The role of desire in the prediction of intention: The case of smoking behavior

Jurnal:

Kova V B & Rise, J. (2011). The Role of Desire in The Prediction of Intention: The Case of Smoking Behavior. Swiss Journal of Psychology, 70 (3), 2011, 141–148 DOI 10.1024/1421-0185/a000049

Kovač, V. B., & Rise, J. (2011). The role of desire in the prediction of intention: The case of smoking behavior. Swiss Journal of Psychology/Schweizerische Zeitschrift für Psychologie/Revue Suisse de Psychologie, 70(3), 141.

Note: Ini hanya sebuah catatan, jangan dijadikan rujukan. Silahkan merujuk ke sumber aslinya

The role of desire in the prediction of intention: The case of smoking behavior

Abstract.This paper is based on the notion that desire represents an important motivational aspect of the decision-making process. Thus, we examined the hypotheses that desire (1) predicts behavioral intentions and(2) mediates the effects of theory of planned behavior (TPB) components and past behavior on an individual’s intention to quit smoking. The analysis is based on three separate conditions in which the intention to quit smoking during the next 1, 4, or 6 months, respectively, was measured; the three conditions contained identical variables. The results of the hierarchical regression analysis show that there are sufficient grounds for including desire as an additional predictor in the TPB model. The results also show that desire mediates the effects of attitudes, norms, and past behavior on intention. However, the mediating role of desire was not obtained for therelationship between PBC and intention. Theoretical implications and recommendations for future research are suggested.

Keywords: TPB, desire, intention, past behavior, decision making

Introduction

The concept of behavioral intention is a well-established predictor of planned behavior (Sheeran, 2002). For instance, the theory of planned behavior (TPB; Ajzen, 1991), arguably the most popular model in the domain of health behavior, posits that the most immediate determinant of actual behavioral performance is the intention to engage in that particular behavior. Behavioral intention reflects an individual’s decision to exert effort to perform the behavior and is assumed to be a function of (1) an individual’s attitude toward the behavior, that is, a positive or negative evaluation of the behavior; (2) subjective norms that refer to an individual’s perceived social pressure to perform the behavior; and (3) perceived behavioral control (PBC) in terms of the perceived ease and difficulty in performing the behavior. These three concepts – attitudes, subjective norms, and PBC – are explicitly taken to encompass not only the reasons for action, evaluation of social pressure, and feasibility of intended actions, but also the motivation for future actions (Ajzen, 1991).

The TPB has been successful in providing a parsimonious account of planned behaviors as evidenced in the metaanalytical literature, which found that TPB components accounted on an average for approximately 40% of the variance in intentions and approximately 30% of the variance in behavior (Armitage & Conner, 2001; Godin & Kok, 1996). Despite the relative success of the TPB components in predicting intentions across various behavioral domains, Ajzen (1991) entertained the possibility that other predictors besides the traditional TPB components might be useful in predicting intentions, by their accounting for additional variance beyond the TPB components. This general approach, which aims to increase the number of predictors in the original model, is commonly termed “theory broadening” in the literature (Perugini & Bagozzi, 2001), and a number of additional predictors have been suggested (see Conner & Armitage, 1998; Conner & Sparks, 2005; O’Keefe, 2002). More specifically, meta-analyses found that descriptive norms (Rivis & Sheeran, 2004), anticipated regret (Sandberg & Conner, 2008), and self-identity (Rise, Sheeran, & Hukkelberg, 2010) account for 5%, 7%, and 6%, respectively, of additional variance across a number of behavioral domains. When it comes to smoking behavior, the TPB has been successfully used to account for intentions to quit smoking (see Rise, Kova2, Kraft, & Moan, 2008, for an overview). Furthermore, additional predictors, such as moral norms, anticipated affect, group identity, and past behavior, have also been shown to have predictive value for the intention to quit smoking (see Falomir & Invernizzi, 1999; Moan & Rise, 2005; Rise & Ommundsen, 2010).

Despite the work of Bagozzi and colleagues (Bagozzi, 1992; Perugini & Bagozzi, 2001, 2004b; Perugini & Conner, 2000), the role of desire as a central psychological process within the general research on TPB across various behavioral contexts remains underexplored. One possible explanation for this general lack of interest in the concept of desire might be related to the established theoretical premise in contemporary literature that desire is either conceptually similar to behavioral intentions (Fishbein & Stasson, 1990) or implied in TPB components (Ajzen, 1991). In contrast, Bagozzi and colleagues (Perugini & Bagozzi, 2004b), building upon philosophical traditions (e.g., Davis, 1984), argue that empirical research should discriminate between desire and intention, based on the idea that the theoretical components of TPB do not explicitly include the motivation to act. The argument is that evaluative appraisals in terms of attitudes, perceived control, and perceived social influence merely provide reasons to act but fail to address more personal motivational commitments to act. Following this line of thinking, the TPB components are taken to be reason-based judgments or evaluations of future actions and as such lack motivational energy in order to “move” the individual in a specific direction (e.g., Perugini & Bagozzi, 2004b). For instance, a passionate smoker may very well agree that quitting is a good idea and feel pressure to do so from important individuals in their innermost social circle; they may also feel capable of quitting yet simultaneously lack the desire to do so. It follows that intention implies desire, while desire does not necessarily imply intention (Bagozzi, 1992; Perugini & Bagozzi, 2004b). As such, desires are conceptually connected with the wishing process, where realistic possibilities as well as pro and cons have not yet been considered, while intentions are per definitionem connected with volition and imply self-regulative efforts at some point in the future. Furthermore, when it comes to deliberative processes, desires are general wishes that by definition are not restricted to specific time spans and do not imply efforts. As such, desires precede intentions in time, and many human desires remain nonrealistic wishes which are never realized. However, if a passionate smoker sets a specific date with respect to their personal desire to quit smoking, he/she automatically translates the desire into behavioral intention and is forced to consider the feasibility aspects of this decision as the quitting date approaches.

The lack of research in including the concept of desire in the equation may sound strange considering its potential dual theoretical role in terms of theory broadening, that is, as an additional predictor in the TPB and as a theory deepener, that is, by illuminating the role of existing theoretical predictors (Perugini & Bagozzi, 2001). Several studies across different contexts have supported the role of desire as an additional predictor in the TPB (Bagozzi & Kimmel, 1995; Leone, Perugini, & Ercolani, 1999, 2004). This line of research indicates that desire also mediates most of the effects of TPB components on intention (Bagozzi, Dholakia, & Basuroy, 2003; Leone et al., 2004; Perugini & Bagozzi, 2001; Shaw, Shiu, Hassan, Bekin, & Hogg, 2007; Taylor, Bagozzi, & Gaither, 2005). In other words, this research shows that, in addition to being an important predictor of intention in terms of theory broadening, desire might also be important in terms of theory deepening by providing a better understanding of the theoretical mechanism underlying the effects of the existing theoretical predictors on behavioral intention (Perugini & Bagozzi, 2001). Nevertheless, these issues have scarcely been explored in the context of smoking behavior. In fact, to our knowledge, this is the first study that specifically aims to explore desire’s usefulness as a predictor of intention and mediator of TPB components on intentions in the context of smoking behavior. The selection of the smoking context might be particularly interesting in order to highlight the conceptual difference between these two concepts when considering the fact that many smokers might simultaneously have a strong desire to quit but are reluctant to form a strong intention to quit based on the belief that the quitting process represents a difficult aim to achieve.

In addition to exploring desire as an additional predictor and mediator, we included a measure of past quitting attempts as an indicator of past behavior. Past behavior has previously been suggested to be an important variable in the intention formation process, and a direct effect of this predictor has been taken as an indication that important, unmeasured behavioral causes have been left out of the equation (see Ouelette & Wood, 1998; but also Ajzen, 1991, 2002b). We used this measurement in two ways. First, we included past behavior as a statistical control taking into consideration that the role of this concept is unclear in the context of TPB (Ajzen, 2002b). Second, we explored whether desire in addition to TPB components also mediates the effect of past behavior on intention. Past quitting attempts have been found to be related to the intention to quit smoking (Kovac, Rise, Moan, 2010; Moan & Rise, 2005; Norman, Conner, & Bell, 1999; Rise et al., 2008), although the direct effect has been noted as weak (Moan & Rise, 2005).

In sum, the present paper sheds some light on these issues in the context of smoking behavior. More specifically, we analyze the relationship between traditional TPB components, desire, past behavior, and intention to quit smoking in three separate timeframe conditions (viz., during the next 1, 4, or 6 months). Following the reasoning of Bagozzi and colleagues (e.g., Taylor et al., 2005), we test the hypotheses that desire (1) represents an important predictor of intention and (2) mediates the effects of the TPB components and past behavior on intention.

Method

Participants and Procedure

The participants in the 1- and 6-month conditions were students approached on campus while smoking and asked to fill out a questionnaire concerning components of the TPB and smoking. The participants in the 4-month condition were daily smokers who responded to an invitation to participate in the study, which was displayed in 15 national online newspapers. Apart from the temporal framing of the intention to quit smoking (viz., during the next 1, 4, or 6 months), all three questionnaires were identical. Respondents were confronted with items measuring their intention to quit smoking at the end of the questionnaire in order to not affect the responses to previous items. Participants in the 1-month condition (N= 96) had a mean age of 23.83 (SD= 3.56); 68% were females and had been smoking for an average of 8.15 (SD= 3.47) years, consuming 9.68 (SD= 4.47) cigarettes per day. The corresponding figures for the 4-month condition were: N= 939, mean age 35.81 (SD= 11.71) years, 49% females, smoking for 18.65 (SD=10.87) years on average, consuming 13.53 (SD= 7.44) cigarettes per day. The corresponding figures for the 6-month condition were: N= 128, mean age 24.43 (SD=5.88) years, 67% females, smoking for 7.90 (SD= 5.95) years on an average, consuming 10.90 (SD= 6.51) cigarettes per day.

Measurements

Attitude

Attitude was measured with 5 items. A sample item is “In my opinion, quitting smoking will be (bad/good).” The 5 items were rated on a 7-point response scale ranging from (–3) bad to (+3) good. Cronbach’s αs were .88 (1 month), .76 (4 months), and .83 (6 months).

Subjective Norm

Subjective norm was assessed by 2 items such as “People who are important to me think I should quit smoking” and “People who are important to me wish that I would quit smoking.” The items were rated on a 7-point response scale ranging from 1 (completely disagree) to7 (completely agree). The correlations (Pearson’sr) between the 2 items were .78,p< .001 (1 month), .85,p< .001 (4 months), and .79, p< .001 (6 months).

Perceived Behavioral Control

Perceived behavioral control was assessed using 5 items, for example, “How much control do you have over quitting smoking?” and “How confident are you that you will quit smoking?” The items were rated on a 7-point response scale ranging from 1 (no control) to7 (a high amount of control). Cronbach’sαs were .88 (1 month), .89 (4 months), and .85 (6 months).

Desire

Desire was assessed by 2 items (“My general desire to quit smoking is usually . . .” and “I often have the desire to quit smoking on a daily basis.”) The first item is designed to capture a general and established desire to quit smoking, the second item to capture the frequency of desire on a daily basis, which in turn is assumed to be related to a more general desire to quit. The items were rated on a 7-point response scale ranging from 1 (very weak/often) to7 (very strong/often). The correlations (Pearson’s r) between the 2 items were .72, p <.001(1 month), .69, p < .001 (4 month), and .72, p < .001 (6 months). Note that, although the measurements of intention were restricted to 1, 4, or 6 months, the items that comprised the measure of desire were not limited to a specific timeframe. This difference is based on the theoretical conceptualizations of desires and intentions (Perugini & Bagozzi, 2004b) and will be further considered in the discussion.

Past Quitting Attempts

Past quitting attempts, a measure of past behavior, was assessed directly in terms of self-reported number of previous attempts to quit smoking.

Intention to Quit Smoking

The participants’ intention to quit smoking was measured by the following items: “During the next (month/4 months/6 months): (1) “I intend to quit smoking,” (2) “I am going to quit smoking.”” The two items were rated on a 7-point response scale ranging from (1) very unlikely to (7) very likely. The correlations (Pearson’sr) between the two items were .93, p< .001 (1 month), .85,p<.001(4 months), and .84,p< .001 (6 months).

Results

Descriptive Statistics

Table 1 presents the means, standard deviations, andp-values that illustrate sample differences regarding background variables such as age, cigarettes smoked per day, smoking duration, and past behavior. The 1-month and 6-month samples differed only with respect to past behavior, which was found to be significantly different on thep< .05 level.

However, the 4-month sample differed significantly from the other two samples with respect to the background variables. This finding was expected considering that the 1- and 6-month samples were taken from student populations, while the 4-month sample stemmed from the general population of smokers. Table 2 presents the correlations between study variables and intention in the three conditions (samples). Intention was significantly correlated with the number of cigarettes smoked per day, attitude, past behavior, and desire in all three samples. Significant correlations between subjective norm and intention were obtained only in the 4- and 6-month conditions, while PBC and intention correlated significantly in the 1- and 4-month conditions.

One unexpected and unusual result was the negative nonsignificant correlation between intention and PBC in the 6-month condition.

Hierarchical Regression Analysis

In order to test the two hypotheses, we conducted the hierarchical linear regression analysis in four steps. Considering the fact that the 4-month sample was significantly different from the other two samples, age, cigarettes smoked per day, and smoking duration were entered in the first step to statistically control the impact of these variables on the overall results. The TPB components were entered in the second step, past behavior in terms of past quitting attempts were entered in the third step, followed by desire in the fourth step. The results of the regression analysis clearly supported the first hypothesis, namely, that desire is an important predictor of the intention to quit smoking (see Table 3 for details). The increase in explained variance in all three conditions after the measure of desire was included in the third step of the regression analysis was considerable (24%, 11%, and 24%, respectively). Furthermore, the impact of desire was statistically significant in all three analyses.

For the second hypothesis, we conducted a series of hierarchical regression analyses to determine whether the effects of the TPB components and past behavior were mediated by desire. Tests for mediation were conducted following the instructions by Baron and Kenny (1986).

Mediation effects are confirmed when a mediating variable accounts for a relationship between two other variables in the sense that the effects of predictor variables are significantly reduced when a hypothesized mediating variable is included in the regression analysis (Baron & Kenny, 1986).

Table 3 shows that the reduction of beta values in the third step, after desire was included, was substantial for attitudes, SN, and past behavior. In order to test whether this reduction was statistically significant, a series of Sobel tests were conducted. The results of these tests were significant for attitudes (z= 3.08,p< .01;z= 12.27,p< .001; z= 4.79,p< .01; for 1, 4, and 6 months, respectively) and past behavior (z= 2.29,p< .05;z= 11.45,p< .001;z= 2.29,p< .05; for 1, 4, and 6 months, respectively) in all three conditions. We also found significant meditation effects of desire on the relationship between SN and intention in the 4-month and 6-month conditions (z= 11.73,p< .001; z= 3.64,p< .001, respectively), while that for the 1-month condition was found to be nonsignificant (z= 1.95,p< .06). In sum, these results clearly indicate that desire, in addition to being an important predictor of intentions in terms of increased explained variance, also serves as an important mediator providing a more comprehensive account for the relation between attitudes, SN, and past behavior, on the one hand, and behavioral intentions on the other. However, there was no meditation effect of desire on the relation between PBC and intention as indicated by the nonsignificant change of beta values of PBC in the third step of the regression analyses when desire was included (Table 3).

Table 1 Descriptive statistics for the three samples

Predictors	1 month			4 months			6 months
	Mean	SD	p	Mean	SD	p	Mean	SD
Age	23.8	3.56	***	35.8	11.71	***	24.3	5.88
CPD	9.6	4.47	***	13.5	7.44	***	10.9	6.50
SD	8.1	3.47	***	18.6	10.87	***	7.9	5.95
PB	2.9	5.59	ns	3.12	2.94	***	1.6	2.78

Note. p-values for the 1- and 6-month conditions are nonsignificant except for PB, which is significant at the .05 level. CPD (cigarettes per day), SD (smoking duration), PB (past behavior). *p< .05, **p< .01, ***p< .001.

Table 2

Correlations between study variables and intention in the three conditions

Predictors	1-month condition	4-month condition	6-month condition
Age	–.05	–.02	–.12
CPD	–.23*	–.15***	–.23*
SD	–.04	–.05	–.18
Attitude	.34**	.48***	.52***
SN	.07	.31***	.31***
PBC	.28**	.37***	–.11
PB	.28**	.25***	.21*
Desire	.43***	.60***	.70***

Note. CPD (cigarettes per day), SD (smoking duration), SN (subjective norm), PBC (perceived behavioral control), and PB (past behavior). *p< .05, **p< .01, ***p< .001.

Discussion

The starting point for the present research is the idea that desire represents a potentially important predictor variable which is frequently overlooked in theoretical models aimed at accounting for people’s behavioral intentions (Bagozzi, 1992; Perugini & Bagozzi, 2001). The results clearly show that desire contributes to theory broadening and hence should be included as an additional predictor of behavioral intentions in extended models of planned behavior. Multiple regression analyses showed that desire enhanced the prediction of quitting intentions quite considerably after the components of the TPB and past behavior were taken into account (24%, 11%, 24% in the 1-, 4- and 6-month conditions, respectively). According to the two criteria proposed by O’Keefe (2002), any potential additional predictor variable in the TPB should contribute to a large amount of explained variance in the outcome variable, and not simply a statistically reliable increment. In our view, values in the range of 11% to 24% additional variance satisfy this criterion. As to the second criterion (viz., that the efficacy of a candidate predictor must be demonstrated across a wide range of behaviors), this needs to be confirmed in further studies.

In terms of theory broadening, the results also show that desire represents an important mediator and may thus account for how traditional TPB variables are theoretically interpreted (Perugini & Bagozzi, 2001). Our findings support the work of Bagozzi and colleagues, who persuasively argued that evaluative aspects of attitudes and normative aspects of SN do not necessarily imply connections to behavioral intention if the object or course of action is not desired (Leone et al., 2004; Shaw et al., 2007). For instance, with increased knowledge about the negative aspects of smoking, it may be fair to say that an increased number of individuals agree that quitting smoking represents a sensible course of action. Hence, very few individuals would argue that the decision underlying personal smoking cessation is bad or report that important people in their social circle would like them to continue smoking. However, the positive attitudes toward quitting smoking and supportive normative opinions of significant others will remain static in terms of action if not first translated into personal desires, which in turn entail a motivation to act. Without desire at work, attitudes and SN might frequently be representatives of the self-detached evaluations on a favor-disfavor continuum, which are not embedded in commitment toward intended action. In other words, the results indicate that positive or negative attitudinal aspects and normative pressures must be connected to personal motivational processes before they can successfully be translated into intention (Shaw et al., 2007). The same logic applies to the finding that the effects of past behavior are mediated by desire in the sense that past behavior (similar to attitudes and SN) is linked to intention to the extent that it is based on a desire to pursue a given action.

Table 3 Hierarchical multiple regression analyses of quitting intention timeframe on Age, CPD, and SD (Step 1), Attitude, SN, and PBC (Step 2), PB (Step 3), and Desire (Step 4) for the three samples

Step	Predictors	1 month			4 months			6 months
		β	R2	F	β	R2	F	β	R2	F
1	Age	–.21			.08			.20
	CPD	–.15			–.15			– .20
	SD	.13			–.09			–.21
			.01	1.12		.02	7.94***		.04	2.35
2	Age	–.25			–.06			.07
	CPD	.05			.02			–.08
	SD	.10			.02			–.10
	A	.29*			.36***			.42***
	SN	.14			.18***			.14
	PBC	.10			.32***			–.11
			.11	3.83*		.34	148.59***		.25	10.38***
3	Age	–.21			–.05			.13
	CPD	–.03			.02			–.12
	SD	.08			–.02			–.16
	A	.24			.33***			.41***
	SN	.11			.15***			.13
	PBC	.12			.34***			–.09
	PB	.17			.20***			.22*
			.13	2.14		.38	51.65***		.29	6.14*
4	Age	–.28*			–.07
	CPD	.01			.00			–.12
	SD	.17			.03			.63
	A	.00			.14***			.19*
	SN	.02			.08**			.02
	PBC	.15			.32***			.03
	PB	.12			.07*			.08
	D	.57***			.45***			.60***
			.37	27.24***		.49	197.24***		.53	48.25***

Note. CPD (cigarettes per day), SD (smoking duration), A (Attitude), SN (subjective norm), PBC (perceived behavioral control), PB (past behavior), and D (desire). *p< .05, **p< .01, ***p< .001.

However, contrary to our hypothesis, we did not find support for a meditational role of desire when it comes to the relation between PBC and intention (for similar findings, see Leone et al., 2004; Shaw et al., 2007). One possible explanation for this finding might be the conceptual difference between the concepts of attitudes and SN, on the one hand, and PBC, on the other (Ajzen, 2002a). Perceptions of control over future actions represent an estimation of people’s sense of control over ease or difficulty at the moment of performance of a given action. In contrast to attitudes and SN, reporting accurate levels of control represents by definition a difficult estimation task, especially when it comes to behaviors that require a great deal of volition, effort, and personal capability, as is the case with the process of quitting smoking. Thus, the concept of PBC, in contrast to attitudes and norms, represents a constrained thinking, which is conceptually closely connected to selfregulation, as it entails perceptions of accurate judgments of performance as well as the prediction of future behavior, that is, a proxy measure for actual control (Ajzen, 1991).

In contrast, the concept of desire, similar to attitudes and SN, is not by definition related to self-regulation, but is mainly based on unconstrained thinking about possible and often unrealistic behavioral alternatives. For instance, Ajzen (2002a) noted that in the case of behaviors which are difficult to perform, PBC might be a strong predictor of intention, even when desire is entered into the equation.

Similarly, Fishbein and Stasson (1990) challenged the idea of making a conceptual distinction between intention and desire, proposing that the distinction is easier to make between desire and behavioral expectations. Thus, when it comes to quitting smoking, a smoker may have a strong desire to quit while at the same time realizing that he or she will not try to stop smoking after taking into account that this might become a difficult project. Furthermore, according to Perugini and Bagozzi (2004a), the concepts of desire and intention are distinct in terms of perceived performability (desired actions are less performable) or action-connectedness (desires are at a higher level of abstraction). By this account, feasibility concerns are less salient with desire than with intention, implying that desire should be a less than optimal mediator of the effect of perceived control on intention, and that perceived control should be a better predictor of intention than of desire. These propositions provide sensible explanations for the lack of support for the meditational role of desire between PBC and intention. Indeed, several other studies have obtained similar results and proposed that when it comes to difficult personal projects in terms of self-regulation, desire should be associated with PBC to a lesser extent and hence be disqualified as a potential mediator between control beliefs and intention (Leone et al., 2004; Shaw et al., 2007). However, it is important to note that these theoretical accounts of the lack of support in favor of the study findings are still speculative and offer a posthoc explanation which need to be formally tested in future research.

Theoretical Implications and

Recommendations

The present results might also shed light on the conceptual distinction between desire and intention (Malle & Knobe, 2001). Perugini and Bagozzi (2004a) noted that the temporal dimension is a key issue with regard to the conceptual difference between these terms, in the sense that behavioral intention, relative to desire, is more closely connected to action in terms of a specified timeframe. Thus, while intention is relatively now-oriented and action-connected, desire tends to be more future-oriented, entailing postponement of action considerations until the actual decision to go through with the action is reached. In other words, desire resides on a level of abstraction where practical aspects of action are not yet considered. This implies that desire by definition should be more strongly related to long-term decisions than to short-term ones. This is the rationale behind why we decided to provide a global measure of desire, that is, not to limit the measure of desire to a specific timeframe.

Although the majority of studies have used specific measures of desire (in which desire was measured at the same level of specificity as the TPB variables), we posit that a specific measure of desire is difficult to distinguish conceptually from a specific measure of intention. The methodological concern was that in the case of a time-limited measurement of desire, we would obtain “intention-like” responses where participants would be forced to limit their desire-based cognitions and consider practical and realistic aspects of future actions. Consistent with this finding, the present data indicate that when the intended targets are situated in the distant future calling for no immediate action, measures of behavioral intention run the risk of being overly abstract and therefore transform themselves into “desirelike” cognitions. More specifically, our data show that correlations between intention and desire tended to become stronger as the timeframe for planned actions increased (r= .43, .60, and .71 for 1, 4, and 6 months, respectively). The differences in correlations between intention and desire were found to be significant for all three conditions (.43, .60, and .71, respectively;p< .05). This pattern of correlation indicates that measurements of desire and intention tend to become more similar over time. Hence, if people are asked to evaluate actions placed in the distant future, they might actually report their desires rather than their behavioral intentions. It further implies that assessing intentions in the very distant future might then be incompatible with the theoretical conceptualization of intention, which per definition is connected with short-term actions (Perugini & Bagozzi, 2004b). In terms of future research, we suggest that measurements of (1) desire should not be timeconstrained or at least not limited to short periods of time, and (2) intention should be time-constrained to relatively short timeframes. Thus, although this issue depends on the behavior in question, the idea is that measuring intention over a period of, say, 2 years would be as useless as measuring desire for, say, 1 month.

Limitations

Some limitations of the present study should be mentioned.

First, some of the measures employed only two items; thus, the study would benefit by including more items in order to reduce possible measurement errors. Second, although measures of behavioral intentions are fairly standardized in TPB research, this can hardly be the case when it comes to measuring desire. Hence, in order to promote the argument that these concepts are distinct, one should first develop a reliable standardized measure of desire, subsequently presenting evidence of discriminating validity between the two measurements (see Perugini & Bagozzi, 2004a). Third, although we argue that desire is conceptually distinct from behavioral intention, the nature of the behavior in question might be of importance. Thus, the distinction between desire and intention might be especially pronounced when it comes to requiring personal projects in terms of self-regulation. In this context, our results might be inflated due to the fact that quitting smoking is a difficult goal to achieve for which the discrepancies between “I want,” “I should,” “I ought,” and “I can” are greater than for behaviors in which wanting and intending have a similar direction (e.g., planning a vacation). Fourth, the samples in the present study were not specifically selected for testing the two study hypotheses. The 1- and 6-month samples were taken from the student population whereas the 4-month sample was from general population (see Table 2). Participants in the 4-month sample had significantly different values regarding common background variables. However, although these differences might have confounded the results, the general pattern supporting the basic reasoning behind the hypotheses is clearly detectable. Finally, there was the unexpected negative effect of PBC on intention in the 6-month sample. This unusual result might have confounded the results as well as represent measurement error. Notwithstanding these limitations, the results of the present study are important in terms of providing support for empirical and theoretical distinctions between the key concepts involved in the decision-making process. This finding also suggests that future research should pay closer attention to the processes and conditions under which desires are successfully translated into behavioral intentions, taking into consideration that intention is generally accepted to be the best predictor of actual performance (Ajzen, 1991). However, our findings also emphasize the importance of developing the theoretically grounded and reliable measurement of desire, which represents a central startingpoint for the inclusion of this concept into decision-making models.

Conclusions

Over the years, considerable empirical evidence has accumulated to support the inclusion of various theoretical constructs to increase the proportion of explained variance in behavioral intention of the TPB (Conner & Sparks, 2005).

Following this line of research, the present results may have important theoretical implications for the inclusion of desire as an additional predictor in the TPB. Although future research needs to confirm the present results before more definite conclusions can be made, these results clearly show that desire (1) has a significant and direct predictive effect on the intention to quit smoking, and (2) represents an important mediator between attitudes, SN, and past behavior, on the one hand, and behavioral intentions, on the other.

References

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

Ajzen, I. (2002a). Perceived behavioral control, self-efficacy, locus of control, and the theory of planned behavior. Journal of Applied and Social Psychology, 32, 1–20.

Ajzen, I. (2002b). Residual effects of past on later behavior: Habituation and reasoned action perspectives. Personality and Social Psychology Review, 6,107–122.

Armitage, C. J., & Conner, M. (2001). Efficacy of the theory of planned behavior: A meta-analytic review. British Journal of Social Psychology, 40, 471–499.

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

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

Bagozzi, R. P., Dholakia, U. M., & Basuroy, S. (2003). How effortful decisions get enacted: The motivating role of decision processes, desires, and anticipated emotions. Journal of Behavioral Decision Making, 16, 273–295.

Baron, R. M., & Kenny D. A. (1986). The moderator-mediator variable distinction in social psychological research: Conceptual, strategic and statistical considerations.Journal of Personality and Social Psychology, 51, 1173–1182.

Conner, M., & Armitage, C. J. (1998). Extending the theory of planned behavior: A review and avenues for further research. Journal of Applied Social psychology, 28, 1429–1464.

Conner, M., & Sparks, P. (2005). The theory of planned behavior and health behaviors. In M. Conner & P. Norman (Eds.),Predicting health behavior (pp. 171–222). Buckingham, UK: Open University Press.

Davis, W. A. (1984). The two senses of desire.Philosophical Studies, 45, 181–195.

Falomir, J. M., & Invernizzi, F. (1999). The role of social influence and smoker identity in resistance to smoking cessation. Swiss Journal of Psychology, 58, 73–84.

Fishbein, M., & Stasson, M. (1990). The role of desires, self-predictions, and perceived control in the prediction of training session attendance.Journal of Applied Social Psychology, 20, 173–198.

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.

Kova2, V. B., Rise, J., & Moan, I. S. (2010). From intentions to quit to actual quitting process: The case of smoking behavior in the light of TPB. Journal of Applied Biobehavioral Research, 14, 181–197.

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

Leone, L., Perugini, M., & Ercolani, A. P. (2004). Studying, practicing, and mastering: A test of the model of goal-directed behavior (MGB) in the software learning domain.Journal of Applied Social Psychology, 34, 1945–1973.

Malle, B. G., & Knobe, J. (2001). The distinction between desire and intention: A folk conceptual analysis. In B. G. Malle, L. J. Moses, & D. A. Baldwin (Eds.),Intentions and intentionality: Foundations of social cognition(pp. 45–67). Cambridge, MA: MIT.

Moan, I. S., & Rise, J. (2005). Quitting smoking: Applying an extended version of the theory of planned behavior in predicting intention and behavior. Journal of Applied Biobehavioral Research, 10,39–68.

Norman, P., Conner, M., & Bell, R. (1999). The theory of planned behavior and smoking cessation.Health Psychology, 18, 89–94.

O’Keefe, D. J. (2002).Persuasion: Theory and research. Thousand Oaks, CA: Sage.

Ouelette, 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.

Perugini, M., & Bagozzi, R. P. (2001). The role of desires and anticipated emotions in goal-directed behaviors: Broadening and deepening the theory of planned behavior.British Journal of Social Psychology, 40, 79–98.

Perugini, M., & Bagozzi, R. P. (2004a). An alternative view of prevolitional processes in decision making: Conceptual issues and empirical evidence. In G. Haddock & G. R. Maio (Eds.), Contemporary perspectives on the psychology of attitudes: The Cardiff symposium(pp. 169–201). Hove, UK: Psychology Press.

Perugini, M., & Bagozzi, R. P. (2004b). The distinction between desires and intentions.European Journal of Social Psychology, 34, 69–84.

Perugini, M., & Conner, M. (2000). Predicting and understanding behavioral volitions: The interplay between goals and behaviors.European Journal of Social Psychology, 30, 705–731.

Rise, J., & Ommundsen, R. (2011). Predicting the intention to quit smoking: A comparative study among Spanish and Norwegian students.Europe’s Journal of Psychology, 7,143–163.

Rise, J., Kova2, V. B., Kraft, P., & Moan, I. S. (2008). Predicting the intention to quit smoking and quitting behavior: Extending the theory of planned behavior.British Journal of Health Psychology, 13, 291–310.

Rise, J., Sheeran, P., & Hukkelberg, S. (2010). The role of selfidentity in the theory of planned behavior: A meta-analysis. Journal of Applied Social Psychology, 40, 1085–1105.

Rivis, A. J., & Sheeran, P. (2004). Descriptive norms as an additional predictor in the theory of planned behavior: A metaanalysis.Current Psychology, 22, 264–280.

Sandberg, T., & Conner, M. (2008). Anticipated regret as an additional predictor in the theory of planned behavior: A metaanalysis.British Journal of Social Psychology, 47, 589–606.

Shaw, D., Shiu, E., Hassan, L., Bekin, C., & Hogg, G. (2007). Intending to be ethical: An examination of consumer choice in sweatshop avoidance.Advances in Consumer Research, 34, 31–38.

Sheeran, P. (2005). Intention-behavior relations: A conceptual and empirical review. In W. Stroebe & M. Hewstone (Eds.), European Review of Social Psychology (Vol. 12, pp. 1–36). Chichester, UK: Wiley.

Taylor, S. D., Bagozzi, R., & Gaither, C. A. (2005). Decision making and effort in the self-regulation of hypertension: Testing two competing theories.British Journal of Health Psychology, 10,505–530.

Velibor B. Kovač

University of Agder Department of Education PO Box 422 4604 Kristiansand Norway bobo.kovac@uia.no