Understanding How Individuals Make Travel and Location Decisions: Implications for Public Transportation (2008)

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

29 One of the objectives of this research project is to explore the TPB as an approach to understanding how individuals make travel and location decisions. This chapter presents key background information from the field of psychology. Liter- ature on TPB, which includes a collection of theories of be- havior, is summarized. After a discussion of those theories, the application of the TPB in transportation is reported. Some of the relevant studies exploring how habit and environmental values influence behavior are described, as are ways of over- coming habit in trying to bring about social change. Literature on the Theory of Planned Behavior An excellent summary of the development and use of the TPB is provided in an article by Icek Aizen in Organizational Behavior and Human Decision Processes (29). The article cov- ers some of the background research behind the TPB, as well as analysis techniques. The article starts by acknowledging the low empirical relationships between personality traits and behavior. Although relationships can be improved by aggre- gating multiple instances of behavior so that random influ- ences specific to a particular occasion can be canceled out, a model that explains behavior at the more disaggregate level would be desirable. The TPB is suggested as such a model for explaining behavior at a more disaggregate level. The TPB grew out of the theory of reasoned action (30, 31), which holds that behavior is the direct result of intent, and that intent is influenced by a person’s ATT and the SN. Because of problems predicting behavior with intent alone, Aizen added PBC as a predictor. Performing a behavior may depend on having requisite opportunities and resources that enable the performance. PBC, as defined by Aizen, is similar to the concept of self-efficacy developed by Albert Bandura (32, 33), the originator of social learning theory. Bandura (34) found that an individual’s behavior is strongly influ- enced by his or her confidence that he or she can perform the behavior. Self-efficacy beliefs influence behavior by influ- encing the choice of activities, preparation for an activity, ef- fort expended, thought patterns, and emotional reactions (29). In general, if the behaviors being investigated pose no seri- ous problems of PBC, there will be a strong relationship between intent and behavior. Aizen illustrates this with a series of 17 studies using the TPB (29). For each of the stud- ies, he shows the results of regression analyses, with behavior as the dependent variable and with intent and PBC as inde- pendent variables. There is a significant coefficient for intent in the prediction of behavior in 15 of the 17 situations. PBC, however, also adds to the prediction of behavior, with 11 of the 17 analyses having significant coefficients for PBC. In most of these studies, the coefficients for intent were greater than the coefficients for PBC. If there is a problem of behav- ioral control, however, intent may not have a strong rela- tionship to behavior. This was the case in two studies on weight loss, where only the PBC was significant. The theory holds that PBC also contributes to intent, as do ATT and SN. Aizen uses a set of studies to illustrate the rela- tionship between ATT, SN, PBC, and intent (29). A consider- able amount of variance in intent is accounted for by the three predictors in the TPB. The coefficients of ATT were significant in 15 of 16 cases, the coefficients of SN were significant in 10 of 16 cases, and the coefficients of PBC were significant in all cases. On the basis of consistent evidence linking ATT and PBC to intent, Aizen concluded that personal factors (ATT and PBC) are more influential in the prediction of behavioral outcomes than are social (or normative) factors (SN). Aizen also discusses attitude formation in the TPB model, including the use of the expectancy-value model of attitudes (30). The expectancy-value model says that, for example, ATT can be indirectly measured by summing the product of belief measures times measures of the belief’s relevance. While results of numerous studies support the expectancy- value model, the magnitude of the relationship between indi- C H A P T E R 3 Background to the TPB and Its Application in Transportation

rect and direct measures of constructs like ATT, SN, and PBC has been only moderate (29). Armitage and Conner (35) provide a metareview of the many research papers that used the TPB. The Theory of Planned Behaviour (TPB) has received consid- erable attention in the literature. The present study is a quantita- tive integration and review of that research. From a database of 185 independent studies published up to the end of 1997, the TPB accounted for 27% and 39% of the variance in behaviour and intention, respectively. The perceived behavioral control con- struct accounted for significant amounts of variance in intention and behaviour, independent of theory of reasoned action vari- ables . . . Attitude, Subjective Norm and [Perceived Behavioral Control] account for significantly more of the variance in indi- viduals’ desires than intentions or self-predictions, but intentions and self-predictions were better predictors of behaviour. The Subjective Norm construct is generally found to be a weak pre- dictor of intentions. This is partly attributable to a combination of poor measurement and the need for expansion of the norma- tive component. (p. 471) The TPB has had broad application in the health field, and more recently in transportation. The breadth of applications of the TPB in health had been described in several articles, including “The Theory of Planned Behavior: A Review of Its Applications to Health-Related Behaviors,” by Godin and Kok (36), whose purpose was “to review applications of Ajzen’s theory of planned behavior in the domain of health and to verify the efficiency of the theory to explain and pre- dict health-related behaviors” (p. 87). The findings of the study included the following: The results indicated that the theory performs well for the explanation of intention; an averaged R2 of .41 was observed. Attitude toward the action and Perceived Behavioral Control were most often the significant variables responsible for this explained variation in intention. The prediction of behavior yielded an averaged R2 of .34. Intention remained the most important predictor, but in half of the studies reviewed Perceived Behavioral Control significantly added to the prediction. (p. 87) Godin and Kok conclude that “the efficiency of the model seems to be quite good for explaining intention, Perceived Behavioral Control being as important as attitude across health- related behavior categories. The efficiency of the theory, how- ever, varies between health-related behavior categories” (p. 87). The Application of the TPB to Transportation The TPB has been applied directly to the issue of mode choice in several studies. The European Union’s ADONIS (Analysis and Development of New Insight into Substitution of Short Car Trips by Cycling and Walking) project applied the theory to the modal choice in short-distance trips in Scandinavia. Bamberg et al. (37) applied the theory to the change in bus ridership in northern Germany as a result of a change in the fare collection method. An issue in both of these studies was the importance of habit in transportation mode choice. This issue is described more fully after descriptions of the two projects. The ADONIS Project The ADONIS Project is described in a report titled A Review of the Effectiveness of Personalized Journey Planning Techniques (38). The report reviews various learning models and notes the extent of application of Aizen’s TPB. The report summarizes the application of the work of Aizen in a survey process undertaken in Scandinavia, as follows: [Aizen’s theory] has recently been used extensively in travel behaviour change analysis (notably in the ADONIS project, For- ward et al., 1998), to explain the likelihood of behavioral change in different circumstances. The theory (through successive adap- tations) currently posits that the intention to change behaviour is related to: • the attitude the person has to the change; • what the person feels others will feel about them if they change; • the extent to which the person feels they are able to change; and • the depth of habit that the person has relating to current behavioral patterns. (paragraph 2.16) The ADONIS studies are important to this project because of their direct application of psychological theories of attitude formation in a planned intervention to alter travel behavior, in this case concerning the short-distance trip. The psychol- ogist who undertook the study, Sonja Forward of the Swedish National Road and Transport Research Institute, described the project as follows: This study analyzed short journeys on foot, cycle and car with the aid of a travel diary and an attitude survey . . . The attitude survey was designed in accordance with an expanded version of the Theory of Planned Behaviour, which included attitudes, sub- jective norm, perceived behavioral control and habit. (39) The ADONIS questionnaire was administered by phone, followed by a second wave which rated “a short imaginary journey.” Based on an analysis of the surveys and the diaries, Forward concluded that the factor of habit was the most pow- erful explanatory variable in understanding the rational for mode change, or the lack of mode change, and that the con- cept of self-efficacy (labeled perceived behavioral control in the TPB) was highly explanatory in interpreting the results. The variables with the highest explanatory value were per- ceived behavioral control and habit. Since perceived behavioral control describes the subjective opinion of a person’s own resources ability, it may be concluded that non-users experience 30

more obstacles than others do. . . . [T]hus, we were able to find that the expanded version of the Theory of Planned Behaviour can advantageously be used in the evaluation of different proj- ects and that it helps to increase our understanding of the best way of motivating road users to select more environmentally friendly modes of transport. (39) The Bamberg/Aizen/Schmidt Study of Mode Change The role of habit in predicting mode change was explored in some depth by Bamberg et al. (37) in an article titled “Choice of Travel Mode in the Theory of Planned Behavior: The Roles of Past Behavior, Habit, and Reasoned Action.” The authors undertook a longitudinal study of attitudes of students before and after the implementation of a prepaid bus pass for all students. Relying on the theory of planned behavior (Aizen, 1991), a longitudinal study investigated the effects of an intervention— introduction of a pre-paid bus ticket—on increased bus use among college students . . . The intervention was found to influ- ence attitudes toward bus use, Subjective Norms, and perceptions of behavioral control and, consistent with the theory, to affect intentions and behavior in the desired direction. Furthermore, the theory afforded accurate prediction of intention and behav- ior both before and after the intervention. (p. 175) The authors found that while habit (past use of a mode) was a significant predictor of mode choice prior to the introduc- tion of a prepaid bus pass, the introduction of a prepaid bus ticket was sufficient to “break the habit” and allow students to reassess their mode choice. That is, habit was not a significant predictor of mode choice following the introduction of the prepaid bus pass. It is concluded that choice of travel mode is largely a reasoned decision; that this decision can be affected by interventions that produce change in attitudes, subjective norms, and perceptions of behavioral control; and that past travel choice contributes to the prediction of later behavior only if circumstances remain relatively stable. (p. 175) The authors also found that the incorporation of a meas- ure of self-efficacy, which they refer to as perceived behavioral control, helped to provide explanatory power in the study of the prepaid bus ticket. As they approached their examination of change in bus ridership, they posited that the TPB could be extended to this transportation issue. The theory of planned behavior has received good empirical sup- port in applications to a wide variety of different domains. . . . However, the study reported in the present article is one of the few attempts to use the theory as a conceptual framework for an inter- vention to effect change in behavior . . . According to the theory, it should be possible to influence intentions and behavior by designing an intervention that has significant effects on one or more of the antecedent factors, i.e., on attitudes toward the behavior, subjective norms, and perceptions of behavioral control. (p. 176) Importantly, the Bamberg et al. article concluded that the theory did indeed help to understand the behavioral implica- tions of the change in attitudes. The results of the present investigation demonstrate the util- ity of the theory of planned behavior as a conceptual framework for the prediction of travel mode choice and for understanding the effects of an intervention on this behavior. Attitude, subjec- tive norm, and perceived behavioral control were found to influence students’ intentions to take the bus to the campus, and these intentions in turn permitted quite accurate prediction of reported behavior. (p. 184) The Role of Routine “Habit” in Transportation A major theme being addressed in the above studies, as well as in others, is the power of habit. To what extent is behavior influenced by reception of new information and new environments, as opposed to the rote repetition of rou- tines that have become habit? Cognitive experts within social psychology have differing viewpoints about the role of habit. In the study of bus use among university students reported above, Bamberg et al. (37) found that choice of travel mode is based more on reason than on habit: Only when circumstances remain relatively stable does prior behavior make a significant contribution to the prediction of later action. Complex human behavior is cognitively regulated and, even after numerous enactments, appears to be subject to at least some degree of monitoring. As a result, new information, if relevant and persuasive, can change behavioral, normative and control beliefs; can affect intentions and perceptions of behav- ioral control; and can influence later behavior. We thus conclude that human social behavior, although it may well contain auto- matic elements, is based on reason. (p. 186) Others, however, emphasize the difficulty of altering behavior away from established routinized behavior, such as the dependence on the automobile for all tripmaking. The question has been explored in depth by three European cog- nitive theorists, Aarts, Verplanken and van Knippenberg, whose article “Habit and Information Use in Travel Mode Choice” is widely referenced in reports about the difficulty of decreasing the use of the automobile (40). Their article . . . focuses on travel mode choice behavior in order to test the- oretical propositions as to habitual decision making. In particular, we are interested in the role of habit in information processing un- derlying daily travel mode choices. Like many behaviors routinely performed in every day life, travel mode decisions are supposed to 31

be often made in a rather ‘mindless’, automatic fashion. . . . In other words, travel behavior is often habitual. (p. 2) The role of habit in mode choice addressed in this article was summed up by Gärling, Gärling, and Loukopoulos in the article titled “Forecasting Psychological Consequences of Car Use Reduction: A Challenge to an Environmental Psychology of Transportation” (41). They describe the effect of habit on mode choice as follows: The frequent use of cars can be partly attributable to the way in which attitudes, beliefs, and choices work together. Work by Gärling, Fujii, and Boe (2001) and by Verplanken, Aarts, and van Knippenberg (1994) has shown that attitudes or preferences guide initial deliberate choices of car for the majority of a person’s activi- ties, but that eventually these choices become a car habit which is difficult to alter. That is, positive attitudes toward driving lead to frequent choices to drive that, in turn, lead to automatised driv- ing choice. Indeed, depending on the type of reduction required, habitual trips may not be reduced at all. Gärling, Gillholm, and A. Gärling (1998) claimed that both planned and habitual trips are equally easy or difficult to reduce in a planning phase, but that such changes in the case of habitual travel would be harder to implement. (p. 97) In an article titled “Habit versus Planned Behaviour: a Field Experiment,” Verplanken et al. (42) concluded that the strength of a habit had a powerful impact on the outcomes that would have been predicted by the cognitive models. Car use during seven days was predicted from habit strength . . . and antecedents of behaviour as conceptualized in the theory of planned behaviour (attitude, subjective norm, perceived behavioral control and behavioral intention). Both habit meas- ures predicted behaviour in addition to intention and perceived control. Significant habit x intention interactions indicated that intentions were only significantly related to behaviour when habit was weak, whereas no intention-behaviour relation existed when habit was strong. . . . The results demonstrate that, although external incentives may increase the enactment of intentions, habits set boundary conditions for the applicability of the theory of planned behaviour. (p. 111) A Swiss researcher, Sylvia Harms, has examined the ten- sion between those who look at transportation decisions as a cognitive activity and those who see it as the result of a rote activity, dominated by habit. In an article titled “From Routine Choice to Rational Decision Making Between Mobility Alternatives,” Harms (43) concludes that the TPB is not inconsistent with the incorporation of acts that are seemingly driven by habit. In a series of studies concerning the relationship between habit and rational decision mak- ing, Harms has placed the TPB into a larger context of understanding the propensity to change one’s transporta- tion (here called mobility). In her model, an individual’s own life situation influences mobility requirements and opportunities, and these influence attitudes and perceived behavioral control. The quantitative studies confirmed that people are more vulnerable to new transportation solutions at a time when their personal lifestyle is changing. The study found that habit is indeed the weakest when people’s behavioral context has recently changed. When the lifestyle context remains stable, the force of habit is stronger. How- ever, during periods of situational change, the influence of attitude and perceived behavioral control grows in relation to the influence of habit. The quantitative finding was consistent with earlier obser- vations about the personal context of individuals who had selected to join car-sharing groups. Harms documents that many who changed their transportation behavior did so because of a change in their personal situation, not in response to some new information about the alternative. In an obser- vation that could have significant implications for this project, Harms noted the following: [A]bout 85% of those people who owned a private car before becoming a car-sharing member reported on significant changes in their personal life situation when being asked about their motivation to join a car-sharing organisation. Only in the second place, the attractiveness of certain product attributes like envi- ronmental friendliness or low car-use costs were mentioned. The reported changes referred to a new working place, moving the own house, the breakdown of the own car or other things that significantly influenced the private mobility context and the availability and/or usefulness of an own car. (p. 7) Harms’ conclusions could be relevant to the selection of key market segments for this study. If routines indeed impose cognitive barriers to information perception and attitude formation . . . marketing efforts for innovative mobility concepts should be adjusted to this phe- nomenon: They should be bundled in moments where routines are the weakest and people are most open to conscious, rational decision-making, i.e., in moments of important context changes (e.g. moving, changing the job). (p. 25) As the result of the quantitative research to confirm (or dis- prove) earlier hypotheses concerning the dominance of the force of habit, Harms concluded that the general structure of the TPB was not inconsistent with the implications of a seri- ous role for habit. At the same time, Harms points out how the subject becomes more vulnerable to incoming informa- tion when the “behavioral context” is upset or changed. [But] under changed context conditions this shortcut doesn’t work anymore and the earlier cognitive elements are consciously activated again and adapted to the new situation. . . . Even rational decision-making approaches like the theory of planned behaviour allow attitudes and control beliefs to be retrieved from memory, without being consciously constructed again each time a similar decision is made. (p. 9) 32

Environmental Values in the Context of the TPB A question for this project is whether environmental val- ues have an impact on intentions and behavior related to choosing a CN or choosing to walk or take public trans- portation. The evidence for relationships among components of the TPB in the context of environmental behavior is pro- vided in a comprehensive review by Kaiser et al. (44). It was found that if only the relationship between environmental at- titude and behavior was examined, then the “relationships appear to be at best moderate across different studies.” The literature also indicates that the relationship between values and intention ranges from weak to strong, and that if it is be- tween values and behavior the relationship is less strong. Kaiser et al. found that “the most striking effect” is between intention and behavior; “ecological behavior intention is strongly related to ecological behavior or at worst moderately related.” They note, however, that the strength of the rela- tionship may vary in different environmental behavior con- texts. The authors conducted a survey of members of two Swiss transportation organizations with different ideologies. They found that “environmental knowledge and environ- mental values explained 40% of the variance of ecological behavior intentions which, in turn, predicted 75% of the vari- ance of general ecological behavior.” Swensen and Wells (45) reviewed the literature on the relationships between demographic characteristics, personal- ity traits, environmental attitudes, and environmental behavior. They reported that past studies indicate that “demographic and personality characteristics correlated with pro-environmental attitudes in one investigation failed to correlate with pro-environmental attitudes in others” and that “attitudes that predicted pro-environmental behavior in one study failed in replications.” (p. 91) Swensen and Wells conduct their own analysis using data from national consumer surveys from the early 1990s. The results indicate that “pro-environmental behavior is correlated with some major demographic variables (education, income, and community size) and with concern for the environment, cosmopolitanism, liberalism, frugality, planfulness, community involvement, health concerns, perceptions of financial distress, and dissatisfaction with life” (p. 91). They conclude that their results, “without negating the value of aspect-specific investiga- tions,” show that “the general concept of pro-environmental behavior is strong enough and consistent enough to provide valuable guidance to theoretical and practical work” (p. 104). Conclusions from the Literature on the TPB The extensive use of the TPB as a model for understanding behavior in the health field, plus many examples in the trans- portation field, indicated that it will be a worthy tool for exploration in this project. Some key lessons from the litera- ture review include the following: • If intent and perceived behavioral control (self-confidence) can be changed, it is likely that behavior can also be changed. • The opportunity for mode change increases when other lifestyle changes are occurring, such as a change in job or residence. • Although mode choice is often habitual, interventions can succeed in changing mode choice. However, the habit of driving is difficult to break. • Although the relationship between environmental values and behavior varies, it will be worthwhile to measure environmental values and their relationship to mode and location decisions. 33