As we have seen, people are complex. They can be consistent, practicing what they preach, or they can surprise you, doing things that you would not expect based on their attitudes. Research sheds light on these phenomena. We know that attitudes frequently guide behavior, though the effects are stronger under certain circumstances, for some individuals, and with some attitudes more than others. The studies offer a patchwork— a pastiche—of conditions under which attitudes are more or less likely to influence action. Social scientists prefer more organized frameworks, such as models that explain and predict behavior. Two major theories of attitude–behavior relations have been proposed: the reasoned action and accessibility models.
The Reasoned Action Model
Fishbein and Ajzen, who brought you the precision of the compatibility principle, also formulated a major approach to attitude–behavior relations: the Reasoned Action
Model (Fishbein & Ajzen, 2010). Originally proposed by Fishbein and Ajzen in 1975, the approach was extended by Ajzen (e.g., 1991) in his planned behavior theory. Planned behavior theory emphasizes that attitudes won’t predict behavior if people don’t think they have control over the action. It adds the notion of perceived behavioral control to the mix. In 2010, Fishbein and Ajzen integrated their early work, planned behavior theory, and newer ideas about norms into the Reasoned Action Model, clarifying, enhancing, and extending the model to take into account new processes.
The Reasoned Action Model offers the most systematic explanation in the field of the processes by which beliefs influence behavior. It offers a roadmap for the journey that thoughts in a person’s head must travel before they can affect the actions he or she performs. In so doing, it generates a series of specific strategies that persuaders should employ to craft communications on any topic you can imagine.
First, a word about the name. The term reasoned action implies that people are rational and deliberate in how they go about deciding whether to perform a particular behavior. Fishbein and Ajzen (2010) are quick to acknowledge that people do not always think their decisions through mindfully. Instead, they argue that once people form a set of beliefs, they proceed to act on these beliefs in a predictable and consistent fashion. Their beliefs provide the cognitive basis from which attitudes, intentions, and behavior subsequently follow. Beliefs, scholar Marco Yzer (2013) emphasizes, are not always rational. However, they are powerful. For example, he explains:
Someone suffering from paranoid personality disorder may lock the door of his office because he believes that his colleagues are conspiring against him. This person acts in a reasoned manner on a belief, even though others would deem his belief irrational. Regardless whether beliefs are irrational, incorrect (because based on false information), or motivationally biased, once beliefs are formed they are the cognitive basis from which behavior reasonably follows.
(p. 121) There are five components of the theory: attitude toward the behavior, the individual’s judgment that performing the action is good or bad; perceived norm, perceived social pressure to perform the action; perceived behavioral control, the degree to which individuals believe they are capable of performing a particular behavior; behavioral intention, the intent or plan to perform a particular behavior; and behavior itself, the action in a particular situation (see Figure 5.5).
As an example, suppose you are doing volunteer work for a local high school, impelled by a recent near-fatal accident a high school friend sustained when he veered off the road, his attention diverted by an urgent text message he was sending to his girlfriend
Beliefs that the behavior leads to certain outcomes Attitude toward the behavior Perceived behavioral control Injunctive normative
beliefs that specific referents think person should or should not perform the behavior Descriptive normative beliefs that specific referents have performed the behavior Motivation to comply with the specific referents Perceived norm Identification with referents Intention Behavior Evaluation of the outcomes
■ Figure 5.5 The Reasoned Action Model
From Fishbein, M., & Ajzen, I. (2010). Predicting and changing behavior: The reasoned action approach. New York: Psychology Press
about a party later that evening. Propelled by knowledge that reckless driving is a major cause of lethal injuries in young adults (Fischer et al., 2011), you agree to help devise persuasive appeals to convince teenagers not to text when they drive. You decide to employ the reasoned action approach, launching your research by trying to understand why teens like to text when they drive. (See Figure 5.6.)
Attitude. Attitude toward the behavior is a highly specific attitude. It consists of two subcomponents: behavioral beliefs (beliefs about consequences of the behavior) and outcome evaluations (evaluations of the consequences). These two elements are combined as they were in the simple expectancy–value model described in Chapter 3. Each behavioral belief is multiplied by the corresponding evaluation, and results are summed across items. Beliefs and evaluations about texting while driving could be measured in this way:
■ Figure 5.6 Texting While Driving Can Increase the Chances of Traffic Fatalities. The Reasoned Action Model suggests strategies to persuade people not to engage in this risky behavior by shedding light on the psychological elements that underlie the behavior
Behavioral Beliefs
Texting While Driving on My Way to School: 1. Helps Me Stay in Touch with Friends
(Likely) 1 2 3 4 5 6 7 (Unlikely)
2. Keeps Me from Getting Bored
(Likely) 1 2 3 4 5 6 7 (Unlikely)
3. Makes Me Feel Less Lonely When I Drive
(Likely) 1 2 3 4 5 6 7 (Unlikely)
4. Reduces My Ability to Drive Safely
(Likely) 1 2 3 4 5 6 7 (Unlikely)
Outcome Evaluations
1. Staying in Touch with Friends
(Good) 1 2 3 4 5 6 7 (Bad)
2. Getting Bored
(Good) 1 2 3 4 5 6 7 (Bad)
3. Feeling Lonely When I Drive
(Good) 1 2 3 4 5 6 7 (Bad)
4. Driving Safely
(Good) 1 2 3 4 5 6 7 (Bad)
Perceived norm. Norms refer to acceptable, recommended behavior in a society. A perceived norm is perceived social pressure to perform a particular action. The perceived norm has two components: injunctive norms, or perceptions of what other people think we should do, and descriptive norms, perceptions of what others have done, are doing, or are likely to do in the future (Fishbein & Ajzen, 2010, p. 151).
Each consists of two subcomponents, which are multiplied together.
Injunctive norms consist of: (a) injunctive normative beliefs, or beliefs that individuals who are important to the person endorse the behavior; and (b) motivation to comply, the person’s motivation to go along with these significant others.
Descriptive norms are composed of: (a) descriptive normative beliefs, or beliefs about how frequently important others engage in the behavior; and (b) identification, the degree to which an individual identifies with these significant others.
Injunctive normative beliefs 1. My Mom Thinks That:
(I Should) 1 2 3 4 5 6 7 (I Should Not Text When I Drive to School)
2. My Dad Thinks That:
(I Should) 1 2 3 4 5 6 7 (I Should Not Text When I Drive to School)
3. My Best Friend Thinks That:
(I Should) 1 2 3 4 5 6 7 (I Should Not Text When I Drive to School)
Motivation to comply
1. When It Comes to Matters of Safety, I Want to Do What My Mom Thinks I Should Do.
(Strongly Agree) 1 2 3 4 5 6 7 (Strongly Disagree)
2. When It Comes to Matters of Safety, I Want to Do What My Dad Thinks I Should Do.
(Strongly Agree) 1 2 3 4 5 6 7 (Strongly Disagree)
3. When It Comes to Matters of Safety, I Want to Do What My Best Friend Thinks I Should Do.
(Strongly Agree) 1 2 3 4 5 6 7 (Strongly Disagree) Descriptive normative beliefs
1. My Mom Texts When She Drives Me to School. (All of the Time) 1 2 3 4 5 6 7 (Never) 2. My Dad Texts When He Drives Me to School.
(All of the Time) 1 2 3 4 5 6 7 (Never) 3. My Best Friend Texts When (S)he Drives Me to School. (All of the Time) 1 2 3 4 5 6 7 (Never)
Identification
1. When It Comes to Matters of Safety, How Much Do You Want to Be like Your Mom?
(A Lot) 1 2 3 4 5 6 7 (Not at All)
2. When It Comes to Matters of Safety, How Much Do You Want to Be like Your Dad?
(A Lot) 1 2 3 4 5 6 7 (Not at All)
3. When It Comes to Matters of Safety, How Much Do You Want to Be like Your Best Friend?
(A Lot) 1 2 3 4 5 6 7 (Not at All)
Descriptive and injunctive norms are then combined to produce an item measuring perceived norms.
Perceived behavioral control. This concept was proposed by Ajzen (1991) in his theory of planned behavior, which extended the theory of reasoned action (TRA) by emphasizing behavioral control perceptions. Perceived behavioral control is the degree to which individuals perceive they are capable of performing a particular behavior, or can control their performance of the action. The perceived behavioral control notion is fascinating, but complicated. It requires some explanation.
Let’s say that a young woman harbors a favorable attitude toward a behavior, and believes that important others endorse the behavior and perform it themselves. You would think that she would then perform the action. However, she might be unable to perform the behavior because she lacks either the necessary skills or confidence that she can enact the behavior in a particular situation.
For example: A woman may feel positively toward losing weight and perceive social pressure to drop 10 pounds. But she just can’t muster the self-control to stay on a diet. A young man may harbor a positive attitude toward quitting smoking and recognize that his friends all want him to quit. But every time he feels nervous, he takes a smoke. He lacks the cognitive skills or self-discipline to quit the habit. So, attitudes and norms will not predict intention or behavior in these cases. But if we add a perceived behavioral control to the mix, we increase capacity to say no to food and cigarettes, and attitude and norm will forecast the intention to lose weight and quit smoking. Among individuals who don’t think they can resist the lure of food and cigarettes, attitude and norm will not predict intention. No matter how positively they feel and how much they seek their friends’ approval, they don’t have the strength to implement the intention or to carry out their plans.
Thus, by adding perceived control to the model, we can come up with more accurate predictions about when attitudes predict intentions and behavior.
Okay, let’s get back to the problem at hand: measuring the degree to which teens plan to text when they drive. Simplifying Fishbein and Ajzen’s measurement strategy just a bit, you might come up with these two questionnaire items to measure perceived behavioral control.
Perceived Behavioral Control
How Much Control Do You Feel You Have Over Whether You Text When You Drive to School?
(No Control) 1 2 3 4 5 6 7 (Complete Control)
How Much Do You Feel That Texting While Driving to School Is under Your Control?
(A Lot) 1 2 3 4 5 6 7 (Not at All)
Behavioral intention. As the name suggests, behavioral intention is the intention to perform a particular action, a plan to put behavior into effect. Intention can be predicted quite accurately from attitude toward the behavior, perceived norm, and perceived behavioral control. Intention is most likely to predict behavior when it corresponds with—is identical to—the behavior in key ways. If you want to predict whether people will text when they drive to school, you should ask them if they intend to do just that. Asking them if they intend to text when they drive around the neighborhood or when they go on a family vacation would not predict texting while driving to school. The model uses a mathematical formula to determine the impact that each of the three factors exerts on intentions.
Intention to text when driving, measured as specifically as possible, could be assessed in this way:
I Intend to Text When I Drive to School.
(Definitely) 1 2 3 4 5 6 7 (Definitely Do Not)
Behavior. In general, intention to perform a particular behavior should predict the actual performance of the act. Thus, to provide an accurate measurement of behavior, you could ask respondents this question:
Over the Past Month, How Often Have You Texted When You Drove to School?
The model says that intention to text while driving—determined by attitude, norm, and perceived behavioral control—should do a pretty good job of predicting behavior. Predicting behavior from attitude. Recall the question guiding this section: Do attitudes predict behavior? The reasoned action approach allows us to specify the precise impact that attitudes exert on behavior. In the present case, young people who strongly believe that texting while driving leads to positive outcomes should be especially likely to intend to text when they drive.
In some cases, though, attitudes will not forecast action. Instead, norms dictate what people do. An adolescent might positively evaluate texting, but refrain because his mom does not think it is a good idea or because it is frowned on in his peer group.
Thus, the theory offers a framework for predicting behavior from attitude. Fishbein and Ajzen caution that behavior can be predicted, but you need to consider likes and dislikes (attitudes), people’s natural propensity to please others (norms), and individuals’ confidence that they can carry out their plans (perceived behavioral control). When strong social pressures are present, attitudes do not accurately forecast behavior (Wallace et al., 2005). When people do not have the skills or lack the confidence to execute their plans, attitudes may not predict intentions and intentions will not predict behavior. Reasoned action theory emphasizes that attitudes are a reasonably accurate indicator of what people will do, provided certain conditions are met. There will always be circumstances in which people, being complex, will behave on the basis of factors other than attitude. But the theory emphasizes that the more specific the correspondence among attitudes, norms, perceived behavioral control, intentions, and behavior, the better one can predict actual behavior. The more one can specify, in the different measures, the behavior that is performed, the context in which it takes place, and the time the behavior is enacted, the better one can forecast behavior.
The Facts on the Ground
The Reasoned Action Model has an outstanding track record in predicting behavior. More than 1,000 published studies have tested the model (Fishbein & Ajzen, 2010). Think about this for a moment: a thousand studies! Each study occurred in a different setting, surveyed different individuals, and yet was carefully executed. The fact that so many empirical studies have confirmed the model should give us pause—and confidence (see model summary in Box 5.2). More than 40 years after the theory was introduced, we know that attitudes, perceived norms, and perceived behavioral control forecast intentions, and that intentions help predict behavior (Ajzen & Fishbein, 2005; Albarracín et al., 2001; Sheeran, Abraham, & Orbell, 1999; Sutton, 1998). There is a long, diverse
list of behavioral correlates that the model has illuminated and usefully predicted. They include:
■ women’s occupational orientations (Sperber, Fishbein, & Ajzen, 1980); ■ breast-feeding and bottle-feeding infants (Manstead, Proffitt, & Smart, 1983); ■ voting in a national election;
■ grabbing meals in fast-food restaurants (Brinberg & Durand, 1983); ■ eating a healthy diet (see Booth-Butterfield & Reger, 2004);
■ condom use among high-risk heterosexual and homosexual adults in the United States and abroad (see Ajzen, Albarracín, & Hornik, 2007 and Morrison, Gillmore, & Baker, 1995);
■ drug use; and
■ hunting (Hrubes, Ajzen, & Daigle, 2001).
Shortcomings. If models could walk, this one looks like it walks on water! Well walk, for sure, but maybe only on land! As good as the model is, like all approaches, it has limitations. Some scholars protest that attitude and behavioral intention measures are virtually the same, making predictions obvious and not so interesting. Others note that, contrary to the assumption that the impact of attitudes on behavior is mediated by intentions, attitudes exert a direct impact on behavior (Bentler & Speckhart, 1979; Fazio, Powell, & Williams, 1989). A number of critics lament that the model assumes that human beings are logical, when much behavior is spontaneous and even impulsive. They note that when people hold a strong implicit prejudice, their conscious expression of the attitude on the survey will not predict negative body language toward a member of the disliked group (Ajzen & Fishbein, 2005).
There has been a lively debate about these issues over the past decades, with Fishbein and Ajzen emphasizing that their model can predict a wide variety of behaviors, including spontaneous and seemingly irrational behavior, like exceeding the legal speed limit and performing sex without a condom. Critics, for their part, argue that the model provides an unwieldy explanation of irrational behaviors and, in any event, not all behavior is preceded by an intention. In order to appreciate an alternative perspective based on these criticisms, the next section describes another approach to attitude–behavior relations.
Accessibility Theory
It’s a humid summer day, and you feel like a cold one. Glancing over the usual suspects—Miller Lite, Coors, Michelob, Bud Lite—your mouth watering, you want to make a quick choice of which six-pack to buy at the convenience store. Suddenly, the
1. The Reasoned Action Model offers a wide-ranging guide to attitudes and behavior.
2. There are five components of the theory: attitude toward the behavior, perceived norm, perceived behavioral control, behavioral intention, and behavior.
3. Attitude toward the behavior consists of behavioral beliefs and outcome evaluations.
4. Perceived norm has two components: an injunctive norm, or motivation to comply with what significant others think one should do in a particular situation; and a descriptive norm, beliefs about how frequently significant others have performed the targeted behavior and identification with these others.
5. These constructs, with their subcomponents, can forecast intention to behave in a particular situation. The degree to which attitude, norm, and perceived behavioral control predict intention depends on the particular context. 6. Perceived behavioral control, or perception of whether one is psychologically
capable of performing the behavior, helps determine whether attitude or norm predicts behavioral intention.
7. Behavioral intention predicts behavior, with prediction more accurate the closer and more specifically that intention corresponds to behavior. 8. There is strong support for the predictions of the model. A weakness is that
the model can be laborious to study, involving so many different steps and a rather arcane phrasing of questions. Another limit is the challenge of forecasting intentions and behavior when attitudes are irrationally or impli - citly held, as in the case of prejudice. This is, after all, a model of reasoned action, and not all behavior involves reasoned analysis. There is debate among researchers as to whether and when the model can account for actions that involve less mental forethought.
9. Persuasion primarily involves changing the core, salient beliefs that under - lie the attitude, normative, and perceived behavioral control aspects of a particular behavior.
BOX 5.2 UNDERSTANDING THEORY: PRIMER ON THE