Initiatives to Motivate Change: A Review of Theory and Practice and Their Implications for Older Adults

Rothman AJ.

Publication Details


Unhealthy behaviors and the disorders they cause pervade modern life. As even a cursory review of Healthy People 2010 (U.S. Department of Health and Human Services, 2001) makes perfectly clear, significant improvements in physical and mental health critically depend on changes in human behavior. The reductions in disease morbidity and premature mortality that would come from increased rates of physical activity; improved compliance with medical recommendations; reduced rates of obesity; and reduced rates of utilization of tobacco, alcohol, and other controlled substances are tantalizing. Moreover, these changes would result in meaningful improvements in quality of life and also increase the potential for dramatic reductions in health care costs.

Yet even though the benefits of these behavioral changes are clear, eliciting consistent changes in people's behavior has proven to be a formidable problem (Baumeister, Heatherton, and Tice, 1994; Rothman, 2000). For example, even though people recognize the health costs posed by smoking, the impact of those costs on behavioral decision making has proven to be quite complex—people downplay the importance of health concerns (Gerrard, Gibbons, Benthin, and Hessling, 1996), underestimate their own personal risk (Weinstein, 1998), or believe their fate is sealed and that cessation will make little difference. Moreover, behavioral decisions unfold in environments that too often hinder people's ability or opportunity to act in a healthy manner.

Although across the life span there are significant shifts in the types of behavioral decisions that people must grapple with, there remains a consistent need for people to recognize and adopt healthy patterns of behavior (Leventhal, Rabin, Leventhal, and Burns, 2001). Given the difficulty people have initiating and maintaining changes in their behavior, there is continued demand for intervention strategies that effectively motivate healthy behavior. To date, efforts to specify and test these strategies have tended not to focus on the behavioral decisions facing older adults. In a similar manner, the theoretical models that have guided these intervention efforts do not address, directly or indirectly, whether the underlying processes generalize to older adults.

In order to stimulate both theoretical and practical initiatives directed at the behavioral practices of older adults, this paper provides an overview of current models of how people decide to initiate and maintain healthy patterns of behavior and the implications of these models for intervention design and implementation. The review is divided into four sections. First, I consider whether people are ready to make changes in their behavior. Second, I examine what can be done to motivate people to initiate changes in their behavior, with a particular focus on communication strategies that have been shown to be effective. Third, I consider whether the factors that guide efforts to initiate a new pattern of behavior are distinct from those that determine whether those actions are sustained over time. Finally, because the development of initiatives to promote healthy behavioral decisions depends on communication between investigators who are working to advance theory and those working to advance practice (Rothman, 2004), I consider what can be done to facilitate this type of collaborative activity. In each section, consideration is given to the implications of the current state of basic and applied science for motivating change in older adults.


Modern society is replete with opportunities for people to make changes in their behavior. Yet the proportion of adults who act on these opportunities has been modest at best (e.g., Schmid, Jeffery, and Hellerstedt, 1989). In response to this observation, investigators have asserted that people differ in their readiness to change and that it is overly simplistic (and optimistic) to expect that most or all people are prepared to avail themselves of the opportunity to modify their behavior (Prochaska, DiClemente, and Norcross, 1992). Thus, initiatives that provide people with effective methods to modify their behavior must be complemented with initiatives that persuade people that it is both important and possible for them to change their behavior.

A rich array of factors has been invoked to explain why people differ in their readiness to make changes in their behavior (Salovey, Rothman, and Rodin, 1998). In some cases, especially those involving new hazards, it may be that people are unaware of the dangers posed by their actions. Alternatively, people's understanding of the health problem may lead them to believe (inaccurately) that their current actions are appropriate (Leventhal, Nerenz, and Steele, 1984). Even if people are reasonably well informed, they may choose not to modify their behavior because they perceive the value of the new pattern of behavior to be insufficient to justify a change in their behavior, or because they may not have confidence in their ability to perform the new pattern of behavior (Bandura, 1997). Structural factors may also constrain people's behavioral options, acting either directly on behavior (e.g., the absence of low-fat snack foods at one's workplace may hamper one from taking action) or indirectly by influencing people's beliefs about the behavior (e.g., the absence of low-fat snacks may lead people to perceive them as unpopular or undesirable).

Given the myriad of factors that determine people's behaviors, investigators have worked to formulate systematic accounts of the factors that regulate people's behavioral decisions with respect to health issues (e.g., health belief model [Rosenstock, Strecher, and Becker, 1988]; precaution adoption process model [Weinstein, 1988]; social cognitive theory [Bandura, 1997]; theory of planned behavior [Ajzen, 1991]; theory of reasoned action [Ajzen and Fishbein, 1980]; transtheoretical model of behavior change [Prochaska et al., 1992]).1 Although these models are remarkably similar in many respects, two general approaches can be identified (see Salovey et al., 1998, for a description of the specific features of each model). Some investigators have proposed that the process by which people decide to make changes in their behavior is best conceptualized as a series of stages and that unique factors determine whether people are able to transition from one stage to the next (e.g., transtheoretical model; precaution adoption process model). The critical implication of the stage-based approach is that intervention strategies are effective only if they target the specific needs of individuals at each stage and therefore interventions need to be stage matched. This approach can be contrasted with the conceptual framework that underlies continuum-based models such as the theory of reasoned action (Ajzen and Fishbein, 1980) or social cognitive theory (Bandura, 1997). In this framework, a single set of factors is identified that affects people's behavioral decisions and, regardless of their readiness to change, people will benefit from information highlighting any or all of these factors (for a detailed discussion of the relative features of stage- and continuum-based models, see Weinstein, Rothman, and Sutton, 1998).

There is little doubt that a continuum-based framework is more parsimonious than a stage-based framework. Stage models require not only a detailed articulation of the determinants of each stage transition, but also empirical evidence of the predicted dissociations among stage predictors. To date, nearly all of the empirical work that has been conducted cannot distinguish between predictions derived from the two frameworks and thus it is difficult to formulate strong arguments for one perspective over the other (Rothman, Baldwin, and Hertel, 2004; Weinstein et al., 1998).2 However, it is critical that investigators recognize explicitly which approach is guiding their efforts to motivate people to change their behavior (e.g., is the intervention expected to work consistently across people or will its impact be contingent on where people are in the behavior change process?).

The observation that most people are not ready to implement changes in their behavior does not appear to be less true for older people (Leventhal et al., 2001). In fact, if anything, it has been suggested that older adults are somewhat less interested in opportunities for new experiences (e.g., Fredrickson and Carstensen, 1990), which would suggest an even lower level of readiness to change. Of the two approaches identified above, a stage-based approach may provide some interesting insights into the behavioral practices of older adults as it more readily affords investigators the opportunity to respond to the practical and psychological demands posed by the situations in which people find themselves. The psychological needs and prior experiences that older adults bring to a decision are likely to have a significant effect on where they are most likely to be in the behavior change process. When younger adults express limited interest in modifying their behavior, it is typically thought to reflect a failure to attend to or think through the behavioral decision. However, among older adults, limited interest in modifying, for example, one's diet may indicate that a decision has been made to not make a change, as opposed to simply failing to consider the behavioral alternative (see Weinstein, 1988). This may be particularly true when examining behavioral decisions that involve issues that have been addressed repeatedly throughout the life span. If true, this has implications for initiatives to encourage behavior change. In this case, the failure to act would represent an active choice and thus initiatives would be needed to persuade people to reconsider their decision. Because a decision to not take action may lead people to reaffirm the set of beliefs that guided that decision, persuading people to alter the decision is likely to be more difficult than persuading them to consider a new behavioral alternative.

Asking for Change: Communication Strategies that Motivate Behavior Change

A primary strategy for motivating people to change their behavior has been to provide them with information that will persuade them to alter their behavior (Eagly and Chaiken, 1993).3 Although in some cases this may involve providing new information about an issue, messages are typically designed to help people recognize or confront issues that are familiar but not seen as important enough to motivate a change in behavior. Thus, the primary challenge in efforts to communicate information is getting people to not only attend to the message but also process it in a manner that maximizes its impact on how they think and feel about the issue (Petty and Wegener, 1998). A critical step in this process is to create a context in which people recognize an issue to be personally relevant. Heightened personal relevance motivates people to not only seek out new information but also process the information in a systematic manner (Petty and Cacioppo, 1990).

There is considerable empirical evidence that people strive to interpret information about themselves in a favorable light (Kunda, 1990). In the health domain, people readily accept information that indicates they are healthy, but are critical of and adopt higher standards of evidence for information that indicates a health problem (e.g., Ditto and Lopez, 1992; Liberman and Chaiken, 1992; see Rothman and Kiviniemi, 1999, for a review). For example, in an elegant series of experiments, Ditto and Croyle (1995) have shown that when people are provided with diagnostic information indicating a health problem, they respond to this information by minimizing its seriousness (“Oh, it's not such a big deal”) and increasing perceptions of its prevalence (“Oh, everyone gets it”). Although these processes have not been studied systematically in populations of older adults, there is evidence that older adults misattribute symptoms to aging and that this tendency both reduces emotional distress about the symptoms and delays interest in seeking care (Prochaska, Keller, Leventhal, and Leventhal, 1987). It is plausible that this tendency reflects a similar underlying interest in construing health information in a self-protective manner.

Despite evidence that people may resist unfavorable information about their health, it must be recognized that people respond to health information in a myriad of ways. As suggested by Leventhal and colleagues (1984), people may adjust their beliefs to alleviate the sense of threat and anxiety posed by a health concern but still choose to act to reduce the danger posed by the health concern. If investigators assess only people's beliefs about the seriousness of the health problem or how worried they are about the problem, they may develop an incomplete description of how people respond to health information. In fact, Croyle, Sun, and Louie (1993) found that even though people who were informed they had a health problem (high cholesterol) minimized the seriousness of the problem, they nonetheless expressed a heightened interest in information about how to treat the problem. It may be that people's efforts to reduce the perceived seriousness of the problem allows them to manage their affective reaction to the threat to a sufficient degree that they have the psychological resources they need to focus on taking action to reduce their risks (e.g., Aspinwall and Brunhart, 1996). However, if people are too quick or too good at minimizing the problem and the appropriate behavioral response to the problem requires sustained action, over time people may find themselves with insufficient motivation to continue in the necessary action. The observation by Carstensen and her colleagues (Carstensen, Isaacowitz, and Charles, 1999) that older adults prioritize actions that address their emotional needs could indicate that older adults will be particularly interested in regulating their affective reactions to the health information. Such an approach might attenuate any behavioral response to the problem. On the other hand, if older adults are able to maintain a positive emotional state, they may find that they have the psychological resources to deal with challenges to their health in an optimal manner (Salovey, Rothman, Detweiler, and Steward, 2000; Taylor, Kemeny, Reed, Bower, and Gruenewald, 2000). Any strong conclusion about the processes that guide how adults (young and old) respond to health information must wait until investigators assess a sufficiently broad set of indicators to provide a complete description of people's psychological and behavioral response.

Given the challenges associated with providing people with information about their health, are there communication strategies that have been shown to be effective ways to motivate behavior change? Message tailoring and message framing represent two conceptual approaches that not only have been shown to alter people's behavior systematically but also have received sufficient empirical scrutiny that investigators have begun to formulate an understanding of the factors that underlie their effectiveness. The state of the science for each of these approaches and their implications for efforts to motivate behavior change in older adults are examined in turn.

Message Tailoring

Message tailoring is guided by the premise that people will pay more attention to and be more persuaded by information that speaks directly to their own personal concerns.4 For example, a smoker who is concerned about the social stigma of smoking would be sent a message focusing on that topic, whereas a smoker who is concerned about how smoking is harming the health of his wife would be sent a message focusing on that topic. There is a growing body of empirical evidence indicating that tailored health messages are more effective than generic messages that provide all individuals with the same information (Brug, Glanz, van Assema, Kok, and van Breukelen, 1998; Dijkstra, De Vries, and Roijackers, 1998a, 1998b; Kreuter and Strecher, 1996; Kreuter, Oswald, Bull, and Clark, 2000; see Skinner, Campbell, Rimer, Curry, and Prochaska, 1999, for review).

Why might tailoring messages be more effective? Contemporary models of attitude change and persuasion may provide a rich conceptual framework for understanding how and when tailored messages are maximally effective (Petty, Barden, and Wheeler, 2002). In particular, to the extent that tailored messages are perceived to be personally relevant, recipients of the message will process the information more extensively, which, in turn, should increase its influence on people's thoughts and feelings about the health issue. All else being equal, greater elaboration of a strong health message is desirable as well-reasoned attitudes are more stable over time and better predictors of behavior (Petty and Wegener, 1998).

Tailoring information has been shown to increase the attention people pay to a message. Tailored messages are more likely to be read and remembered than are nontailored messages (e.g., Skinner, Strecher, and Hospers, 1994), are more likely to be discussed with others (e.g., Brug, Steenhuis, van Assema, and De Vries, 1996), and are perceived as more interesting and engaging (e.g., Brug et al., 1996; Kreuter, Bull, Clark, and Oswald, 1999). To the extent that the information addresses an individual's concerns (i.e., the perception that “it speaks to me”), it may be less likely to elicit feelings of reactance and it may be more difficult for people to assert that the message doesn't apply to them (see Gump and Kulik, 1995). In many ways, tailored messages invoke some of the strengths of a self-persuasion strategy, which puts people in situations in which they themselves bring to mind the relevant arguments or reasons for modifying their behavior (Aronson, 1999; McGuire and McGuire, 1991). As with self-persuasion, people who receive tailored messages may be more likely to construe their motivation to modify their behavior as reflecting intrinsic concerns, which have been shown to be predictive of behavior change (e.g., Williams, Ryan, Rodin, Grolnick, and Deci, 1998).

Advances in technology have made the prospect of tailoring messages to a person's unique set of personal concerns feasible, but it is not yet clear how investigators are to determine or prioritize which information should be tailored. For instance, is it more important that people's understanding of the determinants of a health problem are well matched so that they readily infer that they either suffer from or are at risk for a health problem, or should the recommended response to the problem be well matched so that people more easily infer that they can perform the new pattern of behavior? Although stage-based models of behavior change could provide a theoretical framework for selecting the dimensions on which to tailor a message, to date investigators have only compared tailored messages to generic messages. Any conclusions regarding the value of tailoring messages along specific dimensions (e.g., self-efficacy vs. perceived risk) or to a person's stage of change must wait until investigators systematically compare the impact of differentially tailored messages (see Weinstein et al., 1998, for a general discussion of these issues).

Although message tailoring has not been tested systematically with older adults, it provides a framework that can capitalize on the observation that as people age their primary concerns and goals shift. According to socioemotional selectivity theory, older adults prioritize goals that address emotional as opposed to informational needs and thus are more likely to focus on people and opportunities that can address those needs (Carstensen et al., 1999). This would suggest that tailored messages might prove to be particularly persuasive with older adults if they address how modifying one's behavior can satisfy one's emotional needs. However, research on the interplay between emotion and persuasion has found that people's emotional goals (e.g., desire to maintain a positive mood) affect the manner in which they process a message (Wegener and Petty, 1994; Wegener, Petty, and Smith, 1996). To the extent that older adults prioritize their ability to regulate and optimize their current emotional needs and, in particular, strive to construct a world that both heightens the opportunity for positive emotional experiences and constrains the opportunity for negative emotional experiences (i.e., antecedent emotional regulation; Charles and Carstensen, 1999; Gross, 1998), they may be particularly skilled at minimizing contact with messages about health problems.

A critical assumption implicit in tailoring messages to an individual's set of unique concerns is that the different ways in which people think about a health problem do not differentially predict behavior. For example, when asked to consider cessation, some smokers may focus on the good things they would like to get if they were not smokers (e.g., the extra time and money that they would have available for other interests), whereas other smokers may focus on the unpleasant things about smoking they want to avoid (e.g., the hassle and expenses associated with buying cigarettes and finding a place to smoke). Is it correct to assume that both of these construals are equally likely to motivate behavior?

Andrew Elliot and his colleagues have argued that all goals are not equally motivating and that, in particular, approach goals (i.e., goals that are characterized by a desire to reach a favorable goal state) are more strongly associated with favorable outcomes than are avoidance goals (i.e., goals that are characterized by a desire to stay away from an unfavorable goal state; Elliot and Church, 1997; Elliot and McGregor, 1999). This conclusion rests primarily on data from academic settings that have consistently revealed that the more avoidance goals students report (e.g., desire to not do poorly in the class), the poorer their subsequent academic performance. In a similar vein, contemporary statements of social comparison theory have emphasized that upward social comparisons (i.e., comparing one's performance or condition to that of someone who is doing better) are more likely to elicit improvements in performance than are downward social comparisons (i.e., comparing one's performance or condition to that of someone who is doing worse; Taylor and Lobel, 1989).

Translated into the health domain, this perspective suggests that people who focus on approach goals or upward comparison standards for the behavior (e.g., an interest in dieting predicated on a desire to be thin) will be more successful than those who focus on avoidance goals or downward comparison standards for the behavior (e.g., an interest in dieting predicated on a desire not to be fat), and that tailoring messages to approach goals will be more effective than tailoring them to avoidance goals. Although this prediction follows clearly from empirical findings obtained in academic settings, the extent to which it will be confirmed in the health area is uncertain. In prior studies, avoidance goals have almost always involved the desire to prevent an unpleasant future outcome (e.g., not failing a class). Although similar health goals exist (e.g., not wanting to develop cancer), avoidance goals in the health domain can also involve eliminating the risk of or curing an existing unpleasant state (e.g., shortness of breath). In an initial study of smokers enrolled in a cessation program, my colleagues and I found that smokers whose cessation efforts were motivated by avoidance goals—and, in particular, goals that involved eliminating a current problem—were subsequently more likely to quit smoking (Worth, Sullivan, Hertel, Rothman, and Jeffery, 2005). Specifying the impact of avoidance goals on health behavior change may be particularly important given the observation that older adults frequently possess health-related images of themselves that they want to avoid (Hooker, 1999).

Message Framing

Messages designed to promote a health behavior can be constructed to focus on the benefits of performing the behavior (a gain-framed appeal) or the costs of failing to perform the behavior (a loss-framed appeal). For example, a gain-framed brochure designed to promote prostate cancer screening would emphasize the health benefits afforded by screening, whereas a loss-framed brochure would emphasize the health costs of failing to be screened. According to the framing postulate of prospect theory (Tversky and Kahneman, 1981), how information is framed alters people's preferences. Specifically, people act to avoid risks when considering the potential gains afforded by their decision (they are risk averse in their preferences), but are more willing to take risks when considering the potential losses that may result from their decision (they are risk seeking in their preferences).

What impact might gain- and loss-framed health messages have on people's behavioral practices? In order to specify the relative impact of gain- and loss-framed health appeals, investigators must be able to determine the risk people ascribe to performing the targeted behavior (Rothman and Salovey, 1997). According to prospect theory, the likelihood that a particular frame will effectively motivate behavior is contingent on whether the option under consideration is perceived to reflect a risk-averse or risk-seeking course of action. If a behavior is perceived to involve some risk or uncertainty, loss-framed appeals are more persuasive, but if a behavior is perceived to afford a relatively certain outcome, gain-framed appeals are more persuasive.

In an attempt to apply this framework, Rothman and Salovey (1997; see also Rothman, Kelly, Hertel, and Salovey, 2002) proposed that the function served by a health behavior can operate as a reliable heuristic to predict whether people perceive engaging in a behavior to be risky. Specifically, detection behaviors serve to detect the presence of a health problem, and because they can inform people that they may be sick, initiating the behavior may be considered a risky decision. Although detection behaviors such as mammography provide critical long-term benefits, characterizing them as risky accurately captures people's subjective assessment of these behaviors (e.g., Lerman and Rimer, 1995; Meyerowitz and Chaiken, 1987). In contrast, prevention behaviors such as the regular use of sunscreen or condoms can forestall the onset of a health problem and maintain a person's current health status. In fact, these behaviors are risky only to the extent that one chooses not to engage in them. This distinction in risk perception suggests that loss-framed appeals would be more effective in promoting the use of detection behaviors and gain-framed appeals more effective in promoting the use of prevention behaviors.

The empirical evidence that has been obtained across both laboratory and field studies has provided strong support for the framework set forth by Rothman and Salovey (1997). Efforts to promote the use of detection behaviors have consistently found that loss-framed messages elicit greater interest in or use of behaviors such as breast self-examination (Meyerowitz and Chaiken, 1987; but also see Lalor and Hailey, 1990), mammography (Banks et al., 1995; Cox and Cox, 2001; Schneider et al., 2001; but also see Lerman et al., 1992), HIV testing (Kalichman and Coley, 1995), skin cancer examinations (Block and Keller, 1995; Rothman, Pronin, and Salovey, 1996), and blood-cholesterol screening (Maheswaran and Meyers-Levy, 1990).5

Because prevention behaviors typically afford people the opportunity to maintain their health and minimize the risk of illness, gain-framed messages are predicted to elicit greater interest in and use of prevention behaviors. Although only a few studies have tested the impact of gain- and loss-framed appeals on prevention behaviors, the empirical findings have revealed a consistent advantage for gain-framed appeals (e.g., for the use of condoms [Linville, Fischer, and Fischhoff, 1993] and sunscreen [Detweiler, Bedell, Salovey, Pronin, and Rothman, 1999; Rothman, Salovey, Antone, Keough, and Martin, 1993]). Although the pattern of empirical findings from studies that have targeted prevention or detection behaviors is consistent with the framework outlined by Rothman and Salovey (1997), clear evidence that the function of the behavior determines the relative influence of gain- and loss-framed appeals did not come until my colleagues and I systematically studied whether a given health behavior prevented or detected a health problem (Rothman, Martino, Bedell, Detweiler, and Salovey, 1999).

A compelling body of evidence supports the thesis that gain-framed messages are more effective when promoting a prevention behavior while loss-framed messages are more effective when promoting a detection behavior. However, it must not be forgotten that the distinction between prevention and detection behaviors rests on the premise that people perceive engaging in detection behaviors as posing some risk and engaging in prevention behaviors as posing little to no risk. To the extent that there is systematic variability in how people construe a detection or a prevention behavior, the relative influence of gain- and loss-framed appeals becomes more complex (Rothman et al., 2002). Specifically, people who construe engaging in the behavior to be risky (regardless of its function) will be more responsive to loss-framed information, whereas those who do not perceive it to be risky will be more responsive to gain-framed information. Only a few studies have tested whether people's perceptions of a behavior moderate the influence of framed appeals, and all have focused on perceptions of screening behaviors. However, the pattern of results has tended to provide support for the premise that people's response to framed information is contingent on their beliefs about the behavior (Apanovitch, McCarthy, and Salovey, 2003; Meyerowitz, Wilson, and Chaiken, 1991; Rothman et al., 1996).

To date, investigators have not systematically examined the impact of framed messages on the behavioral decisions of older adults, although some studies such as those targeting mammography have included samples of older participants. In light of the thesis that how people construe a behavior determines their response to gain- and loss-framed information, any efforts to use message framing to motivate older adults must be grounded in a clear understanding of how they perceive the targeted behavior. As people get older, their perceptions of screening behaviors may very likely change and this may be particularly true of those behaviors that have to be performed repeatedly. For example, the experience of having had a series of mammograms, all of which were negative, is likely to influence how women perceive the procedure. My colleagues and I (Rothman et al., 2002) have suggested that there may be two distinct types of reactions. Given the repeated experience that the mammogram did not detect a health problem, some women may begin to view the screening as less risky and even anticipate that subsequent screens will prove to be fine as well. This growing sense of reassurance would suggest that over time these women would become more responsive to a gain-framed appeal. However, other women may conclude based on the same series of events that being screened is at least as risky as not being screened, or even more so, if they believe that the string of favorable outcomes must come to an end. In this case, loss-framed appeals should continue to prove more effective.

People's perceptions of screening procedures are also likely to be affected by the fact that, as they age, they will know more and more people who have had an illness detected. Although it is possible that these experiences heighten concern that one's own procedure might find a problem, it is also quite plausible that the increased rate of health problems renders them more normative, which, according to research by Ditto and Croyle (1995), might work to alleviate people's concern about the problem. The notion that health problems are more normative may also enable older adults to shift their focus from the short-term consequences of screening procedures to the longer-term benefits afforded by the early detection of the disease. Alleviating the distress invoked by the health issue may also serve to increase the attention people will pay to the message (Wegener and Petty, 1994) and allow them to focus their efforts on addressing the danger posed by the health threat (Leventhal et al., 1984).

A critical, unresolved issue in the application of message framing with older adults is the impact that gain- and loss-framed information has on decisions regarding treatments for a health problem. Research on message framing and treatment decisions has almost always been limited to responses to hypothetical scenarios (Rothman and Salovey, 1997). The prototypical finding in these studies is that participants have a more favorable reaction to the proposed treatment when it is described in terms of its benefits (e.g., percentage of people who survive the procedure) than in terms of its costs (e.g., percentage of people who die from the procedure). One study examined the use of framed information in actual doctor-patient interactions regarding treatment for breast cancer (Siminoff and Fettig, 1989). Although doctors were found to have a systematic preference for using either gain- or loss-framed language when describing treatments to their patients, there was no report of the effect of the information frame on treatment decisions. Clearly, more research in this area is needed.

Finally, it may be useful to examine how people's general expectations about their health affect their response to framed appeals. Hooker and her colleagues have shown that as people get older, health concerns become increasingly incorporated into how they think of themselves and of who they might become (Hooker, 1999). To the extent that older adults are chronically concerned about the possibility of a health problem, loss-framed appeals might prove to be particularly persuasive. On the other hand, to the extent that they are focused on sustaining a positive view of themselves as healthy and active, gain-framed appeals might prove to be more persuasive. These perspectives converge with the more general observation that individual differences in how people think about goals and goal pursuits can affect the impact of different message strategies (Cesario, Grant, and Higgins, 2004; Lee and Aacker, 2004).

Although considerable progress has been made in delineating the conditions under which gain- and loss-framed appeals are most effective, the processes that underlie these effects are not sufficiently specified. Research on the influence of message framing on older adults should help to enrich our current understanding of the phenomenon.

Moving from Initiation to Maintenance: Why Do People Sustain Behavior Change?

Even when efforts to motivate people to modify their behavior prove successful, the changes that are elicited more often than not prove transient (e.g., diet and exercise to produce weight loss [Jeffery et al., 2000], smoking cessation [Ockene et al., 2000], recovery from substance abuse [Marlatt and Gordon, 1985]). Although some health behaviors need only be done once, the overwhelming majority of behaviors that people are urged to adopt (such as those identified in Healthy People 2010) require sustained action and their benefits are contingent on maintenance. Why is it that people who are able to successfully initiate changes in their behavior are more often than not unable to sustain those changes over time? One possibility from dual process models of persuasion is that some behavioral changes are ephemeral because the underlying attitudes that support the changes are weak (e.g., based on little cognitive elaboration) and thus unlikely to persist (Petty and Wegener, 1998). In addition, there may be meaningful differences in the processes that underlie the decision to initiate and the decision to maintain a pattern of behavior.

Current models of health behavior, cited above, have focused on elucidating how people determine whether to adopt a given behavior and have assumed, either implicitly or explicitly, that the factors underlying the decision to maintain a pattern of behavior are no different from those that govern its initiation. For example, the health belief model (Rosenstock et al., 1988), theory of reasoned action (Ajzen and Fishbein, 1980), and theory of planned behavior (Ajzen, 1991) make no direct reference to issues regarding behavioral maintenance other than to define it as a course of action sustained over a specified period of time.6 By comparison, stage models have identified maintenance as a distinct stage in the behavior change process. However, the primary focus of these theoretical approaches has been to delineate the processes through which people become ready to initiate a change in their behavior (Prochaska et al., 1992; Weinstein, 1988). The distinction between people in the action and maintenance stages is predicated solely on the length of time the behavior has been adopted, and thus the set of cognitive and behavioral strategies predicted to facilitate initial action are similarly predicted to help sustain that action over time (Prochaska and Velicer, 1997).

Social cognitive theory (Bandura, 1997) does explicitly consider how people regulate their behavior over time and identifies self-efficacy beliefs as a critical determinant of both the initiation and the maintenance of a change in behavior. Confidence in one's ability to take action serves to sustain effort and perseverance in the face of obstacles. The successful implementation of changes in behavior bolsters people's confidence, which, in turn, facilitates further action, whereas failure experiences serve to undermine personal feelings of efficacy. Although the reciprocal relation between perceived self-efficacy and behavior is well documented, this relation needs to be reconciled with the observation that successfully enacted changes in behavior are not always maintained.

I have recently proposed that there are important differences in the decision criteria that guide the initiation and maintenance of behavior change and that these differences may serve to explain why people who are able to make changes in their behavior may subsequently choose not to sustain those changes (Rothman, 2000; Rothman et al., 2004). Behavioral decisions by definition involve a choice between behavioral alternatives. What differentiates decisions concerning initiation from those concerning maintenance are the criteria on which the decision is based. Decisions regarding behavioral initiation involve a consideration of whether the potential benefits afforded by a new pattern of behavior compare favorably to one's current situation, and thus the decision to initiate a new behavior depends both on people holding favorable expectations regarding future outcomes and on their ability to obtain those outcomes. This premise is well grounded by a broad tradition of research (for reviews, see Bandura, 1997; Salovey et al., 1998).

Whereas decisions regarding behavioral initiation are based on expected outcomes, decisions regarding behavioral maintenance involve a consideration of the experiences people have had engaging in the new pattern of behavior and a determination of whether those experiences are sufficiently desirable to warrant continued action. Consistent with Leventhal's Self-Regulatory Model of Illness Behavior (Leventhal and Cameron, 1987), the decision to continue a pattern of behavior reflects an ongoing assessment of the behavioral, psychological, and physiological experiences that accompany the behavior change process. According to the model, people's assessment of these experiences is ultimately indexed by their satisfaction with the experiences resulting from the new pattern of behavior and people will maintain a change in behavior only if they are satisfied with what they have accomplished. The feeling of satisfaction indicates that the initial decision to change the behavior was correct and furthermore provides justification for the continued effort people must put forth to monitor their behavior and minimize vulnerability to relapse (for a more complete description of the model, see Rothman, 2000; Rothman et al., 2004).

Although there is evidence consistent with the premise that satisfaction is associated with behavioral maintenance in the health domain (e.g., Klem, Wing, McGuire, Seagle, and Hill, 1997; Urban, White, Anderson, Curry, and Kristal, 1992) as well as in the area of consumer behavior (e.g., Oliver, 1993), investigators have only begun to test systematically for dissociations between predictors of behavioral initiation and behavioral maintenance (see King, Rothman, and Jeffery, 2002). In order to test predictions regarding the differential determinants of initial and long-term behavior change, investigators have to capture the unique effect that a particular psychological state (e.g., self-efficacy) has on each phase of the behavior change process. To date, claims regarding the determinants of behavioral maintenance have relied on tests of whether a psychological state (e.g., self-efficacy at baseline) can predict a distal behavioral outcome (e.g., smoking status 18 months later). However, this analytic approach is inconclusive regarding the factors that underlie behavioral maintenance as it cannot determine whether, for example, people's initial feelings of self-efficacy influence their ability to maintain a behavior over and above its effect on their initial behavioral efforts. Across a series of intervention studies that have been designed to test predictions from the model in the contexts of smoking cessation and of weight loss, my colleagues and I have begun to accumulate evidence consistent with the model. For example, cessation self-efficacy has been shown to predict people's initial ability to quit smoking but not whether the quit is sustained. People's satisfaction with the outcomes afforded by cessation predicted whether people were able to sustain their quit over an extended period of time (Baldwin, Rothman, Hertel, Linde, Jeffry, Finch, and Lando, in press).

As efforts to understand the distinctions between the processes that guide behavioral initiation and behavioral maintenance move forward, it is critical that investigators examine how these processes operate in older adults. To date, investigators have found in samples of younger adults that people tend to have more difficulty maintaining than initiating a new pattern of behavior. Whether this pattern will hold for older adults is uncertain. Given the premise that older adults are somewhat less interested in new opportunities and find it more difficult to imagine the benefits afforded by a new behavior, it may be that older adults find it particularly hard to initiate new behaviors and thus greater attention must be paid to initiatives that can motivate them to take action.

However, the self-regulatory strategies that are typically employed by older adults may make them better able to maintain a new pattern of behavior (although declines in cognitive functioning may, over time, mitigate the impact of these strategies). To the extent that there may be greater continuity in the environments in which older adults live, it may be easier for them to develop habits. However, little is known about the process by which repeated behaviors become habits. My colleagues and I have suggested that what differentiates habits from behaviors that are maintained over time is that habits are behaviors for which people no longer feel any need to question their value (Rothman et al., 2004). For example, seat belt use may readily become a habit because people feel little need to continually reassess its worth. On the other hand, it may be difficult for a set of dietary behaviors to become a habit as people will continually reassess whether they are satisfied with the outcomes of the behavior. To the extent that habits are regulated by stored representations of the behavior, innovations in how people form and change implicit or automatically activated attitudes (Wilson, Lindsey, and Schooler, 2002) may provide critical insights into the design of initiatives that can create and sustain healthy habits.

If satisfaction proves to be a critical determinant of whether a new pattern of behavior is maintained, investigators need to formulate a better understanding of the factors that cause people to feel more or less satisfied with their actions. This is likely to be a daunting task as one would expect assessments of satisfaction to reflect not only the set of concerns highlighted by the behavioral domain but also the set of priorities specified by the individual. For example, people who want to lose weight in order to improve their social life will focus their assessment of their actions on a different set of criteria than will those who want to lose weight in order to be more physically active.

In order to apply this model to the behavioral practices of older adults, investigators need to develop a rich understanding of what older adults are likely to attend to when assessing the value of their behavior. According to socioemotional selectivity theory (Carstensen et al., 1999), older adults would be expected to base their evaluation on a behavior's ability to address or meet their emotional needs. To the extent that older adults focus their attention to a greater degree on their present psychological needs, they may find it particularly difficult to forestall an assessment of a behavior's value. This would lead to the prediction that older adults would have a particularly difficult time sustaining behaviors for which the initial costs are high and the focal benefits are delayed (e.g., smoking cessation). From the perspective of an intervention, older adults would therefore be likely to benefit from activities that both highlight any immediate benefits and heighten the critical long-term benefits afforded by the new behavior.

Older adults' ability to evaluate the consequences of their actions may also be constrained by cognitive impairments that can cause uncertainty as to whether the behavior has been performed (Purdie and McCrindle, 2002). If people fail to recognize that they have forgotten to perform a behavior (e.g., take medication), when they subsequently find that the benefits of the behavior are diminished, they will mistakenly conclude that the behavior is not effective. People's commonsense model of the health issue can also affect the conclusions they draw about their behavior as their model provides a framework for interpreting both the psychological and the physiological changes attributed to the behavior (Leventhal et al., 2001). For example, investigators have observed that patients misinterpret the feelings elicited by physical exertion as signs of cardiac distress and that this (mis)interpretation is associated with a reduction in activity (e.g., Petrie, Cameron, Ellis, Buick, and Weinman, 2002). Initiatives that shape what people expect to experience when engaging in a particular pattern of behavior may have the ability to increase the likelihood that people are satisfied with the actions they have taken and thus enhance the probability of sustained behavior change.

Taking the Next Steps: Enhancing Collaboration Between Theory and Practice

Given the critical impact that people's behavioral choices have on their health and well-being, the need for intervention strategies that can elicit changes in behavior continues unabated. Successfully meeting this challenge depends on advances in both theory and practice. Thus, we need to increase simultaneously the impact of our interventions and the relevance and precision of our theoretical models. Although there is widespread acknowledgment that theory and practice must progress in tandem, linking these efforts has proven to be a formidable challenge (Rothman, 2004; Suls and Rothman, 2004). One essential strategy is to improve the ways in which basic and applied behavioral scientists communicate with each other. In particular, the two broad groups of investigators must recognize and value the outcomes that can arise from close collaboration between theory and practice. On the intervention side, it is critical not only that intervention strategies be predicated on theoretical principles but also that their impact be assessed in a manner that can, in turn, inform theory. Investigators need to be able to articulate not only why an intervention will be effective but also how it will affect people's behavior. At the same time, theorists must specify the predicted relations in their models to a sufficient degree that the models can inform decisions regarding intervention design. Because interventions focus primarily on altering the status of a particular construct (e.g., perceptions of personal risk), it would be helpful if theorists elaborated more on the processes by which people formulate their beliefs. In this regard, social cognitive theory (Bandura, 1997) serves as an exemplary model as it clearly articulates the determinants of self-efficacy. By actively promoting the reciprocal relation between theory and practice, the information gleaned from intervention activities should lead to improvements in the quality and precision of our theoretical models—advances that, in turn, should enhance the development of new, innovative intervention strategies.

In the present context, efforts to apply current conceptualizations of behavioral decision making to older adults highlight the limits of our knowledge in this area. In particular, the less that is known about the processes underlying a given phenomenon, the more difficult it is to draw firm predictions about how or whether the effect might generalize new populations or situations. For instance, the applicability of message framing as a tool to motivate behavior change among older adults is constrained by limitations in our understanding of the processes that underlie its impact. The ability to specify the processes that underlie a particular intervention strategy also provides investigators with a framework within which to compare different types of intervention protocols. For example, an evaluation of a tailoring intervention might reveal not only that it was effective but also that its effectiveness could be attributed to its ability to heighten people's confidence that they could change their behavior. Although tailoring might prove to be an effective way to raise people's sense of personal efficacy, it is also quite possible that other intervention methods such as those that involve procedural or structural modification might prove to be a more effective way to target people's feelings of efficacy and, in turn, motivate behavior change. In the absence of any clear information about the underlying processes, it can prove to be quite difficult to compare and evaluate different intervention protocols.

If we are to successfully develop methods that can be used to motivate behavioral changes in older adults, investigators must embrace the value of a research program that addresses two distinct goals. The program must include activities that specify and detect both the psychological processes that regulate people's behavioral decisions and those that involve the implementation and evaluation of intervention strategies. It is hoped that the next generation of research in this area will be designed to meet these challenges.


  • Ajzen I. The theory of planned behavior. Organizational Behavior and Human Decision Processes. 1991;50:179–211.

  • Ajzen I, Fishbein M. Understanding attitudes and predicting social behavior. Englewood Cliffs, NJ: Prentice Hall; 1980.

  • Apanovitch AM, McCarthy D, Salovey P. Using message framing to motivate HIV-testing among low-income, ethnic minority women. Health Psychology. 2003;22:60–67. [PubMed: 12558203]

  • Aronson E. The power of self-persuasion. American Psychologist. 1999;54:875–884.

  • Aspinwall LG, Brunhart SM. Distinguishing optimism from denial: Optimistic beliefs predict attention to health threats. Personality and Social Psychology Bulletin. 1996;22:993–1003.

  • Baldwin AS, Rothman AJ, Hertel AW, Linde JA, Jeffery RW, Finch EA, Lando H. Specifying the determinants of behavior change initiation and maintenance: An examination of self-efficacy, satisfaction, and smoking cessation. Health Psychology. (in press) Submitted to. [PubMed: 17014280]

  • Bandura A. Self-efficacy: The exercise of control. New York: W.H. Freeman; 1997.

  • Banks SM, Salovey P, Greener S, Rothman AJ, Moyer A, Beauvais J, Eppel E. The effects of message framing on mammography utilization. Health Psychology. 1995;14:178–184. [PubMed: 7789354]

  • Baumeister RF, Heatherton TF, Tice D. Losing control: Why and how people fail at self-regulation. San Diego: Academic Press; 1994.

  • Block LG, Keller PA. When to accentuate the negative: The effects of perceived efficacy and message framing on intentions to perform a health-related behavior. Journal of Marketing Research. 1995;32:192–203.

  • Brug J, Glanz K, van Assema P, Kok G, van Breukelen GJP. The impact of computer-tailored feedback and iterative feedback on fat, fruit, and vegetable intake. Health Education and Behavior. 1998;25:517–531. [PubMed: 9690108]

  • Brug J, Steenhuis I, van Assema P, De Vries H. The impact of a computer-tailored nutrition intervention. Preventative Medicine. 1996;25:236–242. [PubMed: 8781000]

  • Carstensen LL, Isaacowitz DM, Charles ST. Taking time seriously: A theory of socioemotional selectivity. American Psychologist. 1999;54:165–181. [PubMed: 10199217]

  • Cesario J, Grant H, Higgins ET. Regulatory fit and persuasion: Transfer from “feeling right” Journal of Personality and Social Psychology. 2004;86:388–404. [PubMed: 15008644]

  • Charles ST, Carstensen LL. The role of time in the setting of social goals across the life span. In: Hess TM, Blanchard-Fields F, editors. Social cognition and aging. New York: Academic Press; 1999. pp. 319–345.

  • Cox D, Cox AD. Communicating the consequences of early detection: The role of evidence and framing. Journal of Marketing. 2001;65:91–103.

  • Croyle RT, Sun YC, Louie DH. Psychological minimization of cholesterol test results: Moderators of appraisal in college students and community residents. Health Psychology. 1993;12:503–507. [PubMed: 8293735]

  • Detweiler JB, Bedell BT, Salovey P, Pronin E, Rothman AJ. Message framing and sun screen use: Gain-framed messages motivate beach-goers. Health Psychology. 1999;18:189–196. [PubMed: 10194055]

  • Dijkstra A, De Vries H, Roijackers J. Computerized tailored feedback to change cognitive determinants of smoking: A Dutch field experiment. Health Education Research. 1998;13:197–206. [PubMed: 10181018]

  • Dijkstra A, De Vries H, Roijackers J. Long-term effectiveness of computer-generated tailored feedback in smoking cessation. Health Education Research. 1998;13:207–214. [PubMed: 10181019]

  • Ditto PH, Croyle RT. Understanding the impact of risk factor test results: Insights from a basic research program. In: Croyle RT, editor. Psychosocial effects of screening for disease prevention and detection. Oxford, England: Oxford University Press; 1995.

  • Ditto PH, Lopez DF. Motivated skepticism: Use of differential decision criteria for preferred and nonpreferred conclusions. Journal of Personality and Social Psychology. 1992;63:568–584.

  • Eagly AH, Chaiken S. The psychology of attitudes. Orlando, FL: Harcourt, Brace, Jovanovich; 1993.

  • Elliot AJ, Church MA. A hierarchical model of approach and avoidance achievement motivation. Journal of Personality and Social Psychology. 1997;72:218–232. [PubMed: 10234849]

  • Elliot AJ, McGregor HA. Test anxiety and the hierarchical model of approach and avoidance achievement motivation. Journal of Personality and Social Psychology. 1999;76:628–644. [PubMed: 10234849]

  • Fredrickson BL, Carstensen LL. Choosing social partners: How old age and anticipated endings make people more selective. Psychology and Aging. 1990;5:335–347. [PMC free article: PMC3155996] [PubMed: 2242238]

  • Gerrard M, Gibbons FX, Benthin AC, Hessling RM. A longitudinal study of the reciprocal nature of risk behaviors and cognitions in adolescents: What you do shapes what you think, and vice versa. Health Psychology. 1996;16:344–354. [PubMed: 8891713]

  • Gollwitzer PM. Implementation intentions: Strong effects of simple plans. American Psychologist. 1999;54(7):493–503.

  • Gross JJ. Antecedent- and response-focused emotion regulation: Divergent consequences for experience, expression, and physiology. Journal of Personality and Social Psychology. 1998;74:224–237. [PubMed: 9457784]

  • Gump BB, Kulik JA. The effect of a model's HIV status of self-perceptions: A self-protective similarity bias. Personality and Social Psychology Bulletin. 1995;21:827–833.

  • Hooker K. Possible selves in adulthood: Incorporating telenomic relevance into studies of the self. In: Hess TM, Blanchard-Fields F, editors. Social cognition and aging. New York: Academic Press; 1999. pp. 99–124.

  • Jeffery RW, Drewnowski A, Epstein LH, Stunkard AJ, Wilson GT, Wing RR, Hill DR. Long-term maintenance of weight loss: Current status. Health Psychology. 2000;19:5–16. [PubMed: 10709944]

  • Kalichman SC, Coley B. Context framing to enhance HIV-antibody-testing messages targeted to African American women. Health Psychology. 1995;14:247–254. [PubMed: 7641666]

  • King C, Rothman AJ, Jeffery RW. The challenge study: Theory-based interventions for smoking and weight loss. Health Education Research. 2002;17(5):522–530. [PubMed: 12408197]

  • Klem ML, Wing RR, McGuire MT, Seagle HM, Hill JO. A descriptive study of individuals successful at long-term maintenance of substantial weight loss. American Journal of Clinical Nutrition. 1997;66:239–246. [PubMed: 9250100]

  • Kreuter MK, Bull FC, Clark EM, Oswald DL. Understanding how people process health information: A comparison of tailored and non-tailored weight-loss materials. Health Psychology. 1999;18:487–494. [PubMed: 10519465]

  • Kreuter MK, Oswald DL, Bull FC, Clark EM. Are tailored health education materials always more effective than non-tailored materials? Health Education Research. 2000;15:305–315. [PubMed: 10977378]

  • Kreuter MK, Strecher VJ. Do tailored behavior change messages enhance the effectiveness of health risk appraisals? Results from a randomized trial. Health Education Research. 1996;11:97–105. [PubMed: 10160231]

  • Kunda Z. The case for motivated reasoning. Psychological Bulletin. 1990;108:480–498. [PubMed: 2270237]

  • Lalor KM, Hailey BJ. The effects of message framing and feelings of susceptibility to breast cancer on reported frequency of breast self-examination. International Quarterly of Community Health Education. 1990;10:183–192. [PubMed: 20840929]

  • Lee AY, Aacker JL. Bringing the frame into focus: The influence of regulatory fit on processing fluency and persuasion. Journal of Personality and Social Psychology. 2004;86:205–218. [PubMed: 14769079]

  • Lerman C, Rimer BK. Psychosocial impact of cancer screening. In: Croyle RT, editor. Psychosocial effects of screening for disease prevention and detection. Oxford, England: Oxford University Press; 1995. pp. 65–81.

  • Lerman C, Ross E, Boyce A, Gorchov PM, McLauglin R, Rimer B, Engstrom P. The impact of mailing psychoeducational materials to women with abnormal mammograms. American Journal of Public Health. 1992;82:729–730. [PMC free article: PMC1694134] [PubMed: 1566953]

  • Leventhal H, Cameron L. Behavioral theories and the problem of compliance. Patient Education and Counseling. 1987;10:117–138.

  • Leventhal H, Nerenz DR, Steele DJ. Illness representations and coping with health threats. In: Baum A, Taylor SE, Singer JE, editors. Handbook of psychology and health. Vol. 4. Mahwah, NJ: Lawrence Erlbaum; 1984. pp. 219–252.

  • Leventhal H, Rabin C, Leventhal EA, Burns E. Health risk behaviors and aging. In: Birren JE, Schaie KW, editors. Handbook of the psychology of aging. 5th ed. New York: Academic Press; 2001. pp. 186–214.

  • Liberman A, Chaiken S. Defensive processing of personally relevant health messages. Personality and Social Psychology Bulletin. 1992;18:669–679.

  • Linville PW, Fischer GW, Fischhoff B. AIDS risk perceptions and decision biases. In: Pryor JB, Reeder GD, editors. The social psychology of HIV infection. Mahwah, NJ: Lawrence Erlbaum; 1993. pp. 5–38.

  • Maheswaran D, Meyers-Levy J. The influence of message framing and issue involvement. Journal of Marketing Research. 1990;27:361–367.

  • Marlatt GA, Gordon JR. Relapse prevention: Maintenance strategies in the treatment of addictive behaviors. New York: Guilford Press; 1985.

  • McGuire WJ, McGuire CV. The content, structure, and operation of thought systems. In: Wyer RS Jr, Srull TK, editors. Advances in social cognition. Vol. 4. Mahwah, NJ: Lawrence Erlbaum; 1991. pp. 1–78.

  • Meyerowitz BE, Chaiken S. The effect of message framing on breast self-examination attitudes, intentions, and behavior. Journal of Personality and Social Psychology. 1987;52:500–510. [PubMed: 3572721]

  • Meyerowitz BE, Wilson DK, Chaiken S. Loss-framed messages increase breast self-examination for women who perceive risk. Paper presented at the annual convention of the American Psychological Society; June 13-16; Washington, DC. 1991.

  • Ockene JK, Emmons KM, Mermelstein RJ, Perkins KA, Bonollo DS, Voorhees CC, Hollis JF. Relapse and maintenance issues for smoking cessation. Health Psychology. 2000;19:17–31. [PubMed: 10709945]

  • Oliver RL. Cognitive, affective and attribute bases of the satisfaction response. Journal of Consumer Research. 1993;30:418–430.

  • Petrie KJ, Cameron LD, Ellis CJ, Buick D, Weinman J. Changing illness perceptions after myocardial infarction: An early intervention randomized controlled trial. Psychosomatic Medicine. 2002;64:580–586. [PubMed: 12140347]

  • Petty RE, Barden J, Wheeler SC. The elaboration likelihood model of persuasion: Health promotions that yield sustained behavioral change. In: DiClemente RJ, Crosby RA, Kegler MC, editors. Emerging theories in health promotion practice and research. San Francisco: Jossey-Bass; 2002. pp. 71–99.

  • Petty RE, Cacioppo JT. Involvement and persuasion: Tradition versus integration. Psychological Bulletin. 1990;107:367–374.

  • Petty RE, Wegener DT. Attitude change: Multiple roles for persuasion variables. In: Gilbert D, Fiske S, Lindzey G, editors. The handbook of social psychology. 4th ed. Vol. 1. New York: McGraw-Hill; 1998. pp. 323–390.

  • Prochaska JO, DiClemente CC, Norcross JC. In search of how people change: Applications to addictive behaviors. American Psychologist. 1992;47:1102–1114. [PubMed: 1329589]

  • Prochaska JO, Velicer WF. The transtheoretical model of health behavior change. American Journal of Health Promotion. 1997;12:38–48. [PubMed: 10170434]

  • Prohaska TR, Keller ML, Leventhal EA, Leventhal H. Impact of symptoms and aging attribution on emotions and coping. Health Psychology. 1987;6:495–514. [PubMed: 3691451]

  • Purdie N, McCrindle A. Self-regulation, self-efficacy, and health behavior change in older adults. Educational Gerontology. 2002;28:379–400.

  • Rosenstock IM, Strecher VJ, Becker MH. Social learning theory and the health belief model. Health Education Quarterly. 1988;15:175–183. [PubMed: 3378902]

  • Rothman AJ. Toward a theory-based analysis of behavioral maintenance. Health Psychology. 2000;19:64–69. [PubMed: 10709949]

  • Rothman AJ. “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. 2004;1:11. [PMC free article: PMC514573] [PubMed: 15279674]

  • Rothman AJ, Baldwin AS, Hertel AW. Self-regulation and behavior change: Disentangling behavioral initiation and behavioral maintenance. In: Vohs K, Baumeister R, editors. The handbook of self-regulation. New York: Guilford Press; 2004. pp. 130–148.

  • Rothman AJ, Kelly KM, Hertel A, Salovey P. Message frames and illness representations: Implications for interventions to promote and sustain healthy behavior. In: Cameron LD, Leventhal H, editors. The self-regulation of health and illness behavior. London, England: Routledge; 2002. pp. 278–296.

  • Rothman AJ, Kiviniemi M. “Treating people with health information”: An analysis and review of approaches to communicating health risk information. Journal of the National Cancer Institute Monographs. 1999;25:44–51. [PubMed: 10854457]

  • Rothman AJ, Martino SC, Bedell BT, Detweiler JB, Salovey P. The systematic influence of gain- and loss-framed messages on interest in and use of different types of health behavior. Personality and Social Psychology Bulletin. 1999;25:1355–1369.

  • Rothman AJ, Pronin E, Salovey P. The influence of prior concern on the persuasiveness of loss-framed messages about skin cancer. Presented at the annual meeting of the Society of Experimental Social Psychology; October; Sturbridge, MA. 1996.

  • Rothman AJ, Salovey P. Shaping perceptions to motivate healthy behavior: The role of message framing. Psychological Bulletin. 1997;121:3–19. [PubMed: 9000890]

  • Rothman AJ, Salovey P, Antone C, Keough K, Martin CD. The influence of message framing on intentions to perform health behaviors. Journal of Experimental Social Psychology. 1993;29:408–433.

  • Salovey P, Rothman AJ, Detweiler JB, Steward W. Emotional states and physical health. American Psychologist. 2000;55:110–121. [PubMed: 11392855]

  • Salovey P, Rothman AJ, Rodin J. Health behavior. In: Gilbert D, Fiske S, Lindzey G, editors. Handbook of social psychology. 4th ed. Vol. 2. New York: McGraw-Hill; 1998. pp. 633–683.

  • Schmid TL, Jeffery RW, Hellerstedt WL. Direct mail recruitment to home-based smoking and weight control programs: A comparison of strengths. Preventive Medicine. 1989;18:503–517. [PubMed: 2798372]

  • Schneider TR, Salovey P, Apanovitch AM, Pizarro J, McCarthy D, Zullo J, Rothman AJ. The effects of message framing and ethnic targeting on mammography use among low-income women. Health Psychology. 2001;20:256–266. [PubMed: 11515737]

  • Siminoff LA, Fettig JH. Effects of outcome framing on treatment decisions in the real world: Impact of framing on adjuvant breast cancer decisions. Medical Decision Making. 1989;9:262–271. [PubMed: 2796634]

  • Skinner CS, Campbell MK, Rimer BK, Curry S, Prochaska JO. How effective is tailored print communication? Annals of Behavioral Medicine. 1999;21:290–298. [PubMed: 10721435]

  • Skinner CS, Strecher VJ, Hospers H. Physicians' recommendations for mammography: Do tailored messages make a difference? American Journal of Public Health. 1994;84:43–49. [PMC free article: PMC1614921] [PubMed: 8279610]

  • Suls J, Rothman AJ. Evolution of the psychosocial model: Implications for the future of health psychology. Health Psychology. 2004;23:119–125. [PubMed: 15008654]

  • Taylor SE, Kemeny ME, Reed GM, Bower JE, Gruenewald TL. Psychological resources, positive illusions, and health. American Psychologist. 2000;55:99–109. [PubMed: 11392870]

  • Taylor SE, Lobel M. Social comparison activity under threat: Downward evaluation and upward contact. Psychological Review. 1989;96:569–575. [PubMed: 2678204]

  • Tversky A, Kahneman D. The framing of decisions and the rationality of choice. Science. 1981;221:453–458. [PubMed: 7455683]

  • Urban N, White E, Anderson GL, Curry S, Kristal AR. Correlates of maintenance of a low-fat diet among women in the Women's Health Trial. Preventive Medicine. 1992;21:279–291. [PubMed: 1614990]

  • U.S. Department of Health and Human Services. Healthy people 2010: National health promotion and disease prevention objectives. Washington, DC: U.S. Government Printing Office; 2001.

  • Wegener DT, Petty RE. Mood management across affective states: The hedonic contingency hypothesis. Journal of Personality and Social Psychology. 1994;66:1034–1048. [PubMed: 8046576]

  • Wegener DT, Petty RE, Smith S. Positive mood can increase or decrease message scrutiny: The hedonic contingency view of mood and message processing. Journal of Personality and Social Psychology. 1996;69:5–15. [PubMed: 7643302]

  • Weinstein ND. The precaution adoption process. Health Psychology. 1988;7:355–386. [PubMed: 3049068]

  • Weinstein ND. Accuracy of smokers' risk perceptions. Annals of Behavioral Medicine. 1998;20:1–8. [PubMed: 9989319]

  • Weinstein ND, Rothman AJ, Sutton SR. Stage theories of health behavior. Health Psychology. 1998;17:290–299. [PubMed: 9619480]

  • Williams GC, Ryan RM, Rodin GC, Grolnick WS, Deci EL. Autonomous regulation and long-term medication adherence in adult outpatients. Health Psychology. 1998;17:269–276. [PubMed: 9619477]

  • Wilson TD, Lindsey S, Schooler TY. A model of dual attitudes. Psychological Review. 2002;107:101–126. [PubMed: 10687404]

  • Worth KA, Sullivan HW, Hertel AW, Rothman AJ, Jeffery RW. Avoidance goals can be beneficial: A look at smoking cessation. Basic and Applied Social Psychology. 2005;27(2):107–116.



These models strive to specify the psychological factors that direct behavior, but provide limited information about the processes by which these factors (i.e., attitudes and beliefs) are formed or altered. Contemporary models of attitude formation and change (Petty and Wegener, 1998) delineate these processes and are considered in the next section of the paper.


Even if subsequent empirical work reveals that a stage-based approach provides a more complete account of behavioral decision making, its impact on intervention efforts may be constrained by the logistical and financial costs associated with its implementation.


Interventions targeted at structural or environmental factors (e.g., access to care, availability of facilities) can also afford significant changes in people's behavior but, aside from brief consideration in the final section of this paper, they fall outside the purview of this paper, which focuses on the psychology of change.


It should be noted that tailored messages are distinct from targeted messages. Message tailoring focuses on matching information to individual-level characteristics (e.g., why Alex resists exercising), whereas message targeting focuses on matching information to group-level characteristics (e.g., why men resist exercising).


Loss-framed information about a blood-cholesterol test was effective only when college undergraduates were informed that coronary heart disease was a problem for people less than 25 years old (see Rothman and Salovey, 1997, for a broader analysis of this finding).


To the extent that investigators have actively focused on time, it has been to determine how to increase the likelihood that people act on their intentions (Gollwitzer, 1999). Although instructing people to articulate how they will implement their behavioral intention has been shown to increase the likelihood that people will take action, these approaches do not address whether a change in behavior, once enacted, will be maintained.