Journal of Transport Geography 42 (2015) 98–109

Contents lists available at ScienceDirect

Journal of Transport Geography

journal homepage: www.elsevier .com/locate / j t rangeo

Analysing mobility biographies with the life course calendar: aretrospective survey methodology for longitudinal data collection

http://dx.doi.org/10.1016/j.jtrangeo.2014.12.0010966-6923/� 2014 Elsevier Ltd. All rights reserved.

⇑ Corresponding author at: Goethe-University Frankfurt/Main, Department ofHuman Geography, Grüneburgplatz 1, PEG-Gebäude, D-60323 Frankfurt/Main,Germany. Tel.: +49 69 798 35181; fax: +49 69 798 23548.

E-mail address: schoenduwe@geo.uni-frankfurt.de (R. Schoenduwe).

Robert Schoenduwe a,⇑, Michel G. Mueller b, Anja Peters b,c, Martin Lanzendorf a

a Goethe-University Frankfurt/Main, Department of Human Geography, Frankfurt/Main, Germanyb ETH Zurich, Institute for Environmental Decisions, Natural and Social Science Interface, Switzerlandc Fraunhofer Institute for Systems and Innovation Research, Karlsruhe, Germany

a r t i c l e i n f o

Keywords:Mobility biographyTravel behaviourLife course approachLife course calendarSurvey methods

a b s t r a c t

Transportation research is in need of longitudinal data to better understand travel behaviour. This paperdescribes a new survey method to collect longitudinal data, called the life course calendar. We discuss theneed for and the uses of this instrument with reference to specific case study material and explore towhat extent the life course calendar is suitable for the study of individual life courses and travel behav-iour. In the first part of this article we give an overview of current research on the causal relationshipsbetween certain life events and travel behaviour. Based on the outcomes of the literature review we ana-lyse life course calendars of 646 respondents. Considering the explorative character of our case studymaterial the results are quite promising. Most notably information on car purchases, on relocations,and changes of the work place has been well recorded which indicates that the life course calendar is sui-ted for the study of changes of spatial contexts and accessibility. The data analysis has provided initialresults on the distribution of key events over the life course and on the relation of such events to changesin travel behaviour. Significant effects were found for key events such as relocation, change of job, birth offirst child, separation/divorce, moving in with partner and retirement. The life course calendar can beadvantageous, especially if it is employed to supplement extensive panel studies.

� 2014 Elsevier Ltd. All rights reserved.

1. Introduction

It has become widely acknowledged in the field of transportationresearch that travel behaviour is to a great extent routinized anddevelops over time (e.g. the special issue on routines in Transporta-tion (Gärling and Axhausen, 2003)). Thus, an improved understand-ing of people’s choices and decision-making for travel must take intoconsideration this decision process over time and pay attention tothe development and change of routines. For instance, the role ofspatial determinants of travel could be reconsidered by using longi-tudinal data. It is undisputed that the built environment has a partic-ularly strong impact on travel behaviour (Naess, 2012; Stead andMarshall, 2001) but what happens if people change spatial contextdue to relocations or after a change of work places has not been ade-quately investigated.

A small number of studies have recently put forward the con-cept of mobility biographies (e.g. Scheiner, 2014; Zhang et al.,

2014). This approach aims to conduct an integrative analysis ofindividuals’ mobility decisions in the context of choices and eventsin other domains of individuals’ lives. Moreover, this approachseeks to trace the interrelation of such choices with key eventsfrom an individual’s life course. Key events accompanied bychanges in spatial context and accessibility were shown to havesignificant influence on travel behaviour (see Section 2.2).

The majority of travel behaviour research relates to cross-sectional data and neglects the temporal dimension of traveldecision-making. Thus, many transport researchers ignore theeffect of previous learning experiences and of key events in the lifecourse on people’s and household’s mobility decisions. Instead ofexplaining travel behaviour based on the evolvement of people’sdecision making process, current studies frequently limit their cau-sal understanding to socio-demographics or related personal andhousehold characteristics at the cost of neglecting individualrationalities.

One reason for these limitations in travel behaviour researchmay be the lack of reliable and representative longitudinal data.Likely, the most common method to gather longitudinal data isthe panel study where the same individuals are asked similar ques-tions at different points in time. Thus scholars are able to gain a

R. Schoenduwe et al. / Journal of Transport Geography 42 (2015) 98–109 99

fragmentary overview of the particular development of individu-als. Nevertheless, panel studies have several drawbacks regardingthe selectivity and drop-out rate of respondents (cf. Kitamuraand Bovy, 1987; Zumkeller et al., 2006). Alternative methods togather longitudinal data are repeated cross-sectional surveys(trend studies), cohort and pseudo panel studies as well as retro-spective surveys (cf. Stopher and Stecher, 2006). Those studiesare an option to gain sufficient data but also have various draw-backs. Besides methodological problems the most striking obsta-cles are costs and time.

Besides panel surveys, there have been other methodologiesemployed in transport research for assessing travel behaviourchanges over a longer period of time. Sometimes before and afterstudies or stated preference surveys are conducted to study theimpact of certain events (e.g. residential relocation (Bamberg,2006) or structural changes, for example changes of the transportinfrastructure (Fujii and Gärling, 2005) on travel patterns. Also, tra-vel diaries for longer periods (weeks or even months) are a usefulmethod to capture the variability and rhythms of daily travelbehaviour (Schlich and Axhausen, 2003). Another method, the nar-rative interview, is one of the most commonly practiced retrospec-tive survey methods in social sciences. Qualitative interviewsmainly function as explorative tools in the area of travel research.Such interviews serve to uncover fundamental correlations fromwhich scholars deduce new hypotheses (e.g. Schwanen et al.,2012).

The objective of this paper is to contribute to the further explora-tion of innovative research methodologies for gathering longitudi-nal data. We focus on the life course calendar, a retrospectivesurvey technique, which can be described as a year-by-question-grid with a horizontal time axis and a vertical thematic axis. Weask if new retrospective survey techniques like the life course calen-dar can be used for gathering quantitative data sets on the longitu-dinal development of people’s travel behaviour and, thus, forcreating a tool for assessing people’s mobility biographies more effi-ciently compared to panel data.

For this purpose we focus on four central questions:

– which key events of the life course should be considered in a lifecourse calendar?

– how can we design a life course calendar as an efficient surveyinstrument?

– what kind of data and which empirical results can be expectedfrom this survey instrument?

– where are the limitations of this survey instrument?

We explore these questions in the following sections. In Section2 we review the present state of academic work on mobility biog-raphies and related approaches for explaining travel behaviourfrom a longitudinal perspective. Section 3 contains an overviewon the present scientific knowledge of retrospective data and lifecourse calendars. Section 4 introduces our case study, namely thecase study area and the survey instrument employed. In Section5 we examine selected empirical results of the case study. Forexample, we present the number of key events detected and ana-lyse some of the correlations between key events and other factors.In Section 6 we discuss several strengths and limitations of the sur-vey instrument for further research.

2. Background: previous research on key events and mobilitybiographies

It has been argued that the dominant static perspective in thefield of travel behaviour research and the use of cross-sectionalsurvey data is inadequate to detect trends at the individual level.

This limits our understanding of the dynamics of individual travelbehaviour. Moreover, this approach leads to an insufficient analy-sis of long-term mobility decisions and the role of routine behav-iour. Longitudinal surveys and qualitative methods should befavoured over cross-sectional surveys because only the formermake it possible to analyse cause-impact relationships betweenrelevant factors. For instance, a residential relocation or changeof work place is often accompanied by a change of spatial contextwhich may be followed by an adaptation of travel behaviour. Adeeper understanding of the process of behavioural change andthe interrelationships between spatial contexts, attitudes and tra-vel behaviour can only be achieved by using longitudinal data.

Taking these considerations into account, some theoreticalapproaches assert that a life course perspective is crucial for travelbehaviour research. The concept of mobility biographies is oneexample which especially emphasizes the importance of keyevents. As shown below, several empirical studies proved theimportance of key events for travel behaviour research. However,there is a lack of empirical data that allows for the analysis of rela-tionships between key events, life courses and travel behaviourover longer periods of time.

2.1. Mobility biography approaches

The concept of mobility biography introduces a new perspectiveto the study of travel behaviour. Different approaches exist. Someconcentrate on the importance of habits and context changes inan individual’s mobility biography (Lanzendorf, 2003; Scheiner,2007) and refer to ideas of the life course approach (Giele andElder, 1998). Particular events in a person’s life course (e.g. deci-sions regarding residential locations, change of job, starting a fam-ily) are considered important for shaping daily travel behaviour.Other approaches analyse the interaction of mobility biographiesand social networks (Axhausen, 2008). These approaches empha-size the importance of certain events for the extent of an actor’sactivity space.

Lanzendorf’s (2003) concept draws on approaches from the areaof life course research, in particular Salomon’s (1983) life styleapproach. The concept examines relationships between particularlife events and changes in behaviour. Lanzendorf emphasizes theimportance of routine behaviour. Life events are understood askey triggers for changes in routine behaviour. The analysis dividesthe actors’ life course into three different domains. Particular kindsof life events are assigned to each domain: the so called ‘life styledomain’, the ‘accessibility domain’, and the ‘mobility domain’. Thelife style domain comprises events in context of demography, pro-fession and leisure. Within the accessibility domain spatial con-texts like residential and job location as well as leisure and otherlocations are included. Changes of car and season ticket ownershipas well as changes in distances travelled are included in the mobil-ity domain. Interrelations between the domains are taken intoaccount. The order of the domains is not strictly hierarchical. Eachdomain is further divided into sub-domains. Their number variesaccording to the object of research. Two previous studies haveexamined the applicability of the mobility biography concept(Lanzendorf, 2004, 2010; Prillwitz, 2007).

However, until today existing mobility biography approachesare rather fragmentary. Other studies which do not prescribe a spe-cific biographical approach have also called for a more dynamicunderstanding of travel. Schönfelder and Axhausen (2009) suggestan approach which considers mobility consequences of ‘personalprojects’. Beige and Axhausen (2008) analyse long-term and mid-term mobility decisions during the life course on the basis of datathat has been collected via a quantitative retrospective surveyusing a life course calendar. The focus of their work is the relation-ship between the place of residence, the work place, corresponding

100 R. Schoenduwe et al. / Journal of Transport Geography 42 (2015) 98–109

relocations and the ownership of mobility resources (cf. Beige,2008; Beige and Axhausen, 2012). They show that there existstrong interdependencies between various key events and long-term mobility decisions during the life course. Axhausen (2008)and Ohnmacht et al. (2008) employ a biographical perspective toexamine changes of individual activity spaces in the context ofsocial networks. More specifically, they examine mobility needsduring the life course in regard to patterns of social networks.Finally, mobility biography approaches are used in tourism studiesto analyse the expansion of long-distance travel (Frändberg, 2006,2008).

2.2. Travel behaviour and change of context

While more general approaches may encompass an analysis ofentire mobility biographies or segments thereof, numerous studiesfocus on the effect of single life course events on travel behaviour.Table 1 provides an overview of relevant studies. These studies focuson the change of average distances travelled, the average number oftrips per day, the average travel time per day, modal split changes aswell as changes in the availability of mobility tools such as cars orpublic transport season tickets. The authors employ panel andpseudo-panel data, retrospective surveys as well as other qualitativeand quantitative methods. The changes of context surveyed encom-pass events in the household, the partnership, the family, the educa-tion or the job career. Among other events, the studies take note ofobtaining a driver’s license, changes in the availability of resources,events in relation to children (e.g. births) or relocations. Severalstudies explicitly examine such events as triggers for changes inhabitual behaviour.

It is necessary to distinguish between two fundamental types ofstudies concerned with context changes. On the one hand, studiesexamine relationships between concrete life events and changes intravel behaviour; on the other hand, studies seek to determinemore generally which life events coincide with changes in travelbehaviour. These studies take an inventory of possible events thatmay show relationships with travel behaviour.

2.2.1. Studies focusing on various key eventsAs early as the 1980s, scholars pointed to the relevance of a

dynamic perspective in travel behaviour research. For instance,Goodwin (1989) analysed panel data from 1984 to 1987 with regardto family changes and choice of transportation means. Almosttwenty years later, several exploratory studies focused on an inven-tory of key events that might be important for changing travelbehaviour. Klöckner (2005) examined actor’s subjective perceptionof their mobility decisions. Respondents perceived the following lifeevents as having the greatest influence on their travel behaviour:obtaining a driver’s license, starting university or a vocational train-ing, relocation, purchasing a car, changing to secondary/high school,and entering the job market. Behrens and Mistro (2010) achievedsimilar results. Van der Waerden et al. (2003) conducted interviewsregarding changes in attitude towards transportation means. More-over, they asked respondents about their use of transportationmeans before and after certain life events. The following events weremost commonly stated as being influential on actors’ attitudes:obtaining a drivers’ license, a change of job, a car purchase, a reloca-tion, entering the job market, a change of training post, and the pur-chase of a public transport ticket.

Harms (2003) interviewed 650 people on the subject of carsharing and found that changes in context had a great influenceon whether actors switched over to car sharing. Prillwitz et al.(2006) took up these results and analysed the links between thepossession of a car and particular life events by using panel data.They identified the following life events as having a significantinfluence on car possession: a change in the number of household

members (through marriage, couples moving in together, divorceor separation, children leaving their parents’ home), the birth ofthe first child, a change in the monthly income, and relocation.Using the same data Prillwitz et al. (2007) examined the relationbet-ween life events and distances travelled by passenger car.Their analysis shows that the following events increase the dis-tances travelled by car: relocations from the city centre to theperiphery, an increase of household members, a decrease of thenumber of children belonging to a household, and when the house-hold member with the highest income changes his or her workplace. Dargay et al. (2003) examined the reasons for demotorisa-tion. They found the following reasons to be the most importantfor a decreasing number of passenger cars in the household: adecrease of adult household members, relocations, and a changeof job. A further analysis of the relationship between car ownershipand key events has been conducted by Oakil et al. (2013). Theyfound strong relationships between car ownership and householdformation as well as child birth and residential relocations.

Scheiner and Holz-Rau (2013) widened the focus from car own-ership to the usage of different modes of transport. They analysedchanges in respondents travel mode choice by using data from theGerman Mobility Panel. Covering a period of 14 years (1994–2008)they give a comprehensive analysis on travel behaviour changesinduced by a wide range of key events. Using the same dataOttmann (2010) achieved similar results for three different keyevents: obtaining a driver’s license, child birth and retirement.Finally, Beige and Axhausen (2012) discussed interdependenciesof key events and long-term mobility decisions. Similar to theapproach presented in this paper, they used a life course calendarto gather retrospective data on life course events and mobilitydecisions. They particularly pointed to the fact that a change ofspatial context often triggers a change of travel behaviour ormobility tool ownership.

2.2.2. Studies focusing on the key event ‘relocation’Many studies examined the key event ‘relocation’. An overview

on the influence of residential self-selection on the relationbetween the built environment and travel behaviour has alreadybeen published elsewhere (Bohte et al., 2009). The following pub-lications put particular emphasis on behavioural changes afterrelocation. Scheiner (2005) examined to what extent relocationeffects the number of passenger cars per household. He is able toprove a relation between spatial structure and car ownership andconcludes that car owners choose their area of residency accordingto their previous travel behaviour. Further evidence for changes intravel mode use after residential relocations has been documentedin Scheiner and Holz-Rau (2012). Krizek (2000, 2003) examinedthe effects of relocations between residential areas with differentspatial structure on distances travelled per person per day andon the choice of transportation means. Klinger and Lanzendorf(2012) focus on the effect of different urban mobility culturesand attitudes on mode use changes after relocation. Brunner andHaefeli (2008) observed in an explorative qualitative study thatpeople who relocate question habitual travel behaviour and areable to establish new patterns of travel behaviour. Stanbridgeet al. (2004) surveyed in an explorative qualitative study if personschange travel behaviour after relocating. They observed a changein routine regarding the modal split in the aftermath of relocations.In some cases changes in travel behaviour occurred. However, themajority of habitual car users did not change their behaviour. A fol-low up study confirmed the results of the explorative study(Stanbridge and Lyons, 2006). Cao et al. (2007) compared personswho changed their residence within one year and persons whodid not relocate. One of the conclusions they draw is that travelbehaviour often changes due to a particular influence of the urbanstructure at the new and the former place of residence. Bamberg

Table 1Studies focusing on travel behaviour and context changes.

Author (year) Method DVar Context change B

DVar = Dependent Variable Panel/Pseudo-Panel

Retrospectivesurvey

Otherquantitativesurveymethod

Qualitativesurveymethod

Distances/No.of trips/travel time

Modalsplit

Ownershipof mobilitytools

Household/partnership/family

Education/job

Drivinglicense

Ownershipof mobilitytools

Childbirth

Relocation Othercontextchanges

Breakofhabit

B = Break of habit

Bamberg (2006) X X X XBehrens and Mistro (2010) X X X X X X X XBeige (2008), Beige and Axhausen (2012) X X X X XBrunner and Haefeli (2008) X X X X X XCao et al. (2007) X X X X XDargay et al. (2003) X X X X XDargay and Hanly (2003) X X X X X XDargay (2001) X X X XDavison and Ryley (2013) X X X XFujii et al. (2001), Fujii and Gärling

(2003)X X X X

Fujii and Kitamura (2003) X X XFujii and Gärling (2005) X X X XGoodwin (1989) X X X X X XHarms (2003) X X X X X X X X X X X XHarms and Lanzendorf (2007) X X X X XHeine et al. (2000) X X X X X X XJemelin (2006) X X X X XKlinger and Lanzendorf (2012) X X XKlöckner (2004, 2005) X X X X X X X X X X XKrizek (2000, 2003) X X X XLanzendorf (2010) X X X X X X X X X X X XOakil et al. (2013) X X X X XOttmann (2010) X X X X X X XPrillwitz et al. (2006) X X X X X X XPrillwitz et al. (2007) X X X X X X XRouwendal and Rietveld (1994) X X X XScheiner (2005) X X X XScheiner (2014) X X X X X X X X X X XScheiner and Holz-Rau (2012) X X X XScheiner and Holz-Rau (2013) X X X X X X X X X X XScholl (2002) X X XStanbridge and Lyons (2006) X X X X XStanbridge et al. (2004) X X X X Xvan der Waerden et al. (2003) X X X X X X Xvan Ommeren et al. (1999) X X X X

R.Schoenduw

eet

al./Journalof

TransportG

eography42

(2015)98–

109101

102 R. Schoenduwe et al. / Journal of Transport Geography 42 (2015) 98–109

(2006) assumes that biographical cuts generally increase the sensi-tivity for information on alternative forms of transport. Interven-tions like free public transport tickets distributed shortly afterrelocation could function as a last push to change behaviour. Bam-berg proves his assumption in an intervention study.

2.2.3. Studies focusing on work-related key eventsOther important key events which affect travel behaviour are

changes of workplaces. Such changes are often interrelated withrelocations. Van Ommeren et al. (1999) used panel data to analyseif persons accept an increase of commuting distances or if theychange their place of residence instead. Rouwendal and Rietveld(1994) pointed out that people who changed their job often com-mute over a longer distance after a change occurs. Harms andLanzendorf (2007) analysed changes in travel mode choice andmobility tool ownership among recent university graduates. Theyidentified different directions of change depending on structuraland attitudinal factors. Dargay (2001) examined the effect of incomechanges on car ownership. The analysis shows an asymmetrical rela-tionship between car ownership and income. Among other factors,Dargay ascribes this phenomenon to routine behaviour.

2.2.4. Studies focusing on household-related key eventsKey events in regard to the household primarily encompass

changes of relationship status and changes in the number of house-hold members. Lanzendorf (2010) argues on the basis of qualitativeinterviews that life events affect the ownership of mobilityresources. Secondly, Lanzendorf examined the effect of childbirthon the respondents’ modal split. He has been able to identify typ-ical patterns of travel behaviour. Heine et al. (2000) were able todemonstrate that childbirth greatly changes parents’ time budgetand mobility needs. Often, parents stated to have no choice butto purchase a passenger car.

2.2.5. Studies focusing on other key eventsThe life events mentioned so far are the most commonly

researched context changes. However, other events are also rele-vant. Scholl (2002) conducted a panel-study on adolescent travelbehaviour. Interviews revealed that travel behaviour changes dra-matically upon receipt of a driver’s license. Fujii et al. (2001) inter-viewed respondents from Osaka (Japan) regarding their choice oftransportation means on their journey to and from work. Theauthors conducted these interviews before and during the block ofa highway segment in Osaka. Fujii et al. detected a change fromthe use of passenger cars to the use of public transport. One yearafter the road block, Fujii and Gärling (2003) examined to whatextent the changes in the use of transportation means had endured.Their interviews revealed that the people who had switched to pub-lic transport now used it more often than they did before the roadblock. In a similar study Jemelin (2006) examined the reaction ofcar drivers to long-lasting road works. He was able to demonstratechanges of the modal split. Fujii and Kitamura (2003) conductedan experiment with a public transport ticket free of charge for onemonth. They examined its effect on the choice of transportationmeans. Their analysis encompassed the use of transportation means,the strength of routine, and the attitude towards the transportationmeans in question. As expected, the use of public transport sharplyincreased on average in the month of the intervention. A month laterthe use was still 20% higher than before the intervention.

The results of the studies reviewed in this section indicate thatrelocations, job changes, and changes in household compositionare most likely to evoke changes in travel behaviour. Moreover,evidence suggests that changes in mobility resources often coin-cide with changes in behaviour. This is especially true, if one clas-sifies obtainment of a driver’s license as a change in mobilityresources.

3. The life course calendar as a survey instrument

Retrospective data collection with quantitative survey instru-ments has only rarely been practised in travel behaviour research.A recent example of such a quantitative instrument is the lifecourse calendar. Similar instruments have been previously usedin social sciences, psychological und medical studies. Initially theseinstruments were developed to support and stimulate the autobio-graphical memory in qualitative studies. Alternative terms are lifehistory calendar, life events calendar (Harris and Parisi, 2007:40)and event history calendar (Belli et al., 2001).

Life course calendars have only rarely been used in transport andmobility research. Ohnmacht and Axhausen (2005) used the lifecourse calendar in an explorative qualitative study. It proved to bea helpful supplement for the interviewer and supported the inter-viewee in remembering autobiographical events. Beige (2008) usedthe life course calendar in a quantitative survey to examine long-term and mid-term mobility decisions during the life course.

3.1. Processes of recalling past events

In life course research it is often claimed that the collection oflife course data produces consistent and veridical reconstructionsof event histories (Börsch-Supan et al., 2013; Chen and Lin,2011). However, the reliability of retrospectively collected lifecourse data is often doubted, especially in terms of respondents’ability to correctly recall past events. While having reservationsabout such survey designs is plausible, this concern seems to beempirically unfounded. In fact, survey tools that tackle systematicmemory biases can profoundly improve the quality of retrospec-tive memory (Klein and Fischer-Kerli, 2000). Three different crite-ria have an impact on the reliability of the recall performance.Firstly, there is the influence of the elapsed time: the more distantthe event, the more fragmented the memory. Secondly, the subjec-tively ascribed importance and affective connotation of a pastevent is directly proportional to the individual’s recall perfor-mance, irrespective of the elapsed time. Finally, the influence ofthe general intellectual competence and the age of the respondentcan lead to a strongly selective recall performance. Harris andParisi’s (2007) combined qualitative and quantitative survey indi-cates that quantitative data collection may lead to the over-simpli-fication of life courses. The data of the quantitative survey gives theimpression that a range of people share similar kinds of lifecourses. However, more detailed questions in additional qualita-tive interviews revealed large discrepancies between these lifecourses. On the whole, there is still a lack of adequate documenta-tion and discussion of problems regarding the collection of quanti-tative data for the establishment of representative life courses(Brückner, 1990).

A specific design of the survey instrument should be used toprevent biased memories. Studies in cognitive psychology drawattention to cognitive processes which need to be considered whilesurveying retrospective data. Autobiographical memory in particu-lar depends on certain cues (van der Vaart and Glasner, 2011). Thisleads to the problem that retrospective data is especially influ-enced by the formulation and order of questions (Klein andFischer-Kerli, 2000:297). In contrast to qualitative interviews,quantitative surveys are only able to set a few cues to supportthe respondent’s memory. This is especially problematic becausethe memory of events in general exists separately of the memoryof exact times and dates. Individuals mainly remember dates bylinking them to certain contexts or by using particular recall strat-egies (Reimer, 2001:50).

Life course calendars are designed to enhance memory capacity(Freedman et al., 1988). When respondents fill events into the cal-endar grid they get a direct impression of event sequences and

R. Schoenduwe et al. / Journal of Transport Geography 42 (2015) 98–109 103

therefore are able to link additional events more easily. This‘sequential cueing’ (Belli et al., 2001:48) is one cognitive mecha-nism to build an interrelated personal time frame. Two moremechanisms are used in the life course calendar. The vertical orderof thematic fields leads from important key events which can beremembered easily to less substantial events. In this ‘top-downcueing’, intensely memorable key events are initially set as anchorpoints. These can later be used as a basis for the recall and dating ofless important events. A third recall mechanism called ‘parallelcueing’ has an effect similar to sequential cueing. It aims at associ-ations that exist across different themes. These exist because manyaspects of life and key events are directly linked to another. Forexample, a change in employment which results in the need oflong-distance-commutes may lead to a purchase of a car in a for-merly car free household.

4. Case study: context and methodology

The empirical data for our case study stems from a surveywhich has been conducted as a part of the project ‘How do peoplebuy fuel-efficient cars?’ at the IED-NSSI (Institute for Environmen-tal Decisions – Natural and Social Science Interface) of ETH Zurichin June 2006. The project goal was to understand and modeldecision processes of people who buy cars. A total of ca. 6,000questionnaires were sent to households in the German- andFrench-speaking parts of Switzerland. Participants were randomlychosen from the phone book. A subsample of 1970 questionnaireswas provided with a life course calendar. 752 questionnaires werecompleted and returned, amounting to a response rate of 38%.

Fig. 1. A completed example of the survey’s two main components (translated from Gervehicles were assessed; (B) the life course calendar. In their original form, the list (A) off2005. Source: Mueller (2009:120).

After a data quality check (for details see below) 646 question-naires could be used for further analysis.

4.1. Structure of the life course calendar used in the case study

Two retrospective elements were used to gather longitudinaldata on household composition, life events and vehicle ownership.The central piece of the survey was a life course calendar shown inFig. 1(B). The calendars purpose was to gather retrospective infor-mation on the respondents’ biography and, in particular, on house-hold composition and vehicle ownership for the time period of1991–2005. Its design was selected based on a pre-test comparingthe final version with two alternatives (Peters et al., 2007). In con-trast to the final version of the life course calendar, previous draftsincluded a tabular and a regular questionnaire design. However,these drafts proved to be less feasible than the final version. Thisversion was a year-by-question-grid with a horizontal time axisand a vertical thematic axis. This structure aims to decrease theproblem of incomplete and imperfect personal memory. The gridplaces single events into a structural relation. Hence, respondentsmay find it easier to reflect on the events’ logical order. The verticalaxis (event axis) shows different thematic subjects and categoriesrelevant to the case study. The life course calendar begins withquestions about important autobiographical family events. Giventhe exploratory character of the survey only a few key events wereincluded. Respondents were asked to give information about majorevents that have been proved to have influence on travel behav-iour. Thematic subjects comprise family history (births, deaths,weddings, moving in with/separating from a partner, relocations,retirements, and other events), postal codes or municipality names

man): (A) the list with which technical characteristics of the household’s passengerered space for eight vehicles and the life course calendar assessed the period 1991–

Table 2Key events, change of cars per household and change of mode of transport on trips to work.

EventP

% of all events percategory

Change of transport mode to workwithin same year

Odds change/nochange#

q

Key eventsRelocation 664 38.0 138 .26 ***

Change of job 346 19.8 103 .42 ***

Marriage 155 8.9 11 .08First child 122 7.0 12 .11 **

Second child 111 6.4 6 .06Separation/divorce 81 4.6 8 .11 *

Moving in with partner 78 4.5 14 .22 ***

Retirement 74 4.2 74 1.00 ***

Other (e.g. begin and end of education, death of a family member,employment abroad)

58 3.3 9 .18

Further children 56 3.2 5 .10 *

Job loss 3 0.2 3 1.00

Household eventsIncrease of no. of persons in household (hh.) 510 22.3 63 .14 ***

Decrease of no. of persons in hh. 387 16.9 53 .16 ***

Increase of no. of drivers licences in hh. 414 18.1 52 .14 ***

Decrease of no. of drivers licences in hh. 261 11.4 46 .21 ***

Increase of no. of children in hh. 426 18.6 40 .10 ***

Decrease of no. of children in hh. 287 12.6 21 .08 *

Change in car ownershipIncrease of cars in hh. 1165 60.9 76 .07 **

Decrease of cars in hh. 749 39.1 38 .05

Change of mode of trans-port on trips to workChange from other mode to car 124 18.7 – – –. . . from car to other mode 113 17.0. . . from other mode to public transport 101 15.2. . . from public transport to other mode 80 12.0. . . from other mode to foot 71 10.7. . . from foot to other mode 65 9.8. . . from other mode to bicycle 60 9.0. . . from bicycle to other mode 50 7.5

# Relative probability of a key event with change of transport mode to work within same year against a key event with no change of transport mode to work within sameyear.

* q < 0.01.** q < 0.005.

*** q < 0.001.

104 R. Schoenduwe et al. / Journal of Transport Geography 42 (2015) 98–109

of residences and work places, household details (number of mem-bers, number of children below 18, number of members having adriver’s license), used transportation means on the journey toand from work (private transport, public transport, bike, on foot)and number of cars owned (a maximum of four cars at a time).

The life course calendar had been placed at the end of the ques-tionnaire which may have affected the respondents’ answers posi-tively or negatively. By the end of the questionnaire manystatements which are related to the life course calendar havealready been made. Therefore, the respondents may already be sen-sitized for the respective issues. However, respondents may havebeen less motivated and tired when reaching the life course calen-dar after completing the questionnaire. Particularly, the repetitionof questionnaire questions in the life course calendar could nega-tively influence motivation.

Additionally, the questionnaire contained detailed questions onthe vehicles currently and previously owned (Fig. 1(A)). This infor-mation was assessed in list form and offered space for eight vehi-cles. Respondents were asked to state information about a varietyof technical and non-technical characteristics such as make, model,fuel-type, engine capacity, gear type, year of purchase, and modelyear. By using the Swiss data base of vehicle registrations furthertechnical characteristics, including CO2 emissions, curb weightand power could be assigned to each of the reported vehicles. Intotal detailed information on 1459 vehicles owned by the respon-dents during the time period of 1991–2005 was obtained.

4.2. Sample characteristics

The validity of every empirical study depends on the quality of thesurveyed data. In regard to retrospective data, a quality control isessential before processing the data. After a general assessment ofthe consistency of the life course calendar statements the data wascoded and entered into a database. In the next step, it was checkedfor errors and gaps regarding household characteristics and the tech-nical characteristics of vehicles owned from 1991 to 2005.

We checked integrity and accuracy concerning temporal order ofevents and compared life course calendar data with responses inother questionnaire sections. After the quality control and the con-sistency tests, we were able to use data of 646 respondents for thesubsequent analyses. The resulting sample of N = 646 contains63.2% male respondents. In 2005, the average age was 48.0 years(min. = 18, max. = 95, S.D. = 15.6), the median monthly householdincome was EUR 4001–5350 on a categorical scale, and the averagehousehold size was 2.5 persons (S.D. = 1.25; number of adults:M = 1.9, S.D. = .75). 32.9% of the respondents had children with amean of 1.9 children for this subsample. In 2005 8.5% of the respon-dents’ households possessed no vehicle, 52.6% lived in householdswith one vehicle, 32.3% in households with two vehicles, and 5.6%of the respondents’ households possessed three or more vehicles.In comparison, in 2005 18.8% of all Swiss households possessed novehicle, 50.6% possessed one, 25.1% possessed two and 5.1%possessed three or more vehicles (BFS and ARE, 2007).

0

6

12

18 28 38 48 58 68

0

6

12 First child (Age at which event occurs: M = 31.8; SD = 5.5; N = 122)Corresponding travel related change in same year (Age: M = 31.7; SD = 5.1; N = 47)

Further children (Age at which event occurs: M = 35.3; SD = 7.6; N = 56)Corresponding travel related change in same year (Age: M = 34.0; SD = 5.9; N = 21)

0

6

12 Second child (Age at which event occurs: M = 33.3; SD = 5.5; N = 111)Corresponding travel related change in same year (Age: M = 32.6; SD = 3.9; N = 29)

0

6

12

Change of job (Age at which event occurs: M = 32.5; SD = 11.3; N = 346)Corresponding travel related change in same year (Age: M = 31.6; SD = 10.2; N = 192)

0

6

12 Marriage (Age at which event occurs: M = 32.8; SD = 8.3; N = 155)Corresponding travel related change in same year (Age: M = 31.4; SD = 7.0; N = 43)

0

6

12Separation (Age at which event occurs: M = 40.6; SD = 8.8; N = 81)Corresponding travel related change in same year (Age: M = 38.7; SD = 8.2; N = 26)

18

24

0

6

12

Relocation (Age at which event occurs: M = 32.6; SD = 10.6; N = 664)Corresponding travel related change in same year (Age: M = 32.0; SD = 9.5; N = 302)

18

24

30

36

Fig. 2. Key events and corresponding travel related changes by respondents age.

R. Schoenduwe et al. / Journal of Transport Geography 42 (2015) 98–109 105

5. Key events and their interaction

Several recent studies indicate that major events in a person’slife course can trigger a process of reconsidering current travelbehaviour (cf. Table 1). The life course calendar in our case studyprovides information on the most important of these key events.Interrelations between these key events, changes in the choice oftransportation means for the journey to and from work, andchanges in the number of cars per household will be addressedin the following section. While focussing on work travel and vehi-cle ownership we are not able to obtain a comprehensive picture ofindividual mobility biographies. Nevertheless, using these

important aspects the overall applicability of life course calendarsfor analysing mobility biographies can be demonstrated.

5.1. What can the life course calendar tell us about key events?

We derived information on the number and temporal sequenceof key events. Relocation was the most frequently stated key eventfollowed by change of work place (cf. Table 2). On average 3.5 keyevents per person were reported. 10.2% of the respondents did notreport any event, 32.1% experienced four or more key events in thesurveyed period from 1991 to 2005. The events begin and end ofeducation, death of a family member, employment abroad and

Table 3Interaction of events with change of mode of transport on trips to work in the same year.

Grey table cells indicate the number of key events accompanied by changes of mode of transport on trips to workin the same year. White table cells indicate whether an additional key event took place within the same year.

106 R. Schoenduwe et al. / Journal of Transport Geography 42 (2015) 98–109

others were assigned to the category ‘other’. Table 2 also shows thefrequency of changes of mode of transport on trips to work thathappened in the same year as the key event. Change of job andrelocations seem to be the most important triggers for changes ofmode of transport. These results are in line with the majority ofrecent studies (see for example Klöckner, 2004; van der Waerdenet al., 2003) and are also reflected in the odds indicated in Table 1.The odds show the relative probability of a key event with changeof transport mode to work within same year against a key eventwith no change of transport mode to work within same year.Taking the examples of relocations (.26) and change of job (.42)this means that the odds to change mode of transport to workaccompanied by a change of job within the same year is 1.6 timeshigher than a change of mode of transport to work accompanied bya relocation within the same year.

We defined travel related changes as changes of mode of trans-port on trips to work and changes of the number of cars per house-hold. On average 3.7 travel related changes per respondent werereported. While 55 respondents (8.5%) did not report any change,34.5% reported four or more changes. An increase of cars perhousehold was the most frequently reported event (cf. Table 2).However, only 6.5% (N = 76) of these increases were accompaniedby a change of mode of transport to work within the same year.

5.2. Distribution of key events over life time

Taking a closer look at characteristics of the respondents itbecomes obvious that key events are age-dependent. Generally, per-sons undergo the most changes between the ages of 20 and 40. Fig. 2depicts the respondents’ reported key events by age and travel

related changes that happened in the same year (key events accom-panied by a travel related change are depicted in grey). With a shareof 55.5% changes of work place were most frequently accompaniedby travel related changes within the same year, followed by reloca-tions (45.5%) and the birth of the first child (38.5%). This share isalmost matched by the birth of the third or further children (37%).However, the birth of the second child is accompanied by travelrelated changes less frequently (26%). Marriages show a similar rate(27.7%). A comparison of the survey sample with the Swiss censusindicated that the age distribution is representative for the German-and French-speaking Swiss population.

5.3. Relationship of key events and travel related changes

Tables 3 and 4 provide information on key events which areaccompanied by travel related changes in the same year indicatinginteractions between events and changes. In contrast to Fig. 2, tra-vel related changes are divided into changes of mode of transporton trips to work (Table 3) and changes of the number of cars inhouseholds (Table 4).

Almost every second respondent of our survey (43.2%) reportedchanges of mode of transport on work trips. A total of 495 of thesechanges were reported. Almost half of these changes (N = 264)were accompanied by key events within the same year. Table 3provides information on the interaction of key events and changesof mode of transport on work trips. The grey cells show how often akey event has occurred in the same year as a change in mode oftransport. For example, 22 times the birth of a child took place inthe same year as a change in mode of transport. The cells of onetable row left and right of the grey cells depict the times a second

Table 4Interaction of events with change of number of cars in household in the same year.

Grey table cells indicate the number of key events accompanied by changes of the number of cars per householdin the same year. White table cells indicate whether an additional key event took place within the same year.

R. Schoenduwe et al. / Journal of Transport Geography 42 (2015) 98–109 107

key event additionally took place in the same year. For example, ofthe 22 child births which coincided with a change in mode oftransport, 3 child births additionally coincided with a marriage.In other words, 14% of all births taking place in the same year asa change in mode of transport also coincide with a wedding.

Table 4 shows interactions of key events with changes of thenumber of cars in households. Similar to Table 3 grey cells indicatethe share of changes of the number of cars per household accom-panied by a certain key event. A share of 15.3% (N = 99) of therespondents did not buy or sell a car within the surveyed periodfrom 1991 to 2005. The other 547 respondents reported a totalof 1914 changes of the number of cars in their households. How-ever, only 32.6% (N = 620) of all changes of cars per householdshow interactions with key events.

6. Discussion and conclusion: strengths and limitations of lifecourse calendars

Life course calendars provide detailed data on life events. Relo-cations, changes of the work place, and car purchases werereported very well by the respondents. The data analysis, whichhas been conducted on the basis of a few examples, has providedinitial results on the distribution of key events over the life courseand on the interactions of different key events and travel relatedchanges. Significant interactions were found for the key eventsrelocation, change of job, birth of first child, separation/divorce,moving in with partner and retirement.

Life course calendars help respondents to remember the occur-rence and temporal order of key events by providing a visual

representation of their life course. We share the assumption thata grid form makes it easier for respondents to tap into their contex-tual memories. However, respondents need to fill out life coursecalendars very thoroughly in order to create a helpful visual repre-sentation of their life course. Several respondents remarked thatthe completion of the life course calendar was rather challengingand demanded a lot of concentration. Although these remarks weremerely anecdotal, this issue should be considered when designinglife course calendars.

Retrospective data collected with life course calendars allowsfor the analysis of key event interaction and age-dependency.However, uncertainty regarding the criteria of completeness andaccuracy of recollection of key events may remain. The accuracyof recollection depends on many factors. For example, the charac-teristics of an event play a role. Actors remember emotionallymeaningful, momentous, unique or unexpected events especiallywell. Doubts regarding the accuracy of recollection could not bedispelled with our case study data. However, previous key eventstudies indicate that we have been able to include the most impor-tant key events in our exploratory study.

If one seeks to continue the use of life course calendars in thestudy of mobility biographies a preliminary qualitative studywould be helpful. Such a study could provide important pointersto further relevant key events. Additionally, subsequent interviewscould serve as a means to verify the accuracy of event dates andsequence as recorded in the life course calendars. Moreover, thisstudy would also allow scholars to tackle the question to whatextent the life course calendar is able to capture memories of(quasi-)routine activities.

108 R. Schoenduwe et al. / Journal of Transport Geography 42 (2015) 98–109

The temporal scale which structures the respondents’ state-ments requires further discussion. For example, car acquisitionremains a rather rare event. Thus, the survey of the car acquisitionin the observed time period of 15 years seems to be sufficientlyaccurate at the temporal scale of one year. However, the life coursecalendar in question only vaguely captures travel behaviourchanges. Additionally, the exact temporal order of key events andbehavioural changes remains unclear. For example, one can onlyspeculate whether a respondent’s change of job led him/her tobuy a car in the same year or whether the respondents’ car pur-chase was the pre-requisite for seeking a new job.

Despite this obstacle the results were still quite promising. Witha simple life course calendar on two pages we were able to generatea wealth of information and reproduce results which are prevalentin current research. The use of life course calendars can be advanta-geous, especially if they are employed to supplement extensivepanel studies. For example, panel study data could be comple-mented by data on key events in respondents’ life courses. Thiscombination allows a deeper insight into relationships betweenlong-term decisions like relocations or ownership of mobility tools,changes of household contexts and travel behaviour.

In conclusion, the life course calendar is suited for the study oflife courses in the context of transportation research. It could beused to better understand causal relationships between changesof (spatial) contexts and travel behaviour. Furthermore it couldbe used to study effects of socialisation on travel behaviour. How-ever, a more detailed survey of life courses requires a modificationof the life course calendar. Thus, we call for further improvementsof this instrument as a method of data collection in the study ofmobility biographies. As a next step, we suggest the implementa-tion of life course calendars in existing or planned panel studies.Moreover, it could be productive to investigate electronic versionsof the life course calendar employed by other disciplines. To whatextent are such versions applicable to mobility research? Forexample, it may be possible to employ life course calendars inonline surveys. Finally, cognitive psychology models indicate arange of important memory mechanisms. Future survey designas well as the studies conducted for its purpose should show morein depth engagement with this area of research.

References

Axhausen, K.W., 2008. Social networks, mobility biographies, and travel: surveychallenges. Environ. Plan. B: Plan. Des. 35, 981–996.

Bamberg, S., 2006. Is a residential relocation a good opportunity to change people’stravel behavior? Results from a theory-driven intervention study. Environ.Behav. 38, 820–840.

Behrens, R., Mistro, R.D., 2010. Shocking habits: methodological issues in analyzingchanging personal travel behavior over time. Int. J. Sustain. Transp. 4, 253–271.

Beige, S., 2008. Long-Term and Mid-Term Mobility Decisions during the Life Course:Findings from a Retrospective Survey. Südwestdeutscher Verlag fürHochschulschriften, Saarbrücken.

Beige, S., Axhausen, K.W., 2008. Long-term mobility decisions during the life course:experiences with a retrospective survey. IATSS Res. 32, 16–33.

Beige, S., Axhausen, K., 2012. Interdependencies between turning points in life andlong-term mobility decisions. Transportation 39, 857–872.

Belli, R.F., Shay, W.L., Stafford, F.P., 2001. Event history calendars and question listsurveys: a direct comparison of interviewing methods. Public Opin. Q. 65, 45–74.

BFS – Bundesamt für Statistik (Schweiz), ARE – Bundesamt für Raumentwicklung(Schweiz), 2007. Mobilität in der Schweiz. Ergebnisse des Mikrozensus 2005zum Verkehrsverhalten. Bundesamt für Statistik, Neuchâtel, Bern.

Bohte, W., Maat, K., Van Wee, B., 2009. Measuring attitudes in research onresidential self-selection and travel behaviour: a review of theories andempirical research. Transp. Rev. 29, 325–357.

Börsch-Supan, A., Brandt, M., Schröder, M., 2013. SHARELIFE – one century of lifehistories in Europe. Adv. Life Course Res. 18, 1–4.

Brückner, E., 1990. Die retrospektive Erhebung von Lebensverläufen. Kölner Z.Soziol. Sozialpsychol. 31, 374–403.

Brunner, B., Haefeli, U., 2008. Moving towards sustainability? The consequences ofresidential relocation for mobility and the built environment – methodologicalaspects of our experimental intervention study. In: 8th InternationalConference on Survey Methods in Transport, May 25–31, Annecy, France.

Cao, X., Mokhtarian, P.L., Handy, S., 2007. Do changes in neighborhoodcharacteristics lead to changes in travel behavior? A structural equationsmodeling approach. Transportation 34, 535–556.

Chen, C., Lin, H., 2011. Decomposing residential self-selection via a life-courseperspective. Environ. Plan. A 43, 2608–2625.

Dargay, J.M., 2001. The effect of income on car ownership – evidence of asymmetry.Transp. Res. Part A, 807–821.

Dargay, J.M., Hanly, M., 2003. A Panel Data Exploration of Travel to Work, EuropeanTransport Conference, Straßburg.

Dargay, J., Hanly, M., Madre, J.-L., Hivert, L., Chlond, B., 2003. Demotorisation seenthrough panel surveys: a comparison of France, Britain and Germany. In: 10thInternational Conference on Travel Behaviour Research, Luzern.

Davison, L., Ryley, T., 2013. The relationship between air travel behaviour and thekey life stages of having children and entering retirement. J. Transp. Geogr. 26,78–86.

Frändberg, L., 2006. International mobility biographies: a means to captureinstitutionalisation of long-distance travel? Curr. Issues Tourism 9, 320–334.

Frändberg, L., 2008. Paths in transnational time-space: representing mobilitybiographies of young Swedes. Geogr. Ann.: Ser. B, Human Geogr. 90, 17–28.

Freedman, D., Thornton, A., Camburn, D., Alwin, D., Young-DeMarco, L., 1988. Thelife history calendar: a technique for collecting retrospective data. Sociol.Methodol., 37–68.

Fujii, S., Gärling, T., 2003. Development of script-based travel mode choice afterforced change. Transp. Res. Part F: Traffic Psychol. Behav. 6, 117–124.

Fujii, S., Gärling, T., 2005. Temporary structural change as a strategy to break car-use habit. In: Underwood, G. (Ed.), Traffic and Transport Psychology. Theory andApplication. Proceedings of the ICTTP 2004. Elsevier, Amsterdam, pp. 585–592.

Fujii, S., Kitamura, R., 2003. What does a one-month free bus ticket do to habitualdrivers? An experimental analysis of habit and attitude change. Transportation30, 81–95.

Fujii, S., Gärling, T., Kitamura, R., 2001. Changes in drivers’ perceptions and use ofpublic transport during a freeway closure: effects of temporary structural changeon cooperation in a real-life social dilemma. Environ. Behav. 33, 796–808.

Gärling, T., Axhausen, K.W., 2003. Introduction: habitual travel choice.Transportation 30, 1–11.

Giele, J.Z., Elder, G.H.J., 1998. Methods of Life Course Research. Qualitative andQuantitative Approaches. SAGE Publications, Thousand Oaks, London, GreaterKailash.

Goodwin, P.B., 1989. Family changes and public transport use 1984–1987. Adynamic analysis using panel data. Transportation 16, 121–154.

Harms, S., 2003. Besitzen oder Teilen. Sozialwissenschaftliche Analyse des CarSharings. Verlag Rüegger, Zürich, Chur.

Harms, S., Lanzendorf, M., 2007. From university to working life: Effects of animportant biographic change on travel mode choice. In: European TransportConference, October 17–19, 2007, Noordwijkerhout.

Harris, D.A., Parisi, D.M., 2007. Adapting life history calendars for qualitativeresearch on welfare transitions. Field Methods 19, 40–58.

Heine, H., Mautz, R., Rosenbaum, W., 2000. Möglichkeiten und Grenzen desAutoverzichts. Die Wahl des Verkehrsmittels angesichts der Mobilitätszwängeund normativen Ansprüche der heutigen familiären Lebensweise, Göttingen.

Jemelin, C., 2006. How do motorists react to long road works? The case of GlionTunnel. In: 6th Swiss Transport Research Conference, Monte Verità/Ascona.

Kitamura, R., Bovy, P.H.L., 1987. Analysis of attrition biases and trip reporting errorsfor panel data. Transp. Res. Part A: General 21, 287–302.

Klein, T., Fischer-Kerli, D., 2000. Die Zuverlässigkeit retrospektiv erhobenerLebensverlaufsdaten. Analysen zur Partnerschaftsbiografie desFamiliensurvey. Z. Soziol. 29, 294–312.

Klinger, T., Lanzendorf, M., 2012. Built environment or attitudes – what affectstravel behavior of new residents? In: 13th International Conference on TravelBehaviour Research, Toronto.

Klöckner, C., 2004. How single events change travel mode choice. A life spanperspective. In: 3rd International Conference on Traffic and TransportPsychology, Nottingham.

Klöckner, C., 2005. Das Zusammenspiel von Normen und Gewohnheiten in derVerkehrsmittelwahl – ein integriertes Norm-Aktivations-Modell und seineImplikationen für Interventionen, Fakultät für Psychologie. Ruhr-Universität,Bochum.

Krizek, K.J., 2000. Pretest–posttest strategy for researching neighborhood-scaleurban form and travel behavior. Transp. Res. Rec. 1722, 48–55.

Krizek, K.J., 2003. Residential relocation and changes in urban travel. Doesneighborhood-scale urban form matter? J. Am. Plan. Assoc. 69, 265–281.

Lanzendorf, M., 2003. Mobility biographies. A new perspective for understandingtravel behavior. In: 10th International Conference on Travel BehaviourResearch, Lucerne.

Lanzendorf, M., 2004. Stability and Instability of Mobility Biographies. AnExplorative Analysis, EIRASS Workshop ‘‘Progress in Activity-based Analysis’’,Maastricht.

Lanzendorf, M., 2010. Key events and their effect on mobility biographies: the caseof childbirth. Int. J. Sustain. Transp. 4, 272–292.

Mueller, M.G., 2009. Energy-efficient Cars: Studies on Consumer-technologyInteraction. ETH Zurich, Zurich.

Naess, P., 2012. Urban form and travel behavior: experience from a Nordic context.J. Transp. Land Use 5, 21–45.

Oakil, A., Ettema, D., Arentze, T., Timmermans, H., 2013. Changing household carownership level and life cycle events: an action in anticipation or an action onoccurrence. Transportation, 1–16.

R. Schoenduwe et al. / Journal of Transport Geography 42 (2015) 98–109 109

Ohnmacht, T., Axhausen, K.W., 2005. Entwicklung des Forschungsdesign und derErhebungsinstrumente für das Projekt Mobilitätsbiographien,Mobilitätswerkzeuge und soziale Netze, Arbeitsberichte Verkehrs- undRaumplanung. ETH, IVT, Zürich.

Ohnmacht, T., Frei, A., Axhausen, K.W., 2008. Mobilitätsbiografie undNetzwerkgeografie: Wessen soziale Beziehungen sind räumlich dispers? SwissJ. Sociol. 34, 131–164.

Ottmann, P., 2010. Abbildung demographischer Prozesse in Verkehrsentstehungsmodellen mit Hilfe von Längsschnittdaten, Karlsruhe.

Peters, A., de Haan, P., Müller, M., 2007. Wave 1B der Großbefragung ‘‘Mobilität undAutokauf’’ (Befragung mit Lebensverlaufskalender) – Hintergrundbericht imRahmen des Projekts ‘‘Entscheidungsfaktoren beim Kauf treibstoff-effizienterNeuwagen’’ (überarbeitete Version vom 27. Februar 2007), Berichte zumSchweizer Autokaufverhalten. ETH, IED-NSSI, Zürich, p. 48.

Prillwitz, J., 2007. Der Einfluss von Schlüsselereignissen im Lebenslauf auf dasVerkehrshandeln unter besonderer Berücksichtigung von Wohnumzügen,Fakultät für Physik und Geowissenschaften. Universität, Leipzig.

Prillwitz, J., Harms, S., Lanzendorf, M., 2006. Impact of life-course events on carownership. Transp. Res. Rec., 71–77.

Prillwitz, J., Harms, S., Lanzendorf, M., 2007. Interactions between residentialrelocations, life course events, and daily commute distances. Transp. Res. Rec.,64–69.

Reimer, M., 2001. Die Zuverlässigkeit des autobiographischen Gedächtnisses unddie Validität retrospektiv erhobener Lebensverlaufsdaten. Kognitive underhebungspragmatische Aspekte. Max-Planck-Institut für Bildungsforschung,Berlin.

Rouwendal, J., Rietveld, P., 1994. Changes in commuting distances of Dutchhouseholds. Urban Stud. 31, 1545–1557.

Salomon, I., 1983. Life styles – a broader perspective on travel behaviour. In:Carpenter, S., Jones, P. (Eds.), Recent Advances in Travel Demand Analysis.Gower, Aldershot, Hants, pp. 290–310.

Scheiner, J., 2005. Auswirkungen der Stadt- und Umlandwanderung aufMotorisierung und Verkehrsmittelnutzung. Ein dynamisches Modell desVerkehrsverhaltens.

Scheiner, J., 2007. Mobility biographies: elements of a biographical theory of traveldemand. Erdkunde 61, 161–173.

Scheiner, J., 2014. Gendered key events in the life course: effects on changes intravel mode choice over time. J. Transp. Geogr. 37, 47–60.

Scheiner, J., Holz-Rau, C., 2012. Changes in travel mode use after residentialrelocation: a contribution to mobility biographies. Transportation 40, 431–458.

Scheiner, J., Holz-Rau, C., 2013. A comprehensive study of life course, cohort, andperiod effects on changes in travel mode use. Transp. Res. Part A: Policy Pract.47, 167–181.

Schlich, R., Axhausen, K.W., 2003. Habitual travel behaviour: evidence from a six-week travel diary. Transportation 30, 13–36.

Scholl, W., 2002. Verkehrsmittelnutzung: Jugendliche im Übergang zumErwachsenenalter. In: Scholl, W., Sydow, H. (Eds.), Mobilität im Jugend- undErwachsenenalter. Eine fünfjährige Längsschnittstudie zu Mobilitätsformen,Fahrzeugvorlieben, Freizeit- und Risikoverhalten und deren Abhängigkeit vonUmwelt- und Technikeinstellungen, Werten und Persönlichkeiten. Waxmann,Münster, New York, München, Berlin, pp. 171–249.

Schönfelder, S., Axhausen, K.W., 2009. Travel as a function of (life) projects. In:European Transport Conference, Leiden.

Schwanen, T., Banister, D., Bowling, A., 2012. Independence and mobility in laterlife. Geoforum 43, 1313–1322.

Stanbridge, K., Lyons, G., 2006. Travel behaviour considerations during the processof residential relocation. In: 11th International Conference on Travel BehaviourResearch, Kyoto.

Stanbridge, K., Lyons, G., Farthing, S., 2004. Travel behaviour change and residentialrelocation. In: 3rd International Conference on Traffic and TransportPsychology, Nottingham.

Stead, D., Marshall, S., 2001. The relationship between urban form and travelpatterns. An international review and evaluation. EJTIR 1, 113–141.

Stopher, P.R., Stecher, C., 2006. Travel Survey Methods. Quality and FutureDirections, Amsterdam.

van der Vaart, W., Glasner, T., 2011. Personal landmarks as recall aids in surveyinterviews. Field Methods 23, 37–56.

van der Waerden, P., Timmermans, H., Borgers, A., 2003. The influence of key eventsand critical incidents on transport mode choice switching behaviour: adescriptive analysis. In: 10th International Conference on Travel BehaviorResearch, Luzern.

van Ommeren, J., Rietveld, P., Nijkamp, P., 1999. Job moving, residential moving, andcommuting: a search perspective. J. Urban Econ. 46, 230–253.

Zhang, J., Yu, B., Chikaraishi, M., 2014. Interdependences between householdresidential and car ownership behavior: a life history analysis. J. Transp. Geogr.34, 165–174.

Zumkeller, D., Madre, J.-L., Chlond, B., Armoogum, J., 2006. Panel surveys. In:Stopher, P.R., Stecher, C. (Eds.), Travel Survey Methods. Quality and FutureDirections, Amsterdam, pp. 363–398.