Transportation Research Part F 27 (2014) 90–98

Contents lists available at ScienceDirect

Transportation Research Part F

journal homepage: www.elsevier.com/locate/trf

The role of deliberate planning, car habit and resistance

to change in public transportation mode use
Trond Nordfjærn a,⇑, Özlem Sß imsßekoğlu c, Torbjørn Rundmo a,b
a
NTNU Samfunnsforskning, Studio Apertura, Dragvoll Allé 38 B, 7491 Trondheim, Norway
b
Norwegian University of Science and Technology, Department of Psychology, Dragvoll, 7491 Trondheim, Norway
c
Izmir University of Economics, Department of Psychology, Sakarya Caddesi, NO: 156, 35330 Balcova-Izmir, Turkey

a r t i c l e i n f o a b s t r a c t

Article history: Few studies have examined the role of deliberate planning, car habit and resistance to
Received 14 February 2014 change in relation to transportation mode use. The aim of this study was to examine the
Received in revised form 23 June 2014 relative role of components in the Theory of Planned Behavior (TPB), car habit and resis-
Accepted 21 September 2014
tance to change in use of public transportation. A postal questionnaire survey was carried
Available online 11 October 2014
out in a randomly obtained representative sample (n = 1039) of the Norwegian population
living in the six largest urban areas of Norway. The sample was randomly recruited from
Keywords:
the Norwegian population registry. The results showed that an isolated TPB-model was
Resistance to change
Car
better fitted to the data than an isolated habit-resistance to change model. The isolated
Public transport TPB also explained substantially more of the variance in intentions to use public transport
Norway compared to the habit-based model. A combined model including the TPB, car habit and
Social influence resistance to change was also found to have good fit. Within this model, the most impor-
Environment tant predictor of intentions to use public transport was strong subjective norms of public
transportation mode use. Favorable attitudes towards public transport mode use were
weakly related to intentions, when car habit and resistance to change were accounted
for in the model. Perceived control was not mediated by intentions to use public transport
and solely related directly to use. Car habit was a negative predictor of these intentions. It
is concluded that car habit is not the sole factor related to intentions of using public trans-
portation and that social cognition and social influence are instrumental in promoting use
of such transportation. Use of public transportation seems to partly reflect a planned and
deliberate psychological process.
Ó 2014 Elsevier Ltd. All rights reserved.

1. Introduction

Car use is related to a higher prevalence of accidents and injuries compared to use of public transport (Albertsson &
Falkmer, 2005). In addition to accidents and injuries, the societal costs of car use also include congestion, noise, air pollution,
and substantial use of land. This underlines the importance of promoting use of safe and environmentally friendly public
transportation in the urban population. This is a major challenge for authorities worldwide who in recent years have realized
that it is imperative to reduce car use in order to promote quality of life and to improve the environment in urban areas. The
majority of people in Norway still use private motorized travel modes, especially car (Rundmo, Nordfjærn, Iversen, Oltedal, &

⇑ Corresponding author. Tel.: +47 95 93 47 66; fax: +47 73 59 63 30.

E-mail address: tn@sirus.no (T. Nordfjærn).

http://dx.doi.org/10.1016/j.trf.2014.09.010
1369-8478/Ó 2014 Elsevier Ltd. All rights reserved.
 T. Nordfjærn et al. / Transportation Research Part F 27 (2014) 90–98 91

Jørgensen, 2011). In recent decades, the use of private cars for work travels, shopping, and visits to recreational locations had
beneficial impacts on human mobility and welfare. Simultaneously, the societal costs of car use have reached the limit of
what should be considered tolerable (Greene & Wegener, 1997).
A concurrent discussion in the transport research literature is whether transportation mode use manifests an automatic
and habitual process or a planned deliberate psychological process (or both). The role of habit in car use is guided by
psychological research showing that previous behavior is a strong regressor on future behaviour (Gärling & Axhausen,
2003; Verplanken & Aarts, 1999). Verplanken, Aarts, and Knippenberg (1997) argued that when car use has been repeated
on a frequent basis the ‘choice’ becomes scripted (see also Aarts, Verplanken, & Knippenberg, 1998). This implies that the
‘choice’ does not require deliberate cognitive processing and that car use is a result of mental heuristics, which cause the
individual to be less attentive or ignorant to new information or behavioral alternatives (Bamberg, Rölle, & Weber, 2003a).
The assumption that car use is merely habitual and scripted has been challenged by social psychological theories, such as
the Theory of Reasoned Action (Fishbein & Ajzen, 1975) and its successor the Theory of Planned Behavior (TPB) (Ajzen, 1991).
These models suggest that behaviors are under voluntary control and influenced by deliberate cognitive information
processing. The TPB argues that behavior is predicted by intentions, which in turn is influenced by attitudes towards the
behavior, subjective norms of the behavior and perceived behavioral control. Attitudes are defined as evaluations of the
behavior (e.g. whether the individuals have a positive or negative view on using public transport). Subjective norms are
defined as whether the individual perceives that significant others wish or encourage the person to conduct the behavior,
whereas perceived behavioral control refers to perceptions of control and barriers to conduct the behavior. Further, the cau-
sal assumptions in this model argue that attitudes and subjective norm are mediated by intentions to perform the behavior,
while perceived control has a direct link to behavior, and an indirect mediated relation through intentions (see also Fig. 1).
The TPB has been found to predict behaviors ranging from seat-belt use (S ß imsßekoğlu & Lajunen, 2008), protective sexual
behavior (Sheeran & Taylor, 1999), pro-environmental behavior (Kaiser & Gutscher, 2003) to travel mode use (Bamberg &
Schmidt, 2003; Heath & Gifford, 2002). However, few studies have compared the relative role of the TPB and car habit in
predicting use of public transport. An exception was Bamberg and Schmidt (2003) who compared the TPB, Theory of Inter-
personal Behaviour (TIB) (a partly habit-based theory) (Triandis, 1977) and the Norm Activation model (NAM) (Schwartz,
1977) in predicting car use. The results showed that the TPB explained more of the variance in car use than the NAM, but
car habit was also found to be important for use. Another study was carried out among 241 individuals and showed that
habits were weaker predictors of future mode use behavior than the TPB (Bamberg, Ajzen, & Schmidt, 2003b). The general-
ization potentials in these two citied studies were, however, somewhat mitigated by the fact that the aforementioned study
solely consisted of German university students, and the latter study included a relatively small sample from a delimited area
within Germany.
Moreover, the intention to use public transport could also be influenced by a person’s pre-disposition to resist changes in
routines (i.e. resistance to change) (Oreg, 2003). For instance, a person who has developed a strong car habit may be more
reluctant to consider a change to public transport if the individual holds negative cognitions about changes in their daily
routine and experiences distress and negative emotions when such changes occur (see also Tertoolen, Van Kreveld, &
Verstraten, 1998). To our knowledge no studies have implemented resistance to change in the habit models. Consequently,
we also integrated this psychological construct into the specification of the habit model.
As indicated above, few previous studies have examined the relative role of the TPB, car habit and resistance to change in
relation to use of public transport. This approach is important because it could provide insights into whether transportation
mode use is a planned and deliberate psychological process or more automatic and habitual. If the latter assumption is
supported, a possible implication would be that it is difficult to change transport mode behavior by campaigns that require
a conscious and deliberate processing of the message.
The aim of this urban population-based study was to investigate the relative role of the TPB, car habit and resistance to
change in predicting use of public transportation. The working models of the study are displayed in Fig. 1. To test the isolated
capability of the TPB and car habit-resistance to change in predicting use of public transportation, we first tested these
frameworks separately (Fig. 1a and b). Secondly, we tested an integrated model including all constructs in order to test
the relative contribution of the TPB, car habit and resistance to change constructs (Fig. 1c). This model is in line with the
Extended Theory of Planned Behavior (Åberg, 2001) which suggested that additional constructs, such as habit, are imple-
mented into the TPB taxonomy. On the basis of the assumption that mode use mainly is habitual and scripted (Gärling &
Axhausen, 2003; Verplanken & Aarts, 1999), we hypothesized that the isolated habit-resistance to change model would
be better fitted than the isolated TPB. We also hypothesized that the TPB components would be relatively weakly related
to public transportation mode use when habit and resistance to change were included in the combined model.

2. Methods and materials

2.1. Sampling

A survey was conducted in June and August 2013 in a random representative sample of the Norwegian population
(n = 6200) living in the six largest urban regions of Norway. This random sample was obtained electronically by a firm with
access to the Norwegian population registry. The study protocol was approved by the Norwegian Social Science Data Services
92 T. Nordfjærn et al. / Transportation Research Part F 27 (2014) 90–98

Attitude towards Intention to use public Use of public transport

public transport transport

Subjective norm
public transport

Perceived control
public transport

(a)

Car habit Intention to use public Use of public transport

transport

Resistance to change

(b)

Attitude towards Intention to use public Use of public transport

public transport transport

Subjective norm
public transport

Perceived control
public transport Car habit Resistance to change

(c)
Fig. 1. Hypothesized working models of the study. (a) Theory of Planned Behavior. (b) Habit and resistance to change model. (c) Combined Theory of
Planned Behavior and habit-resistance to change model.

(NSD) before data collection commenced. The sample was restricted to individuals who were 18 years or above. Sampling of
individuals in urban areas was chosen because these individuals are more likely to have a range of public transport options
available, compared to individuals in more remote districts where car may be the sole alternative on mediocre and long
distance travels.
The six urban regions included the central Oslo region in the urbanized south-eastern area of Norway (n = 2000), the
Skien and Porsgrunn area (n = 600), the central Trondheim region in the mid-area of Norway (n = 1000), the central Stavan-
ger area in the south-west region (n = 1000), the central Bergen region at the west coast (n = 1000), and the Tromsø area
(n = 600) in northern Norway. We oversampled regions with rather low population figures and high population figures
(see also Nordfjærn, Lind, S
ß imsßekoğlu, Jørgensen, & Rundmo, 2014 for details). Incentive for a response was enrolment to
 T. Nordfjærn et al. / Transportation Research Part F 27 (2014) 90–98 93

a lottery with the possibility of winning 1900 €. A total of 1039 individuals responded to the enquiry (18% response rate).
Nordfjærn et al. (2014) compared the present sample with the target population in the six urban regions and found that
the sample and population were similar in demographic characteristics. The sample also resembled the characteristics
reported in studies with higher response rates (Backer-Grøndahl, Fyhri, Ulleberg, & Amundsen, 2009; Roche-Cerasi,
Rundmo, Sigurdson, & Moe, 2013). Relatively low response rates are common in transport population studies (e.g.
Backer-Grøndahl et al., 2009; Castanier, Paran, & Delhomme, 2012; Moan, 2013), probably partially due to a low immediate
personal salience of the research topic (Galea & Tracy, 2007).
The sample consisted of 459 males (44%) and 576 females (56%) (n = 4 missing on the gender variable). The average age
was 41.43 years (SD = 12.06) and ranged from 18 to 74 years. A total of 366 individuals (35%) held a basic education of high
school or lower, while 65% (n = 669) had a high education from college or university (n = 4 missing on the education vari-
able). The majority of the respondents in the sample reported that they had access to a car (n = 883, 85%).

2.2. Measures

2.2.1. TPB measures

Attitudes towards transport mode use were measured by a 12 item instrument related to respondents’ evaluations of
public and private transportation mode use. The measure included items such as ‘Time pressure and economic issues make
it impossible for business leaders and management to use public transport’, ‘Public transport is primarily for people with a
low income’, and ‘It is impossible to deliver and pick up children in the kindergarten without using a car’. The respondents
reported their level of agreement to such statements on a seven-point scale ranging from (1) strongly agree to (7) strongly
disagree. Nordfjærn et al. (2014) examined the dimensionality of the instrument in the same sample and found that it
segmented into two dimensions; Social status important for transport mode use (a = 739, average corrected inter-item total
correlation = .50) and self-determined attitudes of car use (a = 770, average corrected inter-item total correlation = .54). The
social status dimension included items related to factors such as using a work position (e.g. holding a business management
position) and economic income as justifications for not using public transport. The self-determined attitude dimension
included items related to innate psychological needs as reasons for not using public transport (e.g. ‘People should use the
mode of transport that suit their needs’ and ‘It is impossible to deliver and pick up children in the kindergarten without using
a car’). The items were reverse-coded to make higher scores reflect favorable attitudes towards public transport.
Subjective norm was measured on a different semantic scale than attitudes by two items related to beliefs about whether
significant others think the respondent should use public transport (Bamberg et al., 2003a). The two items were: ‘Most
people who are important to me would support me in a decision to use public transport from where I live on a daily basis’
and ‘Most people who are important to me would wish that I used public transport from where I live’. The items were scored
on scale ranging from (1) unlikely to (5) likely.
Perceived control was also assessed with two items (Bamberg et al., 2003a); ‘Daily use of public transport from where I
live is’ (1) very difficult (5) very easy and ‘My freedom of choice when it comes to daily use of public transportation from
where I live is’ (1) very low to (5) very high.
Intentions to use public transport was measured by a single item (Bamberg et al., 2003a); ‘My intention to use public
transport on daily travels from where I live is’ (1) very weak to (5) very strong.
Use of public transport was assessed by a single item asking: ‘Think about the last two weeks (the last 14 days). During
this period how often have you used public transport (e.g. metro, bus, train, and tram). Note: One trip fourth and back counts
as two times’. The respondents reported the use frequency of transportation on an open-ended scale. The measure was
limited to the last two weeks in order to reduce the probability that memory bias substantially influenced the responses.

2.2.2. Habit and resistance to change measures

Car habit strength was assessed by a revised nine-item version of the original 12-item Self-Report Habit Index
(Verplanken & Orbell, 2003). The authors of this index concluded that the scale should be adjusted and revised to the specific
habit behavior in question. Verplanken and Orbell (2003) showed that the instrument was well-suited to discriminate
behaviors that vary in frequency. The same study also reported good internal consistency and test–retest reliability. This
instrument included items such as: ‘Car as a travel mode is something I use without having to consciously remember’,
‘Car as a travel mode is something I would find hard not to use’. The instrument was recorded on a scale ranging from
(1) strongly disagree to (5) strongly agree, with high scores reflecting strong car habit strength. Nordfjærn et al. (2014) found
that the car habit measure was uni-dimensional in the current sample.
We measured resistance to change by a previously validated instrument including 19 items (Oreg, 2003). The items relate
to how resistant or willing people are to change across a range of contexts. The instrument includes items such as ‘I like to
experience novelty and change in my daily routine’ and ‘When I am informed of a change of plans, I tense up a bit’. The
respondents recorded their level of agreement on such statements on a scale ranging from (1) strongly disagree to (5)
strongly agree. Oreg (2003) found that the instrument segmented into four factors entitled: Emotional reaction (getting
emotionally tense by changes), Routine seeking (avoiding novelty and finding comfort in routines), Cognitive rigidity (con-
sistency in views over time), and Short-term focus (oppose changes that may be beneficial in the long term). This dimen-
sional structure was also replicated in the present sample and with good internal consistency for all four dimensions
94 T. Nordfjærn et al. / Transportation Research Part F 27 (2014) 90–98

(Nordfjærn et al., 2014). Oreg (2003) also demonstrated that the measure predicted change-resistant behavior across a
variety of different contexts and concluded that the instrument measures resistance to change across different types of
behaviors.

2.3. Statistical procedures

IBMÒ SPSSÒ Statistics 21 was used to obtain descriptive statistics of sample characteristics and scores on the psycholog-
ical constructs entered into the three models. Structural Equation Modelling (SEM) was applied to test the three models (see
also Fig. 1) using IBMÒ SPSSÒ AMOS 20.0.0. The fit indices used as criteria for model fit were the Root Mean Square Error of
Approximation (RMSEA) and Comparative Fit Index (CFI). It has been argued that an RMSEA value equal to or less than 0.05
and a CFI higher than 0.95 reflect good correspondence between the estimated model and data (Kim & Bentler, 2006;
Tabachnick & Fidell, 2000). Modification indices were also studied in order to examine whether the models would improve
by specifying direct relations between car habit and resistance to change to use of public transport. Because people who
manifest change-resistance may already use public transport quite frequently and thus not influence intentions to use public
transport in any direction, we also conducted sensitivity analyses excluding those who either reported use of public
transport above the median (MDN = 3) or had missing values on the transport use variable.

3. Results

Means and standard deviations on all dimensions included in the three models are displayed in Table 1. The respondents
reported relatively favourable self-determined attitudes related to public transport, whereas the sample mostly did not use
social status as an explanation for not using public transport. Overall the sample reported relatively high support and
preferences from their significant others (i.e. subjective norm) for using public transportation. The overall perceived difficul-
ties of using public transportation were relatively low, whereas the freedom of choice component of perceived control and
intentions to use public transport were more moderate. The use of public transportation the last two weeks ranged from 0 to
60 (M = 6.62, SD = 1.25). The respondents reported overall moderate car habit strength. Resistance to change was also rela-
tively modest, with somewhat higher scores on cognitive rigidity.
SEM showed that the isolated TPB had good fit to the data (v2 = 50.70, df = 14, p < .001, RMSEA = .050, CFI = .99) and
explained 47% of the variance in intentions to use public transport and 34% in use of public transport. Favorable attitudes
towards public transport had a slight, but significant, relation to a stronger intention of using public transport (B = .12,
p < .001). Increased perceived control was also associated with use of public transportation (B = .16, p < .001), whereas a
mediated relation through intentions of using public transport was not supported by the data (B = .02, n. s.). The strongest
factor associated with intentions of using public transport was subjective norms related to public transportation (B = .62,
p < .001). Strong intention of using public transportation modes was a significant predictor of public transportation mode
use (B = .52, p < .001), as expected.
An additional SEM suggested that a model solely including car habit and resistance to change had relatively poor fit to the
data (v2 = 135.10, df = 13, p < .001, RMSEA = .097, CFI = .84) and explained a substantially lower proportion of the variance in
intentions to use public transport (20%) than the isolated TPB. The model explained 31% of the variance in use of public trans-
port. Car habit had a negative relation with intentions to use public transport (B = .44, p < .001), whereas resistance to
change was not significantly related to such intentions ((B = .05, n.s.). An examination of modification indices showed no
improvement potentials in the model by adding direct paths between car habit and resistance to change to use of public
transport. The sensitivity analysis excluding people who either had missing (n = 14) values or scores above the median

Table 1
Means and standard deviations of the constructs included in the models.

Dimension Min Max Mean SD

Attitude – self-determination 1 7 3.69 1.33
Attitude – social status 1 7 5.46 1.18
Subjective norm – important people would support 1 5 3.55 1.24
Subjective norm – important people would wish 1 5 3.22 1.13
Perceived control – perceived difficulties 1 5 3.91 1.17
Perceived control – perceived freedom of choice 1 5 3.10 1.25
Intentions to use public transport 1 5 3.09 1.25
Public transportation mode use 0 60 6.62 9.29
Car habit strength 1 5 2.89 1.05
Resistance to change – routine seeking 1 5 2.67 .62
Resistance to change – emotional reaction 1 5 2.60 .73
Resistance to change – short-term focus 1 5 2.46 .63
Resistance to change – cognitive rigidity 1 5 3.17 .66

Higher scores reflect public-transport favourable attitudes, higher subjective norm and perceived control, stronger intentions to use public transport, more
use of public transport, stronger car habit strength and resistance to change.
 T. Nordfjærn et al. / Transportation Research Part F 27 (2014) 90–98 95

(n = 484) on the transportation mode use variable left 541 individuals who reported low use of public transport. This caused
slight improvements in the fit of the model (RMSEA = .091, CFI = .91), but resistance to change remained weakly related to
intentions to use public transport (B = .07, n.s.). Also excluding resistance to change entirely from the model did not have a
substantial impact on model-data fit.
The SEM testing the relative role of the TPB, car habit and resistance to change revealed a good correspondence between
the hypothesized model and the data (v2 = 220.90, df = 51, p < .001, RMSEA = .057, CFI = .96). This model explained 50% of the
variance in intention to use public transportation and 34% in use of such transportation. Within this model (Fig. 2), subjective
norm was the most important predictor of increased intentions to use public transport, whereas transport attitudes and
perceived control were weakly related to intentions. Perceived control had a significant direct positive relation to public
transportation mode use. Car habit also had a negative relation to intention of using public transport, whereas resistance
to change was weakly related to such intention. Intention to use public transport was significantly related to use of public
transport. Modification indices did not suggest improvement potentials by adding direct paths between additional con-
structs and public transportation mode use. Neither did the model fit improve by excluding frequent users scoring above
the median on the public transportation mode use variable.

4. Discussion

The aim of the study was to investigate the relative role of the TPB, car habit and resistance to change in predicting use of
public transportation. In opposition to our hypothesis, the findings showed that an isolated TPB was well-fitted to the data,
while an isolated model based on car habit and resistance to change had weak fit. The isolated TPB framework also explained
substantially more of the variance in intentions to use public transport (47%) compared to the habit model (20%). Moreover,
the isolated habit model showed a low improvement potential (e.g. by adding direct paths between habit and resistance to
change to public transportation use) and a sensitivity analysis reflected that the model did not improve substantially by
excluding those who already used public transport frequently from the model estimation. A combined model was also found
to have good fit. Contradicting our hypothesis, this model suggested that subjective norms and perceived control were the

Cognitive E9 = .69
rigidity

.53 Routine
E10 = .70
seeking
.54
Emotional
E11 = .26
Resistance .86 reaction
Social status
e1 = .77 .48 to change
.90
Short-term E12 = .18
Transport focus
attitudes .15
.82
.00
Self- Car habit
e2 = .33
determined -.02

.41 e7 = .50 e8 = .66

R2 = .50 -.22 R2 =.34

Important
e3 = 34 .35
people would .81
support Intention to use .51 Use of public
.60
Subjective public transport transport
norm public
.85
transport
e4 = .27
Important
people wish

.32 .18
-.01

Perceived .90
e5 = .18 Perceived
difficulties control
public
.62 transport
Perceived
e6 = .61
freeedom of
choice

Fig. 2. Combined TPB and habit-model predicting public transportation mode use (standardized coefficients, significant p < .001 coefficients in bold).
v2 = 220.90, df = 51, p < .001, RMSEA = .057, CFI = .96.
96 T. Nordfjærn et al. / Transportation Research Part F 27 (2014) 90–98

two more important social cognitive constructs of the TPB for use of public transportation, while car habit and resistance to
change were included in the model.
It would be premature to conclude that habits are negligible in transport behavior on the basis of the present findings.
Both within the isolated habit and resistance to change model, as well as in the combined TPB and habit model, car habit
had a significant negative relation to intentions of using public transport, as could be expected. Meanwhile, the isolated
TPB performed better than the habit-based model in explaining use of public transportation. Also, subjective norm and per-
ceived control were important for public transportation mode use while car habit was adjusted for in a combined model. This
is contradicting previous work (e.g. Gärling & Axhausen, 2003; Verplanken & Aarts, 1999) which suggested that habit is a
substantially stronger predictor of mode use than social cognition. Our findings suggest that social cognitive constructs,
as operationalized by the TPB, are overall more strongly related to public transportation mode use than the habit constructs.
The results are also in line with the Extended Theory of Planned Behavior arguing that habit should be incorporated in social
cognitive models rather than investigated as an isolated factor in relation to mode use (Åberg, 2001).
Our findings also contradict previous assertions (e.g. Tertoolen et al., 1998) suggesting that people’s resistance to change
may render it ineffective to influence mode use by rational arguments. The present results suggest that resistance to change
has weak relations to cognitive intentions of using public transport, which could support the assumption that a pre-
dispositional resistance to change in routines is not a strong barrier for influencing cognitions about mode use. The findings
suggest that subjective norms are more strongly associated with intentions to use public transport, and as such it could be
that rational arguments from significant others aimed at promoting public transport could be efficient irrespective of the
overall resistance to change in the target groups. An alternative explanation, however, is that resistance to change was mea-
sured on a rather general level rather than specifically at the level of transport mode use. Although previous studies have
shown that the present measure of resistance to change generalizes across different behaviors (Oreg, 2003), future studies
could examine whether a specific transport-related measure of resistance to change yields other findings.
Several previous studies of the TPB reported that attitudes and perceived control were equally or more important than
subjective norm in predicting intentions to use public transport (e.g. Bamberg, Hunecke, & Blöbaum, 2007; Heath &
Gifford, 2002). Intriguingly, subjective norm was the most important factor for intentions to use public transport in the pres-
ent study, also when the role of car habit was adjusted for. This suggests that solely aiming to amend a car habit is not suf-
ficient in promoting use of public transportation. Social influence may also exert influences on public transportation mode
use. An implication is that it is possible to influence transport use by, for instance, exerting social pressure and to point out
the advantages of use to others (e.g. environmental benefit, avoiding traffic jams and promoting safety). In the light of this
finding, future longitudinal studies could investigate whether transportation mode use among parents and future transport
use among their children correspond over time. It is likely that parents who favor public transport also tend to choose these
modes more frequently for their children. As such, the children may adopt their parents’ transport mode use behaviors. The
socialization element in transport mode use should probably not be underestimated and is an important avenue in need of
more research.
The TPB argues that perceived behavioral control has a mediated relation through intentions on behavior in addition to a
direct association with behavior. In the present study we did not find support for a mediated relation. A potential reason is
that when there are few public transportation modes available, the individuals are to a low extent exposed to the alterna-
tives of using a car. In line with the availability heuristic (Tversky & Kahneman, 1973) people may start to associate travels
with the specific modes they have available (e.g. a car). Therefore, information about public transport could be difficult to
retrieve from memory in lack of sufficient retrieval cues. A consequence could be that the perceived difficulties and lack
of different available public transport options may not influence cognitive intentions of using these modes, but is manifested
directly in behavior.
The present study focused on the role of car habit and resistance to change as habit constructs related to use of public
transportation. This does not exclude the possibility that there could be additional habit concepts that should be tested in
this line of enquiry. For instance, previous studies included a measure of transport mode behavior in the past in addition
to a measure of habit (e.g. Bamberg et al., 2003a). Nevertheless, the habit measure utilized in the present study has been
found to correlate strongly with past behavior (Verplanken & Orbell, 2003) and it is likely that multicollinearity issues would
have been introduced in the model were we to include a measure of past behavior in the model taxonomy. We acknowledge
the challenges related to measuring habits on a questionnaire (e.g. not possible to control for deliberate cognitive processing
of the test items). Future studies should attempt to measure habits in more controlled settings, where it could be more fea-
sible to control the level of deliberate cognitive processing of stimuli than in a population-based questionnaire survey.
These issues notwithstanding, however, we agree with the suggestion of Bamberg et al. (2003b) who argued that even
well-established habits involve an element of deliberate cognition. For instance, in the Western world it is common social
practice to shake hands with women and handshaking behavior is a scripted response when people are introduced in social
contexts. If a person from Western countries is a guest in countries where it is not customary to shake hands with women
(e.g. countries with Sharia regulations, such as Iran and Saudi Arabia), and the person is aware of this difference, the person
may regulate this behavior and respond with a friendly nod rather than a handshake when introduced to women in these
settings. In a similar manner both habits and deliberate planning may underlie behaviors such as transportation mode
use, but our study does not support that habit and resistance to change are sufficient components to explain use of public
transportation. Bamberg et al. (2003b) argued that habits could have a stronger role when the context and information
 T. Nordfjærn et al. / Transportation Research Part F 27 (2014) 90–98 97

stream are stable. Consequently, in order to promote public transport mode use it is important for governments and policy
makers alike to provide information that challenges established car habits in the population.
Some limitations of the present study should be pointed out. A cross-sectional design does not allow for conclusions
about causality, but the causal assumptions in the present study are driven by theories arguing the links between the pre-
dictors and outcomes (Ajzen, 1991; Verplanken et al., 1997). The response rate of the survey was low. However, comparisons
with the urban target population did not reveal substantial differences. Previous studies that examined the relative role of
the TPB and habit-based constructs in predicting transportation mode use mostly recruited quite small non-randomized
samples (e.g. Bamberg et al., 2003b) or rather homogenous samples, such as university students (Bamberg & Schmidt,
2003). To our knowledge this is the first study to investigate the relative role of the TPB, car habit and resistance to change
in a large randomly obtained sample of an urban public, with substantial geographic spread across a country. In order to
reduce the impact of memory bias we measured transportation mode use the last two weeks. The study was conducted dur-
ing the summer and autumn months in Norway. We cannot exclude the possibility that more people would have reported
use of public transportation during the winter months, when it is more inconvenient to use a car due to snow conditions.

4.1. Conclusions and future directions

The present study investigated the relative role of the TPB, car habit and resistance to change in predicting public
transportation mode use. The TPB was found to have better fit than the habit model, when these models were compared
in isolation. Subjective norm was found to be more important for public transportation mode use than both habits and
resistance to change in a combined model. This challenges the assumption that mode use is merely habitual, and that social
cognition is less important in promoting use of public transport.
The results could suggest that public transportation mode use is partly a deliberate and planned psychological process.
The important role of subjective norm for public transportation mode use revealed in the present study, suggests that use
could be promoted by encouragement from significant others. Therefore, it is important to include a person’s significant
others (e.g. family members and friends) in interventions aimed to promote public transportation mode use. Future research
should examine the role of habit on mode use and preferences in controlled settings. Longitudinal efforts that examine the
role of socialization on public transportation mode use could also provide useful information to be utilized by interventions
and campaigns.
Moreover, the findings should be replicated by research conducted during different seasons (e.g. winter time vs. summer
time) as transportation mode use may differ during summer and winter. Car use may be more inconvenient in winter and
some individuals may choose to use more public transport in this particular period. The results also need replication in
different countries and cultures. Norway has strong socio-democratic political traditions aiming to minimize differences
in socio-economic status (SES) across segments in the population. In some countries, however, public transport vs. car
use could be highly correlated with SES and the quality of the transportation system or concerns over personal security could
also induce a desire to change to motorized transportation modes. The relative explanatory power of TPB and habit may
thereby vary according to cultural contexts. Comparative studies including countries with more pronounced SES differences
would be useful.

Acknowledgement

The study was funded by the Research Council of Norway as part of the Safety and Security in Transport (TRANSIKK)
Programme Grant No. 224754.

References

Aarts, H., Verplanken, B., & Knippenberg, A. V. (1998). Predicting behaviour from actions in the past: Repeated decision making or a matter of habit? Journal
of Applied Psychology, 28, 1355–1374.
Åberg, L. (2001). Attitudes. In P. E. Barjonet (Ed.), Traffic psychology today (pp. 119–137). USA: Kluwer Academic Publisher.
Ajzen, I. (1991). The theory of planned behavior. Organizational Behavior and Human Decision Processes, 50, 179–211.
Albertsson, P., & Falkmer, T. (2005). Is there a pattern in European bus and coach incidents? A literature analysis with special focus on injury causation and
injury mechanisms. Accident Analysis & Prevention, 37, 225–233.
Backer-Grøndahl, A., Fyhri, A., Ulleberg, P., & Amundsen, A. H. (2009). Accidents and unpleasant incidents: Worry in transport and prediction of travel
behavior. Risk Analysis, 29, 1217–1226.
Bamberg, S., Ajzen, I., & Schmidt, P. (2003b). Choice of travel mode in the theory of planned behavior: The roles of past behavior, habit, and reasoned action.
Basic and Applied Social Psychology, 25, 175–187.
Bamberg, S., Hunecke, M., & Blöbaum, A. (2007). Social context, personal norms and the use of public transportation: Two field studies. Journal of
Environmental Psychology, 27, 190–203.
Bamberg, S., Rölle, D., & Weber, C. (2003a). Does habitual car use not lead to more resistance to change of travel mode? Transportation, 30, 97–108.
Bamberg, S., & Schmidt, P. (2003). Incentives, morality or habit? Predicting students’ car use for university routes with the models of Ajzen, Schwartz, and
Triandis. Environment and Behavior, 35, 264–285.
Castanier, C., Paran, F., & Delhomme, P. (2012). Risk of crashing with a tram: perceptions of pedestrians, cyclists, and motorists. Transportation Research Part
F: Traffic Psychology and Behaviour, 15, 387–394.
Fishbein, M., & Ajzen, I. (1975). Belief, attitude, intention, and behavior: An introduction to theory and research. Reading, MA: Addison-Wesley.
Galea, S., & Tracy, M. (2007). Participation rates in epidemiological studies. Annals of Epidemiology, 17, 643–653.
Gärling, T., & Axhausen, K. W. (2003). Introduction: Habitual travel choice. Transportation, 30, 1–11.
Greene, D. L., & Wegener, M. (1997). Sustainable transport. Journal of Transport Geography, 5, 177–190.
98 T. Nordfjærn et al. / Transportation Research Part F 27 (2014) 90–98

Heath, Y., & Gifford, R. (2002). Extending the theory of planned behavior: Predicting the use of public transportation. Journal of Applied Social Psychology, 32,
2154–2189.
Kaiser, F. G., & Gutscher, H. (2003). The proposition of a general version of the theory of planned behavior: Predicting ecological behavior. Journal of Applied
Social Psychology, 33, 586–603.
Kim, K. H., & Bentler, P. M. (2006). Data modeling: Structural equation modeling. In J. Green, L., Camilli, G., Elmore, P. B., Skukauskaite, A., & Grace, E. (Eds.),
Handbook of complementary methods in educational research (pp. 161–175). New Jersey: Lawrence Erlbaum Association Publishers.
Moan, I. S. (2013). Whether or not to ride with an intoxicated driver: Predicting intentions using an extended version of the theory of planned behaviour.
Transportation Research Part F: Traffic Psychology and Behaviour, 20, 193–205.
Nordfjærn, T., Lind, H. B., Sßimsßekoğlu, Ö., Jørgensen, S. H., & Rundmo, T. (2014). Perceived thresholds for transport mode change and tolerance of push measures
in an urban Norwegian public. Trondheim, Norway: Norwegian University of Science and Technology, Department of Psychology.
Oreg, S. (2003). Resistance to change: Developing an individual differences measure. Journal of Applied Psychology, 88, 680–693.
Roche-Cerasi, I., Rundmo, T., Sigurdson, J. F., & Moe, D. (2013). Transport mode preferences, risk perception and worry in a Norwegian urban population.
Accident Analysis & Prevention, 50, 698–704.
Rundmo, T., Nordfjærn, T., Iversen, H. H., Oltedal, S., & Jørgensen, S. H. (2011). The role of risk perception and other risk-related judgements in transportation
mode use. Safety Science, 49, 226–235.
Schwartz, S. H. (1977). Normative influences on altruism. In L. Berkowitz (Ed.). Advances in experimental social psychology (Vol. 10, pp. 221–279). San Diego:
Academic Press.
Sheeran, P., & Taylor, S. (1999). Predicting intentions to use condoms: A meta-analysis and comparison of the theories of reasoned action and planned
behaviour., 29, 1624–1675.
Sßimsßekoğlu, Ö., & Lajunen, T. (2008). Social psychology of seat belt use: A comparison of theory of planned behavior and health belief model. Transportation
Research Part F: Traffic Psychology & Behaviour, 11, 181–191.
Tabachnick, B. G., & Fidell, L. S. (2000). Using multivariate statistics (4th ed.). United States: Allyn and Bacon.
Tertoolen, G., Van Kreveld, D., & Verstraten, B. (1998). Psychological resistance against attempts to reduce private car use. Transportation Research Part A:
Policy and Practice, 32, 171–181.
Triandis, H. C. (1977). Interpersonal behavior. Monterey, CA: Brooks/Cole.
Tversky, A., & Kahneman, D. (1973). Availability: A heuristic for judging frequency and probability. Cognitive Psychology, 5, 207–232.
Verplanken, B., & Aarts, H. (1999). Habit, attitude, and planned behaviour: Is habit an empty construct or an interesting case of goal-directed automaticity?
European Review of Social Psychology, 10, 101–134.
Verplanken, B., Aarts, H., & Knippenberg, A. V. (1997). Habit, information acquisition, and the process of making travel mode choices. European Journal of
Social Psychology, 27, 539–560.
Verplanken, B., & Orbell, S. (2003). Reflections on past behavior: A self-report index of habit strength. Journal of Applied Psychology, 33, 1313–1330.