An alternative view of pre-volitional processes in decision making

An alternative view of pre-volitional processes in decision making: Conceptual issues and empirical evidence

Perugini, M., & Bagozzi, R. P. (2004). 8 An alternative view of pre-volitional processes in decision making. Contemporary perspectives on the psychology of attitudes, 169.

 Sumber: 

http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.120.5428&rep=rep1&type=pdf 

An alternative view of pre-volitional processes in decision making: Conceptual issues and empirical evidence Marco Perugini Department of Psychology, University of Essex, UK and Richard P. Bagozzi Rice University, USA

Running Heading: PRE-VOLITIONAL PROCESSES

Address correspondence to: Dr Marco Perugini Department of Psychology University of Essex Wivenhoe Park Colchester, CO4 3SQ United Kingdom E-mail: mperug@essex.ac.uk

To appear in: Haddock, G., & Maio, G.R. (Eds.), Contemporary perspectives on the psychology of attitudes: The Cardiff symposium. Hove, UK: Psychology Press.

The last twenty years have seen a series of developments within attitude theory that are likely to shape the field in the next decades. Two main approaches to decision making within this field have emerged: automatic and deliberative (Eagly & Chaiken, 1998). These two approaches have been repeatedly sketched in the so-called dual theories, for instance in models of attitude change and persuasive messages (e.g., Petty & Wegener, 1999; Chen & Chaiken, 1999), whereby both routes are considered as possible, depending on variables such as motivation and ability, as well as in theories where the emphasis has been placed on one of the two approaches, such as the MODE (Motivation and Opportunity as Determinants) model proposed by Fazio (1990) and the Theory of Planned Behavior (TPB) developed by Ajzen (1991).

 In this chapter we will briefly examine the two positions. We will argue that, although recent evidence emphasizes the importance of automatic processes and momentum appears to be with this approach, the automatic approach still falls short from offering a comprehensive view and a satisfactory prediction of behavior. In particular, we will argue that, whereas it is clear that automatic processes play an important role and should be taken into account in any attempt to explain and predict human behaviors, they alone cannot suffice to fully explain behavior. We will then review the mainstream approach within the deliberative camp and identify areas for improvement. The core of this chapter will be on proposing an alternative conceptualization of attitude theory that offers a deeper understanding and improved prediction of prevolitional processes. We also review empirical evidence supporting our conceptualization. Finally, we will discuss possible extensions of the models and future directions of research.

Automatic approaches to attitudes

The automatic, effortless, and implicit aspects of human information processing are currently at the center of attention in attitude research. Several recent studies have shown that implicit attitudes can be activated automatically and guide behavior directly (e.g., Bargh, Chen, & Burrows, 1996; Chen & Bargh, 1999; Dovidio, Kawakami, Johnson, Johnson, & Howard, 1997; Fazio & Dunton, 1997). Other studies have found that attitude accessibility moderates the link between attitudes and behavior (e.g., Fazio, Sanbonmatsu, Powell, & Kardes, 1986; Fazio & Williams, 1986; Posovac, Sanbonmatsu, & Fazio, 1997). Fazio’s MODE model encapsulates this empirical evidence by proposing that attitudes that are automatically accessed, via strength of the object-evaluation association, bias perceptions of the object and lead directly to behavior without any conscious reasoning processes occurring (Fazio, 1990). Still other studies have emphasized implicit attitudes that are thought to direct people’s reactions to attitude objects outside of conscious awareness (Greenwald & Banaji, 1995). The Implicit Association Test (Greenwald, McGhee, & Schwartz, 1998) has been specifically developed to measure implicit attitudes and used in several studies (e.g., Cunningham, Preacher, & Banaji, 2001; Dasgupta, McGhee, Greenwald, & Banaji, 2000; Greenwald & Farnham, 2000; Greenwald, Banaji, Rudman, Farnham, Nosek, & Mellott, 2002). Altogether, there is growing and convincing evidence that automatic processes play an important role in human cognition and that they can direct behavior even when it is complex (e.g., Bargh, Gollwitzer, Lee-Chai, Barndollar, & Trötschel, 2001; see also Vargas’s chapter in this book).

However, the extent to which these studies imply that reasoned processes are only marginally or not at all relevant for the prediction of human behavior is a more controversial issue. It is outside the scope of the present chapter to deal with this issue. Interested readers are referred to Ajzen (2001) and Eagly & Chaiken (1998), who review the theoretical and practical importance of considering reasoned or deliberative decision making models. We will briefly focus here merely on one issue not dealt with previously: namely, the empirical evidence produced so far falls well short of showing that automatic processes explain much of the richness of human behaviors (cf. Bargh & Chartrand, 1999). If one analyzes the empirical evidence in more detail, the role of automaticity in predicting behavior appears to be more limited than claimed.

For purposes of illustration, we will consider three among the most famous studies in the field1 . The study by Bargh, Chen, & Burrows (1996) is a widely cited example of automaticity in action. The experiments show the influence of subliminally priming (1) the concept of rudeness on increasing the likelihood with which participants interrupt the experimenter (experiment 1), (2) stereotypes of elderly people on the speed of walking (experiments 2a and 2b), and (3) stereotypes of African American on producing hostile behaviors (experiment 3). Equally well known is the paper by Chartrand and Bargh (1999) studying the chameleon effect: that is, the nonconscious mimicry of the postures of the experimenter displayed by participants. This mimicry effect was investigated in three experiments with several dependent variables. Finally, the study by Djikstheruis & Van Knippenberg (1998) presents four experiments showing that priming a stereotype or a trait (professors or soccer hooligans) leads to improved or decreased performance in a Trivial Pursuit game. All the above studies have been often cited as exemplary evidence for the automaticity of behavior² .

Using the method suggested by Rosenthal and Rubin (1982), we have calculated the variance explained in behavior from the effects found in the experiments (see Table 1). The amount of variance explained by the experimental manipulations that are due to the automaticity effect ranges from .06 to .37, with a simple average value of .13 across the eighteen dependent variables. That is, the above studies finding automaticity in action predict only 13% of the variance in behavior on average.

 [Insert Table 1 about here]

In closing this section, we would like to emphasize that our goal is not to trivialize the results obtained within the automaticity paradigm. We truly believe that the paradigm and findings to date represent important contributions to the field of attitude research. The studies clearly show that people need not necessarily form an intention in order to perform a behavior. Thus, the conscious deliberative link between attitude and behavior can be bypassed.

Nevertheless, we argue that this does not imply that intentions do not play any role or that intentions are simply epiphenomenal attributions made by people to explain to themselves why they do what they do (Wegner & Wheatley, 1999). On logical grounds, the argument claiming that intentions are epiphenomena is self-contradictory (cf. Yanchar, 2000). On pragmatic grounds, we believe it would be wiser and scientifically more productive to acknowledge that reasoned and deliberative approaches cannot be dismissed out of hand, if one wants truly to explain and predict a realistic range of ecologically relevant behaviors. As we argue below, the empirical evidence clearly supports the importance of constructs such as intention and volition in predicting human behavior, over and above the contributions provided by automatic attitudinal effects.

Deliberative approaches to attitudes

Expectancy-value approaches have been dominant within the deliberative tradition, and the leading contemporary model in this tradition is Ajzen’s Theory of Planned Behavior (TPB, Ajzen, 1991, 2001). Under the TPB, the proximal cause of behavior is an intention to perform that given behavior. Moreover, attitude is only one of the determinants of intention. Subjective norms (i.e., the perceived social pressure to perform a given behavior) and perceived behavioral control (i.e., the perceived ease or difficulty of performing the given behavior) are also hypothesized to influence one’s intention. Both constructs are formed in parallel summative manners: normative beliefs times motivations to comply for subjective norms and control beliefs times perceived power for perceived behavioral control. Perceived behavioral control is also posited to have a direct influence on behavior, to the extent that it corresponds to actual behavioral control (Ajzen & Madden, 1986).

From an empirical point of view, the performance of the TPB has been assessed in a number of recent meta-analytic studies. Godin and Kok (1996) reviewed 76 applications of the TPB in the domain of health-related behaviors, and found an average of 41% of variance explained in intentions and 34% in behavior. Similar findings were obtained by Armitage and Conner (2001) who examined 185 empirical tests of the TPB and found that the TPB accounted on average for 39% of the variance in intention and 27% of the variance in behavior. A recent meta-analysis of meta-analyses by Sheeran (forthcoming) found that over a total sample of N=82,107, whereby the TPB as well as other theoretical models that include intention as predictors of behavior were scrutinized, the average sample-weighted correlation between intention and behavior was .53, corresponding to 28% of variance in behavior explained for by the intention to behave. This latter value should be compared with the average value of 13% across the experiments by Bargh and colleagues (Table 1). Despite the differences across studies which limit direct comparisons, it is obvious that explanations in terms of automaticity of behavior that exclude any role played by intentions do not offer pragmatically superior alternatives. These data support the notion that, at least on pragmatic grounds, the construct of intention cannot be dismissed as irrelevant. On the contrary, given that the empirical evidence supports the important role played by intentions and volitions in explaining and predicting behavior, the question of how people come to form intentions or volitions remains a key issue.

An alternative view of pre-volitional processes in decision making

The TPB provides a robust explanation of the pre-volitional processes in decision making and predicts a remarkable amount of variance in intention, as testified by the average value of 39% in Armitage & Conner’s meta-analysis (2001). Yet, the findings suggest that there is still room for improvement in both explanatory and predictive senses. Several proposals to improve the TPB have been offered, often with a mix of new theoretical and empirical contributions. The weak point of the TPB, it has been often argued, is its sufficiency. Researchers have proposed additions to the TPB addressing self-identity processes (Sparks & Shephard, 1992), moral norms (Beck & Ajzen, 1991; Parker, Manstead, & Stradling, 1995), the distinction between perceptions of control and self-efficacy (Armitage & Conner, 1999; Terry & O’Leary, 1995), anticipated emotions (Parker et al., 1995; Richard, van der Pligt, & de Vries, 1995), and habit or past behavior (Bagozzi, 1981; Ouellette & Wood, 1998). The focus of these proposals is on the addition of independent variables as parallel predictors of behavioral intentions and behaviors, along with the established predictors. The rationale is that more variance can be accounted for by specifying processes formally contained in error terms in tests of the theory and that the resulting theoretical understanding of the decision making process can be broadened. Four main areas of improvement over the TPB can be identified: a) automatic processes, b) affective processes, c) motivational processes, and d) means-end analyses. Below, we review these areas and present two recently developed models that offer alternative views of pre-volitional decision making processes.

Automatic processes

 “Consuetudine quasi altera quandam naturam”

[Habit is practically our second nature] Cicero, 45 BC, De finibus bonorum et malorum, Lib. 5, Cap. 25, Par. 74 The role of automatic processes in decision making can hardly be exaggerated. As we have argued above, there is impressive experimental evidence in this regard. Additional evidence can be found in previous studies whereby past behavior and habit was shown to explain additional variance in behavior, over and above that explained by intentions (e.g., Bagozzi, 1981; Bentler & Speckart, 1979, 1981). Ajzen (1991) criticised the use of past behavior on grounds that it offers no explanatory content, although he allowed that past behavior provides a methodological control in tests of any theory: “…past behavior can be used to test the sufficiency of any model designed to predict future behavior” (Ajzen, 1991, p. 202). Further, Ajzen (1991) argued that the inclusion of perceived behavioral control (PBC) in the TPB should preclude the need for past behavior, in that PBC should mediate any residual effects of past behavior. This was one of the main reasons why the TPB superseded its predecessor Theory of Reasoned Action (Fishbein & Ajzen, 1975). However, recent empirical work finds that past behavior still predicts intentions and behavior in tests of the TPB (e.g., Bagozzi & Kimmel, 1995; Beck & Ajzen, 1991; Leone, Perugini, & Ercolani, 1999).

  

The question remains what the effects of past behavior might represent. A meta-analysis recently found robust evidence for the impact of frequency of past behavior on both intentions and behavior (Ouellette & Wood, 1998). Ouellette and Wood (1998) proposed two processes through which frequency of past behavior guides future behavior. When a behavior is well-learned and practiced in a non-changing environment, frequency of past behavior reflects habit strengths and therefore shows a direct effect on future behavior. When a behavior is novel or performed in nonstable contexts, frequency of past behavior influences intentions on the grounds that people like to do things that they have done in the past.

Another proposal has been to partition the effects of past behavior into frequency and recency effects (Bagozzi & Warshaw, 1990). Although seemingly related, frequency and recency effects are conceptually distinct and therefore might carry independent information. For instance, one may have a long history of performing a given behavior without having performed it recently or one may have recently taken up an activity with no prior experience with it. Recency of behavior performance should influence future behavior to the degree that availability and anchoring/adjustment biases occur in information processing (e.g., Tversky & Kahneman, 1974) and to the degree that an activity, whether established or not, has been recently initiated. The recent initiation of an activity may carry implicit information about intentions over and above the degree to which intentions are accessible to conscious awareness. Recency may therefore serve as an indirect indicator that an intention has been activated and therefore be positively associated with subsequent performance of the behavior.

PRE-VOLITIONAL PROCESSES 9

The concept of habit as the key explanatory variable underlying the role of past behavior has also been proposed by other researchers (for a review, see Verplanken & Aarts, 1999). For instance, Aarts and Dijksterhuis (2000) define habit as a form of goal-directed automatic behavior, which is activated automatically by the presence of relevant environmental cues (cf. the automotive model, Bargh, 1990), provided that the relevant goal is activated. Measures relying on habit strength and on aspects of automaticity have been proposed by Verplanken and colleagues. A different view has been recently put forward by Ajzen (in press). Ajzen maintains that the empirical evidence produced so far does not offer sound theoretical explanations of the role of past behavior in terms of habitualized responses. He argues that other explanations, such as the instability of intentions or the presence of unrealistic optimism and inadequate planning, should instead be considered before claiming that past behavior (habit) provides a theoretical basis for predictions.

To sum up, the evidence of the role played by past behavior in predicting future behavior is very strong. However, there is a range of opinions concerning its role in explaining future behavior. Either way, we believe that any comprehensive decision making model in this field should include past behavior, if for no other reason than to improve the prediction of intentions and behavior and/or control for unmeasured determinants.

Affective processes

 “...I shall endeavour to prove first, that reason alone can never be a motive to any action of the will; and secondly, that it can never oppose passion in the direction of the will” Hume, 1739, A Treatise of Human Nature, Section III, p.413

The TPB has traditionally emphasized the cognitive-evaluative component of attitude and neglected the affective component. Attitude is conceived as “a psychological tendency that is expressed by evaluating a particular entity with some degree of favour or disfavor” (Eagly & Chaiken, 1993, p. 1, emphasis in original removed). Under the TPB, attitude is usually formulated to refer to a target behavior and its overall measure consists of bi-polar semantic differential items such as good-bad, harmful-beneficial, rewarding-punishing, and unpleasantpleasant (Ajzen, 1991). Therefore, depending on the specific pair of adjectives used in a study, an overall measure of attitude in practice can be characterized mostly by cognitive aspects, by affective aspects or, more frequently, by a variable mix of the two. In this sense, the original conception of the TPB did focus mostly on the evaluative aspect without explicitly dealing with the specific contribution of the affective and cognitive components. Some authors have instead argued for an explicit distinction between the evaluative and affective components of attitude and collected empirical support for this distinction (e.g., Breckler & Wiggins, 1989; Trafimow & Sheeran, 1998; Verplanken, Hofstee, & Janssen, 1998). The affective component has been defined as an affective disposition to respond in a favorable or unfavorable manner. A person is expected to acquire over a period of time or through repeated contact accompanied by reinforcement an affective orientation toward an object or act that, once learned, can be triggered automatically when one is exposed to the object or act or thinks about it (Fazio, 1995).

An alternative approach has been proposed by Bagozzi, Baumgartner, and Pieters (1998), who introduced the concept of anticipated emotions as predictors of intentions to act. An attitude, either affective or cognitive, is typically constant over reasonable periods of time and it is not formulated as a response contingent on the occurrence of particular happenings to be appraised. By contrast, the processes behind the functioning of anticipated emotions are more dynamic and entail self-regulation in response to feedback (Bagozzi, 1992; Carver & Scheier, 1998). They are specifically contingent on one’s appraisal of achievement or failure, which changes from time to time, depending on the context. An implicit comparison is made between one’s goal as a standard or reference value and achieving and failing to achieve that goal, with anticipated emotions as consequences.

This functioning of anticipated emotions is related to counterfactual thinking. Counterfactuals are conditional statements that contain both an antecedent expressed as a hypothetical (e.g., “If X happens”) and a consequent expressed as an implication (“then Y will occur”). Quite a bit of research has been conducted in recent years with respect to the cognitive processes underlying “backward looking” counterfactuals, wherein people consider what might have been, if certain things had happened differently than they did (e.g., Roese & Olson, 1995). Less frequently studied has been the function of anticipated counterfactuals, where people consider future prospects and evaluate the consequences of alternative, contingent outcomes. Gleicher et al. (1995, p. 294) call such anticipated counterfactuals, “prefactuals”.

With regard to negative outcomes, two mechanisms were hypothesised by Gleicher et al. (1995) for serving as the basis by which counterfactuals operate to influence intentions and behavior:

“First, when a person generates a counterfactual that reverses a negative outcome, he or she is likely to make the attribution that there is an effective action that can be taken in the future… (Second) when an individual thinks about a counterfactual in advance, the motivation to avoid this negative affect influences behavioural choices.” (pp. 294-295).

Parker, Manstead, and Stradling (1995) were among the first to find that anticipated regret (measured with “make me feel sorry” and “make me feel good” items) affected expectations that one would commit certain automotive driving violations. Likewise, Richard et al. (1995) found that negative anticipated emotions (called anticipated postbehavioral affective reactions and measured with “worried-not worried”, “regret-no regret”, and “tense-relaxed” items) increased expectations that people would refrain from sexual intercourse or would use condoms in casual encounters. In these, as well as in other works, the emphasis has been placed on a few negative anticipated emotions, especially on regret.

However, the range of negative anticipated emotions potentially affecting one’s intention to perform a given behavior is much greater (e.g., shame, disappointment, anger, guilt, sadness).

Moreover, parallel processes can be hypothesised for positive outcomes. People should be motivated to make choices promoting positive affects such as gladness, satisfaction, happiness, and pride. Whether positive or negative anticipated emotions play a more important role in decision making processes may depend on the kind of behavior (for instance, whether it is a preventive or a detection behavior, Rothman, & Salovey, 1997), or on the kind of person (for instance, whether one is promotion or prevention focused, Higgins, 1996). Within each cluster of positive and negative emotions, even more specific predictions concerning which emotion is more likely to play an important role can be made by relying on emotional appraisal theories such as the one proposed by Roseman, Antoniou, and Jose (1996).

A question that can be raised is whether anticipated emotions overlap with the affective component of attitude and its measures. Although both conceptualizations attempt to tap affective processes in decision making, they do so in fundamentally different ways.

Attitude focuses upon “…the evaluation of an object, concept, or behavior, along a dimension of favor or disfavor, good or bad, like or dislike” (Ajzen & Fishbein, 2000, p. 3). It has also been argued that attitude should be sharply distinguished from affect, given that the operational definition of attitude assesses an evaluative dimension, which may include affective evaluations, rather than a properly affective or emotional component (Ajzen & Fishbein, 2000, p.3). By contrast, anticipated emotions are dynamic, self-regulatory, and based on contingent appraisals of the consequences of alternative outcomes. They refer to anticipated affective consequences of failure or success rather than on static evaluations. Anticipated emotions are hypothesized also to affect the intention to perform a given behavior (through their effect on desires, as developed below). Finally, another difference occurs at the level of measurement. When people are asked to respond with their attitudes, they are forced to make a choice of favourability or unfavourability. This is a consequence of the recommended practice of using bipolar items to indicate respondents’ attitudes. By contrast, anticipated emotions are better

measured by using uni-polar items (e.g., the experience of disappointment along a “not at all” to “very much” continuum). In a study of affect, Bagozzi, Wong, and Yi (1999) recently found that positive and negative affect can be positively related, negatively related, or unrelated to each other, depending on conditions such as the independence vs. interdependence of the culture where people had their upbringing and live. This finding is in line with related evidence for basic differences between positive and negative emotional reactions, in terms of static structural representations of affect (e.g., Russell & Barrett, 1999), basic functioning of physiological systems (e.g., Gray, 1990), behavioral strategies (e.g., Higgins, 1996), anticipation of future consumption (e.g., Loewenstein, 1987), decision making (e.g., van der Pligt, Zeelenberg, van Dijk, de Vries, & Richard, 1998) and effects on behavior (e.g., Cacioppo & Berntson, 1994). To use a bi-polar scale to measure affective features (as it is routinely done in attitude theory) would make positive and negative affect mutually exclusive by definition and not permit respondents the opportunity to express their differential relevance.

To sum up, the TPB does not pay sufficient attention to the role played by affective processes in decision making. One possibility is to construe affect as another component of attitudes. An alternative possibility, that we believe to be more fruitful, is to examine affective processes as instances of pre-factual thinking where people are asked to anticipate what their emotional responses would be to the occurrence and non-occurrence of a future event. Analogous to the notion of prospect-based emotions (e.g., Ortony, Clore, & Collins 1988), anticipated emotions can be hypothesized to be integrated with attitudes, subjective norms, and perceived behavioral control as a basis for intention formation.

Motivational processes

 “I believe that I will die someday, and make a mistake today, but I intend to do neither. What is lacking is desire, the motivational element in intention” Davis, 1984, A Causal Theory of Intending, American Philosophical Quarterly, 21, p. 45.

Another weak area of the TPB concerns the motivational processes in decision making. It has been argued that the TPB fails to consider how intentions become energized (Bagozzi, 1992; Fazio, 1995, pp. 271-272). Attitudes, subjective norms, and perceived behavioral control provide reasons for acting, but do not incorporate explicit motivational content needed to induce an intention to act. Bagozzi (1992, pp. 184-6) early on proposed that desires provide the motivational impetus for intentions and suggested that attitudes work through desires to influence intentions. Expanding on Bagozzi’s (1992) arguments, Perugini and Bagozzi (2001) proposed that desire is a stronger predictor of intentions than attitudes, subjective norms, and perceived behavioral control, and it plays a key mediational role transforming reasons to perform a given behavior into a motivation to do so. From this perspective, the usual TPB predictors do not directly determine behavioral intentions, but rather indirectly through desires.

The concept of desires is rarely mentioned in contemporary attitude theory presentations. Mainstream attitude models (e.g., Ajzen, 1991; Eagly & Chaiken, 1993; Fazio, 1990) do not consider desires at all or else seemingly lump them together with intentions. On the other hand, philosophers make a clear distinction between desires and intentions (Audi, 1986; Bratman, 1987; Davidson, 1963; Davis, 1984; Mele & Moser, 1994; Schueler, 1995). Although theoretical disagreements exist concerning the specific way in which desires and intentions are related, the distinction is one of the central tenets of philosophical treatments of action, dating back at least to Aristotle. The dominant model in this stream of work within the so-called philosophy of action is some variant of a three-facet model of decision making, based on a distinction between beliefs and desires and their influence on corresponding intentions. For philosophers, desires are thought to have a fundamental role in decision making. Likewise people in everyday life distinguish between desiring to act or obtain something and intending to do or to achieve it, respectively, placing much emphasis on the concept of desire in one’s thoughts, interpersonal relations, and actions. That desires and intentions are distinct can be witnessed in familiar or conventional problem solving: John feels a desire to take a vacation but for various reasons intends not to do so; Janet lacks a desire to attend a lecture yet plans to attend it nevertheless.

Recent empirical evidence supports this important distinction. Theoretically, the distinction between the two constructs has been developed by Malle and Knobe (2001) and by Perugini and Bagozzi (in press). Malle and Knobe (2001) approach this issue from a social perceiver perspective and argue that social perceivers are well aware of the distinction embedded in these two mental states. For instance, people are relying more on someone else’ intentions than desires when trying to make accurate predictions, because they are aware that desires can be highly conflicting whereas intentions are less so. Moreover, they are aware that an intention is a further step in the decision making process and it does imply some form of even partial commitment to action, whereas a desire can be more fleeting and it might not involve any decision to act upon it.

Perugini and Bagozzi (in press) approach the issue from a decision maker’s perspective and proposed three criteria to distinguish between desires and intentions: perceived performability (desires being typically perceived as less performable than intentions), action-connectedness (desires are less connected to actions), and temporal frame (desires are relatively more futureoriented than intentions). In two studies, they have found evidence that people perceive their own desires as less connected to specific actions, as more framed in longer temporal horizons, and as perceived as less obtainable than their own intentions. In other words, people do appreciate that, in their own decision making flow, a desire comes before an intention, it does imply a more limited commitment to action, and it does not always lead to an intention to act. This appeared to be the case even when focusing on action rather than goal content, which is a special situation whereby the differences between desires and intentions should be minimized (cf. Malle & Knobe, 2001).

Empirically, there is recent correlational and experimental evidence supporting the distinction between desires and intentions. In their meta-analysis of TPB studies, Armitage and Conner (2001) showed that attitudes, subjective norms and perceived behavioral control predicted significantly more variance in desires than in intentions, and that intention was a better predictor of behavior than desires. We will review additional correlational evidence below. Malle and Knobe (2001) reported the results of a series of scenario-based experiments finding clear empirical evidence for the distinction. Finally, Perugini and Bagozzi (in press) provided empirical support for this distinction in two experiments.

Interestingly, whereas desire is rarely mentioned3 in the literature, the concept of desirability is often used in social psychology to explain a wide range of phenomena. Particularly interesting is the temporal construal theory proposed by Liberman and Trope (1998) which focuses on the mental construction of events at different levels of abstraction and under different time frames. Liberman and Trope (1998) propose that feasibility and desirability are key elements in choosing between temporally near and distant alternative courses of action. Desirability, reflecting the superordinate “why” aspects of an action (cf. Vallacher & Wegner, 1987), was defined as the valence of an action’s end state and hypothesized to be a main reason to act, especially for distant future scenarios; feasibility, defined as the ease or difficulty of reaching the end state, was argued to be more relevant for near-future actions (Liberman & Trope, 1998, p. 7).

 The emphasis of research on desirability has been more on the value of a particular goal outcome (e.g., the valence of an action’s end state) than on the role of the agent and on the processes of valuation and their impact on action initiation, per se. However, whether an agent will act in relation to a “desirable” end-state depends on the agent’s desire to achieve that end state. A definition of desire resting on the valence attributed to an object or action would seem to be empty without reference to the motivational state of mind of the agent. In other words, a key causal element in action is the personal motivation (desire) to achieve an end-state and not an inherent objective property of the end state itself (i.e., its desirability) 4 . Otherwise, we might end up believing that it was the nice car that caused me to buy it, and not my desire to have it.

PRE-VOLITIONAL PROCESSES 17

Furthermore, whether one desires to perform a certain action does not necessarily imply that he/she will intend to do it. We often have desires, fleeting or otherwise, that we never intend to act upon. However, an intention to perform an action generally follows a volitional desire (i.e., behavioral desire or desire to act) which represents the motivational state of mind wherein appraisals and reasons to act are transformed into a personal motivation to do so.

The Model of Goal-directed Behavior

 An alternative account of the decision making processes whereby these three areas (automatic, affective, motivational) are integrated is given by the Model of Goal-directed Behavior (MGB, Perugini & Bagozzi, 2001). The main focus of the MGB is on the pre-volitional stages of decision making which involve a number of new components. The MGB begins with the basic constructs of the TPB but focuses additionally on the three relatively overlooked areas. The MGB is depicted in Figure 1. Note that for the sake of simplicity the figure does not include behavior as an outcome, given that the MGB focuses mostly on pre-volitional processes.

[Insert Figure 1 about here]

The MGB does not simply add other variables to the TPB but rather redefines the decision making process. A key construct introduced in the MGB is desire. As argued above, desires are distinct from intentions and represent the motivational state of mind in which appraisals and reasons to act are reconciled. Desires perform energizing and transformative functions for the antecedents of decision making and represent the most proximal determinants of intentions. Unlike Ajzen (1991, p. 181) who maintains that “…intentions are assumed to capture the motivational factors that influence a behavior”, we argue that intentions in and of themselves do not have specific motivational content. Rather, intentions perform planning (Bratman, 1987) and directive functions. The motivational content in decision making is constituted by the desire to perform a certain behavior. It is this desire which energizes intentions.

Among the determinants of desires are positive and negative anticipated emotions which refer to anticipated affective reactions to failure and success. Unlike recent proposals where some affect-related variables have been added to improve the prediction of intention and behavior (e.g., Conner & Abraham, 2002; Parker, Manstead & Stradling, 1995; Richard, Van der Pligt & De Vries, 1995), the MGB stipulates that anticipated emotions represent a form of counterfactual (prefactual) thought processes where the emotional consequences of achievement and failure are appraised. Moreover, a more comprehensive set of both positive and negative emotions, assessed separately for perspectives of success and of failure, is introduced. Based on a previous study by Bagozzi, Baumgartner, and Pieters (1998), seventeen emotions (seven positive and ten negative) were originally proposed by Perugini and Bagozzi (2001). However, the number and the content of the anticipated emotions could be different depending on the specific behavior under investigation.

Finally, past behavior is introduced as a proxy for habit and also given a substantive interpretation. As argued above, past behavior has been repeatedly found to have independent effects on intentions (see Conner & Armitage, 1998 for a review). In contrast to Ajzen’s conclusion that “although past behavior may well reflect the impact of factors that influence later behavior, it can usually not be considered as a causal factor in its own right” (Ajzen, 1991, p. 203; see also Ajzen, in press), recent thinking provides a theoretical rationale for the causal influence of past behavior or habit on current behavior (e.g., Aarts & Dijksterhuis, 2000; Albarracin & Wyer, 2000; Bargh, 1990; Oulette & Wood, 1998; Verplanken & Aarts, 1999). A similar rationale can be extended to explain the influence of past behavior on the motivational and volitional stages in decision making, as proposed under the MGB. Perugini and Bagozzi (2001) distinguished between frequency and recency as two related but separate mechanisms through which the influence of past behavior can be conceptualized. Other possible explanations are selfidentity (Oullette & Wood, 1998) and habit (Verplanken, Aarts, & van Knippenberg, 1997).

PRE-VOLITIONAL PROCESSES 19

Whatever the specific mechanisms behind past behavior, the general point is that automatic processes can have an unmediated impact on pre-volitional states, especially on desire and intention, and should be included in any comprehensive modeling of decision making.

The MGB proposes that volitive desires mediate the effects of deliberative processes on intentions. However, there are two ways that reasons for acting can have direct effects on intentions, one methodological, one substantive. Methodologically, to the extent that the measures of desires fail to capture its full nature or scope, empirical tests of the MGB may find direct paths from the reasons for acting to intentions. Substantively, to the extent that the reasons for acting reflect non-deliberative processes, we might expect direct effects from the reasons to intentions. The latter direct effects would reflect automatic activation of intentions. In sum, although we expect that desires generally mediate the effects of the reasons for acting on intentions, it is possible to find partial mediation or direct effects on intentions depending on the presence of random and systematic error in measures of desires and of automatic processes.

To sum up, the MGB redefines the theoretical mechanisms of the decision making process and incorporates three important, new theoretical areas (affective, motivational, and automatic processes) that are overlooked by the TPB. While the MGB originally conceived as an integration of goal and behavioral criteria, as implied by the focus of anticipated affective reactions on goals instead of behavior (cf. Perugini & Bagozzi, 2001), we believe that it is more useful to conceive it as a model at a behavioral level. The focus of anticipated emotional reactions can be defined at the level of behavioral outcomes or goal outcomes, but to fully explain goal-directed behavior the MGB must be extended, as developed below.

Means-end analysis

“The origin of action—its efficient, not its final cause—is choice, and that of choice is desire and reasoning with a view to an end.” Aristotle, c. 340 BC, Nicomachean Ethics, 1139a 31-2

PRE-VOLITIONAL PROCESSES 20

Goals play a central role in the explanation of many behaviors because these behaviors are chosen as means to goal achievement (e.g., Gollwitzer & Moskowitz, 1996). Yet, in attitude theory only rarely have behaviors been studied in relation to the goals for which they are performed. The focus has been restricted to specific behaviors in isolation from the broader context which justify their performance (Eagly & Chaiken, 1998).

Vallacher and Wegner (1987) have suggested that actions may be represented at different hierarchical levels, in terms of superordinate or subordinate goals, and have found that people tend to define their actions mostly in terms of higher levels of abstraction (the why aspects of an action). Gollwitzer (1990) proposed a model of action phases, whereby the first stage (predecisional stage) consists in the contemplation of different goals and the setting of preferences among them (goal setting). The culmination of this stage is the formation of a goal intention, which is initiated and executed in the following two stages (pre-actional and actional, goal striving; see also Bagozzi & Edwards, 1998). Other research (e.g., Deci & Ryan, 1991; Locke & Latham, 1990) also shows that goals play a central role in understanding behavioral intentions. Despite these findings, work on goals has had limited impact on mainstream attitude research. The implicit assumption is that goals represent distal determinants of behavior whose influence is fully mediated by more proximal determinants of behaviors (cf. Eagly & Chaiken, 1993). Therefore, the analysis of goal related constructs is assumed to not be useful to improve the prediction of behavioral intentions, once the proximal determinants are taken into account. However, no relevant empirical data support this mediational hypothesis. On the contrary, to the extent that a goal can be realized in different ways through different behaviors, it is not obvious that the cognitive determinants of a goal-directed behavior would be sufficient to predict the volition to perform it. The distinction between goals and behaviors, their simultaneous analysis, and the specification of how goals translate into behavioral volitions may add to the prediction of these volitions over and above their most proximal determinants.

 In choosing a goal, two variables have been identified as central, goal desirability and goal feasibility (cf. Gollwitzer & Moskowitz, 1996). Actually, goal desirability and goal feasibility, interpreted as desire and belief, are classical concepts in the philosophy of action (Mele, 1997), dating back at least to Hume, and rediscovered periodically every since. As we argued above, goal desirability emphasizes the personal value which is attached to a certain goal outcome, while at the same time neglecting the personal motivation to achieve that goal outcome, which is central in philosophical analysis (e.g., Davis, 1984; Mele, 1995). Moreover, research within the goal tradition emphasizes the direct link between goal intentions and behavior. In other words, the standard assumption is that a goal is a direct organizer of actions.

To sum up, while research within the attitude tradition has usually focused on the behavioral level alone, research within the goal tradition has focused on the goal level alone. Rarely has an explicit interplay between goal and behavioral levels been articulated. Yet, this seems to be a common situation: when people behave, they often do so because they want to achieve a goal and they have preferences and evaluations both for the behavior and the goal. Expressions such as “I want to get a good degree [goal] but I do not fancy studying hard [behavior]” make perfect sense in everyday conversation and convey different evaluations and motivations at the goal and behavioral levels. Whether a student will study hard depends on the balance between how much he/she wants to achieve a good degree and how much he/she dislikes studying hard. By focusing on one level alone, we would miss important aspects of the decision making process.

The Extended Model of Goal-directed Behavior

Building on the MGB, Perugini and Conner (2000) recently proposed the Extended Model of Goal-directed Behavior (EMGB), depicted in Figure 2. As in the MGB, the focus of the EMGB is on the pre-volitional stages of decision making. Under the EMGB, however, the concept of desires is further developed, and a distinction is drawn between goal (or outcome) and behavioral (or action) desires. This distinction reflects a refinement in the conceptualization of desires and their differences from intentions. Building on distinctions between intrinsic and extrinsic desires made in philosophy (Mele, 1995), Perugini and Conner (2000) proposed that both goal desires and behavioral desires play a role in decision making concerning goal-directed actions. This follows from the assumption that most relevant behaviors can be better understood in the light of the interplay between the goal and behavioral levels. It is maintained that the influence of a desire to achieve a certain goal will influence the desire to perform a certain behavior that is subjectively felt to be instrumental for goal attainment. As a result, the behavioral desire will be the most proximal determinant of the intention to perform the behavior in question, and goal desire will have an indirect effect on intentions through behavioral desire.

[Insert Figure 2 about here]

This distinction between goal and behavioral desires is consistent with the previously described criteria differentiating desires from intentions. In particular, the distinction between goal and behavioral desires accommodates the notion that an intention has an action-content, whereas a desire can have goal- or action- content. Furthermore, the notions that goal desires are different from behavioral desires and have different roles in decision making allows us to understand motivation and its manifestation at the goal and behavioral levels of analysis. In this way, goal desires represent the wishes of the decision maker as they guide further planning and elaboration concerning how to fulfill desires. The “how” (i.e., behavioral) level in turn requires behavioral desires for initiation, which resemble the concept of extrinsic desire, that is, “...a desire for something for its believed conduciveness to something else that one desires” (Mele, 1995, p. 391).

PRE-VOLITIONAL PROCESSES 23

Why two models?

The MGB and the EMGB are two models that can increase theoretical understanding and, as we will show after, the empirical predictions of pre-volitional processes. A question might be posed on the need for both models. We believe that the best way to conceptualize them is to consider the EMGB as providing an addition ladder over the MGB, which is therefore the basic model. In other words, one way to compare and interpret the models is in terms of the level of analysis. To the extent that one wants to emphasize both the behavioral and goal aspects of decision making, the EMGB is the most comprehensive model, because provides an additional ladder. In this context, anticipated emotions are re-defined and anchored in terms of goals, as proposed by Bagozzi, Baumgartner, and Pieters (1998) and initially incorporated in the MGB (Perugini & Bagozzi, 2001). Quite a bit of research has focused on the importance of goals in influencing action (for a review, see Gollwitzer and Moskiwitz, 1996). We believe that attitudebased models can and will be enriched by focusing on the more distal determinants of action, such as the desire to achieve a given goal by means of some more specific behaviors. However, this would imply a more detailed and richer analysis of both the goal and the behaviors that can be functional to achieve it.

We acknowledge that practical considerations might prevent a researcher from including both levels of analysis in a specific study. For instance, there are some applied contexts in which the key question is to predict and understand the behavior itself, rather than its distal determinants. Moreover, the links between goals and behaviors are multi-determined: a given goal can be achieved in multiple ways (i.e., with different behaviors) and a given behavior can be functional to achieve different goals, both within and between persons. Therefore, the simultaneous examination of both the goal and the behavioral levels requires a simplification of both levels, which occasionally might not be the best strategy given the specific research question. Indeed, there are circumstances in which the theoretical interest of the researcher will focus solely on a specific behavior that happens to be functional to achieve a number of goals at the same time. In these circumstances, it might become too complicated to enlarge the analysis to both levels simultaneously. Finally, there also specific cases whereby the distinction between goal and behavior is difficult to draw, and therefore the analysis of their interplay would not add much. In all these cases, the MGB can be the best alternative.

 In short, there are pros and cons for using either model. As a general rule, when the focus and the interest is on both goals and behaviors, the EMGB is the best choice, whereas when the focus is, or must, be restricted to behavior only, the MGB is the best choice.

Empirical evidence

Distinction between desires and intentions.

Besides the previously mentioned experimental evidence, correlational evidence also supports a distinction between desires and intentions for actions or goals. We located 13 studies where both desires and intentions were measured (Bagozzi & Edwards, 1998; Bagozzi & Kimmel, 1995; Capozza, Cesco, Dazzi, Voci, & Martines, 1999; Leone, Perugini, & Ercolani, 1999, 2002; Perugini, 2002; Perugini & Bagozzi, 2002; Perugini & Conner, 2000). The domains of application were varied, including bodyweight regulation, studying, helping charities, and dieting. Multiple measures were available for each construct (see Perugini & Conner, 2000, for typical examples of items). The relation between desires and intentions was investigated using Confirmatory Factor Analysis (CFA). The models tested formally the discriminant validity of the constructs and provided unbiased (corrected for measurement error) estimates of their correlation. The average correlation between the constructs of desires and intentions over the 13 studies was 0.75 (raw correlation= 0.68), with values ranging from 0.52 to 0.94. Formal tests of discriminant validity showed that the hypothesis that desires and intentions are different constructs cannot be rejected5 . Furthermore, it should be emphasized that these correlations are likely to be inflated due to common method biases and that they are upward corrected for measurement error. In sum, the estimated correlations at the latent variable level suggest that discriminant validity between observed measures is achieved, even after correcting for measurement error and using a common method.

Comparison between MGB, EMGB, and TPB. In addition to theoretical arguments supporting the MGB and the EMGB, it is also possible to provide empirical evidence to support the predictive power of the models and to compare them with the TPB. This empirical evidence has started to accumulate in a series of studies (Perugini & Bagozzi, 2001; Perugini & Conner, 2000; Leone, Perugini, & Ercolani, 2002; Taylor, Bagozzi, & Gaither, 2001). In Table 2, we summarize a set of studies where all three models were tested. The set include such varied goals and behaviors as studying efforts, weight control, learning SPSS, helping charities, and keeping fit. Additional studies are ongoing and include socializing, dental flossing, drinking, eating with friends, and nutrition choices.

[Insert Table 2 about here]

 In the first set of seven studies, the TPB and the MGB were tested (five in Italy, two in the United States), whereas in the remaining four studies, performed in the United Kingdom, the EMGB was also tested. Considering first the five Italian and two U.S. studies, the average prediction for the TPB was 32% of the variance for intentions and 20% for behavior. These values are comparable to results reported in previous meta-analytic reviews. The MGB produced a dramatic improvement in predictive power over the TPB, with averages of 58% of variance for desires, 58% for intentions, and 26% for behavior explained. Note that the improvement is especially strong for pre-volitional and volitional constructs, due to the main focus of the MGB on these aspects. On average, an additional 26% of the variance in intentions was gained by introducing the concept of desires. One might also wish to compare the prediction of intentions under the TPB with that of desires under the MGB, given that they are functionally equivalent under the two theories: in both cases they are meant to represent the outcome of a decision maker’s weighting of various reasons for acting. In this case, the MGB explains 26% more variance in desires than the TPB does in intentions. The EMGB improves further the predictive power of the MGB.

 PRE-VOLITIONAL PROCESSES 26

Focusing on the average values of the second set of four studies, where all three models were tested, the TPB explains 32%, the MGB 78%, and the EMGB 79% of the variance in intentions. Furthermore, the MGB explains 63% and the EMGB 71% of the variance in behavioral desires. Again the differences in predictive power are substantial, and both the MGB and the EMGB clearly outperform the TPB. The EMGB shows a moderate improvement in the prediction of behavioral desires, with an increase of 8% in explained variance. As argued elsewhere (Perugini & Conner, 2001), it is fair to note that this is not the ideal arena to test the TPB, especially because items in the questionnaires are explicitly worded as functional to goals (e.g., “I intend to do X in order to achieve Y” vs. “I intend to do X”). Nevertheless, the differences in predictive power are so great as to suggest that the MGB and EMGB capture essential determinants of decisions neglected by the TPB. It is important to note, too, that these differences are equally great if we base our comparisons on the results of recent meta-analytic reviews which included “traditional” TPB studies.

A closer look at the estimates for structural paths specified by the MGB (Table 3) shows interesting results. The average values for estimates, which are based on a total sample of 1155 participants, suggest that all paths are likely to play a significant role. However, it can be seen that which specific path is significant varies considerably depending on the study. For instance, attitudes show a significant path on desires in seven of the nine studies, whereas subjective norms reveal it in five studies. Never is it the case that all predictors are significant in the same study. However, which specific construct plays a more important role varies among studies, and this therefore suggests the importance of considering them all. In other words, the antecedents function in a compensatory way and do so differentially, depending on the context.

Similar to expectations for the predictors of intentions in the TPB, there is no claim that all predictors of desires under the MGB should always be significant. But of course the implicit expectation is that a considerable portion of variance should be explained by the set of predictors, leaving it an empirical matter which predictor should be significant in any particular study. Note that, in 8 of the 10 studies, either positive or negative anticipated emotions have significant influences on desires, and in one study both are significant. Past behavior also plays an important role in predicting both desires and intentions. This result is in line with recent emphasis on automatic processes in decision making and behavior. The influence of desires on intentions is always strong, but never such that one could claim that they are indistinguishable constructs, even after correcting for measurement error.

[Insert Table 3 about here]

Likewise, the values of the structural paths specified by the EMGB (Table 4) show that all variables play an important role. The introduction of goal desires proves important, as the construct always significantly predicts behavioral desires. Note, however, that the measures of goal and behavioral desires achieve discriminant validity, even after correcting for measurement error. These results support the view that the interplay between goal and behavioral levels of analysis represents an important element to understand the decision making process. Note that, although anticipated emotions have a significant impact on behavioral desires in 3 out of 4 studies, their average impact is substantially lower than that found under the MGB. It is possible that the introduction of goal desires reduces the impact of anticipated emotions on behavioral desires, as both sets of constructs have goals as their referents. As for the MGB, past behavior shows a substantial impact on both behavioral desires and intentions. Finally note that, considering both the MGB and the EMGB, the impact of perceived behavioral control on desires is moderate but often smaller in magnitude than the path from past behavior to desires.

[Insert Table 4 about here]

Conclusions

The MGB and EMGB provide a fuller understanding of decision making than the TPB. On the face of it, it seems that the advantages of the MGB and EMGB come at the cost of loss of parsimony. However, there is a sense in which the MGB and EMGB are more parsimonious than the TPB. If we focus on the proximal determinants of intentions, we see that the MGB and EMGB actually specify fewer direct antecedents to intentions than the TPB. More importantly, it should be pointed-out that the theories are based on fundamentally different theoretical processes. The TPB assumes that intentions summarize the effects of attitudes, subjective norms, and perceived behavioral control and that these antecedents each provide motivational impetus for intending to act. The processes are deterministic, and it is left to empirical research to ascertain which of the antecedents are operative in any particular setting. By contrast, under the MGB and EMGB the antecedents are regarded as reasons for acting that do not, in and of themselves, contain motivations to act. Rather, a decision maker must personally decide and accept these as inner, private concerns and experience their meaning as a desire to act. Desire, then, introduces an element of self-regulation into human behavior, an element missing from the TPB (see also Orbell’s chapter on self-regulation in this book). Desire transforms the reasons for acting into intentions and is the proximal cause of intentions.

PRE-VOLITIONAL PROCESSES 29

 It remains an open question whether other reasons for acting influence intentions and action. Under the MGB and EMGB, as noted above, we acknowledge that automatic processes can have direct effects on action. It has been shown that implicit evaluations can trigger behavior automatically. However, it might be argued that the power for automatic activation of action resides also in learned motives or desires that become triggered or primed in the proper situation. That is, it can be hypothesized that alongside the automatic activation effect on behavior due to the triggering of an implicit evaluation, implicit desires can be automatically triggered as well. We have argued extensively and provided empirical evidence that desires are broader than evaluations (i.e., attitudes) and represent the integration of different reasons to act into a personal motivation to do so. Hence, they are more important than attitudes in determining behavior via intention. It might be argued that desires are more important than attitudes also at an implicit or automatic level. This would imply that people have not only associative networks organized in terms of valence (e.g., Greenwald et al., 2002), but also in terms of wishes. In other words, it looks reasonable that people create associations not only between objects and evaluations, but also between objects and desires or wishes. The activation by proper cues of these latter associations could directly lead to action bypassing intentions.

Another key related issue concerns the interplay between implicit and explicit determinants of action. The two traditions have developed largely in isolation and few attempts have been made to develop a comprehensive framework. Wilson, Lindsey, and Schooler (2000) have recently proposed a model of dual attitudes, defined as different evaluations, implicit and explicit, of the same attitude object. They suggest the co-existence in memory of implicit and explicit attitudes toward the same attitude object. In this framework, implicit attitudes are assumed to direct behavior that people do not monitor consciously or that they do not see as an expression of their attitude, whereas explicit attitudes tend to predict controlled behaviours or behaviours that they see as expressive of their attitudes. The empirical evidence supporting these claims comes mainly from research on prejudice and stereotyping (e.g., Fazio, Jackson, Dunton & Williams, 1995; see also Vargas’s chapter in this book). Despite the useful framework provided by Wilson and colleagues, there are still important questions that are unanswered. One of these questions concern the interplay between implicit and explicit attitudes. The theoretical elaboration of Wilson and colleagues is focused only on cases of conflicts between implicit and explicit attitudes. But, what happens when implicit and explicit attitudes are congruent? Are implicit attitudes interacting with explicit attitudes in affecting behavior? Is the influence of implicit attitudes mediated, at least partly, by deliberative intentions or by desires? These and similar questions represent an important avenue for future investigations. Both the MGB and the EMGB might represent a comprehensive theoretical framework that could be expanded by incorporating implicit determinants of action and might ultimately provide a common ground for both traditions.

The TPB, MGB, and EMGB are basically psychological explanations of action and accommodate social processes only through the impact of subjective norms. Bagozzi (2000) and Bagozzi and Lee (2002) recently reconceptualized intentions to apply in social or group situations. Heretofore, psychologists have studied what might be called personal intentions to perform an individual act (e.g., “I intend to read a novel this evening”). A personal intention is a “conscious plan to exert effort to carry out a behavior” (Eagly & Chaiken, 1993, p. 168) by oneself, where the behavior is the performance of an individual action. There are two types of social intentions found frequently in group situations. One is a personal intention to perform one’s part of a group action (e.g., “I intend to play bridge with my friends on Thursday evening”). A second is a shared intention to engage in a group act (e.g., “We intend to have a vacation in the Canadian Rockies this summer”). Under a shared intention, a person plans to participate in a joint activity, but he/she conceives it, not so much as an individual performing a personal act that contribute to a group performance, but rather as a group action where one is a member of the group. Under a personal intention to perform a group act, the person thinks of a group activity in an atomistic sense such that members of the group, and especially the self, are distinct entities who come together but act individually to contribute to the group activity. Under a shared intention, in contrast, the person thinks of a group activity holistically, such that one sees the self as part of a social representation (e.g., a family, club, team, work organization), and it is the group that acts.

 Bagozzi and Lee (2002) differentiated between three social reasons for acting as explanations of social intentions: compliance (e.g., interpersonal pressure), internalization (e.g., group norms), and social identity (i.e., cognitive self-awareness of membership in a group, affective commitment to the group, affective commitment to the group, and collective selfesteem). This has been tested in the context of the MGB as applied to friendship groups that were organized around ownership of Harley Davidson motorcycles (Bagozzi & Dholakia, 2002).

Thus, the MGB and EMGB can be expanded to explain social behavior by reconceptualizing individual intentions to accommodate social intentions and introducing compliance, internalization, and social identity processes as additional antecedents of desire. Other antecedents or reasons for acting might be relevant to consider, depending on the particular action and person and situation in which the action is performed. For example, Bagozzi and Lee (2002) found that culture moderates the effects of reasons for acting on intentions with social identity more important in group-or interdependent-based cultures and group norms more salient in individual-or independent-based cultures. It is important to stress, however, that desire is the proximal determinant of intentions, and additional reasons for acting will be translated through desire, if they have significance in decision making. In conclusion, we believe that researchers should take into considerations both deliberative and automatic processes in decision making. An approach focused exclusively on the automatic links between attitudes and behavior, which seems very fashionable these days, does not provide a theoretically and pragmatically superior alternative. An approach focused only on the deliberative decision making processes would exclude important theoretical aspects and neglect a key direct route to action whereby explicit intentions and volitions are bypassed. More research is needed into the conditions governing the operation of automatic and deliberative processes, including inquiry into when one or the other functions or when both are at work.

The MGB and the EMGB provide an alternative conceptualization of decision making processes developed within the deliberative camp but that tries to accommodate also automatic processes. The main focus of both models is on pre-volitional processes, yet they can be easily integrated with approaches that focus on post-volitional and pre-actional processes. By their very nature, both models are open to improvements and extensions. The data collected so far are sufficient to claim that the MGB and the EMGB deserve closer scrutiny and investigation and that they are real alternatives to the TPB. Future research should aim at testing the two models in an extensive range of domains and at refining some of the mechanisms that underlie them.

ACKNOWLEDGMENTS

We wish to thank Geoff Haddock and an anonymous reviewer for useful comments and suggestions.

References

Ajzen, I. (1991). The theory of planned behavior. Organizational Behaviour and Human Decision Processes, 50, 179-211.

Ajzen, I. (2001). Nature and operation of attitudes. Annual Review of Psychology, 52. Ajzen, I. (in press). Residual effects of past on later behavior: Habituation and reasoned perspectives. Personality and Social Psychology Review.

Ajzen, I., & Fishbein, M. (2000). Attitudes and the attitude-behavior relation: Reasoned and automatic processes. In W. Stroebe & M. Hewstone (Eds.), European Review of Social Psychology (vol. 11, pp. 1-33). London: Wiley & Sons.

Ajzen, I., & Madden, T.J. (1986). Prediction of goal-directed behavior: Attitudes, intentions, and perceived behavioral control. Journal of Experimental Social Psychology, 22, 453-474.

Albarracin, D., Wyer, R.S. (2000). The cognitive impact of past behavior: Influences on beliefs, attitudes, and future behavioral decisions. Journal of Personality and Social Psychology, 79, 5-22.

Armitage, C.J., & Conner, M. (1999). Distinguishing perceptions of control from self-efficacy: Predicting consumption of a low fat diet using the theory of planned behaviour. Journal of Applied Social Psychology, 29, 72-90.

Armitage, C.J., & Conner, M. (2001). Efficacy of the theory of planned behaviour: A metaanalytic review. British Journal of Social Psychology, 40, 471-799.

Aarts, H., & Dijksterhuis, A. (2000). Habit as knowledge structure: Automaticity in goal-directed behavior. Journal of Personality and Social Psychology, 78, 53-63. Audi, R (1986). Acting for reasons. Philosophical Review, 95, 511-546.

Bagozzi, R.P. (1981). Attitudes, intentions and behavior: A test of some key hypothesis. Journal of Personality and Social Psychology, 41, 607-627.

Bagozzi, R.P. (1992). The self-regulation of attitudes, intentions, and behavior. Social Psychology Quarterly, 55, 178-204.

Bagozzi, R.P. (2000). On the concept of intentional social action in consumer behavior. Journal of Consumer Research, 27, 388-396.

Bagozzi, R.P., Baumgartner, H., & Pieters, R. (1998). Goal-directed emotions. Cognition and Emotion, 12, 1-26. Bagozzi, R.P., & Dholakia, U.M. (2002). Brand community behavior: Psychological and social antecedents plus purchase consequences. Unpublished working paper.

Bagozzi, R.P., & Edwards, E.A. (1998). Goal setting and goal pursuit in the regulation of body weight. Psychology and Health, 13, 593-621.

PRE-VOLITIONAL PROCESSES 35

Bagozzi, R.P., & Kimmel, S.K. (1995). A comparison of leading theories for the prediction of goal-directed behaviours. British Journal of Social Psychology, 34, 437-461.

Bagozzi, R.P. & Lee, K-H. (2002). Multiple routes for social influence: The role of compliance, internalization, and social identity. Social Psychology Quarterly, in press. Bagozzi, R.P. & Warshaw, P.R. (1990). Trying to consume. Journal of Consumer Research, 17, 127-140.

 Bagozzi, R.P., Wong, N., & Yi, Y. (1999). The role of culture and gender in the relationship between positive and negative effect. Cognition and Emotion, 13, 641-672.

Bargh, J.A. (1990). Auto-motives: Preconscious determinants of social interaction. In R.M. Sorrentino & E.T. Higgins (Eds.), Handbook of motivation and cognition (pp.93-130). New York: Guilford Press.

Bargh, J.A., & Chartrand, T.L. (1999). The unbearable automaticity of being. American Psychologist, 54, 462-479.

Bargh, J.A., Chen, M., & Burrows, L. (1996). Automaticity of social behavior: Direct effects of trait construct and stereotype activation on action. Journal of Personality and Social Psychology, 71, 230-244. Bargh, J.A., Gollwitzer, P.M., Lee-Chai, A., Barndollar, K., & Troetschel, R. (2001). The automated will: Nonconscious activation and pursuit of behavioral goals. Journal of Personality and Social Psychology, 81, 1014-1027.

Beck, L., & Ajzen, I. (1991). Predicting dishonest actions using the theory of planned behaviour. Journal of Research in Personality, 25, 285-301.

Bentler, P.M. & Speckart, G. (1979). Models of attitude-behavior relations. Psychological Review, 86, 452-464.

Bentler, P.M. & Speckart, G. (1981). Attitudes "cause" behaviors: A structural equation analysis. Journal of Personality and Social Psychology, 40, 226-238.

Bratman, M.E. (1987). Intentions, plans, and practical reason. Cambridge, MA: Harvard University Press.

 Breckler, S.J., & Wiggins, E.C. (1989). Affect versus evaluation in the structure of attitudes. Journal of Experimental Social Psychology, 25, 253-271.

Cacioppo, J.T., & Berntson, G.G. (1994). Relationship between attitudes and evaluative space: A critical review, with emphasis on the separability of positive and negative substrates. Psychological Bulletin, 115, 401-423.

 PRE-VOLITIONAL PROCESSES 36

Capozza, D., Cesco, V., Dazzi, C., Voci, A., & Martines, D. (1999). Verifica di teorie delle intenzioni tramite modelli di equazioni strutturali [Test of intention theories through structural equation models]. TPM, 6, 133-146.

Carver, C.S., & Scheier, M.F. (1998). On the self-regulation of behavior. Cambridge, UK: Cambridge University Press.

Chartrand, T.L., & Bargh, J.A. (1999). The chameleon effect: The perception-behavior link and social interaction. Journal of Personality and Social Psychology, 76, 893-910.

Chen, M., & Bargh, J.A. (1999). Consequences of automatic evaluation: Immediate behavioral predispositions to approach or avoid the stimulus. Personality and Social Psychology Bulletin, 25, 215-224. Chen, S., & Chaiken, S. (1999). The heuristic-systematic model in its broader context. In S. Chaiken & Y. Trope (Eds.), Dual-process theories in social psychology (pp. 73-96). New York: The Guilford Press.

Conner, M., & Abraham, C. (2001). Conscientiousness and the theory of planned behavior: Toward a more complete model of the antecedents of intentions and behavior. Personality and Social Psychology Bulletin, 27, 1547-1561.

 Conner, M., & Armitage, C.J. (1998). Theory of planned behaviour: A review and avenues for future research. Journal of Applied Social Psychology, 28, 1430-1464.

Cunningham, W.A., Preacher, K.J., & Banaji, M.R. (2001). Implicit attitude measures: Consistency, stability, and convergent validity. Psychological Science, 12, 163-170.

Dasgupta, N., McGhee, D.E., Greenwald, A.G., & Banaji, M.R. (2000). Automatic preference for White Americans: Eliminating the familiarity explanation. Journal of Experimental Social Psychology, 36, 316-328.

Davidson, D. (1963) Actions, reasons, and causes. Journal of Philosophy, 60, 685-700. Davis, W.A. (1984). A causal theory of intending. American Philosophical Quarterly, 21, 43-54.

Deci, E.L., & Ryan, R.M. (1991). A motivational approach to self: Integration in personality. In R. Dienstbier (Ed.), Nebraska Symposium on Motivation (vol. 38, pp. 199-255). Lincoln: University of Nebraska Press.

Dovidio, J.F., Kawakami, K., Johnson, C., Johnson, B., & Howard, A. (1997). On the nature of prejudice: Automatic and controlled processes. Journal of Experimental Social Psychology, 33, 510-54.

Eagly, A.H., & Chaiken, S. (1993). The psychology of attitudes. Fort Worth, TX: Harcourt Brace Jovanovich.

PRE-VOLITIONAL PROCESSES 37

Eagly, A.H., & Chaiken, S. (1998). Attitude structure and function. In D.T. Gilbert, S.T. Fiske, & G. Lindzey (Eds.), The handbook of social psychology (4th ed, Vol 1, pp. 269-322). New York: McGraw-Hill.

Fazio, R.H. (1990). Multiple processes by which attitudes guide behavior: The MODE model as an integrative framework. In M.P. Zanna (Ed.), Advances in experimental social psychology, Vol. 23, 75-109. New York: Academic Press.

Fazio, R.H. (1995). Attitudes as object-evaluation associations: Determinants, consequences, and correlates of attitude accessibility. In R.E. Petty & J.A. Krosnick (Eds.), Attitude strength: Antecedents and consequences, Vol. 4, pp. 247-282. Hillsdale, NJ: Lawrence Erlbaum Associates.

 Fazio, R.H., & Dunton, B.C. (1997). Categorization by race: The impact of automatic and controlled components of racial prejudice. Journal of Experimental Social Psychology, 33, 451-470.

Fazio, R.H., & Williams, C.J. (1986). Attitude accessibility as a moderator of the attitudeperception and attitude-behavior relations: An investigation of the 1984 presidential election. Journal of Personality and Social Psychology, 51, 505-514. Fazio, R.H., Sanbonmatsu, D.M, Powell, M.C. & Kardes, F.R (1986). On the automatic activation of attitudes. Journal of Personality and Social Psychology, 50, 229-238.

Fishbein, M., & Ajzen, I. (1975). Belief, attitude, intention and behaviour: An introduction to theory and research. Reading, Mass.: Addison-Wesley.

Gleicher, F., Boninger, D.S., Strathman, A., Armor, D., Hetts, J., & Ahn, M. (1995). With an eye toward the future: The impact of counterfactual thinking on affect, attitudes, and behaviour. In N.J. Roese & M.M. Olson (Eds.), What might have been: The social psychology of counterfactual thinking (pp. 283-304). Mahwah, NJ: Lawrence Erlbaum Associates.

Godin, G., & Kok, G. (1996). The theory of planned behavior: A review of its applications to health-related behaviors. American Journal of Health Promotion, 11, 87-98.

Gollwitzer, P.M. (1990). Action phases and mind-sets. In E.T. Higgins & R.M. Sorrentino (Eds.), Handbook of motivation and cognition (Vol. 2, pp. 53-92). New York: Guilford.

Gollwitzer, P.M., & Moskowitz, G.B. (1996). Goal effects on action and cognition. In E.T. Higgins & A.W. Kruglanski (Eds.), Social psychology: Handbook of basic principles, pp. 361-399. New York: Guilford.

Gray, J.A. (1990). Brain systems that mediate both emotion and cognition. Cognition and Emotion, 4, 269-288.

PRE-VOLITIONAL PROCESSES 38

Greenwald, A.G., & Banaji, M.R. (1995). Implicit social cognition: Attitudes, self-esteem, and stereotypes. Psychological Review, 102, 4-27.

Greenwald, A.G., Banaji, M.R., Rudman, L.A., Farnham, S.D., Nosek, B.A., & Mellott, D.S. (2002). A unified theory of implicit attitudes, stereotypes, self-esteem, and self-concept. Psychological Review, 109, 3-25.

Greenwald, A.G., & Farnham, S.D. (2000). Using the Implicit Association Test to measure selfesteem and self-concept. Journal of Personality and Social Psychology, 79, 1022-1038.

Greenwald, A.G., McGhee, D.E., & Schwartz, J.K.L. (1998). Measuring individual differences in implicit cognition: The implicit association test. Journal of Personality and Social Psychology, 74, 1464-1480.

Higgins, E.T. (1996). Ideals, oughts, and regulatory focus. In P.M. Gollwitzer & J.A. Bargh (Eds.), The Psychology of Action (pp. 119-145). New York: Guilford Press.

Karpinski, A., & Hilton, J.L. (2001). Attitudes and the Implicit Association Test. Journal of Personality and Social Psychology, 81, 774-778.

Leone, L., Perugini, M., & Ercolani, A.P. (1999). A comparison of three models of attitudebehaviour relationships in studying behaviour domain. European Journal of Social Psychology, 29, 161-189.

Leone, L., Perugini, M., & Ercolani, A.P. (2002). A test of the Model of Goal-directed Behavior (MGB). Unpublished manuscript.

Liberman, N., & Trope, Y. (1998). The role of feasibility and desirability considerations in near and distant future decisions: A test of temporal construal theory. Journal of Personality and Social Psychology, 75, 5-18.

 Locke, E.A., & Latham, G.P. (1990). Work motivation and satisfaction: Light at the end of the tunnel. Psychological Science, 1, 240-246.

Loewenstein, G. (1987). Anticipation and the valuation of delayed consumption. The Economic Journal, 97, 666-684.

Malle, B.F., & Knobe, J. (2001). The distinction between desire and intention: A folk-conceptual analysis, In B.F. Malle, L.J. Moses, & D.A. Baldwin (Eds.), Intentions and intentionality: Foundations of social cognition. Cambridge, Mass.: MIT Press.

Mele, A.R., & Moser, P.K. (1994). Intentional action. Nous, 28, 39-68.

Mele, A.R. (1995). Motivation: Essentially motivation-constituting attitudes. The Philosophical Review, 104, 387-423. Mele, A.R. (Ed.) (1997). The philosophy of action. New York: Oxford University Press.

Ortony, A., Clore, G.L., & Collins, A. (1988). The cognitive structure of emotions. Cambridge, UK: Cambridge University Press.

Ouellette, J.A., & Wood, W. (1998). Habit and intention in everyday life: The multiple processes by which past behavior predicts future behavior. Psychological Bulletin, 124, 54-74.

 Parker, D., Manstead, A.S.R., & Stradling, S.G. (1995). Extending the theory of planned behaviour: The role of personal norm. British Journal of Social Psychology, 34, 127-137.

Perugini, M. (2002). A test of the Extended Model of Goal-directed Behavior in three domains. Unpublished raw data.

Perugini, M., & Bagozzi, R.P. (2001). The role of desires and anticipated emotions in goaldirected behaviours: Broadening and deepening the Theory of Planned Behaviour. British Journal of Social Psychology, 40, 79-98.

Perugini, M., & Bagozzi, R.P. (in press). The distinction between desires and intentions. European Journal of Social Psychology. Perugini, M., & Conner, M.T. (2000). Predicting and understanding behavioral volitions: The interplay between goals and behaviors. European Journal of Social Psychology, 30, 705- 731.

Petty, R.E., & Wegener, D.T. (1999). The elaboration likelihood model: Current status and controversies. In S. Chaiken & Y. Trope (Eds.), Dual-process theories in social psychology (pp. 37-72). New York: The Guilford Press. Posavac, S.S, Sanbonmatsu, D.M., & Fazio, R.H. (1997). Considering the best choice: Effects of the salience and accessibility of alternatives on attitude-decision consistency. Journal of Personality and Social Psychology, 72, 253-261.

Richard, R., van der Pligt, J., & de Vries, N. (1995). Anticipated affective reactions and prevention of AIDS. British Journal of Social Psychology, 34, 9-21.

Roese, N.J., & Olson, J.M. (Eds.) (1995). What might have been: The social psychology of counterfactual thinking. Mahwah, NJ: Lawrence Erlbaum Associates.

Roseman, I.J., Antoniou, A.A., & Jose, P.E. (1996). Appraisal determinants of emotions: Constructing a more accurate and comprehensive theory. Cognition and Emotion, 10, 241- 277.

Rosenthal, R., & Rubin, D.B. (1982). Comparing effect sizes of independent studies. Psychological Bulletin, 92, 500-504. Rothman, A. J. & Salovey, P. (1997). Shaping perceptions to motivate healthy behavior: The role of message framing. Psychological Bulletin, 121, 3-19.

Russell, J.A., & Barrett, L.F. (1999). Core affect, prototypical emotional episodes, and other things called emotion: Dissecting the elephant. Journal of Personality and Social Psychology, 76, 805-819.

Schueler, G.F. (1995). Desire: Its role in practical reasons and the explanation of action. Cambridge, Mass.: MIT Press.

Sparks, P., & Shepherd, R. (1992). Self-identity and the theory of planned behaviour: Assessing the role of identification with green consumerism. Social Psychology Quarterly, 55, 388- 399.

Taylor, S.D., Bagozzi, R.P., & Gaither, C.A. (2001). Gender differences in the self-regulation of hypertension. Journal of Behavioral Medicine, 24, 469-487.

Terry, D.J., & O’Leary, J.E. (1995). The theory of planned behaviour: The effects of perceived behavioural control and self-efficacy. British Journal of Social Psychology, 34, 199-220.

Trafimow, D., & Sheeran, P. (1998). Some tests of the distinction between cognitive and affective beliefs. Journal of Experimental Social Psychology, 34, 378-397.

Tversky, A. & Kahneman, D. (1974). Judgment under uncertainty: Heuristics and biases. Science, 185, 1124-1131.

Vallacher, R.R., & Wegner, D.M. (1987). What do people think they’re doing? Action identification and human behavior. Psychological Review, 94, 3-15.

van der Pligt, J., Zeelenberg, M., van Dijk, W.W., de Vries, N.K., & Richard, R. (1998). Affect, attitudes and decisions: Let’s be more specific. In W. Stroebe & M. Hewstone (Eds.), European Review of Social Psychology (vol. 8, pp. 33-66). London: Wiley & Sons.

Verplanken, B. & Aarts, H. (1999). Habit, attitude, and planned behaviour: Is habit an empty construct or an interesting case of goal-directed automaticity? In W. Stroebe & M. Hewstone (Eds), European Review of Social Psychology (vol. 10, pp. 101-134). London: Wiley & Sons.

Verplanken, B., Aarts, H., & van Knippenberg, A. (1997). Habit, information acquisition, and the process of making travel mode choices. European Journal of Social Psychology, 27, 539- 560.

Verplanken, B., Aarts, H., van Knippenberg, A., & Moonen, A. (1998). Habit versus planned behaviour: A field experiment. British Journal of Social Psychology, 37, 111-128.

Verplanken, B., Hofstee, G., & Janssen, H.J.W. (1998). Accessibility of affective versus cognitive components of attitudes. European Journal of Social Psychology, 28, 23-35.

Wegner, D.M., & Wheatley, T. (1999). Apparent mental causation: Sources of the experience of will. American Psychologist, 54, 480-492.

Yanchar, S.C. (2000). Some problems with Humean causality. American Psychologist, 55, 767- 768.

PRE-VOLITIONAL PROCESSES 42

FOOTNOTES

1 It is outside the scope of this chapter to embark into a systematic analysis of all studies conducted on automaticity. The selectivity of our review is exclusively done for purposes of exemplification. However, we believe that by analyzing three among the most famous and widely cited examples of “automaticity in action”, general considerations can still be reasonably made.

 2 The three studies have been cited a total of 184 times by other published works (August 2002).

3 One exception is given by the model of action phases (Heckhausen & Gollwitzer, 1987; Gollwizter, 1996). The first stage of this model (predecisional stage) consists of the contemplation of different wishes and desires and the setting of preferences among them. The culmination of this stage is the formation of a goal intention, which is initiated and executed in the following two stages (pre-actional and actional). However, Gollwitzer does not consider extensively how these goal intentions come about (e.g., on the basis of which criteria a goal is chosen) but rather focuses on subsequent action phases.

 4 Obviously there are circumstances where it might be a reasonable approximation to assume that a certain end state is desirable. Yet, strictly speaking, even in extreme cases this would generally be a matter of interpretation. Most people would agree that it is more desirable to be alive than to die, or to be rich than to be poor, yet some individuals prefer to die instead of live (euthanasia, martyrdom) or to choose poverty rather than wealth (ascetics, religious people).

 5 At least three procedures can be used to test the discriminant validity of two constructs using CFA. One approach is to inspect the confidence interval of the estimated correlation between the latent constructs. If the confidence interval, built by considering ± 2 standard errors from the point-estimate, does not include 1, then the constructs are discriminated. A second procedure is to compare two models, the first with the observed variables loading on two correlated factors and the second with the variables loading on a single overall factor. A third procedure, which perhaps is superior to the first two, is to compare two models, the first with the correlation between latent constructs set free and the second with the correlation fixed to unity. If the latent variables are not significantly different, one cannot reject the hypothesis that the two constructs are the same. Hence, a significant difference is needed to provide evidence of discriminant validity.

PRE-VOLITIONAL PROCESSES 44

Table 1. Variance explained in behavior in the experiments of Bargh, Chen & Burrows (1996), Chartrand & Bargh (1999), and Dijksterhuis & Van Knippenberg (1998).

Study DV Value df R r2

Bargh, Chen, & Burrow (1996) Exp 1 (Rudeness) Counting of rude behaviora % 73.5 vs. 26.5 1 .36 .13 Exp. 2 (Elderly people) a) Walking speed t 2.86 28 .48 .23 b) Walking speed t 2.16 28 .38 .14 Exp. 3 (African American) Hostility rating (by coders) F 6.95 39 .39 .15 Chartrand & Bargh (1999) Exp 1 (mimicry between strangers) 1) Facial expression (by coders) F 20.31 34 .61 .37 2) Face rubbing F 5.71 34 .38 .14 3) Foot shaking F 3.76 34 .32 .10 4) Smiling F 4.16 34 .33 .11 Exp 2 (adaptive function) 1) Liking F 5.55 70 .27 .07 2) Smoothness F 4.08 70 .23 .06 Exp 3 (individual differences) For high perspective-takers 1) Combined face rubbing and foot shaking F 3.85 48 .27 .07 Dijksterhuis & Van Knippenberg (1998) Exp. 1 (Professors) 1) Correct answers (Prof. vs. Secret.) F 10.45 57 .39 .15 2) (Prof. vs. No-prime) F 5.84 57 .30 .09 Exp. 2 (Length of priming) 1) Correct answers (2 min vs. No-prime) F 4.83 55 .28 .08 2) (9 min vs. No-prime) F 16.36 55 .48 .23 Exp. 3 (Soccer hooligans) 1) Correct answers (2 min vs. No-prime) F 1.35 92 .12 .01 2) (9 min vs. No-prime) F 10.58 92 .32 .10 Exp. 4 (Direction of prime) Correct answers F 7.12 39 .39 .15 Average .35 .13

Notes. a= The primary dependent variable of this experiment (number of seconds the participants waited before interrupting the experimenter) could not be used to calculate the variance because the authors do not report the t-values for pairwise comparisons. We used the secondary dependent variable (percentage of participants in each priming condition who interrupted at all) that could be approximately inferred from Figure 1.

PRE-VOLITIONAL PROCESSES 45

Table 2. Summary of predictive power of TPB, MGB, and EMGB STUDY TPB MGB EMGB N Goal Behavior R2 Int. R2 Beh. R2 Des. R2 Int. R2 Beh. R2 Des. R2 Int. Perugini & Bagozzi (2001) 108 Weight control Dieting .34 .19 .54 .74 .25 na na Exercising .58 .38 .72 .78 .46 na na 122 Studying Free choice .34 .15 .49 .53 .24 na na Leone, Perugini, & Ercolani (2002) 102 Learning SPSS Handbook .37 .18 .52 .58 .27 na na Practice .24 .16 .41 .38 .27 na na Bagozzi & Dholakia (2002) 154 Group participation Riding, Shopping, Socializing .12 .21 .80 .64 .22 na na 255 Group participation Riding, Shopping, Socializing .23 .14 .58 .41 .14 na na Average .32 .20 .58 .58 .26 - - Perugini & Conner (2000) 104 Weight/Studying Free choice .30 na .46 .74 na .62 .76 Perugini (2002) 107 Keeping fit Free choice .37 na .69 .79 na .79 .80 101 Look after health Free choice .31 na .66 .81 na .70 .82 Perugini & Bagozzi (2002) 102 Helping charity Free choice .28 na .72 .79 na .74 .79 Average .32 - .63 .78 - .71 .79 Total Average .35 .21 .58 .68 .30 .71 .79 Notes. Int=Intention, Beh=Behavior, Des=Desire; na=not available.

PRE-VOLITIONAL PROCESSES 46

PB1 PB2 LPE1 LPE2 PC P1 P2 PB3 BD1 BD2 Average PAE → Desires .28* -.14 .11 .35* .47* -.01 -.05 -.05 .23** .28*** .15 NAE → Desires .12 .32* .12 .04 .08 .14 .35* .39* .10* .01 .17 Attitudes → Desires .17 .41* .39* .01 .13 .49* .20* .19* .13* .24*** .24 Subjective Norms → Desires .24* .16* .15 .17 .14 .07 .26* .03 .24*** .11* .16 PBC → Desires -.08a, .61b .37* .19* .21* .34* -.11 .08 .23* .05 .07 .18 Past behavior → Desires .24* .24* .12c .29*c .10 .42* .39* .51* na na .31 Desires → Intentions .77* .66* .71* .60* .85* .62* .62* .76* .78*** .78*** .72 Past behavior → Intentions .16* .14 .12c .04c .03 .34* .35* .17 na na .17

Table 3. Summary of structural paths of MGB
											Average
PAE → Behavioral desires
NAE → Behavioral desires
Attitudes → Behavioral desires
Subjective Norms → Desires
PBC → Desires
Past behavior → Desires
Behavioral desires → Intentions
Past behavior → Intentions
Notes. PAE= Positive anticipated emotions; NAE=Negative anticipated emotions; PBC=Perceived behavioral control. PB1=Perugini & Bagozzi, 2001, Weight control; PB2=Perugini & Bagozzi, 2001, Studying; LPE1=Leone, Perugini & Ercolani, 2002, Learning SPSS, Handbook; LPE2=Leone, Perugini & Ercolani, 2002, Learning SPSS, Practice; PC=Perugini & Conner, 2000, Weight/Studying; P1=Perugini, 2002, Keeping fit; P2=Perugini, 2002, Look after health; PB3=Perugini & Bagozzi, 2002, Helping charity; BD1=Bagozzi & Dholakia, 2002, Study 1; BD2 = Bagozzi & Dholakia, Study 2. a= Dieting; b= Exercising; c= Recency

PRE-VOLITIONAL PROCESSES 47

Table 4. Summary of structural paths of EMGB
	PC	P1	P2	PB3	Average
Goal desires → Behavioral desires	.52*	.44*	.28*	.27*	.38
PAE → Behavioral desires	.38*	-.04	-.13	-.10	.03
NAE → Behavioral desires	-.28	-.03	.24*	.25*	.04
Attitudes → Behavioral desires	.17*	.44*	.15	.15	.23
Subjective Norms → Behavioral desires	.13	.07	.25*	.03	.12
PBC → Behavioral desires	.28*	-.07	.04	.17*	.11
Past behavior → Behavioral desires	.12	.26*	.34*	.41*	.28
Behavioral desires → Intentions	.85*	.64*	.63*	.76*	.72
Past behavior → Intentions	.03	.32*	.34*	.18	.22
Notes. PAE= Positive anticipated emotions; NAE=Negative anticipated emotions; PBC=Perceived behavioral control. PC=Perugini & Conner, 2001, Weight/Studying; P1=Perugini, 2002, Keeping fit; P2=Perugini, 2002, Look after health; PB3=Perugini & Bagozzi, 2002, Helping charity.

Figure Captions

Figure 1. The Model of Goal-directed Behavior.

Attitude Subjective norms Intention Perceived behavioral control Past Behavior Positive anticipated emotions Negative anticipated emotions Behavioral desire Affective processes Motivational processes Automatic processes

Figure 2. The Extended Model of Goal-directed Behavior.

PRE-VOLITIONAL PROCESSES 50

Attitude Subjective norms Intention Perceived behavioral control Past Behavior Positive anticipated emotions Negative anticipated emotions Behavioral desire Affective processes Motivational processes Automatic processes Goal desire Means/end processes

.