indirect measures yields a different picture. Responses that are difficult to control (e.g., physiological reactions, reaction times following racial primes, etc.) uncover more negative feelings and beliefs. Which set of findings more closely represents “true” attitudes? Results from direct measures must be viewed with some skepticism, but social desirability biases more problematic in certain contexts than others (atts toward fat and toward gay/lesbian people vs. race, gender, ethnicity). Some general conclusions: Some general conclusions 4. A combination of indirect and direct measures may be needed to fully understand people’s attitudes toward some groups and other attitude objects. Need such an approach to detect attitudinal subtleties (e.g., Fazio et al, 1995) Some general conclusions: Some general conclusions Fazio et al: Whites can be divided into three categories with respect to attitudes towards Blacks: “truly nonprejudiced”: no negative beliefs or feelings about Blacks; low prejudice scores on both direct and indirect measures. “truly prejudiced”: high scores on both direct and indirect measures; do not try to hide their negative feelings (either because prejudice is OK or because they fail to recognize that their attitudes are prejudiced) Some general conclusions: Some general conclusions “mixed” group: have automatic negative reactions to Blacks, but are motivated to control their prejudiced responses; look nonprejudiced on direct, self-measures, but negative attitudes more apparent on more subtle measures (e.g., some behavioral measures) or measures that tap uncontrollable responses (e.g., RTs). Point: one needs to adopt a multi-method approach and compare responses to both direct and indirect measures to detect these differences. Some general conclusions: Some general conclusions 5. Different instruments are designed to measure different aspects of intergroup and other kinds of attitudes. Physiological measures tap affective component; stereotype measures more cognitive; unobtrusive behavioral measures and social distance measures intended to assess discriminatory tendencies. 6. Tempting to always think that affective, cognitive, and behavioral measures are equivalent; but in fact only modestly correlated. Some general conclusions: Some general conclusions e.g., (Dovidio, Brigham, Johnson, & Gaertner, 1996): Some people hold negative beliefs about outgroups, yet believe it is wrong to act on them. Others deny having negative beliefs about outgroups, yet experience negative feelings toward those groups. Can’t assume that tripartite attitude model holds all or even most of the time (Schneider, 2004, pp.29-30). Point: Researchers’ measurement strategies will be shaped by the particular facets of the attitudes of greatest interest to them, and often they will find it necessary to use more than one type of measure. Intra-attitudinal Structure Matters: The allure of “safer” tobacco products: Intra-attitudinal Structure Matters: The allure of “safer” tobacco products “The more successfully a cigarette reduces risk, the more it might encourage smokers not to quit. Or lure ex-smokers to resume their habit. Or make kids smokers. It might, in other words, do exactly the opposite of what it is intended to do. In a worst-case scenario, it could reverse a half-century of antismoking education, policy and litigation in a flash.” (Gertner, 2005, p.46) Slide46: Tobacco harm reduction • Availability of “low-yield cigarettes” has led to public perceptions of “safer” cigarettes, but with no resultant decrease in morbidity (Fairchild & Cosgrove, 2004, American Journal of Public Health, “Out of the Ashes”; Myers, 2000, NEJM). • Similar concerns have been raised about the marketing of reduced harm products, underscoring need for science to fill the information gap on attitudes toward harm reduction and federal regulation of reduced harm products (H.R. 140, “Title V – FDA Regulations of Tobacco Products, referred to House Subcommittee on Health, 2/14/03) Two Key Background Concepts: Two Key Background Concepts Harm reduction: relates to actually seeing a reduction in mortality or morbidity with the use of a product. Potentially reduced-exposure products (PREPs): tobacco products that have been modified or designed in some way to reduce users’ exposure to tobacco toxins. Two categories – variants of traditional tobacco cigarettes (e.g., smokeless tobacco; new cigarettes that heat rather than burn tobacco), or pharmaceutical agents that are meant to aid in smoking cessation (e.g., nicotine gum, lozenges, nicotine patch). Slide48: The Psychology of Attitudes: Role of Attitude Structure: Cognitive versus affective bases Experience with smoking / harm reduction Knowledge about smoking / harm reduction Stark, Borgida, Kim, & Pickens (2006): The psychological bases of attitudes may influence the way consumers respond to ads about reduced harm/reduced risk products. Consistent with prior research in other social issue domains. Slide49: Method Survey: Minnesota Center for Survey Research •sent to 1,300 randomly selected households in 5-state Upper Midwest region (Minnesota, Iowa, North Dakota, South Dakota, and Wisconsin); Fall 2003 •438 adult participants returned the survey (38% return rate) 58.9% Male, 95.7% Caucasian Mean age: 54.2 years 21.9% smoked in last 30 days. Slide50: Predicting attitudes toward harm reduction: Affective score and experience significantly predict More positive feelings, being a smoker, lead to more positive harm reduction attitudes Affective Score: b=.394, p<.0001 Cognitive Score: b=.163, p<.099 Knowledge: b=.055, p<.346 Consistency: b=.163, p<.099 Experience: b= -.551, p<.006 Slide51: Predicting attitudes toward harm reduction by level of experience: Smokers’ attitudes are best predicted by their feelings, non-smokers’ attitudes are best predicted by their thoughts and beliefs Smokers: Affective score: b=.477, p<.0001 Cognitive score: b= -.122, p<.44 Non-smokers: Affective score: b=.079, p<.534 Cognitive score: b=.414, p<.0001 Does structure matter and for whom?: Does structure matter and for whom? For smokers, their feelings about harm reduction were the primary predictor of overall attitudes toward harm reduction; for non-smokers, thoughts and beliefs were the primary predictor. Feelings associated with smoking (taste, reduction of cravings, relaxation) may create positive attitudes that are difficult to counter with information on the health risks of these products. Slide53: Structural bases of attitudes may matter when: Educating people about these products and their associated risks. Persuading smokers to use these products to reduce their health risk. Strong feelings toward harm reduction might result in resistance towards some types of health messages – increased interest in resistance processes in persuasion field.