رضایت با سفر و بهزیستن ذهنی: توسعه و آزمون یک ابزار اندازه گیری
|کد مقاله||سال انتشار||مقاله انگلیسی||ترجمه فارسی||تعداد کلمات|
|37992||2011||9 صفحه PDF||سفارش دهید||6528 کلمه|
Publisher : Elsevier - Science Direct (الزویر - ساینس دایرکت)
Journal : Transportation Research Part F: Traffic Psychology and Behaviour, Volume 14, Issue 3, May 2011, Pages 167–175
Abstract Subjective well-being (SWB) that includes individuals’ cognitive and affective evaluations of life in general is proposed to be a more appropriate measure capturing the benefits individuals derive from travel improvements. We develop and test a measure of travel-related SWB, the nine item self-report satisfaction with travel scale (STS). In a survey of 155 undergraduates, STS, mood ratings, and ratings of SWB were collected for three hypothetical weekdays differing in travel mode, travel time, access to bus stops, and daily activity agenda. The results showed that STS is reliable and differentiates between changes in travel conditions. STS, mood, and to some extent SWB were shown to be affected by travel mode (bus vs. car), travel time, access to bus stops, and the number of activities in the daily agenda.
. Introduction Car-use reduction is on the political agenda in many countries due to the accelerating adverse effects of motorized traffic on human environments and public health (e.g., Hensher, 1998, Vlek, 2007 and OECD, 2008). Successful car-use reduction depends on individual changes of travel, in particular that car users switch to more environmentally friendly travel modes. Over the past years various studies have investigated how car users’ mode choice can be influenced. Many of these studies have focused on the effects of soft transport policy measures (Richter et al., in press and Richter et al., 2010), also referred to as travel feedback programs (Fujii and Kitamura, 2003, Fujii and Taniguchi, 2006 and Taniguchi and Fujii, 2007), which aim at making car users voluntarily switch travel mode by providing customized information about sustainable alternatives, primarily public transport services. Yet, the effectiveness of soft transport policies to influence car use is limited if the existing level-of-service of public transport cannot compete with the car (Beale and Bonsall, 2007 and Brög et al., 2009). Therefore, improvements of public transport services are needed in order to achieve a more substantial car-use reduction. This raises the question of what improvements of public transport services are attractive to car users such that they switch to public transport. Various previous studies provide some information in this respect. In Eriksson, Friman, and Gärling (2008) interviews with employees who commuted by car to work showed that the further to work or bus stops, the more the participants desired increased frequencies and shorter travel times but less often lower fares. In another study Eriksson, Friman, Ettema, Fujii, and Gärling (2010; see also Horeni, Gärling, Loukopoulos, & Fujii, 2007) employed an experimental simulation to further investigate car users’ switching to public transport. The main results showed that shorter travel times as well as high access to bus stops led to greater bus use. However, constraints imposed by a busy daily activity agenda led to larger car-use, in particular if car costs were low. The present study is a continuation of this study focusing on how switching from car to public transport affects satisfaction with travel and as a consequence satisfaction with the daily routines. The focus on satisfaction with travel and daily routines is relevant both from a viewpoint of the implications of policies for well-being, but also since travel that is experienced as more satisfactory is more likely to be sustained over a longer period. Currently, utility theory is the dominant account of how travel-related choices of activity, destination, and travel mode are made (McFadden, 2001). Assessments of people’s satisfaction with their travel are similarly based on this theory (e.g. in cost-benefit analyses). Utility measures used for the appraisal of policies, or for investigating the consequences of people’s travel choices in general, are thus derived from observed choices. Yet, it is doubtful whether utility derived in this way is a valid measure of travellers’ satisfaction, since choices are frequently made under constraints and lack of complete information. The current application of utility theory in cost-benefit analysis is based on the utility derived from observed choices (termed decision utility). However, it has been shown ( Wilson & Gilbert, 2003) that anticipated and actual experiences may differ considerably. Specifically, the intensity of both positive and negative experiences is usually underestimated. As a consequence, to properly assess the benefits of travel improvements, one should measure experienced utility rather than rely on decision utility (see Ettema, Gärling, Olsson, & Friman, 2010, for a more elaborate discussion). As an alternative to experienced utility, subjective well-being (SWB) has been proposed as a measure of people’s satisfaction with outcomes of choices (Kahneman, 1999). SWB encompasses judgments of availability of material and immaterial resources that are important for being satisfied with one’s life as well as enduring feelings of pleasantness (e.g. Diener and Biswas-Diener, 2008 and Diener and Suh, 1997). Since SWB refers to satisfaction with life in general, it is assumed to be relatively stable across time. Yet, there is still an interest in understanding how changes in SWB depend on context-specific factors, including various forms of consumption (Diener & Seligman, 2004). An important research question raised by Ettema et al. (2010) and Jakobsson Bergstad et al. (2009c) is whether and how changes in travel context (e.g. switching travel mode or improved level-of-service of public transport) will cause changes in SWB. If it is possible to find a relationship with travel context, SWB has the potential of becoming a new powerful tool for policy evaluations (Diener, 2006). Diener, Emmons, Larsen, and Griffen (1985) posited that SWB consists of three components: (1) intensity, frequency, and duration of positive affect (PA), (2) intensity, frequency, and duration of negative affect (NA), and (3) a cognitive judgment of satisfaction with life as a whole (see also Arhaud-Day, Rode, Mooney, & Hear, 2005). The affective components may be assessed by self-reports of specific immediate emotions or moods (Stone, Shiffman, & DeVries, 1999). In such self-reports, participants report their affective experience during an episode or activity. An alternative is to obtain retrospective self-reports on rating scales such as the Positive and Negative Affect Scale (PANAS, Watson, Clark, & Tellegen, 1988) or the Swedish Core Affect Scale (SCAS, Västfjäll et al., 2002 and Västfjäll and Gärling, 2007) derived from the affect circumplex defined by a valence (pleasantness–unpleasantness) and activation (activation–deactivation) dimension ( Russell, 1980 and Russell, 2003; see also Diener & Emmons, 1984). While retrospective self-reports refer to how people feel in general or have felt during a past period (last day, week, month, or year), the same scales may be used to measure current mood, that is how people feel at the moment. Immediate approaches measure direct affective responses, retrospective approaches measure the memory of such responses, which may be biased due to memory distortions ( Kahneman, 2000). Still, Kahneman, Krueger, Schkade, Schwarz, and Stone (2004) found high correlations between immediate and retrospective measurements. Cognitive SWB may be assessed by means of the five-item Satisfaction with Life Scale (SWLS) ( Diener et al., 1985 and Pavot and Diener, 1993) or a single-item question ( Eurobarometer, 2008 and World Values Survey, 2005). At an individual level, both the affective and cognitive components of SWB are partly explained by stable, genetically influenced personality traits (Tkash & Lyubomirsy, 2006). It is estimated that about 50% of the variance in SWB is accounted for in this way (Lyubomirsky, King, & Diener, 2005). Specifically, extraversion and low emotional arousability lead to higher SWB than do introversion and high emotional arousability (Weiss, Bates, & Luciano, 2008). Other research has demonstrated that SWB depends on life circumstances, explaining about 10% of the variance (Lyubomirsky et al., 2005). People with a higher income have higher SWB (e.g., Clark and Oswald, 1996 and Ferrer-i-Carbonell, 2005). SWB usually has a U-shaped relationship to age (Diener & Suh, 1997), being at its lowest when people are around the age of 40, and then gradually increasing. Marriage tends to increase, divorce or death of spouse to decrease SWB (Diener, Suh, Lucas, & Smith, 1999). Education increases SWB, although more strongly in less wealthy countries. Unemployment reduces SWB, in particular if there is little social support (see Argyle, 1999). The results for sex are mixed, some studies showing that women have higher SWB than men, others that there are no differences, and still others that sex differences vary across the life course (Tesch-Römer, Motel-Klingebiel, & Tomasik, 2008). Previous research also provides evidence for the impact of activity performance on SWB. Pychyl and Little (1998) proposed that SWB significantly depends on progress towards life goals, and that activities organised in projects are instrumental for such progress. Oishi, Diener, Lucas, and Suh (1999) showed that to some extent daily life satisfaction is explained by the types of activities that people perform. Gadermann and Zumbo (2007) found that daily hassles increase negative mood and have a negative impact on SWB. Further support for a negative impact on SWB was provided by van Emmerik and Jawahar (2006) who found that work-related stress associated with time pressure negatively affects current mood both in the work and personal domains. McLeod, Coates, and Hetherton (2008) found that progress towards goals has a positive effect on SWB, and that if this progress is speeded up by improved planning skills, the positive effect increases. Satisfaction may also be assessed for life domains including work, family, and leisure. Schimmack (2008) reviews research showing that SWB is influenced by such assessments of domain satisfaction. A form of domain-specific SWB is customer satisfaction (Oliver, 2010), usually defined as the extent to which goods or services fulfil customers’ specific needs. Customer satisfaction with transport services may similarly be defined as the extent to which the services fulfil travel needs. Customer satisfaction is less general than many other life domains and only applies to those using the service. Similar to SWB, customer satisfaction include both a cognitive and an affective component. Travellers’ satisfaction is influenced by events experienced when travelling. Friman et al., 1998 and Friman and Gärling, 2001, and Friman, Edvardsson, and Gärling (2001) demonstrated that both single critical incidents (events deviating from users´ expectations) and memory for their frequencies affect satisfaction with public transport services. Friman (2004) also found that critical incidents elicited affective responses (described as changes in valence and activation). On the basis of critical incidents frequently encountered during repeated travel, customers develop an accumulated satisfaction with the transport service. In a similar vein, research has recognised that travellers value travel per se (Mokhtarian and Salomon, 2001 and Mokhtarian et al., 2001). For instance, Jakobsson Bergstad et al. (2009a) and Steg (2005) investigated psychological reasons for car use including emotions evoked by driving a car (e.g. pleasure-to-use). Thus, driving affects people’s mood and for some people explains why the car is perceived to be attractive. It has also been found that symbolic (self-presentation) aspects significantly contribute to the positive utility of driving (Mokhtarian & Salomon, 2001). Stradling, Anable, and Carreno (2007) demonstrated that satisfaction with bus services depends on a variety of non-instrumental factors including cleanliness, privacy, safety, convenience, stress, social interaction, and scenery. They furthermore found that pedestrians evaluated their walking trips in terms of non-instrumental factors, such as crowdedness, air quality, presence of trees and flowers, presence of beggars, and type of pavement. Taken together, it is suggested that, independently of travel mode, a trip will result in a variety of affective responses, which will eventually impact on satisfaction with travel. Customer satisfaction also has a cognitive component related to cost, travel time, and punctuality. Travellers will likewise develop a cumulative satisfaction with these aspects based on their repeated daily encounters with the service. Previous research thus suggests that daily travel is likely to affect individuals’ mood and lead to a cumulative satisfaction. As a consequence, changes in public transport services, either at a functional or affective level, may influence SWB in the domain of travel, that is the affective appreciation as well as cognitive evaluation of travel. Hence, measuring travel-specific SWB may yield an indication of the effect of travel on people’s SWB which is useful information when evaluating different transport policies. Measurement of SWB in the context of travel has so far been limited to the study by Jakobsson Bergstad et al. (2009c) proposing a five-item scale (satisfaction with travel scale or STS) to measure travel-specific SWB. The items include four cognitive evaluations and a general affective item to measure how good travel made the respondents feel. This existing STS scale worked well to investigate the relationships between general subjective well-being and domain specific well-being in the context of activities and travel. However, since it consists mainly of cognitive items, it is limited in measuring the affective components of travel. This paper proposes an improved STS that extends the existing STS especially in the affective domain. Specifically, it combines cognitive judgments of travel satisfaction with measures of the activation and valence dimensions of mood. As such, it is consistent with measurement of affective well-being according to the Swedish core affect scale (SCAS) (Västfjäll et al., 2002). The improved STS is tested in a survey in which respondents evaluate their travel during hypothetical days. In addition to a direct effect on SWB, improved travel conditions will possibly lead to a higher SWB because of increased access to the daily activities that facilitate progress to life goals (White & Dolan, 2009). This hypothesis proposed by Ettema et al. (2010) was empirically supported by a survey conducted by Jakobsson Bergstad et al., 2009b and Jakobsson Bergstad et al., 2010. Thus, the impact of travel improvements on SWB may materialize via a more pleasant (or less displeasant) experience of travel (as measured by STS) and via improved access to activities that add to SWB. Assessing the relative role of each is however beyond the scope of the present study. Here the focus is on the effect of travel improvements on STS and how mood and satisfaction are influenced by STS. A survey is undertaken describing to participants three hypothetical days entailing travel to and from work and asking them to report for each how they would evaluate the day. Measures of STS, affective SWB and cognitive SWB are made in order to investigate the relationships between these variables. The paper is organised as follows. Section 2 presents the employed measures, including the improved satisfaction with travel scale (STS). Section 3 outlines the survey that was undertaken to test the improved STS. The survey results are presented in Section 4. A discussion of the results follows in Section 5.
نتیجه گیری انگلیسی
4. Results 4.1. Satisfaction with travel (STS) In the analyses on STS and mood the different sub-scales correlated highly. A composite measure was therefore constructed by averaging across all items. In Table 4 the average STS scores are given for each condition and agenda. As may be seen, STS is higher for the car condition than for any of the bus conditions and for the bus conditions STS decreases with travel time and increases with access to bus stops. In all conditions STS is lower for the agendas with more activities. Significance tests were performed by means of an overall analysis of variance (ANOVA) followed by separate t-tests. Only significant effects at α = .05. No non-significant effect was of medium or large size (d ⩾ 0.50 or ω2 ⩾ 0.06, see Kirk, 1995). Table 4. Mean STS related to travel mode, travel time, walking time, and agenda. Condition Agenda 1 Agenda 2 Agenda 3 Mean Bus, short travel time, high access to bus stops 0.54 0.07 −0.12 0.16 Bus, long travel time, high access to bus stops −0.04 −0.41 −0.43 −0.69 Bus, short travel time, low access to bus stops −0.26 −0.93 −0.86 −0.29 Bus, long travel time, low access to bus stops −0.43 −1.20 −0.92 −0.85 Car 1.01 0.44 0.24 0.56 Table options A five (condition) by three (agenda) mixed factorial analysis of variance (ANOVA) with repeated measures on the last factor yielded a significant main effect of condition, F(4, 150) = 5.96, p < .001, ω2 = .11, and of agenda, F(2, 300, Greenhouse-Geisser ε = .94) = 30.86, p = .005, ω2 = .28. Independent t-tests showed that except for the most attractive bus condition, STS for the car condition differed significantly from all bus conditions, t(60) = 2.49, p = .016, d = 0.63 (long travel time, high access to bus stops), t(60) = 3.69, p < .001, d = 0.94 (short travel time, low access to bus stops), and t(60) = 3.94, p < .001, d = 0.99 (long travel time, low access to bus stops). The most attractive bus condition differed significantly from the two least attractive bus conditions, t(60) = 2.65, p = .010, d = 0.67 (short travel time, low access to bus stops), and t(60) = 2.96, p = .004, d = 0.75 (long travel time, low access to bus stops). Paired t-tests showed that agenda 1 with the fewest activities had a higher STS (M = −0.16) than agenda 2 (M = −0.41), t(154) = 6.99, p < .001, d = 0.37, and agenda 3 (M = −0.42), t(154) = 6.24, p < .001, d = 0.38. 4.2. Mood (SCAS) Table 5 shows that mood (affective SWB) is most positive for the car condition and the most attractive bus condition (short travel times and high access to bus stops). Within the bus conditions mood decreases with travel time and decreases with access to bus stops. In some but not all conditions, mood is more positive for the agenda with the fewest activities than the other agendas. Table 5. Mean mood related to travel mode, travel time, walking time, and agenda. Questionnaire version Agenda 1 Agenda 2 Agenda 3 Mean Bus, short travel time, high access to bus stops 0.88 0.84 0.89 0.87 Bus, long travel time, high access to bus stops 0.39 0.07 0.21 0.22 Bus, short travel time, low access to bus stops 0.54 0.57 0.47 0.52 Bus, long travel time, low access to bus stops −0.04 −0.44 −0.55 −0.34 Car 1.25 0.84 0.91 1.00 Table options A parallel five (condition) by three (agenda) mixed factorial analysis of variance with repeated measures on the last factor yielded a significant main effect of condition, F(4, 150) = 3.43, p = .010, ω2 = .06, and of agenda, F(2, 300, Greenhouse-Geisser ε = .99) = 4.41, p = .013, ω2 = .04. The independent t-tests showed that mood for the car condition only differed from the two bus conditions in which mood was lowest, t(60) = 2.05, p = .045, d = 0.52 (long travel time, high access to bus stops), and t(60) = 3.16, p = .002, d = 0.80 (long travel time, low access to bus stops). The bus condition with the highest mood also differed significant from the bus condition with the lowest mood, t(60) = 2.92, p = .005, d = 0.74. A higher mood was observed for agenda 1 with the fewest activities than the other agendas, F(1, 101) = 5.87, p = .017, ω2 = .56. Paired t-tests showed that agenda 1 (M = 0.60) differed significantly from agenda 2 (M = 0.38), t(154) = 2.66, p = .009, d = 0.13, and agenda 3 (M = 0.38), t(154) = 2.48, p = .014, d = 0.12. 4.3. Satisfaction with day (SWLS) The means of the average of the SWLS items are displayed in Table 6 for each condition and agenda. As may be seen, SWLS is higher in the car condition than in the bus conditions. Within the bus conditions, SWLS decreases with travel time and decreases with access to bus stops. No differences are observed between agendas. Table 6. Mean SWLS related to travel mode, travel time, walking time, and agenda. Questionnaire version Agenda 1 Agenda 2 Agenda 3 Mean Bus, short travel time, high access to bus stops 2.72 3.13 3.01 4.14 Bus, long travel time, high access to bus stops 2.16 2.08 2.07 3.65 Bus, short travel time, low access to bus stops 2.23 2.29 1.77 3.71 Bus, long travel time, low access to bus stops 1.75 1.60 1.93 3.17 Car 3.77 3.75 3.22 3.58 Table options A parallel five (condition) by three (agenda) mixed factorial analysis of variance with repeated measures on the last factor only yielded a significant main effect of condition, F(4, 150) = 3.62, p = .008, ω2 = .06. The independent t-tests revealed that SWLS differed significantly from the three bus conditions with lowest SWLS, t(60) = 2.85, p = .006, d = 0.72 (long travel time, high access to bus stops), t(60) = 2.44, p = .018, d = 0.62 (short travel time, low access to bus stops, and t(60) = 4.12, p < .001, d = 1.05 (long travel time, low access to bus stops). Of the bus conditions only those with highest and lowest SWLS differed significantly from each other, t(60) = 2.76, p = .008, d = 0.70. No differences between the agendas (Ms = 3.80, 3.85, and 3.82 for agendas 1, 2, and 3) were significant.