Maio, G. R., Haddock, G., & Verplanken, B. (2018). The Psychology of

Maio, G. R., Haddock, G., & Verplanken, B. (2018). The Psychology of Attitudes and Attitude Change (3rd Edition). SAGE Publications, Ltd. (UK).
Chapter 2 noted that attitude content, structure, and function are not independent. We have structured much of this textbook by focusing on attitude content and noting the relevance of the other attitude structure and function as we went along. Now that the bulk of the evidence has been covered, it is worth revisiting connections between these three properties of attitude.
As noted in Chapter 2, attitude function and content may be closely connected in many everyday attitudes. For example, value-expressive attitudes toward an automobile might be more strongly based on feelings than are utilitarian attitudes (Ennis & Zanna, 2000). In general, value-expressive attitudes may be more emotional, because of the strong emotional response that is thought to underlie cherished social values (Bernard et al., 2003a, 2003b; Maio & Olson, 1998b). Attitudes serving an ego-defensive function may possess a similarly strong foundation in emotion, which is consistent with classic psychoanalytic theory about the emergence of neurotic emotional behavior (Freud, 1966). It would be intriguing to see more tests of these ideas.
In addition, it would be interesting to learn more about the role of attitude function in attitude structure. Ambivalent attitude structures should be aversive to individuals because they subsume dissonance between cognitions (Maio et al., 1996; Newby-Clark et al., 2002; Nordgren, van Harreveld, & van der Pligt, 2006), though recent evidence impacts that this consequence varies as a function of mindfulness (Haddock et al., 2017). At the same time, these attitudes may be less useful for guiding behavior (Armitage & Conner, 2000). Because they fail to provide a single, useful summary of current knowledge, ambivalent attitudes might be less likely to serve a knowledge function.
Megan Thompson and Mark Zanna (1995) indirectly studied this issue by testing whether there are specific personality characteristics that are held by individuals who frequently possess ambivalent attitudes. The results were somewhat surprising. Specifically, these scientists found that participants who possessed ambivalent attitudes across many attitude issues were lower in need for cognition and higher in fear of invalidity. Put simply, people who are low in the desire to exert mental effort on complex problems (i.e., low in need for cognition) or high in the fear of being wrong (i.e., high fear of invalidity) are more content with ambivalent attitudes. The ambivalence may help them to save time and energy in reaching a decisive opinion and help them “sit on the fence” enough to avoid being completely wrong. This evidence implies that there can be a tactical laziness to ambivalence, and it contradicts the idea that ambivalent attitudes are less likely to serve a knowledge function. To the contrary, the people who should be most motivated to adopt attitudes serving this function (e.g., people high in fear of invalidity) are more likely to adopt ambivalent attitudes.
Perhaps, however, there is a vital distinction between having attitudes that accurately summarize information and having attitudes that are useful guides for behavior, even though both roles are often seen as characteristic of the knowledge function. Ambivalence may be useful for summarizing information, but less useful for behavior (van Harreveld et al., 2009). If this is true, then we may need to think of the knowledge function as involving separate information and action functions, which are not distinguished in current theories of attitude function. This issue is further highlighted by evidence that “wanting” and “liking” are different. For instance, people who fail to get a prize want it more and are more likely to choose it later (Litt, Khan, & Shiv, 2010). At the same time, they evaluate it more negatively after they fail to get it, and discard it relatively quickly after receiving it. Our appetites for action and our attitudes can be different, although they may harmonize in the long run.
Another interesting question is whether the positive and negative dimensions of attitudes fulfill different content and functions. Most measures of attitude allow for different content to influence the positive and negative dimensions of evaluation. For example, we might like ice cream because it is tasty, but dislike it because it is fattening. In contrast, the Individual Association Questionnaire (IAQ) is a new technique (Sedek, Piber-Dabrowksa, Maio, & von Hecker, 2011) that measures positivity and negativity in a manner that uses the same content to describe an attitude object (e.g., Germans). The IAQ uses synonyms of the same characteristic, but which vary in their valence (e.g., hardworking vs. workaholic). Such an approach can be used to examine attitude valence in a way that is at least somewhat separable from effects of content and function.
Related to this possibility, the bidimensional view indicates that positive and negative affective responses to attitude objects have distinct properties (Cacioppo et al., 1997), and we previously described neurological and behavioral evidence for this point of view (Chapters 2 and 9). This evidence should make us wonder whether the positive and negative dimensions of attitudes serve different functions. For example, positivity might be more strongly connected to needs for belongingness with others and creativity, while negativity might be more strongly connected with needs for security and self-protection. At the same time, we might wonder whether the functions of each dimension simply add together or does one dimension’s functions tend to override the other dimension’s functions? Negative reactions tend to be stronger than positive ones (Cacioppo et al., 1997), so it is plausible that the functions underlying negative responses might dominate attitudes.
The potential for one dimension’s functions to dominate the other is just one example of many issues that arise if we take seriously the view that attitudes subsume different positive and negative dimensions. However, there is an interesting overlap between this idea and recent evidence that people have dispositional attitudes; that is, people vary in the extent to which they like things in general and dislike things in general (Eschleman, Bowling, & Judge, 2015; Hepler & Albarraćin, 2013). This personality difference presumably reflects differences in tendencies to utilize positive versus negative content, perhaps helping to address psychological needs to see the world more or less positively or more or less negatively. For some people, the functions served by the positive dimensions of their attitudes may be more important than those served by the negative dimensions of their attitudes. In contrast, for other people, the functions served by the negative dimensions of their attitude may dominate.
In works of fiction, we hardly ever read anything about how witches come to be who they are. Were they abused as children? Like The Joker from the Batman comic books and movies, did they fall in a vat of toxic waste? The focus is always on how the characters affect others. Similarly, we have focused on understanding the effects of our three ‘witches’ and not so much on the factors that influence them.
The major exception to this was our discussion of how cognitive, affective, and behavioral processes shape attitudes in Chapters 5, 6, and 7. Attitudes should come to take on content that reflects the way they were formed. For instance, attitudes shaped by mere exposure should be based initially on affective information, with little accompanying cognitive and behavioral content. Similarly, attitudes shaped by self-perception of behavior should be based initially on the behavioral information, with little accompanying affective and cognitive content. Thus, these chapters potentially reveal important ways in which attitude content is shaped.
We have not said as much about the ways in which attitude structure and function are shaped because there is less evidence on these mechanisms. Yet there are many relevant issues. One interesting issue is that research has not systematically tested whether attitude structure changes after persuasive interventions. We learned in Chapter 9 that messages might cause discrepancies to emerge between explicit measures of attitude and implicit measures of attitude and that this evidence fits the idea that ambivalence emerges between the positive and negative dimensions of attitude. However, there is no model (that we know of) about how ambivalence is affected by messages and the ways different types of ambivalence emerge (e.g., potential ambivalence, felt ambivalence, cognitive-affective ambivalence).
A similar issue exists for attitude function. How do persuasive messages shape the functions of attitudes? Chapter 8 described evidence that messages are more likely to change attitudes when the messages match the function of recipients’ attitude than when the messages mismatch the function of recipients’ attitude. But what happens to the function of the attitude itself? Perhaps a mismatching message causes the attitude to partly change the function that it serves, leads to an attitude that has a different function, or produces an attitude that serves multiple functions. If multiple functions do arise, the attitude might become stronger because of its ability to serve diverse motives (see also Cialdini, 2008, for his discussion of “consistency”).
These speculations illustrate just some of the reasons why it is important to no longer treat attitude valence as though it were the only aspect of attitudes that can change after exposure to a persuasive intervention. Most research on persuasion has simply used explicit, self-report questions of attitude to assess the effects of messages. Although research on persuasion has begun to include implicit measures of attitude (Petty et al., 2006; Smith, De Houwer, & Nosek, 2012), it is worth considering that the content, structure, and function of attitudes can vary as well. These properties of attitude are too important to neglect as dependent variables in their own right, and we anticipate that future research will help us to better understand their development.