Show Summary Details

Page of

Printed from Oxford Research Encyclopedias, Communication. Under the terms of the licence agreement, an individual user may print out a single article for personal use (for details see Privacy Policy and Legal Notice).

date: 14 May 2021

Reasoned Action As an Approach to Understanding and Predicting Health Message Outcomesfree

  • Marco YzerMarco YzerSchool of Journalism and Mass Communication, University of Minnesota


The reasoned action approach is a behavioral theory that has been developed since the 1960s in a sequence of reformulations. It comprises the theory of reasoned action; the theory of planned behavior; the integrative model of behavioral prediction; and its current formulation, the reasoned action approach to explaining and changing behavior. Applied to health messages, reasoned action theory proposes a behavioral process that can be described in terms of four parts. First, together with a multitude of other potential sources, health messages are a source of beliefs about outcomes of a particular health behavior, about the extent of social support for performing that behavior from specific other people, and about factors that may hamper or facilitate engaging in the behavior. Second, these beliefs inform attitude toward performing the behavior, perceptions of normative influence, and perceptions of control with respect to performing the behavior. Third, attitude, perceived norms, and perceived control inform the intention to perform the behavior. Fourth, people will act on their intention if they have the required skills to do so and if there are no environmental obstacles that impede behavioral performance.

The theory’s conceptual perspective on beliefs as the foundation of behavior offers a theoretical understanding of the role of health messages in behavior change. The theory also can be used as a practical tool for identifying those beliefs that may be most promising to address in health messages, which makes the theory useful for those designing health message interventions. Reasoned action theory is one of the most widely used theories in health behavior research and health intervention design, yet is not without its critics. Some critiques appear to be misconceptions, such as the incorrect contention that reasoned action theory is a theory of rational, deliberative decision making. Others are justified, such as the concern that the theory does not generate testable hypotheses about when which variable is most likely to predict a particular behavior.

The Development of Reasoned Action Theory

The reasoned action approach to the understanding and prediction of human behavior, or in short, reasoned action theory, proposes that a small set of psychological variables explains any given behavior. Since its inception in the 1960s (e.g., Fishbein, 1963, 1967), the theory has been refined in a sequence of reformulations, including the theory of reasoned action (Fishbein & Ajzen, 1975), the theory of planned behavior (Ajzen, 1985), the integrative model of behavioral prediction (Fishbein, 2000), and its current formulation, the reasoned action approach to explaining and changing behavior (Fishbein & Ajzen, 2010). The different formulations build on one another in a developmental fashion and reflect improvements in conceptualization and measurement of the theory’s key constructs for the purpose of improving the precision with which behavior can be explained. Because the better one can explain behavior, the more information one has for designing a message that can effectively change that behavior, each next formulation of reasoned action theory also further improves the usefulness of the theory for understanding health message effects.

The Beginnings: A Theory of Attitude-Relationship Relationships

To understand the development of reasoned action theory, we should go back to the early 20th century, a time when social psychological scholarship strongly focused on the hypothesis that attitude is the basis for human behavior. Toward the second half of the 20th century, enthusiasm for this hypothesis waned, primarily because empirical findings did not consistently confirm that people act on their attitude toward a particular object. For example, Wicker (1969) reviewed 42 studies and concluded that “(t)aken as a whole, these studies suggest that it is considerably more likely that attitudes will be unrelated or only slightly related to overt behaviors than that attitudes will be closely related to actions” (p. 65).

In his review, Wicker next addressed possible explanations for the weak observed attitude-behavior relationship. Summarizing arguments made by a number of scholars, he noted that contemporary attitude measures reflected a range of concepts that purportedly all reflected attitude, yet in fact might conflate attitudinal antecedents and consequences, as well as attitude itself. He then observed that only one scholar, Martin Fishbein, had “attempted to combine several factors into a systematic formulation” (p. 74). Fishbein’s “theory of attitude-behavior relationships” (as Wicker referred to the unnamed theory) thus was designed to clarify the conceptual meaning of attitude, in part by mapping relations among a number of variables that together could explain behavior. Whereas the conceptual definitions of these variables would further develop across publications in the late 1960s, we can recognize the basis for what would become reasoned action theory. For example, Fishbein proposed that a person’s attitude “. . . is a function of his beliefs about the object (i.e., the probability that the object is associated with other objects, concepts, values, or goals) and the evaluative aspect of those beliefs. . .” (1963, p. 233). Drawing on Dulany’s (e.g., 1962) theory of propositional control over verbal responses, which explained intentions to elicit a conditioned verbal response as a function of variables that were conceptually similar to expectancy-value perceptions and social normative perceptions, Fishbein (1967) further argued that in addition to attitude, personal normative beliefs (“what do I feel I should do”) and social normative beliefs (“what do significant other people think I should do”) as well as the motivation to comply with those norms are the basic determinants of behavioral intentions, and that behavior is a function of the intention to perform that behavior. Any other variable, Fishbein argued, might have an effect on behavior, but if it does, it would do so indirectly, that is, by influencing attitudinal and/or normative beliefs.

The Theory of Reasoned Action

In the next stage of the theory’s evolvement, Fishbein and Icek Ajzen further formalized the theory as an explanation of an individual’s behavior as a function of behavioral intention, attitude, subjective norm, and ultimately, beliefs underlying attitude and subjective norm. Formally named the theory of reasoned action in 1979 (Fishbein, 1979), the theory was introduced in detail in a much cited 1975 book (Fishbein & Ajzen, 1975), which itself was the result of years of research that informed definitions, measurement, and relations between the theory’s primary variables. (In the process, personal norms lost their central position in the theory.)

In the theory of reasoned action, people’s intention to perform a behavior is the single best predictor of whether people will or will not perform the behavior. Behavioral intention is conceptualized as how probable people think it is that they will perform a particular behavior, and has typically been assessed by asking to what extent people think they will, plan, expect. and intend to perform the behavior. The intention to engage in a particular behavior was proposed to be shaped by attitude toward performing the behavior and by the subjective norm toward the behavior. Attitude was defined as one’s overall evaluation of performing the behavior in terms of favorability ratings such as bad or good, and subjective norm as the extent to which a person feels other people from his or her social network will approve or disapprove the person’s engaging in the behavior. Measurement recommendations argued for semantic differential scales.

According to the theory of reasoned action, attitude and subjective norm are themselves a function of a potentially rich set of beliefs. These beliefs were conceptualized as highly specific thoughts and feelings that represent what people have learned over time about performing a behavior. Similar to Bandura’s social learning theory, according to the theory of reasoned action, beliefs are formed through direct experience with the behavior but can also be learned vicariously or through inference. As in Fishbein’s initial theory (e.g., 1963), attitude was proposed to be a function of behavioral beliefs—sometimes also referred to as outcome expectancies—which in the theory of reasoned action refer to beliefs about the likelihood that performing a particular behavior will have certain outcomes. Theoretically, behavioral beliefs are thought to be weighed by how people evaluate each possible behavioral consequence in terms of good or bad. For example, people may expect that having a few drinks may make them feel tipsy, but whether this results in feeling favorable toward having a few drinks depends on whether people evaluate feeling tipsy as a desirable or undesirable outcome. Although in empirical research outcome evaluations are seldom measured, the idea of behavioral beliefs weighed by outcome evaluations has important conceptual and applied merit. For example, it makes clear that health messages can change attitude by successfully addressing beliefs about the likelihood of particular consequences of a behavior, evaluations of those consequences, or both.

Subjective norm was proposed to be a function of a person’s beliefs about whether specific individuals or specific groups of people, such as friends, colleagues, or one’s religious community, among others, will support or oppose the person’s engaging in a behavior. According to the theory of reasoned action, these normative beliefs are weighed by the extent to which a person is motivated to comply with each normative referent. Thus, for example, a person may expect that her physician will approve her getting vaccinated against HPV, but also believe that her parents will disapprove. If it is more important for her to do what her parents want her to do than what her physician wants her to do, she will experience an overall sense of normative pressure against getting HPV vaccination shots. See Figure 1 for a visual presentation of the theory of reasoned action.

Figure 1. The theory of reasoned action. Note that each formulation of the theory, including the theory of reasoned action, proposed external factors as sources of beliefs. Because diagrams are visual simplifications of the full array of theoretical propositions, they have typically not been included in visual presentations, however. Similarly, feedback loops from behavior to beliefs are proposed in all iterations of reasoned action theory, but seldom included in diagrams of the theory. Figures 1–4 present the diagrams most often used to visualize the theory.

Fishbein and Ajzen developed a number of propositions that are as relevant for predicting behavior today as when they were first introduced with the theory of reasoned action. Similarly, these propositions have remained the basis for each of the theory’s next formulations. All of these propositions are rules that explain when prediction of behavior is most precise. How behavior is defined is central to reasoned action’s propositions, which is more complicated than one would perhaps think. Consider, for example, weight loss. Weight loss seems an appropriate health research and message focus, because, for example, being overweight is a cancer risk factor. But weight loss is not a behavior; it is an outcome of behavior (and biological, social, and other factors). Many different behaviors may potentially lead to weight loss, and these different behaviors may vary in terms of the beliefs that underlie them. If weight loss is an outcome of behavior, would it be better to focus on, say, exercise? This, too, is problematic. Exercise is not a behavior either; it is a behavioral category, which can include a range of single actions.

For these reasons, Fishbein and Ajzen proposed that specific, single behaviors should be defined in behavioral research and health interventions alike. Levels of specificity vary by the extent to which a behavioral definition includes each of four components, that is, action (e.g., walking), target (e.g., two miles), context (e.g., at work), and time (e.g., every weekday during lunch breaks). Clearly, “walking” can be interpreted more broadly than “walking two miles at work every weekday during lunch breaks,” and when thinking about “walking” people may therefore think about quite different behaviors, each possibly associated with different, behavior-specific beliefs. Following the same logic, the correspondence proposition says that once a behavior is defined, all measures of all theory variables (i.e., intention, attitude, subjective norm, and behavioral and normative beliefs) should correspond to this behavioral definition in order to maximize the extent to which reasoned action variables can predict the behavior. Importantly, these propositions apply to health message interventions in a similar fashion: Health messages that address a single, specific behavior are more effective than messages that address outcomes, behavioral categories, or nonspecific behaviors (Fishbein, 2000).

The Theory of Planned Behavior

The theory of reasoned action was designed to explain behaviors that are under people’s volitional control. It is plausible, however, that for many behaviors people will not universally feel that performing the behavior is under volitional control. To broaden the explanatory scope of the theory, Ajzen proposed the concept of perceived behavioral control. He argued that in addition to attitude and subjective norm, perceptions of control over performing a particular behavior also shape behavioral intentions. Conceptually similar to Bandura’s conceptualization of self-efficacy (e.g., 1997), perceptions of control over behavioral performance were defined as the extent to which one feels capable of successfully performing a particular behavior. Ajzen argued that perceived behavioral control should strengthen intention to engage in the particular behavior, and, if perceptions of control match actual control, directly guide behavior. In the theory’s subsequent formulation, named the theory of planned behavior, Ajzen thus proposed perceptions of control over performing the behavior as a third determinant of intention, and as a direct determinant of behavior.

Perceived behavioral control was proposed to be shaped by control beliefs. In the theory of planned behavior, control beliefs pertain to how likely people believe it is that there will be certain resources and opportunities for, or obstacles and barriers to behavioral performance. These beliefs were argued to be weighed by perceptions of how powerful those resources and opportunities facilitate or obstacles and barriers obstruct behavioral performance. While theoretically plausible and practically important, very little research has attended to control, and in particular, power beliefs. In contrast, the structural part of the theory of planned behavior that explains intention as a function of attitude, subjective norm, and perceived behavioral control (see Figure 2) has been tested across a very wide range of domains on a truly massive scale. Several meta-analyses have supported the theory of planned behavior’s ability to explain substantial parts of the information in intention, and to somewhat lesser extent, behavior.

Figure 2. The theory of planned behavior.

The Integrative Model of Behavioral Prediction

The next iteration of the theory, named the integrative model of behavioral prediction (see Figure 3), was in part the result of a “theorists workshop”—a conference of leading behavioral theorists organized by the U.S. National Institute of Mental Health that sought to find agreement on factors that can best explain behavior in the context of HIV and AIDS risk and prevention. Building on the workshop, Fishbein (2000) expanded the normative influence concept by proposing that perceptions of normative pressure involve both subjective norms, henceforth labeled in reasoned action theory as injunctive norms, and, drawing on Cialdini and others’ work (e.g., Cialdini, Reno, & Kallgren, 1990), descriptive norms, which pertain to the extent to which a person believes other people from one’s social network engage in the behavior themselves. Fishbein further included Bandura’s concept of self-efficacy instead of the theory of planned behavior’s perceived behavioral control, in large part because he was concerned by the wide range of interpretations of perceived control in empirical research. He was particularly concerned about the practice of using perceived ease or difficulty as a measure of perceived control, which, similar to Bandura (1997), Fishbein argued, is not a marker of one’s sense of capability regarding performing a behavior. In fact, research has shown that perceived difficulty in some instances can be better understood as a measure of attitude (e.g., Yzer, Hennessy, & Fishbein, 2004).

In addition, Fishbein emphasized two propositions that had already been advanced in earlier iterations of the theory, but had not been recognized by other researchers as critically important for understanding behavior. The first was the idea that positive intentions do not always translate to behavior. In the integrative model two factors explain why people may not engage in a particular behavior despite their intention to do so. The first factor involves environmental barriers. For example, a heavy thunderstorm would make it hard if not impossible to act on one’s intention to go outside for some vigorous inline skating. The second factor involves required skills. One may have the intention to go inline skating out of a desire to adopt it as a part of one’s exercise regimen, but intention will not lead to behavior if one lacks the necessary skating skills. Note that this implies that if people already hold positive intentions to engage in a particular health behavior but do not do so, an intervention should address factors that impede acting on one’s intentions, and not messages that encourage behavioral performance by addressing attitudinal, normative, or control beliefs (Fishbein & Yzer, 2003).

Second, the integrative model emphasized that the role of any other variable can best be understood as a possible source of beliefs. Initially labeled external, then, in the integrative model, distal, and more recently background factors, these variables in essence include all variables outlined in social ecological models, including, among many others, cultural, political, religious variables; individual difference variables; demographic variables; attitudes toward a target (rather than toward a behavior—for example, attitude toward the environment compared to attitude toward my always recycling batteries); and media and interpersonal messages. The theory was not designed to explain the interrelations between all these background variables. Rather, its inclusion as background variables served to clarify belief-based behavioral processes: Exposure to a wide range of variables over time shapes beliefs about a particular behavior through a learning process.

Figure 3. The integrative model of behavioral prediction.

The Reasoned Action Approach to Predicting and Changing Behavior

In 2010 Fishbein and Ajzen published a book in which they reviewed five decades of reasoned action scholarship. They built on their review to advance the current formulation of the theory, which they named the reasoned action approach to predicting and changing behavior. (This was an accurate but perhaps not optimally effective name, as it does not seem to be universally evident that the current formulation of the theory is a next step in the development of the theory, and thus, for example, that the reasoned action approach builds on but is not identical to the theory of reasoned action.) New to the theory was the contention that attitude, perceived normative pressure, and perceived behavioral control each have two aspects. The two attitudinal aspects are instrumental attitude—an evaluation in terms of positive or negative attributes (e.g., foolish or wise)—and experiential attitude—an evaluation in terms of positive or negative experiences (e.g., pleasant or unpleasant). As in the integrative model, perceived normative pressure has injunctive norm and descriptive norm aspects. The two perceived behavioral control (or self-efficacy—the two labels refer to the same concept) aspects are perceived capacity—one’s perceived ability to perform the behavior—and perceived autonomy—the extent to which a person perceives that the decision to engage in the behavior is up to him or her. See Figure 4.

The dual aspect interpretation of attitude, perceived normative pressure, and perceived behavioral control is arguably the most significant modification of the theory relative to previous formulations. A 2016 meta-analysis by McEachan and colleagues supports the potential of reasoned action’s dual aspect propositions for better understanding behavior formation. Because the dual aspect conceptualization of attitude, perceived normative pressure, and perceived behavioral control offers a more finely tuned breakdown of behavioral determinants, it promises better understanding of how health messages work. Specifically, health message objectives can be set and their effects examined in terms of experiential and instrumental attitude, injunctive and descriptive normative pressure, and perceived autonomy and capacity, which is an improvement over omnibus attitudinal, normative and control objectives, and outcome measures that are not sensitive to the possibility that message effects on attitudinal, normative, and control aspects differ in magnitude and direction.

Figure 4. The reasoned action model.

In conclusion, the evolvement of reasoned action theory has been the result of decades of conceptual and empirical research that sought to maximize the precision with which behavior can be predicted. For the fathers of reasoned action theory, Martin Fishbein and Icek Ajzen, the guiding objective in this work has been the real-world value of the theory in terms of contributions to the solution of social problems. Health issues have been dominant among those.

The Reasoned Action Perspective on Health Message Design and Outcomes

To make clear how reasoned action theory explains and predicts health message outcomes and how it can inform the design of such messages, it is useful to first consider the multifaceted nature of health messages. One way of characterizing the message design process is by differentiating between strategy development and creative development. Strategy development involves maximizing understanding of the health issue, that is, why a health problem exists for a particular audience, for the purpose of determining what should be conveyed, or in short, message content. Next, creative development involves considering how message content can best be presented, which involves a plethora of decisions on, among many others, appeal, framing, source, and modality (Cappella, 2006; Hornik & Woolf, 1999). The interpretation of the message design process in terms of a strategy development stage and a creative development stage makes clear that any given health message has both content and format aspects, which represent quite different questions that need very different (yet complementary) theories in order to answer them (Cappella, 2006; NCI, 2005). This makes clear that no single theory exists that on its own can inform all possible design questions, or fully explain why a message did or did not work. The usefulness of reasoned action theory for understanding and predicting health message outcomes lies in its ability to identify optimal message content, conceptualized as the beliefs that most importantly guide a particular behavior.

To see how this works, consider that in the course of their lives, people’s experiences lead to the formation of a possibly large number of beliefs about a particular behavior. Those beliefs will not necessarily all be consistent with each other. In fact, it is likely that many beliefs that people hold about a particular behavior conflict with each other. For example, people may simultaneously believe that if they would have five or more alcoholic drinks they would have a good time with friends, but also that they might do something embarrassing. Or they may believe that their friends would approve their having five or more drinks but their spouse would not. Given that people’s actual behavior cannot be consistent with each of these beliefs—continuing the current example, people will either have five or more drinks or they will not—some beliefs must be more influential in guiding people’s behavior than others. This idea is consistent with the reasoned action proposition that even when people hold a very large number of beliefs about performing a particular health behavior, only a small number of these beliefs determine people’s behavior. Health messages that address those determinant beliefs have the potential to move people to behavior change. Equally important, health messages that address beliefs that are not related to the decision to engage in a particular behavior will not produce behavior change, even if messages successfully change those beliefs. Reasoned action theory thus argues that addressing those beliefs that most strongly guide behavior is necessary (but not sufficient) for health messages to be effective.

But how can reasoned action be used to identify such beliefs? A sequence of research steps has been recommended. First, the behavior of interest needs to be identified and defined at a level of specificity appropriate for the health issue at hand. Second, a belief elicitation study is conducted. This serves to identify beliefs about the particular behavior that exist in a population of interest, or in reasoned action terms, that are accessible in people’s memory. Open-ended questions are used that ask study participants to report which positive and negative outcomes, which referents who will approve or disapprove, and which factors that will hinder or facilitate what they associate with their performing the behavior. Answers to these questions are content analyzed and categorized into specific behavioral, normative, and control beliefs. These beliefs are typically then rank-ordered by the frequency with which they were reported, and those beliefs that are mentioned most often are retained for the third step in the research sequence. In this third step, the beliefs identified in the belief elicitation study are included in a closed-ended questionnaire that also asks about attitude, perceived normative pressure, perceived behavioral control, and behavioral intention. A regression analysis is used to determine the extent to which attitude, perceived normative pressure, and perceived behavioral control explain intention. In a last analysis, the beliefs underlying those variables that the regression analysis showed to most strongly determine intention are correlated with behavioral intention. According to reasoned action theory, those beliefs that most strongly correlate with intention are good candidates to consider as the content of a health message, following the logic that change in those beliefs should result in the greatest gain in intention.

Empirical work generally supports the effectiveness of reasoned action-based interventions. For example, Sheeran and colleagues (2016) analyzed 155 studies to test whether changing attitudes, norms, or perceived control leads to subsequent intention and behavior change. They found support for medium effect sizes (ranging from d = .36 to .51). As an illustration of how interventions brought about such change, consider the AIDS Community Demonstration Projects (ACDP). The ACDP was a large-scale, multiyear intervention aimed to increase a number of HIV-preventive behaviors—for example, using bleach to clean needles for intravenous drug use and condom use for vaginal and anal sex with steady and casual partners—among several at-risk communities—for example, homeless youth, commercial sex workers, and men who have sex with men but who are not gay identified. Through extensive formative research in the project communities, the intervention team identified beliefs about the various HIV-preventive behaviors that were next translated in role model stories distributed through small media materials such as newsletters. Because formative research had shown that self-efficacy was importantly associated with the various HIV-preventive behaviors, the stories emphasized self-efficacy, and in addition to the stories bleach kits and condoms were made available, thus further addressing self-efficacy. Initial research tested intervention effects on behavior change, which found that compared to comparison communities, bleach use and particularly condom use had increased in communities that received the intervention (CDC ACDP Research Group, 1999). In later research, Yzer, Fishbein, and Hennessy (2008) demonstrated that intervention effects on self-efficacy were responsible for these behavior changes.

The outcome evaluation research that was mentioned in the ACDP example brings us to the issue of what a health message can be expected to achieve. It is true that a health message’s ultimate objective is to move message recipients toward adopting, increasing, or maintaining a particular health behavior, or move message recipients toward ceasing or preventing them from initiating a particular behavior that puts their health at risk. Because of this, health messages are often considered to be a success if they result in a demonstrable change in behavior in the population under consideration, but a failure if they cannot be directly associated with behavioral outcomes. Behavior change should not be the only or even primary message objective or outcome variable, however, as an exclusive focus on behavioral objectives implies the erroneous assumption that health messages can directly change behavior. In fact, according to the research action approach, health messages can only directly affect beliefs about a particular behavior. Change in beliefs brought about by a health message should affect attitude, perceived normative pressure, and/or perceived behavioral control, which in turn should affect intention and behavior. Health messages thus improve health behavior by setting in motion a behavioral determinant process. An appropriate analysis strategy involves testing direct message effects on those beliefs that a message sought to introduce, improve, or reinforce, and next mediation analyses that test belief effects on attitude, perceived normative pressure, and perceived behavioral control, and last on intention and behavior. This strategy is consistent with reasoned action propositions, but is in practice seldom used. Instead, health message effects often are operationalized as direct effects on behavior, as in the initial ACDP outcome evaluation research. Such an analytical strategy cannot demonstrate whether the message was effective in achieving what it was designed to do—affecting beliefs. Whereas behavior change is an appropriate ultimate objective of health messages, immediate health message objectives and outcome evaluations should focus on a message’s ability to create, reinforce, or change particular beliefs.

Does this mean that reasoned action theory proposes that a message that successfully introduces, changes, or reinforces beliefs that guide a particular health behavior will always move people to behavior change? It does not. Because health messages are only one of possibly many sources of beliefs, the effectiveness of health messages in part depends on the importance of other sources of beliefs. These sources, which were discussed as background variables in reasoned action terminology, can include interpersonal communication, supporting and rival media content, previous behavioral experience, culture, and message recipients’ personality, among many others. Some of these may facilitate health message effects on beliefs, whereas others may hamper message effects. To fully understand health message effects on beliefs it therefore is necessary to consider message effects in the context of other sources that might affect belief formation and change. For example, work on exposure to health messages in the public sphere implies that the little exposure to health messages that campaigns can achieve is dwarfed by exposure to all other relevant media information that often goes against what health messages seek to achieve (e.g., Hornik, 2002). This idea is consistent with a series of studies by Sargent and colleagues, who found that watching movies importantly moved adolescents to a positive stance toward smoking and alcohol use, even though health messages that discourage smoking and alcohol use are available as well. The implication is that the effects of a health message are a function of the content and creative format of the message itself and the amount and nature of competing and facilitating information that message audiences are also exposed to.

In addition to competing sources of beliefs, several other factors might hamper or facilitate message effectiveness. Two of these are particularly important. First, the nature of the behavior at hand has a direct effect on what a message can achieve. Health messages are better able to contribute to behavior change when those behaviors do not require complex skills, resources, or a change in lifestyle. For example, a Reye’s syndrome prevention message that encourages parents to use acetaminophen instead of aspirin to treat their infants’ flu symptoms asks for a simple replacement behavior, which does not require complicated actions or lifestyle change. Compare this with a message that encourages lifelong heavy smokers to quit tobacco use. This is not a simple behavior; for one reason, the addictive nature of tobacco use makes quitting much more challenging than replacing one product with another, as in the Reye’s syndrome example. Second, personal involvement also determines how much can be expected from a health message. Continuing the smoking example, there is good evidence that the more smoking is central to how people see themselves, that is, if smoking is part of people’s identity, the less likely it is that smoking cessation messages can motivate them to quit.

These examples and the preceding discussion underscore that reasoned action theory offers a potentially useful tool for the identification of health message content and for understanding health message outcomes, yet that for full understanding of health message outcomes complementary theory is needed. Useful among those are exposure, information processing, and self-concept theories.

Discussion of the Literature

Reasoned action theory has received a vast amount of research attention in the domains of health, consumer, political, and environmental behaviors, among others. This research can be categorized in three broad groups. A first body of work has applied the theory to explain intention and, less frequently, behavior, in a particular behavioral domain. A second has used reasoned action to inform the design and evaluation of message-based interventions. A third group consists of research on reasoned action propositions. Among this latter work are studies that tested factors that determine the strength of the intention-behavior relation, the sufficiency of attitude, perceived norm, and perceived behavioral control as determinants of intention, the position of non-model variables such as self-identity, and differences between reasoned action and other behavioral theories in the extent to which these theories can predict intention and behavior. For example, in a series of studies, Conner and colleagues examined possible moderators of the intention-behavior relationship, including the possibility that the source of intention affects the extent to which intention predicts behavior (e.g., Conner, McEachan, Lawton, & Gardner, 2016).

In terms of research volume, research that applied reasoned action theory to determine how well intention to engage in a particular behavior can be explained by attitude, perceived norms, and perceived behavioral is perhaps the largest of these groups. Although important from an applied value perspective, these tests are unlikely to advance further development of the theory. Also noteworthy is that research that only tests intention as a function of attitude, perceived norms, and perceived behavioral control cannot speak to arguably the two most important parts of reasoned action theory, namely, beliefs and behavior. Research on beliefs and prospective research on behavioral outcomes is scarce. Moreover, research that sought to explain intention has relied on a single methodological approach, which makes it difficult to assess the extent to which research findings are affected by method variance. Specifically, a common approach is to use cross-sectional survey data to regress behavioral intention on attitudinal, normative, and perceived control measures to determine to what extent intention to engage in a particular behavior is shaped by attitudinal, normative, and/or control perceptions.

The lack of methodological diversity in reasoned action research brings with it a number of unique problems. For example, the reliance on small sample survey data makes it mathematically near impossible to demonstrate interactions between reasoned action variables, even if they in fact exist. In addition, cross-sectional data do not offer opportunities for testing reasoned action hypotheses. Although some researchers have tested reasoned action propositions experimentally, there is a great need for more experimental research to move reasoned action away from being an empirical tool only and toward a hypothesis-generating theory. Because interaction effects and experimental hypothesis testing offer important conceptual advancement, such tests are needed and it is unfortunate that conceptually meritorious ideas remain underexplored because of a lack of methodological creativity.

It is promising that increasing research attention is being paid to the dual aspect conceptualization of attitude, perceived normative pressure, and perceived behavioral control, which was the most novel modification in the current formulation of reasoned action theory. For example, McEachan and colleagues (2016) submitted 62 studies to a meta-analysis, which provided support for the idea that differential effects can be expected from different attitudinal, normative, and control aspects. Interestingly, their findings showed that in contrast to other model variables, perceived autonomy did not meaningfully contribute to intention across the 62 studies that McEachan and colleagues analyzed.

The McEachan meta-analysis allows another important observation. Of 374 studies that passed initial eligibility for meta-analysis screening, 292 were excluded because those studies did not include measures of instrumental and experiential attitude, injunctive and descriptive norms, or perceived capacity and autonomy. This signals that research that purportedly all apply reasoned action theory often differ in the measures used to capture reasoned action variables. It would be beneficial for replication as well as comparability purposes if reasoned action research would use similar measures for the same concepts as outlined in the current formulation of the theory. This is not yet the case—quite different measures are being used that make the reasoned action empirical literature unnecessarily heterogeneous.

A reasoned action approach suggests that the extent to which attitude, perceived normative pressure, and perceived behavioral control shape intention and behavior varies across behaviors and populations. A recent promising line of research has examined whether the extent to which these determinant variables shape intention and behavior can be made predictable, rather than treating it as an empirical question. For this purpose insights from other theories have been used in order to understand what it is about a behavior or a population that can inform predictions about intention and behavior formation. For example, construal level theory has been applied to argue that attitudinal perceptions may play an important role in shaping behavioral decisions that take place in the distant future, whereas perceived behavioral control may play an important role in shaping behavioral decisions that take place in a more proximal time frame.

The use of complementary theory may be among the most promising next steps in the development of reasoned action theory, because, as the construal level example illustrated, it has the potential to generate hypotheses about behavioral prediction. Reasoned action theory describes intention and behavior formation well, but treats the question which beliefs and variables will predict a particular behavior in a particular population as an empirical question. Indeed, this has been a concern for many scholars (e.g., Rothman, 2004).

Other concerns about reasoned action theory have been raised throughout the theory’s existence. (For an illustrative overview of opponent and proponent perspectives, see the 2015, volume 9, issue 2 of Health Psychology Review.) The most widely voiced concerns have to do with the sufficiency of reasoned action theory as an account of intention and behavior, and with the question whether reasoned action assumes rational, deliberative action.

Reasoned action theory proposes that behavior is a function of intention, and intention is a function of attitude, perceived norms, and perceived control. The theory’s sufficiency assumption holds that adding any other variable to this set of behavioral determinants should not improve the prediction of behavior. Many disagree with this assumption. In fact, a substantial amount of research has focused on a possible extension of the theory by adding a fourth determinant of intention, such as personal norms, anticipated affect, past behavior, and self-identity, among others. Some of these variables have been included on the basis of careful conceptual arguments, but most tests of additional predictors of intention relied on increments in variance explained in intention as an inclusion criterion, which is not a compelling argument. In response, it has been argued that many of the variables added as an extension of reasoned action theory are in fact background variables; measures of attitude, perceived norms, or perceived control; or explaining variance because of flawed measurement of reasoned action variables. Fishbein and Ajzen (2010) do offer the idea, however, that consistent with how the theory evolved, the search for additional variables that lead to a more accurate prediction of intention and behavior is in principle justified.

The reasoned action proposition that behavior follows reasonably from beliefs has met with arguably most criticism. Critics have offered at least two reasons for their concerns. First, some have interpreted “reasoned” as rational, in the sense of deliberative, systematic, objective, and judicious. Reasoned action theorists have argued that this is a misconception (Fishbein & Ajzen, 2010; see Ajzen, 2015, for a particularly powerful rebuttal). In their view, “reasoned” means that a person’s behavior is logically consistent with his beliefs, even if other people deem those beliefs irrational, biased, or based on incorrect information and thus objectively false. For example, individuals afflicted with severe depression often believe that they are in large part responsible for the symptoms they are experiencing, and for this reason are less likely to seek help for their depressive symptoms than individuals who experience less severe symptoms. These individuals act in a reasoned manner on their belief about personal responsibility, even though friends, family, and mental health professionals would deem the responsibility belief inaccurate, even irrational. Seen through a reasoned action lens, once beliefs are formed they are the cognitive basis from which behavior reasonably follows, regardless whether others deem those beliefs rational or irrational, accurate or inaccurate, objective or motivationally biased.

In a similar vein, the proposition that, for example, attitude is a function of the sum of belief-evaluation multiplicative products have led some to voice concerns that the theory incorrectly sees people as computers who analytically process potentially large numbers of beliefs to determine how they feel about performing a particular behavior. This indeed seems very unlikely. A few moments of reflection on the many behavioral choices we make each day makes clear that people do not consciously retrieve and then review all beliefs they hold about performing a particular behavior. In fact, it would be cognitively very demanding if not impossible to consciously process every specific belief that has been formed in the course of one’s life when making a decision about whether to engage in a particular behavior. It is true that the more personally relevant and otherwise significant a behavior is—and many health behaviors qualify as personally relevant—the more likely it is that one scrutinizes available information about that behavior. According to reasoned action theorists, even in those circumstances a general sense of favorability toward the behavior can come to mind spontaneously, which means that to draw on one’s evaluation of behavioral options one does not need to systematically review each and every specific belief one holds about those behaviors. Fishbein and Ajzen (Fishbein, 1963; Fishbein & Ajzen, 2010) argued that this phenomenon reflects the attitude formation process. In reasoned action theory, attitude formation is the process by which the specific beliefs about a behavior that one has formed over time inform an overall, generalized sense of favorableness toward engaging in the behavior. In turn, attitude, rather than its antecedents, that is, the complex set of beliefs one has learned in the course of one’s life, comes to mind, often automatically and unconsciously, in situations that are relevant for the attitude. For example, a message reminding us of an annual dentist visit prompts an immediate attitudinal response (i.e., negative or positive toward visiting the dentist), and not the potentially great many specific beliefs that over the years have shaped our attitude toward visiting the dentist. This learning process is fully consistent with reasoned action theory, and does not imply that beliefs are processed in computer-like fashion in decision-making situations.

Primary Sources

Primary sources relevant for reasoned action theory include textbooks dedicated to the theory. For the most recent version of the theory, Fishbein and Ajzen’s 2010 book (published by Psychology Press) on the reasoned action approach is an excellent choice. The book includes a detailed conceptual explanation of all reasoned action propositions, as well as a review of research on the theory and detailed recommendations for measurement and analysis. Also see Ajzen’s detailed recommendations for theory of planned behavior research at In addition, the theory has been described in multiple journal articles and textbook chapters that have appeared in many domains, but primarily in psychology and communication literatures. As one example, Sheeran and colleagues’ (2016) meta-analysis of the question whether changes in attitude, norms, and perceived control lead to changes in intention and behavior is useful by itself, but also because it includes clear references to 18 meta-analyses of reasoned action theory’s ability to explain health intention and behavior. A search for reasoned action theory in scholarship search engines using reasoned action, planned behavior, and/or integrative model will identify additional sources. Last, an annotated bibliography of reasoned action theory has been published in the Oxford Bibliographies series and can be found at

Further Reading

  • Ajzen, I. (2005). Attitudes, personality, and behavior (2d ed.). Maidenhead, U.K.: Open University Press.
  • Fishbein, M. (2008). A reasoned action approach to health promotion. Medical Decision Making, 28, 834–844.
  • Fishbein, M., & Ajzen, I. (2010). Predicting and changing behavior: The reasoned action approach. New York: Psychology Press.
  • Hennessy, M., Bleakley, A., Fishbein, M., Brown, L., DiClemente, R. R., Romer, D., . . . Salazar, L. (2010). Differentiating between precursor and control variables when analyzing reasoned action theories. AIDS and Behavior, 14, 225–236.
  • Robbins, R., & Niederdeppe, J. (2015). Using the integrative model of behavioral prediction to identify promising message strategies to promote healthy sleeping behavior among college students. Health Communication, 1, 26–38.
  • Yzer, M., Weisman, S., Mejia, N., Hennrikus, D., Choi, K., & Desimone, S. (2015). Informing tobacco cessation benefit use interventions for unionized blue-collar workers: A mixed methods reasoned action approach. Prevention Science, 16, 811–821.
  • Yzer, M. C. (2012). The integrated model of behavioral prediction as a tool for designing health messages. In H. Cho (Ed.), Designing messages for health communication campaigns: Theory and practice (pp. 21–40). Thousand Oaks, CA: SAGE.


  • Ajzen, I. (1985). From intentions to actions: A theory of planned behavior. In J. Kuhl & J. Beckman (Eds.), Action-control: From cognition to behavior (pp. 11–39). Heidelberg, Germany: Springer.
  • Ajzen, I. (1991). The theory of planned behavior. Organizational Behavior and Human Decision Processes, 50, 179–211.
  • Ajzen, I. (2015). The theory of planned behaviour is alive and well, and not ready to retire: A commentary on Sniehotta, Presseau, and Araújo-Soares. Health Psychology Review, 9, 131–137.
  • Bandura, A. (1997). Self-efficacy: The exercise of control. New York: Freeman.
  • Cappella, J. N. (2006). Integrating message effects and behavior change theories: Organizing comments and unanswered questions. Journal of Communication, 56, S265–S279.
  • Conner, M., McEachan, R., Lawton, R., & Gardner, P. (2016). Basis of intentions as a moderator of the intention-health behavior relationship. Health Psychology, 35, 219–227.
  • CDC ACDP Research Group (1999). Community-level HIV intervention in 5 cities: Final outcome data from the CDC AIDS Community Demonstration Projects. American Journal of Public Health, 89, 336–345.
  • Cialdini, R. B., Reno, R. R., & Kallgren, C. A. (1990). A focus theory of normative conduct: Recycling the concept of norms to reduce littering in public places. Journal of Personality and Social Psychology, 58, 1015–1026.
  • Dulany, D. E. (1962). The place of hypotheses and intentions: An analysis of verbal control in verbal conditioning. Journal of Personality, 30, 102–129.
  • Fishbein, M. (1963). An investigation of the relationships between beliefs about an object and the attitude toward that object. Human Relations, 16, 233–240.
  • Fishbein, M. (1967). Attitude and the prediction of behavior. In M. Fishbein (Ed.), Readings in attitude theory and measurement (pp. 477–492). New York: Wiley.
  • Fishbein, M. (1979). A theory of reasoned action: Some applications and implications. In H. E. Howe & M. M. Page (Eds.), 1979 Nebraska symposium on motivation (pp. 65–116). Lincoln: University of Nebraska Press.
  • Fishbein, M. (2000). The role of theory in HIV prevention. AIDS Care, 12, 273–278.
  • Fishbein, M., & Ajzen, I. (1975). Belief, attitude, intention, and behavior: An introduction to theory and research. Reading, MA: Addison-Wesley.
  • Fishbein, M., & Ajzen, I. (2010). Predicting and changing behavior: The reasoned action approach. New York: Psychology Press.
  • Fishbein, M., & Yzer, M. C. (2003). Using theory to design effective health behavior interventions. Communication Theory, 13, 164–183.
  • Hornik, R. (2002). Exposure: Theory and evidence about all the ways it matters. Social Marketing Quarterly, 8, 31–37.
  • Hornik, R., & Woolf, K. (1999). Using cross-sectional surveys to plan message strategies. Social Marketing Quarterly, 5, 34–42.
  • McEachan, R., Taylor, N., Harrison, R., Lawton, R., Gardner, P., & Conner, M. (2016). Meta-analysis of the Reasoned Action Approach (RAA) to understanding health behaviors. Annals of Behavioral Medicine, 50, 592–612.
  • National Cancer Institute. (2005). Theory at a glance. A guide for health promotion practice (2d ed.). Bethesda, MD: U.S. Department of Health and Human Services.
  • Rothman, A. J. (2004). Is there nothing more practical than a good theory?: Why innovations and advances in health behavior change will arise if interventions are used to test and refine theory. International Journal of Behavioral Nutrition and Physical Activity, 1, 11.
  • Sheeran, P., Maki, A., Montanaro, E., Avishai-Yitshak, A., Bryan, A., Klein, W. M. P., Miles, E., & Rothman, A. J. (2016). The impact of changing attitudes, norms, and self-efficacy on health-related intentions and behavior: A meta-analysis. Health Psychology, 35, 1178–1188.
  • Wicker, A. W. (1969). Attitudes versus actions: The relationship of verbal and overt behavioral responses to attitude objects. Journal of Social Issues, 25, 41–78.
  • Yzer, M., Fishbein, M., & Hennessy, M. (2008). HIV interventions affect behavior indirectly: Results from the AIDS community demonstration projects. AIDS Care, 20, 456–461.
  • Yzer, M. C., Hennessy, M., & Fishbein, M. (2004). The usefulness of perceived difficulty for health research. Psychology, Health and Medicine, 9, 149–162.