Determining people’s ease and difficulty of movement based on observed travel behavior

Determining people’s ease and difficulty of movement based on observed travel behavior | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article Determining people’s ease and difficulty of movement based on observed travel behavior Diana Saadi, Karel Martens This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-4450289/v1 This work is licensed under a CC BY 4.0 License Status: Published Journal Publication published 17 Apr, 2025 Read the published version in Transportation → Version 1 posted 9 You are reading this latest preprint version Abstract This paper presents an approach to use GPS-based travel behavior surveys to determine who is being served well and who is being served poorly by the transport system. We draw on the extensive literature on travel behavior, which has shown that people’s travel behavior is at least in part shaped by the travel barriers they experience. Starting from this basic insight, we define 17 parameters that may provide insight into a person’s relative ease of movement. These ease of movement parameters cover dimensions related to trip frequency (e.g., overall and in evening hours), transport mode use (e.g., as driver or passenger), travel speed (e.g., for public transport legs), distance (e.g., trip detour ratio), and effort (e.g., ratio between trip legs and out-of-home activities). None of these parameters by themselves is sufficient to determine whether someone is served well or poorly by the transport system, as behaviors may be the result of choice as well as constraint. However, we argue that jointly the parameters are likely to differentiate well-served from poorly-served people. We apply our approach to data from six GPS-based travel behavior surveys conducted in Israel’s four main metropolitan areas (N = 62,981). We calculate z-scores for all ease of movement parameters, with negative values suggesting mobility problems and positive values relative ease of movement compared to the entire sample. We conduct four known-group analysis, comparing mean z-scores by level of access to a private motorized vehicle, age, gender, and disability. Results are systematically in line with expectations: population segments identified in the literature as experiencing (more severe levels of) transport disadvantage show systematically lower composite mobility scores. These outcomes are particularly striking, taking into account the short observation period of only one day per respondent. Taken together, these findings provide a first indication that revealed travel behavior patterns can be used to identify population segments poorly served by the transport system and thus to determine both success and failure of the existing transport system. While more research is needed, the approach holds promise to determine the impacts of transport investments on people’s ease of movement. ease of movement travel behavior travel behavior surveys mobility problems transport disadvantage GPS Figures Figure 1 Figure 2 Figure 3 Figure 4 Introduction This paper presents an approach to use GPS-based travel behavior surveys to determine who is being served well and who is being served poorly by the transport system. Travel behavior surveys have typically not been used for this purpose, even though an increasing number of papers have analyzed data collected through such surveys to understand the travel behavior of people deemed transport disadvantaged, i.e. groups that tend to be poorly served by the current transport system (see below). This paper is inspired by this work, but has a different purpose. Rather than focusing on population segments who are considered or expected to be transport disadvantaged, we seek to ascribe an ‘ease of movement score’ to all people included in a travel behavior survey, in order to understand who is served well and who is served poorly by the current transport system. Travel behavior surveys date back to at least the 1930s (Chapin 1974 ) and have been part-and-parcel of transport research and practice since the 1960s (de Dios Ortuzar and Willumsen 2011 ). Over the past six decades, they have evolved from the original face-to-face interview or paper-based questionnaire to GPS-based surveys (Zmud, Lee-Gosselin et al. 2013 ). The outputs of travel behavior surveys are widely used in transport practice, in particular as a key source of information to identify transport problems. For this purpose, the travel behavior patterns obtained from surveys are used as input for travel demand models, which in turn are used to obtain insights into the functioning of the transport system, usually for some year in the future. The models typically result in the identification of transport links – road segments or public transport connections – that perform poorly, i.e. links where congestion is expected to occur as predicted travel flows (demand) exceed the existing capacity (supply). The transport problem is then typically equated with these congested links in the (road or public transport) network, which drives the search for ‘solutions’ and shapes policy-makers’ attention. We argue that this is an unnecessarily indirect and highly problematic way to identify transport problems, for at least two distinct reasons. First, the approach overlooks many transport problems that have been identified in the broad literature on transport disadvantage. Second, we argue that transport planning should not focus on the functioning of the transport system itself but instead on the service people receive from that transport system. The challenge is to develop approaches that can identify who is served well by the transport system and who is served poorly, so that the policy attention will be directed to improving the situation of the latter. The analysis of the functioning of the transport system does not provide the required insight, as it does not specify how multiple issues, including a congested transport network, may accumulate for different persons into experienced mobility problems. Against this background, the purpose of the paper is to explore whether it is possible to identify who is served well and who is served poorly by the transport system based on people’s revealed travel behavior. The extensive literature analyzing people’s travel behavior suggests that people more likely to experience transport problems show distinctly different behaviors from people who can travel with relative ease (see below). Hence, in this paper we develop an approach to use data collected through GPS-based travel behavior surveys to identify the service different people receive from the transport system. The paper is structured as follows. In the next section, we present our conceptual approach of using information on people’s travel behavior to gain an understanding of their ease of movement. We then provide a brief review of the vast literature describing and explaining differences in travel behavior, highlighting how travel behavior is shaped by people’s ease of movement. We subsequently operationalize our approach and describe data and methods. We then present the results, using a set of known-group analyses to provide evidence regarding the validity of the proposed approach. We end the paper with a brief conclusion and a call for further research. Conceptual approach The purpose of this paper is to propose an approach to estimate a person’s ease of movement based on a systematic analysis of their travel behavior. A person’s ease of movement is shaped by the presence or absence of mobility problems, with the latter term referring to any difficulty a person may experience in reaching necessary or desired destinations due to the (perceived) poor functioning of the transport system. This preliminary definition is deliberately broad to encompass a range of mobility problems, varying in character, frequency and impact on people’s life (Shay, Combs et al. 2016 ). It includes long travel times for relatively short distances (for instance, due to traffic congestion or long waiting and transfer times on public transport), high travel costs, concerns over traffic or social safety, dependence on others for travel, and so on. The claim underlying our analysis is that observed travel behavior may provide, direct or indirect, evidence of the extent to which a person experiences mobility problems and thus of their ease of movement. The conceptual approach starts from the widely accepted understanding that travel behavior is the result of both choice and constraint (Hägerstrand 1970; Chapin 1974 ), with the balance between both differing from trip to trip, from day to day, and from person to person. Moreover, an identical type of observed behavior, such as a low trip frequency during a particular time period, may be the result of choice (e.g., a preference for a sedentary lifestyle) or constraint (e.g., forgone trips due to a lack of feasible transport options to reach desired destinations). Yet, drawing on the extensive literature briefly discussed in what follows, we posit that persons experiencing systematic constraints limiting their ease of movement will likely show a pattern of travel behavior that is distinct from people who enjoy a high ease of movement. Hence, we postulate that a systematic analysis of a person’s travel behavior may provide solid evidence of the extent to which they may experience mobility problems. More precisely, while travel behavior itself can never provide complete proof that someone is experiencing mobility problems, we hypothesize that this behavior can provide a robust indicator of a person’s ease of movement and thus for inverse: the probability that someone is experiencing mobility problems. Figure 1 provides a simplified representation of the conceptual approach. It starts from the understanding that a person’s ease of movement (C in the figure) is the result of the interaction between the features of the transport system (B) – across all modes – and the characteristics, abilities and circumstances of the person (A). In the ideal case, a person can make use of all available transport modes at any moment in time, without any concerns. This will rarely be the case and many people will experience some restrictions in their use of distinct transport options, for reasons of costs, regulations, impairments, personal safety concerns, racial profiling, and so on. Hence, the interplay between a person and the transport system shapes a person’s ease of movement. Persons with few or no restrictions enjoy the highest possible ease of movement in light of the qualities of the available transport system; persons with multiple restrictions will experience a low ease of movement or, in other words, will be facing multiple transport problems and, in the extreme case, transport poverty (Lucas 2012 ). In combination with the prevailing land use system (D) – the spatial and temporal distribution of activities – ease or difficulty of movement translates into ease or difficulty of access (E) (Levinson and Krizek 2005 ). Persons with limited ease of movement residing in a location with unfavorable land use patterns (e.g., few nearby activities or activities poorly served by the parts of the transport system the person can use) are likely to face accessibility problems and, in the extreme case, accessibility poverty (Jeekel and Martens 2017 ). Ease of access, in turn, interacts with a person’s needs and desires for out-of-home activity participation (A), ultimately shaping a person’s observed travel patterns (F). Not shown in the figure, but obviously of importance, is the societal context shaping a person’s particular ‘role’, required activities, legitimate desires, legitimate (travel) behaviors, and so on (e.g., Law 1999 ; Uteng and Cresswell 2016 ). The basic proposition of the proposed approach is that if travel patterns are shaped by ease of movement and ease of access, also the reverse will hold, at least to some extent: travel patterns will reflect a person’s ease of movement and ease of access. It will only hold to some extent, because a person’s needs and desires for out-of-home activity participation are an unknown ‘intervening variable’, which may strongly affect observed travel behavior. Yet, while preferences and choices obviously shape travel behavior, even for people experiencing severe restrictions on ease of movement, we assert that data on a person’s observed travel behavior over an extended period of time is highly likely to deliver solid evidence of a person’s ease of movement and ease of access. This conceptual, statistical, relationship is highlighted by the dashed black lines in Fig. 1 . In the paper, we seek to provide the first evidence of the viability of this approach. We will specify it in more detail in what follows. Given the aim to generate estimates of a person’s ease of movement – in its essence a latent variable – we are not interested in analyzing separate dimensions of travel behavior, such as trip frequencies, trip distances, or mode use, as is commonly done in the literature (see below). Rather, we view these and other dimensions as indicators of a person’s ease of movement. For example, we view a high trip frequency as an indication of a high level of ease of movement. We acknowledge that a single indicator is insufficient to determine whether a person experiences a high ease of movement. Yet, we expect that by extracting multiple indicators from observed travel behavior, it will be possible to obtain quite reliable estimates of a person’s ease of movement relative to the general population, thus making it possible to differentiate people served well by the transport system from people who are poorly served. We will operationalize the approach and specify selected ease of movement indicators below. Literature review Our approach builds on the vast literature analyzing the various relationships between the key variables in Fig. 1 : people’s characteristics and circumstances, the specifics of the transport and land use systems, ease of movement and ease of access, and people’s travel behavior. This literature includes, amongst others, the broad body of work describing and explaining differences in travel behavior, the research strand on the interrelationship between the built environment and travel behavior, the growing number of studies exploring how accessibility affects travel behavior patterns, and the multiple research strands into the transport challenges experienced by particular (disadvantaged) population segments such as low-income households, people with impairments, women, and the elderly. This broad body of literature analyzes (some of) the relationships represented by the solid lines in Fig. 1 . In contrast, our approach takes an inverse perspective in that it seeks to derive estimates of people’s ease of movement through a systematic analysis of their travel behavior patterns (with the reverse direction of this analysis represented by the black dashed line in the figure). While different, our approach does rely on the extensive evidence collected in the various bodies of literature analyzing and explaining differences in people’s travel behavior, as this evidence provides the basis for the selection of ease of movement indicators and their interpretation. This vast body of work shows that people who typically experience more restrictions on movement and thus on access, such as low-income, unemployed, and car-less individuals, people with impairments, and minorities, show distinctly different travel behavior from their more advantaged counterparts. On average, people belonging to disadvantaged population segments make less trips, travel over shorter distances, have a smaller activity space, travel more as a passenger, make more walking trips, and make more use of public transport and less of private motorized vehicles (e.g., Hine and Grieco 2003 ; Delbosc and Currie 2011 ; Lucas 2012 ; Nordbakke and Schwanen 2015 ; Lucas, Bates et al. 2016 ; Shay, Combs et al. 2016 ; Tao, He et al. 2020 ; Hidayati, Tan et al. 2021 ; Kim and Ulfarsson 2021 ). The travel behavior pattern of other population segments known to experience a range of mobility challenges, such as women and people living in rural areas, deviate to some extent from this overall pattern. For instance, the literature suggests that women tend to make more trips overall then men, even if they do tend to travel over shorter distances and more often make trips as a passenger (e.g., Law 1999 ; Priya Uteng 2021 ). Rural residents, in turn, tend to travel over relatively large distances, although this applies mostly to car-owning households, with rural residents with no or restricted car access showing small activity spaces (e.g., Nutley 2005 ; Kolodinsky, DeSisto et al. 2013 ; Chen and Akar 2016 ; van Dülmen, Šimon et al. 2022 ). Elderly people in general show lower trip frequencies, in part because they are no longer in the labor force, yet trip frequencies tend to further drop with increasing age, in part because of increasing (impairment-related) restrictions on movement (e.g., Giuliano 2004 ; Corran, Steinbach et al. 2018 ). Another line of research relevant to, but again different from, our proposed approach seeks to use data on travel behavior patterns, accessibility, and socio-economic indicators to identify transport disadvantaged population segments. For instance, Pyrialakou et al. (Pyrialakou, Gkritza et al. 2016 ) combine a composite accessibility measure with eight socio-economic characteristics to identify population segments and neighborhoods with a high ‘transport need’ yet low levels of accessibility. Comparable approaches have been adopted in other studies (e.g., Casas 2007 ; Scott and Horner 2008 ; Delbosc and Currie 2011 ; Carroll, Benevenuto et al. 2020 ). Our approach is different from this work, for three reasons. First, we are not solely interested in identifying transport disadvantaged population segments, but in the situation of the entire (sample) population. Second, we seek to estimate relative ease of movement of individuals, going beyond the disadvantaged-advantaged binary. Third, for estimating ease of movement we rely solely on people’s travel behavior patterns (again, represented by dashed black lines in Fig. 1 ), while only making use of data on personal characteristics and the transport and land use system to verify the validity of the approach (represented by the dashed grey lines in the figure). Operationalizing the approach The purpose of our approach is to estimate people’s ease of movement based on their observed travel behavior. The vast body of literature analyzing travel behavior patterns, briefly described above, suggests the feasibility of the approach. It has provided ample evidence that people more likely to experience restrictions on ease of movement, due to income, gender, car access, (age-related) impairments, residential location, and so on, show distinctly different travel behavior patterns. At the same time, this literature also underscores that the patterns are not uniform within and across population segments. For instance, while difficulties in movement do shape low trip frequencies among low-income households, women may show high trip frequencies in spite of experiencing barriers to travel. This and other such instances imply that the estimation of people’s ease of movement has to rely on multiple indicators to avoid many ‘false positives’ and ‘false negatives’. Against this background, we have identified a set of travel behavior indicators that, in combination, can provide an indication of the probability that a person is facing mobility problems. In what follows, we use the term ease of movement (EoM) parameters for these travel behaviors. We prefer the term ease of movement over ease of access parameters, because all selected indicators relate to some dimension of movement, but only a subset of them provides evidence of relative ease of access to destinations. Yet, since ease of movement strongly shapes ease of access, we would argue that our approach also provides insight into people’s ease of access. We distinguish four distinct categories encompassing a total of 17 EoM parameters: trip frequency (5 parameters), mode use (4 parameters), travel speed (4 parameters), travel distance (3 parameters), and effort (1 parameter). Table 1 gives a detailed description of the 17 selected parameters, which can all be extracted from a typical GPS-based travel behavior survey. Parameters belonging to a single category are partly overlapping, yet each parameter provides distinct information. For instance, the ease of movement parameter ‘trips per day’ is closely related to the parameter ‘activities per day’, but they are not identical. Trips are defined as a one-directional movement between a single origin and a single destination. A person who on a particular day is only going to a shop while remaining at home the rest of the day, is conducting two trips yet only one activity. The literature shows that transport disadvantaged persons tend to conduct less activities and even less trips, as they tend to combine multiple activities in a chain of one-directional trips. By including both parameters in our set, we are thus more likely to identify people who experience difficulty of movement. Table 1 Description of ease of movement parameters employed in the analyses Ease of movement parameter Description Assumed link with mobility problems Trip frequency Number of trips per day The number of one-directional trips between a single origin and a single destination conducted during the 24-hour reporting period. People who conduct more trips are more likely to enjoy ease of movement. Number of trips in evening hours (19.00-01.00h) The number of one-directional trips between a single origin and a single destination conducted between 19.00-23.00h during the 24-hour reporting period. People who conduct more trips in the evening are more likely to enjoy ease of movement, as they are more likely to have adequate transport service. Number of trips in night hours (01.00-05.00h) The number of one-directional trips between a single origin and a single destination conducted between 01.00-05.00h during the 24-hour reporting period. People who conduct more trips in the night are more likely to enjoy ease of movement, as they are more likely to have access to private transport means (car, motorbike, bicycle, other). Number of activities per day The number of stationary activities conducted at another location than the respondent's home location during the 24-hour reporting period. People who conduct more activities are more likely to enjoy ease of movement. Number of distinct activity types visited Number of distinct activity types visited during the 24-hour observation period People who visit a variety of activity types (work, shopping, leisure) are more likely to enjoy ease of movement. Mode use Number of independent motorized trip legs Number of trip legs made independently with a motorized vehicle (car, motorbike, public transport, taxi, other). People who make more independent motorized trips are more likely to enjoy ease of movement. Number of trip legs by private motorized vehicle as a driver Number of trip legs by private motorized vehicle (car, motorbike, other) as a driver during the 24-hour reporting period. People who conduct more trip legs by private motorized vehicle as a driver have higher levels of access to motorized modes and are thus more likely to enjoy ease of movement. Number of motorized trip legs as a chauffeur Number of trip legs made in a private motorized vehicle in order to bring or take someone else somewhere, during the 24-hour reporting period. People who make more trips as a chauffeur are assumed to be able to travel with relative ease and thus more likely to enjoy ease of movement. Number of motorized trip legs as a passenger Number of trip legs made as a passenger in a private motorized vehicle driven by someone else (excluding taxi rides). People who make more trips as a passenger are less likely to enjoy ease of movement. Speed Average aerial speed across all motorized trips Aerial speed is the calculated by dividing the aerial distance between origin and destination by the travel time extracted from GPS traces; the average aerial speed is obtained by averaging the values across all motorized trips conducted during the 24-hour reporting period. A trip with at least one motorized trip leg is considered a motorized trip. People who travel at higher aerial speed are more likely to enjoy ease of movement. Average aerial speed across all trips by private motorized vehicle as a driver Aerial speed is the calculated by dividing the aerial distance between origin and destination by the travel time extracted from GPS traces; the average aerial speed is obtained by averaging the values across all trips by private motorized vehicle as a driver, conducted by a respondent during the 24-hour reporting period. People who travel at higher aerial speed as a driver are more likely to enjoy ease of movement. Average aerial speed for all public transport trips Aerial speed is calculated by dividing the aerial distance between start and end point of a trip containing at least one public transport leg as extracted from GPS traces; the average aerial speed is obtained by averaging the values across all public transport trips conducted during the 24-hour reporting period. People who make public transport trips with high aerial door-to-door speeds are more likely to enjoy ease of movement. Average aerial speed for all public transport trip legs Aerial speed is calculated by dividing the aerial distance between start and end point of a public transport trip leg as extracted from GPS traces; the average aerial speed is obtained by averaging the values across all public transport trip legs conducted during the 24-hour reporting period. People who with high aerial speeds on the public transport leg of a trip are more likely to enjoy ease of movement. Distance Total aerial distance traveled Total aerial distance traveled across all trips, defined as summation of the aerial distance between origins and destinations as extracted from GPS traces for each trip. People who travel over longer (aerial) trip distance overall are more likely to enjoy ease of movement. Directness of travel Average ratio between road distance and air (= Euclidian) distance across all motorized trip legs. People with a lower ratio (i.e., people who can travel on more direct routes for each motorized trip leg) are more likely to enjoy ease of movement. Long walking trips Number of walking trips and walking trip legs over 2 kilometers in length. People who make less long walking trips are more likely to enjoy ease of movement. Effort Trip leg ratio Ratio between the total number of trip legs a person conducts and the total number of out-of-home activities conducted by the person (as specified above). People with a lower trip leg ratio (i.e., less trip legs per conducted activity) are more likely to enjoy ease of movement. Composite mobility score Composite measure of all OeM parameters Average respondent z-score derived from z-scores for OeM parameters available for the respondent People with a higher mobility score are more likely to enjoy ease of movement. Methods Data We rely on GPS-based travel behavior surveys conducted in the four metropolitan areas of Israel: Tel Aviv, Jerusalem, Haifa and Be'er Sheva (Table 2 ). It includes a total of six distinct surveys conducted between 2010 and 2018. While the surveys were conducted in different contexts across space and time, with different transport, land use, population and economic conditions, these differences are relatively modest so as not to prevent their combined use for our analyses. The travel behavior surveys combined GPS location tracking enriched with information provided by the respondents through a dedicated app. Additionally, they included a questionnaire to obtain key socio-economic characteristics of respondents and their household. The time period for which respondents reported on their trips varied between a ‘day’ (starting at home in the morning and ending at home in the evening) and up to 48 hours. We acknowledge that this short observation period is a major limitation for obtaining reliable estimates of a person’s ease of movement. Yet, since we will conduct our analyses at the level of population segments rather than individuals, we argue that the data are adequate to obtain proof-of-concept of the proposed approach. Jointly, the six surveys contain data for 106,049 respondents who made 699,785 one-directional trips during the days of observation. For the purpose of the study, the data set was cleaned to eliminate respondents for which essential socio-economic or trip data was missing. Moreover, professional drivers and respondents younger than 18 years old were excluded from the sample, as their travel behavior may be expected to be distinctly different from the ‘regular’ adult population and may thus distort the calculation of EoM parameters. The resulting cleaned data set consists of 68,424 respondents (65%) who made 620,973 trips (89%) during the days of observation. Given that the reporting period varied between respondents, the EoM parameters have been calculated for each respondent for a 24-hour period. In case a respondent reported on more than 24 hours, we used only the first 24-hour period to calculate the parameters, to avoid problematic transposing of observations across a longer time period towards a 24-hour period. Respondents reporting for less than 24 hours but still for an entire day (i.e., who explicitly marked their first activity and last activity of the day in the app) were fully included in the sample. In contrast, respondents who reported for less than 24 hours on their travel patterns and did not explicitly mark their first activity and last activity through the app, were excluded from the sample. Table 2 Overview of raw and cleaned data obtained from GPS-based travel behavior surveys (see also Appendix A). Metropolitan area Data collection period Number of households Number of respondents Number of trips Raw Cleaned Raw Cleaned Raw Cleaned Jerusalem 2010–2011 3,632 3,174 16,961 11,139 113,294 98,337 2014–2017 4,193 3,387 17,918 11,087 113,394 96,748 Tel Aviv 2014 4,586 4,287 16,149 11,287 112,698 97,859 2016–2017 4,239 4,145 16,937 10,247 108,447 98,398 Haifa 2016–2017 5,108 4,772 18,547 12,186 125,476 113,976 Be'er Sheva 2014–2015 6,309 5,967 19,537 12,478 126,476 115,655 Total 28,067 25,732 106,049 68,424 699,785 620,973 Calculating ease of movement parameters and mobility risk score The underlying assumption of the approach taken in this paper is that ease of movement is a relative phenomenon. That is, someone can move with ease in comparison to what is common in a particular society. In line with this understanding and given the fact that the units and values vary widely between the 17 EoM parameters, we made use of z-scores for each parameter. A z-score describes the position of a raw score in terms of its distance from the mean in standard deviation units. The z-score is positive if the value lies above the mean and negative if it lies below the mean. For each travel parameter, we thus calculate the mean for the entire sample of respondents and subsequently ascribe a z-score to each respondent based on the observed value, using the following widely-known equation: where x is the observed value for a respondent, μ is the mean for the sample, and σ is the standard deviation for the sample. For ease of interpretation, all 17 ease of movement parameters have been transposed, so that higher positive z-scores always imply a higher ease of movement, while negative z-scores imply relative difficulty in movement. By using z-scores in this way, it becomes straightforward to compare across respondents and EoM parameters. Moreover, it makes it possible to calculate the composite mobility score with ease. This score is calculated by first taking, for each respondent, the average value of the z-scores for the relevant EoM parameters. The resulting average values are then again normalized for the entire sample using z-scores, so that the average composite mobility score for the entire sample will be zero. Lacking theoretical arguments for ascribing weights to particular EoM parameters, no weighing is applied when calculating a respondent’s composite mobility score. Note that the use of z-scores does not imply that the approach will always identify people at risk of mobility problems, even when none exist. The size of the differences in z-scores matters. In cases where differences in ease of movement between people are small, we would expect small differences in z-scores and weak relationships between z-scores and respondents’ socio-economic profiles. In contrast, in contexts with substantial differences in ease of movement, we would expect large differences in z-scores, as well as systematic patterns in z-scores between distinct population segments. Thus, while the approach relies on a comparison across people and is thus relative in nature, it does not inevitably imply that substantial risks of mobility problems will be identified even if none exist. Pre-assessment of ease of movement parameters Before conducting the analyses, we conducted a range of tests to determine which ease of movement parameters and which respondents to include in the analyses. These tests are described in detail in Appendix B. Based on the tests, we decided to exclude respondents with zero trips during the observation period (N = 5,443; 8%), while including all ease of movement parameters in the analyses. Obviously, making no trips at all may be a strong indicator of difficulties in movement, which is indeed confirmed in part by the socio-economic profile of the relevant respondents (Table B.5 in Appendix B). Yet, given the short observation period and the limited number of ease of movement parameters that can be calculated for this sub-segment of the sample, it was decided to exclude them from the analysis. The number of ease of movement parameters that can be calculated for the respondents included in the analyses (N = 62,981; 92%) depends on their observed travel behavior. Since all included respondents made at least one trip during the observation period, it was possible to calculate at least 12 parameters for each respondent. This minimum was calculated for only 15% of respondents (N = 9,525), while for 48% of the respondents it was possible to calculate 15 or more parameters (N = 30,088) (Appendix C). Results The purpose of the current paper is to provide a first proof-of-concept. Hence, we have conducted four known-group analyses to determine whether the proposed approach can deliver the expected results: by level of access to a private motorized vehicle; by age; by gender; and by disability. For the latter three groups, we conduct the analyses for the entire sample and by vehicle access level. The first known-group analysis distinguishes between respondents based on their access to private motorized vehicles (cars or motorbikes). We distinguish between four distinct groups: (1) respondents who do not have a private motorized vehicle in the household (N = 15,653, 24.9%); (2) respondents who have a private motorized vehicle in the household and share it with more than one other adult (N = 15,272, 24.3%); (3) respondents who have a private motorized vehicle in the household and share it with only one other adult (N = 16,047, 25.5%); and (4) respondents who are the sole user of their own private motorized vehicle (N = 15,919, 25.3%). Figure 2 provides the average scores for all 17 EoM parameters as well as for the composite mobility score. The results are nearly perfectly in line with expectations: as access to a motor vehicle goes up, z-scores on all but one parameter and on the composite mobility score go up (the only exception is ‘number of trips in night hours’, where respondents who share a motor vehicle with one adults have a higher z-score than respondents with their own vehicle). Among others, respondents without private vehicle access conduct much more long walking trips, less overall trips and less trips in evening hours, more trips as a passenger, and travel at lower speeds when using public transport. Particularly striking are the differences in number of independent motorized trip legs and average aerial speed across all independent motorized trips. The composite mobility score for each individual respondent separately is also surprisingly in line with expectations, in spite of the short observation period (Fig. 3 ). The vast majority of respondents without access to a motor vehicle show composite mobility scores well below the average, with 65.4% of respondents having a z-score below − 2. Moreover, only 3.1% of this population segment has an above-average composite mobility score. The exact opposite holds for respondents who have their own motor vehicle: 61.7% has a composite mobility score above + 2, while only 5.7% has a below-average score. The situation for the other two segments of respondents falls between these extremes and exactly in line with expectations: respondents who share a motor vehicle with two or more adults show lower composite mobility scores than respondents who share their vehicle with only one other adult. Interestingly, the range of observed composite mobility score for these two segments is substantially smaller than that for the respondents without or with their own motor vehicle. The second known-group analysis analyzes the composite mobility score by age, for the entire sample and for the four vehicle access levels. The literature shows that both younger and older people are more likely to experience mobility problems. For the younger segment is related in part to relatively low income levels and limited vehicle access (e.g., Ralph 2015 ; Klein and Smart 2017 ; Ralph 2017 ), while for the older segment mobility problems are partly the result of increasing travel-related impairments and related (gradual) driving cessation, lower income levels, as well as concerns over social safety (e.g., Páez, Scott et al. 2007 ; Luiu, Tight et al. 2017 ; Corran, Steinbach et al. 2018 ; Palm, Nakshi et al. 2024 ). Our results are perfectly in line with these results, with below-average composite mobility scores for respondents up to age ~ 33 and from age ~ 64 onwards. The pattern is remarkably similar for each of the four segments by vehicle access, even though the age range within which the average respondent experiences an above-average ease of movement varies. This range is largest for respondents who have their own motor vehicle (ranging from ages ~ 24 to ~ 70), and the smallest for respondents without access to a motor vehicle (ranging from ages ~ 33 to ~ 54). For all motor vehicle segments, average composite mobility score drops sharply at older age, again in line with expectations (e.g., Siren and Hakamies-Blomqvist 2004 ). The fact that respondents without access to a motor vehicle have above-average composite mobility scores between ages ~ 33 and ~ 54, in combination with the declining scores for all population segments irrespective of vehicle access at older age, raises the question whether the composite mobility score is not shaped too strongly by parameters related to trip frequency. The above-average composite mobility scores occur especially in the age range when respondents may be both employed and are taking care of children, which may result in relatively high trip rates, positively affecting the composite mobility score even if respondents may experience substantial mobility problems. Hence, we conducted an additional analysis and calculated the composite mobility score based on 12 EoM parameters only, excluding the five trip frequency parameters (number of trips per day; number of trips in evening hours; number of trips in night hours; number of activities per day; and number of distinct activity types visited). This analysis results in virtually identical patterns for all population segments (Fig. 4 ). Only for respondents with their own motor vehicle does the elimination of frequency-related EoM parameters result in relatively higher composite mobility scores for respondents from age ~ 36 onwards, suggesting that their relatively high ease of movement only partially translates into a higher trip frequency in comparison to people with lower levels of access to a motor vehicle. For all other vehicle segments, the analysis shows that the observed lower composite mobility scores at younger and older ages are hardly shaped by trip frequency parameters, suggesting that the composite mobility score indeed captures ease of movement. The third known-group analysis explored the impact of gender. Extensive research shows that women have distinct travel patterns from men and many studies also show that women experience more mobility problems, in part due to limited access to a motor vehicle, lower incomes, and concerns over social safety (e.g., Law 1999 ; Siren and Hakamies-Blomqvist 2006 ; Uteng and Cresswell 2008 ; Priya Uteng 2021 ; Zhang, Zhao et al. 2022 ). Our results are in line with the literature: women have a somewhat lower, statistically significant, composite mobility score than men (-0.31 versus + 0.22; p < 0.05). Comparable modest differences in composite mobility score are found for all four population segments by vehicle access, with the differences being significant for all four segments except respondents who share a vehicle with three or more adults (Table 3 ). These results underscore that women’s ease of movement is not only hindered because of lower vehicle access, but by other factors as well, as suggested by the literature. A further analysis shows that women tend to show lower scores on most EoM parameters (Appendix D). The situation is particularly striking for women with access to a car, who show significantly lower z-scores on all parameters, except for number of activities per day and number of long walking trips, where not significant differences between women and men are observed (Appendix D). The pattern is more diverse for respondents without a car. In line with the literature, women without a car show much lower z-scores for number of trips in the evening and night hours (p < 0.001). Surprisingly, they also perform a smaller diversity of activities during the day than men without a motor vehicle. Also somewhat at odds with the literature is that women show higher z-scores for a number of speed parameters and number of long walk trips, suggesting that they travel on average at higher speeds and make less long walking trips than their male counterparts. Table 3 Composite mobility score by gender for each vehicle segment separately, for the entire sample (N = 62,981). Population segment Number of respondents Composite mobility score Women Men Women Men Statistical test n % n % Mean SD Mean SD p-value t value Full sample 33,842 100% 29,139 100% -0.31 1.85 0.22 1.72 2.58* 0.04 No motor vehicle 8,938 26.4% 6,715 23.0% -2.71 2.41 -2.42 1.18 2.58* 0.03 Vehicle shared with multiple adults 8,582 25.4% 6,780 23.2% -0.84 2.36 -0.78 1.31 0.16 1.07 Vehicle shared with one adult 8,564 25.3% 7,483 25.6% 0.75 2.75 0.87 1.63 2.38* 0.04 Sole user of vehicle 7,758 22.9% 8,161 27.9% 2.46 2.42 2.71 2.31 2.69* 0.01 *p < .05 The final known-group analysis compares ease of movement for people with and without impairments. The literature shows that people with physical, sensory, or cognitive impairments experience a range of restrictions on movement and are less mobile as a result (e.g., Imrie 1996 ; Shoval, Wahl et al. 2011 ; Sammer, Uhlmann et al. 2012 ; Ratering, Van der Heijden et al. 2024 ). People with impairments not only experience more mobility problems than others due to a mismatch between their abilities and the specifics of the transport system, but also because disability affects economic opportunity, resulting in lower incomes and thus in lower vehicle access, both of which affect ease of movement. Since only the travel behavior surveys conducted in the Jerusalem metropolitan area included a question about respondents’ disability status, the known-group analysis is limited to this sub-sample (N = 20,986; 33% of the sample for all four metropolitan areas). From this group, 37% (7,764 respondents) indicated that they are disabled. This share is much higher than commonly seen in the literature, which typically reports shares ranging between 6% and 10% (Martens 2018 ). Hence, this set of respondents probably also includes people with only modest impairments, so that the results of our analysis should be treated with some caution (see Appendix E). Yet, in spite of these concerns, the results are in line with expectations. While the Jerusalem sub-sample without impairments enjoys a slightly higher ease of movement than the entire set of respondents (which, of course, also includes people with impairments in the other metropolitan areas), the sub-sample with impairments has a composite mobility score well below the average: +0.12 versus − 0.48. The difference between the two sub-segments is statistically significant (p < 0.01). Significant differences comparable in size can also be observed when conducting the analysis by vehicle access level. In all cases, the composite mobility score for people without impairments is higher than for people with impairments (Table 4 ). Moreover, in line with expectations, for both respondents with and without impairments the composite mobility score increases substantially as the level of vehicle access goes up. Interestingly, the differences by vehicle access are much larger than the differences between respondents with and without impairments. For instance, while people with impairments with their own private vehicle have a substantially lower composite mobility score than their non-impaired counterparts (+ 1.87 versus + 2.69; p < 0.05), they still have a much higher score than non-impaired respondents without vehicle access (+ 1.87 versus − 2.66). The large differences across level of car access in combination with the relatively modest differences related to impairment, suggest that vehicle access level is a much stronger predictor of ease of movement than impairment itself. To determine whether this holds, we conducted a stepwise regression with backward elimination. The initial model included a broad range of socio-economic and land use variables, among which a disability variable (see explanation below). The final model retained six significant predictors: age, household size, disability, car ownership, population density, and employment opportunity. These six factors collectively explained 46.4% of the variance in the composite mobility score, with a statistically significant model fit (F (6, 20979) = 58.64, p < .001). The analysis confirms the dominant importance of car ownership level in explaining the composite mobility score, even if disability was also found to have a significant impact (Table 4 ). Table 4 Stepwise regression for composite mobility score for respondents in the Jerusalem region (N = 20,986). Variable β SEB t-value p-value Age -0.26 0.17 -2.51** .006 Household size -0.29 0.18 -3.29*** < .001 Disability -0.27 0.16 2.27** 0.008 Car ownership level 0.57 0.29 8.62*** < .001 Population Density 0.28 0.19 2.62** .004 Employment opportunity 0.25 0.35 2.41** .005 *p < .05, **p < .01, ***p < .001. Finally, we employed another stepwise regression with backward elimination for the entire sample, to identify the factors that most significantly predict a respondent’s composite mobility score. The included factors in the initial model are age, gender, years of education, household size, number of children in the household, household type, religion, driving license (binary: yes/no), motor vehicle access level (four distinct categories, as specified above), population density in zone of residence, and employment opportunity in zone of residence. Table 5 presents the final model. Five significant variables were retained: age, household size, motor vehicle access level, population density, and employment opportunity. These five factors collectively explained 44.6% of the variance in mobility score, with a statistically significant model fit (F (5, 62976) = 56.73, p < .001). Most importantly, all retained variables show the expected sign, supporting the claim that the composite mobility score indeed captures ease of movement. As shown in the table, motor vehicle access level (β = 0.53, p < .001) and household size (β = -0.38, p < .001) emerged as the strongest predictors of composite mobility score. Age also emerged as a relevant predictor, confirming the bivariate analyses presented above showing that especially at older age composite mobility score drops substantially. Gender was not retained in the final model. While at odds with the literature, this result is in line with the finding above that women and men show relatively modest differences in composite mobility score. Table 5 Stepwise regression for composite mobility score for the entire sample (N = 62,981). Variable β SEB t-value p-value Age -0.28 0.16 -2.44** .007 Household size -0.38 0.14 -4.07*** < .001 Car ownership level 0.53 0.31 8.43*** < .001 Population density 0.29 0.17 2.81** .003 Employment opportunity 0.26 0.37 2.83** .002 *p < .05, **p < .01, ***p < .001. Conclusion This paper presented an approach to use data on observed travel behavior to determine who is being served well and who is being served poorly by the transport system. Drawing on the extensive literature analyzing and explaining travel behavior, we identified 17 parameters assumed to capture a person’s relative ease of movement. We included parameters regarding trip frequency, speed, travel distance, and effort. We acknowledge that none of these parameters by themselves is sufficient to determine whether someone is served well or poorly by the transport system, as travel behavior may be the result of choice as well as constraint or, perhaps more precisely, choices made within constraints, with the strength of the latter depending on a person’s specific circumstances. However, we argue that jointly the parameters are likely to differentiate between well-served and poorly-served people. We applied our approach to data from six GPS-based travel behavior surveys conducted in Israel’s four main metropolitan areas (N = 62,981 valid respondents). Given the variety in measurement units and values, we relied on z-scores for all ease of movement parameters (EoM parameters), with negative values suggesting low levels of ease of movement and thus mobility problems, and positive values relative ease of movement, both in comparison to the entire sample. In addition to the 17 ease of movement parameters, we calculated a composite mobility score for each respondent based on the respondent’s z-scores for all 17 EoM parameters. We subsequently conducted four known-group analyses to verify the validity of the proposed approach. Results from these analyses are in line with expectations and thus confirm the potential of the approach. The first analysis, comparing four population segments differing in their level of access to private motorized vehicles (car or motorbike), showed that z-scores systematically increase for all EoM parameters as access to private motorized vehicles improves. Respondents without a motor vehicle scored poorest on all parameters on average. Among others, they conduct much more long walking trips, less overall trips and less trips at night, more trips as a passenger, and travel at lower speeds when using public transport. The second known-group analysis examined the composite mobility score across age. In line with the literature, we found that both younger and older population segments experience lower ease of movement, irrespective of whether the composite mobility score was based on the complete set or a reduced set of EoM parameters. Results also show a sharp decline in ease of movement as people age for all car access segments. The third known-group analysis by gender shows a somewhat lower ease of movement for women, irrespective of vehicle access, again in line with the literature. Finally, a comparison between people with and without impairments for the Jerusalem sub-sample (N = 20,986) also showed differences in line with expectations: people with impairments have a significantly lower ease of movement than people without impairments. Regression analyses for the entire sample and the Jerusalem sub-sample confirmed that vehicle access, age, and disability shape ease of movement, but did not corroborate the relevance of gender. Taken together, these results thus provide first evidence in support of the fundamental claim of this paper: an estimate of a person’s ease of movement can be derived from a systematic analysis of their observed travel behavior. This outcome is particularly striking, taking into account the short observation period per respondent. Clearly, more research is needed to refine and test the approach. Several directions for further research can be highlighted. First, the approach should be tested on datasets that include observations on people’s travel behavior across multiple days, as studies have shown that people’s travel behavior fluctuates substantially over time (Schlich and Axhausen 2003 ; Raux, Ma et al. 2016 ; Deschaintres, Morency et al. 2022 ). Second, a systematic assessment of the relevance of different ease of movement parameters can strengthen the approach. The current set is suitable for contexts in which car-based travel provides superior ease of movement in most cases. In the ideal case, only ‘mode-agnostic’ ease of movement parameters would be included, so that the approach can be applied across a range of contexts, including metropolitan areas with outstanding public transport service. Such a systematic assessment of potential EoM parameters should also take into account the data that are typically collected in GPS-based travel behavior surveys across the world, so that an approach can be developed that would allow for comparative analyses across geography. Finally, and perhaps most essential, the approach should be scrutinized for external validity, by collecting, among the same population sample, both quantitative data on respondents’ travel behavior and qualitative data on their travel experiences and challenges, either through bespoke surveys or in-depth interviews (Murphy, Gould-Werth et al. 2021 ; Singer and Martens 2023 ), and subsequently assessing whether ease of movement estimates derived from the former indeed correlate systematically with the latter. Declarations Author Contribution Diana Saadi: Data cleaning and preparation; Data analysis; Methodology; Software; Validation; Visualization.Karel Martens: Conceptualization; Funding acquisition; Methodology; Project administration; Supervision; Validation; Writing - original draft; Writing - review & editing. All authors reviewed the manuscript. Acknowledgement The authors would like to thank Sami Adri of Netivei Ayalon Highways Company for generously sharing the processed travel behavior survey data. Data Availability Data used in the paper are managed by the Netivei Ayalon Highways Company, a government-owned company. Data may be requested from the company. References Carroll, P., R. Benevenuto and B. Caulfield (2020). "Identifying Hotspots of Transport Disadvantage and Car Dependency in Rural Ireland." Transport Policy . Casas, I. (2007). "Social Exclusion and the Disabled: An Accessibility Approach." The Professional Geographer 59 (4): 463-477. Chapin, F. S. (1974). Human activity patterns in the city: things people do in time and in space . New York, Wiley. Chen, N. and G. Akar (2016). "Effects of neighborhood types & socio-demographics on activity space." Journal of Transport Geography 54 : 112-121. Corran, P., R. Steinbach, L. Saunders and J. Green (2018). "Age, disability and everyday mobility in London: An analysis of the correlates of ‘non-travel’ in travel diary data." Journal of Transport & Health 8 : 129-136. de Dios Ortuzar, J. and L. G. Willumsen (2011). Modelling Transport . Chichester, John Wiley & Sons. Delbosc, A. and G. Currie (2011). "Transport problems that matter - social and psychological links to transport disadvantage." Journal of Transport Geography 19 (1): 170-178. Deschaintres, E., C. Morency and M. Trépanier (2022). "Cross-analysis of the variability of travel behaviors using one-day trip diaries and longitudinal data." Transportation Research Part A: Policy and Practice 163 : 228-246. Giuliano, G. (2004). "Land use and travel patterns among the elderly." Transportation in an Aging Society 27 : 192-210. Hidayati, I., W. Tan and C. Yamu (2021). "Conceptualizing Mobility Inequality: Mobility and Accessibility for the Marginalized." Journal of Planning Literature 36 (4): 492-507. Hine, J. P. and M. S. Grieco (2003). "Scatters and clusters in time and space: implications for delivering integrated and inclusive transport." Transport Policy 10 (4): 299-306. Imrie, R. (1996). "Disability and discourses of mobility and movement." Environment and Planning A 32 : 1641-1656. Jeekel, H. and K. Martens (2017). "Equity in transport: Learning from the policy domains of housing, health care and education." European Transport Research Review 9 (53). Kim, S. and G. F. Ulfarsson (2021). "Staying home or going places: Mobility factors of older minority women's daily trip making in the United States." Journal of Transport & Health 21 : 101031. Klein, N. J. and M. J. Smart (2017). "Millennials and car ownership: Less money, fewer cars." Transport Policy 53 : 20-29. Kolodinsky, J. M., T. P. DeSisto, D. Propen, M. E. Putnam, E. Roche and W. R. Sawyer (2013). "It is not how far you go, it is whether you can get there: Modeling the effects of mobility on quality of life in rural New England." Journal of Transport Geography 31 : 113-122. Law, R. (1999). "Beyond 'women and transport': towards new geographies of gender and daily mobility." Progress in Human Geography 23 (4): 567-588. Levinson, D. M. and K. J. Krizek, Eds. (2005). Access to destinations . Lucas, K. (2012). "Transport and social exclusion: Where are we now?" Transport Policy 20 : 105-113. Lucas, K., J. Bates, J. Moore and J. A. Carrasco (2016). "Modelling the relationship between travel behaviours and social disadvantage." Transportation Research Part A: Policy and Practice 85 : 157-173. Luiu, C., M. Tight and M. Burrow (2017). "The unmet travel needs of the older population: a review of the literature." Transport Reviews 37 (4): 488-506. Martens, K. (2018). "Ageing, impairments and travel: Priority setting for an inclusive transport system." Transport Policy 63 : 122-130. Murphy, A. K., A. Gould-Werth and J. Griffin (2021). "Validating the Sixteen-Item Transportation Security Index in a Nationally Representative Sample: A Confirmatory Factor Analysis." Survey Practice 14 (1): 27185. Nordbakke, S. and T. Schwanen (2015). "Transport, unmet activity needs and wellbeing in later life: exploring the links." Transportation 42 (6): 1129-1151. Nutley, S. (2005). "Monitoring rural travel behaviour: a longitudinal study in Northern Ireland 1979-2001." Journal of Transport Geography 13 (3): 247-263. Páez, A., D. Scott, D. Potoglou, P. Kanaroglou and K. B. Newbold (2007). "Elderly mobility: demographic and spatial analysis of trip making in the Hamilton CMA, Canada." Urban Studies 44 (1): 123-146. Palm, M., P. Nakshi, E. Yousefzadeh Barri, S. Farber and M. Widener (2024). "Uncovering suppressed travel: A scoping review of surveys measuring unmet transportation need." Travel Behaviour and Society 36 : 100784. Priya Uteng, T. (2021). Chapter Two - Gender gaps in urban mobility and transport planning. Advances in Transport Policy and Planning . R. H. M. Pereira and G. Boisjoly, Academic Press. 8: 33-69. Pyrialakou, V. D., K. Gkritza and J. D. Fricker (2016). "Accessibility, mobility, and realized travel behavior: Assessing transport disadvantage from a policy perspective." Journal of Transport Geography 51 : 252-269. Ralph, K. M. (2015). Stalled on the road to adulthood? Analyzing the nature of recent travel changes for young adults in America, 1995 to 2009 , University of California, Los Angeles. Ralph, K. M. (2017). "Multimodal Millennials? The Four Traveler Types of Young People in the United States in 2009." Journal of Planning Education and Research 37 (2): 150-163. Ratering, C., R. Van der Heijden and K. Martens (2024). "Moving around with Anxiety Disorder." Transportation Research Part F: Psychology and Behaviour 100 : 493-506. Raux, C., T.-Y. Ma and E. Cornelis (2016). "Variability in daily activity-travel patterns: the case of a one-week travel diary." European Transport Research Review 8 (4): 26. Sammer, G., T. Uhlmann, W. Unbehaun, A. Millonig, B. Mandl, J. Dangschat and R. Mayr (2012). "Identification of mobility-impaired persons and analysis of their travel behavior and needs." Transportation Research Record 2320 (1): 46-54. Schlich, R. and K. W. Axhausen (2003). "Habitual travel behaviour: Evidence from a six-week travel diary." Transportation 30 (1): 13-36. Scott, D. M. and M. W. Horner (2008). "The role of urban form in shaping access to opportunities: An exploratory spatial data analysis." Journal of transport and land use 1 (2): 89-119. Shay, E., T. S. Combs, D. Findley, C. Kolosna, M. Madeley and D. Salvesen (2016). "Identifying transportation disadvantage: Mixed-methods analysis combining GIS mapping with qualitative data." Transport Policy 48 : 129-138. Shoval, N., H.-W. Wahl, G. Auslander, M. Isaacson, F. Oswald, T. Edry, R. Landau and J. Heinik (2011). "Use of the global positioning system to measure the out-of-home mobility of older adults with differing cognitive functioning." Ageing & Society 31 (5): 849-869. Singer, M. E. and K. Martens (2023). "Measuring travel problems: Testing a novel survey tool in a natural experiment." Transportation Research Part D: Transport and Environment 121 : 103834. Siren, A. and L. Hakamies-Blomqvist (2004). "Private car as the grand equaliser? Demographic factors and mobility in Finnish men and women aged 65+." Transportation Research Part F: Traffic Psychology and Behaviour 7 (2): 107-118. Siren, A. and L. Hakamies-Blomqvist (2006). "Does gendered driving create gendered mobility? Community-related mobility in Finnish women and men aged 65+." Transportation Research Part F: Traffic Psychology and Behaviour. Special Issue on Older drivers' safety and mobility: Current and future issues 9 (5): 374-382. Tao, S., S. Y. He, M.-P. Kwan and S. Luo (2020). "Does low income translate into lower mobility? An investigation of activity space in Hong Kong between 2002 and 2011." Journal of Transport Geography 82 : 102583. Uteng, T. P. and T. Cresswell, Eds. (2008). Gendered mobilities . Transport and Society. Aldershot, Ashgate. Uteng, T. P. and T. Cresswell, Eds. (2016). Gendered mobilities . London/New York, Routledge. van Dülmen, C., M. Šimon and A. Klärner (2022). "Transport poverty meets car dependency: A GPS tracking study of socially disadvantaged groups in European rural peripheries." Journal of Transport Geography 101 : 103351. Zhang, M., P. Zhao and X. Tong (2022). "Constructing women's immobility: Fear of violence and Women’s constricted nocturnal travel behaviour." Travel Behaviour and Society 26 : 178-192. Zmud, J., M. Lee-Gosselin, M. Munizaga and J. A. Carrasco, Eds. (2013). Transport Survey Methods: Best Practice for Decision Making . Bingley, Emerald. Additional Declarations No competing interests reported. Supplementary Files AppendixADataset.docx AppendixBTestingparametersandrespondentsv01km.docx AppendixCParametersversusrespondents.docx AppendixDComparisonacrossgender.docx AppendixERespondentswithoutimpairments.docx Cite Share Download PDF Status: Published Journal Publication published 17 Apr, 2025 Read the published version in Transportation → Version 1 posted Editorial decision: Revision requested 04 Dec, 2024 Reviews received at journal 13 Nov, 2024 Reviews received at journal 03 Nov, 2024 Reviewers agreed at journal 13 Oct, 2024 Reviewers agreed at journal 12 Oct, 2024 Reviewers invited by journal 10 Oct, 2024 Editor assigned by journal 01 Jul, 2024 Submission checks completed at journal 21 May, 2024 First submitted to journal 20 May, 2024 You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-4450289","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":324871640,"identity":"116051e5-9f9c-478c-9ea4-907928934b19","order_by":0,"name":"Diana Saadi","email":"","orcid":"","institution":"Technion – Israel Institute of Technology","correspondingAuthor":false,"prefix":"","firstName":"Diana","middleName":"","lastName":"Saadi","suffix":""},{"id":324871641,"identity":"6c32dcb7-f8fd-48be-b202-23983b983b15","order_by":1,"name":"Karel Martens","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA7ElEQVRIiWNgGAWjYBACCRiDH4iZEXwDC/xaDgCxZBuqFgnCWgyOgbVgWo8BJKcdfvb5Q8Ude+P7zQc/F+6wkDNnYH74gaEAtxZp6TTjGQfOPEvcdowtWXrmGQljywY2Ywl8DpOTTjBmONh2OMHsGI8ZM2+bROKGAwxmeP0iJ53+meHgv8P2xm3836Ba2L/h1SItnQO0peEw4wY2HjaoFh78tkjOzilmOHPscOKMY2nG0kAtxgaHeYolEvAF8u30zQwVNYft+ZsPP/zM21YnZ3C8feOHD39scGrBAkCxk0CKhlEwCkbBKBgFGAAAKqlKNc+rGaQAAAAASUVORK5CYII=","orcid":"","institution":"Technion – Israel Institute of Technology","correspondingAuthor":true,"prefix":"","firstName":"Karel","middleName":"","lastName":"Martens","suffix":""}],"badges":[],"createdAt":"2024-05-20 16:27:31","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-4450289/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-4450289/v1","draftVersion":[],"editorialEvents":[{"content":"https://doi.org/10.1007/s11116-025-10604-x","type":"published","date":"2025-04-17T15:58:02+00:00"}],"editorialNote":"","failedWorkflow":false,"files":[{"id":60994177,"identity":"b7fc8757-bcd1-440d-bc20-21b865f5aace","added_by":"auto","created_at":"2024-07-24 11:52:14","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":47087,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eConceptual model underpinning the study.\u003c/strong\u003e\u003c/p\u003e","description":"","filename":"1.png","url":"https://assets-eu.researchsquare.com/files/rs-4450289/v1/84aea6fbe65ef837cb776ab0.png"},{"id":60994180,"identity":"f54ca587-204f-4790-85ef-a88130ecffcc","added_by":"auto","created_at":"2024-07-24 11:52:14","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":61599,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eAverage z-scores for all 17 ease of movement parameters and the composite mobility score, for four distinct groups of respondents.\u003c/strong\u003e\u003c/p\u003e","description":"","filename":"2.png","url":"https://assets-eu.researchsquare.com/files/rs-4450289/v1/bc9a1b8c6891b4c030f92548.png"},{"id":60994174,"identity":"ccaa927b-c3c8-4d62-88cb-5922bfb3a0cf","added_by":"auto","created_at":"2024-07-24 11:52:12","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":71681,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eDistribution of respondents’ composite mobility score, distinguished by level of access to a motor vehicle.\u003c/strong\u003e\u003c/p\u003e","description":"","filename":"3.png","url":"https://assets-eu.researchsquare.com/files/rs-4450289/v1/8f65e55212bc23bad2a47e46.png"},{"id":60994183,"identity":"85119b15-81ea-4823-9743-0fc8fefd24de","added_by":"auto","created_at":"2024-07-24 11:52:15","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":61222,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eEase of movement as captured by the composite mobility score, for the entire sample and the four population segments by access to a motor vehicle. Composite mobility scores are calculated using a three-year rolling average for age, for ease of interpretation.\u003c/strong\u003e\u003c/p\u003e","description":"","filename":"4.png","url":"https://assets-eu.researchsquare.com/files/rs-4450289/v1/1c5018c45ace53a291726276.png"},{"id":81051517,"identity":"6ad425af-f90e-4224-906f-99f49e290146","added_by":"auto","created_at":"2025-04-21 16:10:57","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":1702735,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-4450289/v1/c4d7be9d-5b10-47b3-a6eb-32226ba30763.pdf"},{"id":60994173,"identity":"fd086a28-f988-46e7-b7fd-524a10b8227f","added_by":"auto","created_at":"2024-07-24 11:52:12","extension":"docx","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":18638,"visible":true,"origin":"","legend":"","description":"","filename":"AppendixADataset.docx","url":"https://assets-eu.researchsquare.com/files/rs-4450289/v1/ea96471e84379dda5f1080f3.docx"},{"id":60994176,"identity":"cc63e25b-5751-49d2-afdc-f54cb244fc02","added_by":"auto","created_at":"2024-07-24 11:52:14","extension":"docx","order_by":2,"title":"","display":"","copyAsset":false,"role":"supplement","size":44733,"visible":true,"origin":"","legend":"","description":"","filename":"AppendixBTestingparametersandrespondentsv01km.docx","url":"https://assets-eu.researchsquare.com/files/rs-4450289/v1/96c376c16ad494bede43fd68.docx"},{"id":60994181,"identity":"afff5222-746b-4ce8-b6b9-b9683f6e3311","added_by":"auto","created_at":"2024-07-24 11:52:14","extension":"docx","order_by":3,"title":"","display":"","copyAsset":false,"role":"supplement","size":15699,"visible":true,"origin":"","legend":"","description":"","filename":"AppendixCParametersversusrespondents.docx","url":"https://assets-eu.researchsquare.com/files/rs-4450289/v1/9ad491cf327a43cee0ca1d81.docx"},{"id":60994179,"identity":"d383da3d-21c8-4754-8e15-678eefc151e0","added_by":"auto","created_at":"2024-07-24 11:52:14","extension":"docx","order_by":4,"title":"","display":"","copyAsset":false,"role":"supplement","size":24150,"visible":true,"origin":"","legend":"","description":"","filename":"AppendixDComparisonacrossgender.docx","url":"https://assets-eu.researchsquare.com/files/rs-4450289/v1/ab561f8f54c8df255bcb9115.docx"},{"id":60994973,"identity":"30ffaba6-8d1c-4a77-b2da-d6c3a2b51ecb","added_by":"auto","created_at":"2024-07-24 12:00:14","extension":"docx","order_by":5,"title":"","display":"","copyAsset":false,"role":"supplement","size":24619,"visible":true,"origin":"","legend":"","description":"","filename":"AppendixERespondentswithoutimpairments.docx","url":"https://assets-eu.researchsquare.com/files/rs-4450289/v1/647d8419bfec34cbcf070188.docx"}],"financialInterests":"No competing interests reported.","formattedTitle":"Determining people’s ease and difficulty of movement based on observed travel behavior","fulltext":[{"header":"Introduction","content":"\u003cp\u003eThis paper presents an approach to use GPS-based travel behavior surveys to determine who is being served well and who is being served poorly by the transport system. Travel behavior surveys have typically not been used for this purpose, even though an increasing number of papers have analyzed data collected through such surveys to understand the travel behavior of people deemed transport disadvantaged, i.e. groups that tend to be poorly served by the current transport system (see below). This paper is inspired by this work, but has a different purpose. Rather than focusing on population segments who are considered or expected to be transport disadvantaged, we seek to ascribe an \u0026lsquo;ease of movement score\u0026rsquo; to all people included in a travel behavior survey, in order to understand who is served well and who is served poorly by the current transport system.\u003c/p\u003e \u003cp\u003eTravel behavior surveys date back to at least the 1930s (Chapin \u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e1974\u003c/span\u003e) and have been part-and-parcel of transport research and practice since the 1960s (de Dios Ortuzar and Willumsen \u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e2011\u003c/span\u003e). Over the past six decades, they have evolved from the original face-to-face interview or paper-based questionnaire to GPS-based surveys (Zmud, Lee-Gosselin et al. \u003cspan citationid=\"CR47\" class=\"CitationRef\"\u003e2013\u003c/span\u003e). The outputs of travel behavior surveys are widely used in transport practice, in particular as a key source of information to identify transport problems. For this purpose, the travel behavior patterns obtained from surveys are used as input for travel demand models, which in turn are used to obtain insights into the functioning of the transport system, usually for some year in the future. The models typically result in the identification of transport links \u0026ndash; road segments or public transport connections \u0026ndash; that perform poorly, i.e. links where congestion is expected to occur as predicted travel flows (demand) exceed the existing capacity (supply). The transport problem is then typically equated with these congested links in the (road or public transport) network, which drives the search for \u0026lsquo;solutions\u0026rsquo; and shapes policy-makers\u0026rsquo; attention.\u003c/p\u003e \u003cp\u003eWe argue that this is an unnecessarily indirect and highly problematic way to identify transport problems, for at least two distinct reasons. First, the approach overlooks many transport problems that have been identified in the broad literature on transport disadvantage. Second, we argue that transport planning should not focus on the functioning of the transport system itself but instead on the service people receive from that transport system. The challenge is to develop approaches that can identify \u003cem\u003ewho\u003c/em\u003e is served well by the transport system and \u003cem\u003ewho\u003c/em\u003e is served poorly, so that the policy attention will be directed to improving the situation of the latter. The analysis of the functioning of the transport system does not provide the required insight, as it does not specify how multiple issues, including a congested transport network, may accumulate for different persons into experienced mobility problems.\u003c/p\u003e \u003cp\u003eAgainst this background, the purpose of the paper is to explore whether it is possible to identify who is served well and who is served poorly by the transport system based on people\u0026rsquo;s revealed travel behavior. The extensive literature analyzing people\u0026rsquo;s travel behavior suggests that people more likely to experience transport problems show distinctly different behaviors from people who can travel with relative ease (see below). Hence, in this paper we develop an approach to use data collected through GPS-based travel behavior surveys to identify the service different people receive from the transport system.\u003c/p\u003e \u003cp\u003eThe paper is structured as follows. In the next section, we present our conceptual approach of using information on people\u0026rsquo;s travel behavior to gain an understanding of their ease of movement. We then provide a brief review of the vast literature describing and explaining differences in travel behavior, highlighting how travel behavior is shaped by people\u0026rsquo;s ease of movement. We subsequently operationalize our approach and describe data and methods. We then present the results, using a set of known-group analyses to provide evidence regarding the validity of the proposed approach. We end the paper with a brief conclusion and a call for further research.\u003c/p\u003e\n\u003ch3\u003eConceptual approach\u003c/h3\u003e\n\u003cp\u003eThe purpose of this paper is to propose an approach to estimate a person\u0026rsquo;s ease of movement based on a systematic analysis of their travel behavior. A person\u0026rsquo;s ease of movement is shaped by the presence or absence of mobility problems, with the latter term referring to any difficulty a person may experience in reaching necessary or desired destinations due to the (perceived) poor functioning of the transport system. This preliminary definition is deliberately broad to encompass a range of mobility problems, varying in character, frequency and impact on people\u0026rsquo;s life (Shay, Combs et al. \u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e2016\u003c/span\u003e). It includes long travel times for relatively short distances (for instance, due to traffic congestion or long waiting and transfer times on public transport), high travel costs, concerns over traffic or social safety, dependence on others for travel, and so on. The claim underlying our analysis is that observed travel behavior may provide, direct or indirect, evidence of the extent to which a person experiences mobility problems and thus of their ease of movement.\u003c/p\u003e \u003cp\u003eThe conceptual approach starts from the widely accepted understanding that travel behavior is the result of both choice and constraint (H\u0026auml;gerstrand 1970; Chapin \u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e1974\u003c/span\u003e), with the balance between both differing from trip to trip, from day to day, and from person to person. Moreover, an identical type of observed behavior, such as a low trip frequency during a particular time period, may be the result of choice (e.g., a preference for a sedentary lifestyle) or constraint (e.g., forgone trips due to a lack of feasible transport options to reach desired destinations). Yet, drawing on the extensive literature briefly discussed in what follows, we posit that persons experiencing systematic constraints limiting their ease of movement will likely show a pattern of travel behavior that is distinct from people who enjoy a high ease of movement. Hence, we postulate that a systematic analysis of a person\u0026rsquo;s travel behavior may provide solid evidence of the extent to which they may experience mobility problems. More precisely, while travel behavior itself can never provide complete proof that someone is experiencing mobility problems, we hypothesize that this behavior can provide a robust indicator of a person\u0026rsquo;s ease of movement and thus for inverse: the \u003cem\u003eprobability\u003c/em\u003e that someone is experiencing mobility problems.\u003c/p\u003e \u003cp\u003eFigure \u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e provides a simplified representation of the conceptual approach. It starts from the understanding that a person\u0026rsquo;s ease of movement (C in the figure) is the result of the interaction between the features of the transport system (B) \u0026ndash; across all modes \u0026ndash; and the characteristics, abilities and circumstances of the person (A). In the ideal case, a person can make use of all available transport modes at any moment in time, without any concerns. This will rarely be the case and many people will experience some restrictions in their use of distinct transport options, for reasons of costs, regulations, impairments, personal safety concerns, racial profiling, and so on. Hence, the interplay between a person and the transport system shapes a person\u0026rsquo;s ease of movement. Persons with few or no restrictions enjoy the highest possible ease of movement in light of the qualities of the available transport system; persons with multiple restrictions will experience a low ease of movement or, in other words, will be facing multiple transport problems and, in the extreme case, transport poverty (Lucas \u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e2012\u003c/span\u003e). In combination with the prevailing land use system (D) \u0026ndash; the spatial and temporal distribution of activities \u0026ndash; ease or difficulty of movement translates into ease or difficulty of access (E) (Levinson and Krizek \u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e2005\u003c/span\u003e). Persons with limited ease of movement residing in a location with unfavorable land use patterns (e.g., few nearby activities or activities poorly served by the parts of the transport system the person can use) are likely to face accessibility problems and, in the extreme case, accessibility poverty (Jeekel and Martens \u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e2017\u003c/span\u003e). Ease of access, in turn, interacts with a person\u0026rsquo;s needs and desires for out-of-home activity participation (A), ultimately shaping a person\u0026rsquo;s observed travel patterns (F). Not shown in the figure, but obviously of importance, is the societal context shaping a person\u0026rsquo;s particular \u0026lsquo;role\u0026rsquo;, required activities, legitimate desires, legitimate (travel) behaviors, and so on (e.g., Law \u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e1999\u003c/span\u003e; Uteng and Cresswell \u003cspan citationid=\"CR44\" class=\"CitationRef\"\u003e2016\u003c/span\u003e).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eThe basic proposition of the proposed approach is that if travel patterns are shaped by ease of movement and ease of access, also the reverse will hold, at least to some extent: travel patterns will reflect a person\u0026rsquo;s ease of movement and ease of access. It will only hold to some extent, because a person\u0026rsquo;s needs and desires for out-of-home activity participation are an unknown \u0026lsquo;intervening variable\u0026rsquo;, which may strongly affect observed travel behavior. Yet, while preferences and choices obviously shape travel behavior, even for people experiencing severe restrictions on ease of movement, we assert that data on a person\u0026rsquo;s observed travel behavior over an extended period of time is highly likely to deliver solid evidence of a person\u0026rsquo;s ease of movement and ease of access. This conceptual, statistical, relationship is highlighted by the dashed black lines in Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e. In the paper, we seek to provide the first evidence of the viability of this approach. We will specify it in more detail in what follows.\u003c/p\u003e \u003cp\u003eGiven the aim to generate estimates of a person\u0026rsquo;s ease of movement \u0026ndash; in its essence a latent variable \u0026ndash; we are not interested in analyzing separate dimensions of travel behavior, such as trip frequencies, trip distances, or mode use, as is commonly done in the literature (see below). Rather, we view these and other dimensions as \u003cem\u003eindicators\u003c/em\u003e of a person\u0026rsquo;s ease of movement. For example, we view a high trip frequency as an indication of a high level of ease of movement. We acknowledge that a single indicator is insufficient to determine whether a person experiences a high ease of movement. Yet, we expect that by extracting multiple indicators from observed travel behavior, it will be possible to obtain quite reliable estimates of a person\u0026rsquo;s ease of movement relative to the general population, thus making it possible to differentiate people served well by the transport system from people who are poorly served. We will operationalize the approach and specify selected ease of movement indicators below.\u003c/p\u003e"},{"header":"Literature review","content":"\u003cp\u003eOur approach builds on the vast literature analyzing the various relationships between the key variables in Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e: people’s characteristics and circumstances, the specifics of the transport and land use systems, ease of movement and ease of access, and people’s travel behavior. This literature includes, amongst others, the broad body of work describing and explaining differences in travel behavior, the research strand on the interrelationship between the built environment and travel behavior, the growing number of studies exploring how accessibility affects travel behavior patterns, and the multiple research strands into the transport challenges experienced by particular (disadvantaged) population segments such as low-income households, people with impairments, women, and the elderly. This broad body of literature analyzes (some of) the relationships represented by the solid lines in Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e. In contrast, our approach takes an inverse perspective in that it seeks to derive estimates of people’s ease of movement through a systematic analysis of their travel behavior patterns (with the reverse direction of this analysis represented by the black dashed line in the figure).\u003c/p\u003e \u003cp\u003eWhile different, our approach does rely on the extensive evidence collected in the various bodies of literature analyzing and explaining differences in people’s travel behavior, as this evidence provides the basis for the selection of ease of movement indicators and their interpretation. This vast body of work shows that people who typically experience more restrictions on movement and thus on access, such as low-income, unemployed, and car-less individuals, people with impairments, and minorities, show distinctly different travel behavior from their more advantaged counterparts. On average, people belonging to disadvantaged population segments make less trips, travel over shorter distances, have a smaller activity space, travel more as a passenger, make more walking trips, and make more use of public transport and less of private motorized vehicles (e.g., Hine and Grieco \u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e2003\u003c/span\u003e; Delbosc and Currie \u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e2011\u003c/span\u003e; Lucas \u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e2012\u003c/span\u003e; Nordbakke and Schwanen \u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e2015\u003c/span\u003e; Lucas, Bates et al. \u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e2016\u003c/span\u003e; Shay, Combs et al. \u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e2016\u003c/span\u003e; Tao, He et al. \u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e2020\u003c/span\u003e; Hidayati, Tan et al. \u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e2021\u003c/span\u003e; Kim and Ulfarsson \u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e2021\u003c/span\u003e). The travel behavior pattern of other population segments known to experience a range of mobility challenges, such as women and people living in rural areas, deviate to some extent from this overall pattern. For instance, the literature suggests that women tend to make more trips overall then men, even if they do tend to travel over shorter distances and more often make trips as a passenger (e.g., Law \u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e1999\u003c/span\u003e; Priya Uteng \u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e2021\u003c/span\u003e). Rural residents, in turn, tend to travel over relatively large distances, although this applies mostly to car-owning households, with rural residents with no or restricted car access showing small activity spaces (e.g., Nutley \u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e2005\u003c/span\u003e; Kolodinsky, DeSisto et al. \u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e2013\u003c/span\u003e; Chen and Akar \u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e2016\u003c/span\u003e; van Dülmen, Šimon et al. \u003cspan citationid=\"CR45\" class=\"CitationRef\"\u003e2022\u003c/span\u003e). Elderly people in general show lower trip frequencies, in part because they are no longer in the labor force, yet trip frequencies tend to further drop with increasing age, in part because of increasing (impairment-related) restrictions on movement (e.g., Giuliano \u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e2004\u003c/span\u003e; Corran, Steinbach et al. \u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e2018\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eAnother line of research relevant to, but again different from, our proposed approach seeks to use data on travel behavior patterns, accessibility, and socio-economic indicators to identify transport disadvantaged population segments. For instance, Pyrialakou et al. (Pyrialakou, Gkritza et al. \u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e2016\u003c/span\u003e) combine a composite accessibility measure with eight socio-economic characteristics to identify population segments and neighborhoods with a high ‘transport need’ yet low levels of accessibility. Comparable approaches have been adopted in other studies (e.g., Casas \u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2007\u003c/span\u003e; Scott and Horner \u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e2008\u003c/span\u003e; Delbosc and Currie \u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e2011\u003c/span\u003e; Carroll, Benevenuto et al. \u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e2020\u003c/span\u003e). Our approach is different from this work, for three reasons. First, we are not solely interested in identifying transport disadvantaged population segments, but in the situation of the entire (sample) population. Second, we seek to estimate relative ease of movement of individuals, going beyond the disadvantaged-advantaged binary. Third, for estimating ease of movement we rely solely on people’s travel behavior patterns (again, represented by dashed black lines in Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e), while only making use of data on personal characteristics and the transport and land use system to verify the validity of the approach (represented by the dashed grey lines in the figure).\u003c/p\u003e \u003cdiv id=\"Sec4\" class=\"Section2\"\u003e \u003ch2\u003eOperationalizing the approach\u003c/h2\u003e \u003cp\u003eThe purpose of our approach is to estimate people’s ease of movement based on their observed travel behavior. The vast body of literature analyzing travel behavior patterns, briefly described above, suggests the feasibility of the approach. It has provided ample evidence that people more likely to experience restrictions on ease of movement, due to income, gender, car access, (age-related) impairments, residential location, and so on, show distinctly different travel behavior patterns. At the same time, this literature also underscores that the patterns are not uniform within and across population segments. For instance, while difficulties in movement do shape low trip frequencies among low-income households, women may show high trip frequencies in spite of experiencing barriers to travel. This and other such instances imply that the estimation of people’s ease of movement has to rely on multiple indicators to avoid many ‘false positives’ and ‘false negatives’.\u003c/p\u003e \u003cp\u003eAgainst this background, we have identified a set of travel behavior indicators that, in combination, can provide an indication of the probability that a person is facing mobility problems. In what follows, we use the term ease of movement (EoM) parameters for these travel behaviors. We prefer the term ease of movement over ease of access parameters, because all selected indicators relate to some dimension of movement, but only a subset of them provides evidence of relative ease of access to destinations. Yet, since ease of movement strongly shapes ease of access, we would argue that our approach also provides insight into people’s ease of access.\u003c/p\u003e \u003cp\u003eWe distinguish four distinct categories encompassing a total of 17 EoM parameters: trip frequency (5 parameters), mode use (4 parameters), travel speed (4 parameters), travel distance (3 parameters), and effort (1 parameter). Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e gives a detailed description of the 17 selected parameters, which can all be extracted from a typical GPS-based travel behavior survey. Parameters belonging to a single category are partly overlapping, yet each parameter provides distinct information. For instance, the ease of movement parameter ‘trips per day’ is closely related to the parameter ‘activities per day’, but they are not identical. Trips are defined as a one-directional movement between a single origin and a single destination. A person who on a particular day is only going to a shop while remaining at home the rest of the day, is conducting two trips yet only one activity. The literature shows that transport disadvantaged persons tend to conduct less activities and even less trips, as they tend to combine multiple activities in a chain of one-directional trips. By including both parameters in our set, we are thus more likely to identify people who experience difficulty of movement.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e\u003cdiv class=\"gridtable\"\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e\u003ctable float=\"Yes\" id=\"Tab1\" border=\"1\"\u003e\u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 1\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eDescription of ease of movement parameters employed in the analyses\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e\u003ccolgroup cols=\"4\"\u003e\u003c/colgroup\u003e\u003cthead\u003e\u003ctr\u003e\u003cth align=\"left\" colspan=\"2\" nameend=\"c2\" namest=\"c1\"\u003e \u003cp\u003eEase of movement parameter\u003c/p\u003e \u003c/th\u003e\u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eDescription\u003c/p\u003e \u003c/th\u003e\u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003eAssumed link with mobility problems\u003c/p\u003e \u003c/th\u003e\u003c/tr\u003e\u003c/thead\u003e\u003ctbody\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\" morerows=\"4\" rowspan=\"5\"\u003e \u003cp\u003eTrip frequency\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eNumber of trips per day\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eThe number of one-directional trips between a single origin and a single destination conducted during the 24-hour reporting period.\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003ePeople who conduct more trips are \u003cb\u003emore likely\u003c/b\u003e to enjoy ease of movement.\u003c/p\u003e \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eNumber of trips in evening hours (19.00-01.00h)\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eThe number of one-directional trips between a single origin and a single destination conducted between 19.00-23.00h during the 24-hour reporting period.\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003ePeople who conduct more trips in the evening are \u003cb\u003emore likely\u003c/b\u003e to enjoy ease of movement, as they are more likely to have adequate transport service.\u003c/p\u003e \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eNumber of trips in night hours (01.00-05.00h)\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eThe number of one-directional trips between a single origin and a single destination conducted between 01.00-05.00h during the 24-hour reporting period.\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003ePeople who conduct more trips in the night are \u003cb\u003emore likely\u003c/b\u003e to enjoy ease of movement, as they are more likely to have access to private transport means (car, motorbike, bicycle, other).\u003c/p\u003e \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eNumber of activities per day\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eThe number of stationary activities conducted at another location than the respondent's home location during the 24-hour reporting period.\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003ePeople who conduct more activities are more likely to enjoy ease of movement.\u003c/p\u003e \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eNumber of distinct activity types visited\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eNumber of distinct activity types visited during the 24-hour observation period\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003ePeople who visit a variety of activity types (work, shopping, leisure) are \u003cb\u003emore likely\u003c/b\u003e to enjoy ease of movement.\u003c/p\u003e \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\" morerows=\"3\" rowspan=\"4\"\u003e \u003cp\u003eMode use\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eNumber of independent motorized trip legs\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eNumber of trip legs made independently with a motorized vehicle (car, motorbike, public transport, taxi, other).\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003ePeople who make \u003cb\u003emore independent motorized trips\u003c/b\u003e are \u003cb\u003emore likely\u003c/b\u003e to enjoy ease of movement.\u003c/p\u003e \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eNumber of trip legs by private motorized vehicle as a driver\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eNumber of trip legs by private motorized vehicle (car, motorbike, other) as a driver during the 24-hour reporting period.\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003ePeople who conduct more trip legs by private motorized vehicle as a driver have higher levels of access to motorized modes and are thus \u003cb\u003emore likely\u003c/b\u003e to enjoy ease of movement.\u003c/p\u003e \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eNumber of motorized trip legs as a chauffeur\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eNumber of trip legs made in a private motorized vehicle in order to bring or take someone else somewhere, during the 24-hour reporting period.\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003ePeople who make more trips as a chauffeur are assumed to be able to travel with relative ease and thus \u003cb\u003emore likely\u003c/b\u003e to enjoy ease of movement.\u003c/p\u003e \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eNumber of motorized trip legs as a passenger\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eNumber of trip legs made as a passenger in a private motorized vehicle driven by someone else (excluding taxi rides).\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003ePeople who make \u003cb\u003emore trips\u003c/b\u003e as a passenger are \u003cb\u003eless likely\u003c/b\u003e to enjoy ease of movement.\u003c/p\u003e \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\" morerows=\"3\" rowspan=\"4\"\u003e \u003cp\u003eSpeed\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eAverage aerial speed across \u003cb\u003eall motorized trips\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eAerial speed is the calculated by dividing the aerial distance between origin and destination by the travel time extracted from GPS traces; the average aerial speed is obtained by averaging the values across all motorized trips conducted during the 24-hour reporting period. A trip with at least one motorized trip leg is considered a motorized trip.\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003ePeople who travel at higher aerial speed are \u003cb\u003emore likely\u003c/b\u003e to enjoy ease of movement.\u003c/p\u003e \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eAverage aerial speed across all trips by \u003cb\u003eprivate motorized vehicle as a driver\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eAerial speed is the calculated by dividing the aerial distance between origin and destination by the travel time extracted from GPS traces; the average aerial speed is obtained by averaging the values across all trips by private motorized vehicle as a driver, conducted by a respondent during the 24-hour reporting period.\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003ePeople who travel at higher aerial speed as a driver are \u003cb\u003emore likely\u003c/b\u003e to enjoy ease of movement.\u003c/p\u003e \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eAverage aerial speed for all public transport trips\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eAerial speed is calculated by dividing the aerial distance between start and end point of a trip containing at least one public transport leg as extracted from GPS traces; the average aerial speed is obtained by averaging the values across all public transport trips conducted during the 24-hour reporting period.\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003ePeople who make public transport trips with high aerial door-to-door speeds are \u003cb\u003emore likely\u003c/b\u003e to enjoy ease of movement.\u003c/p\u003e \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eAverage aerial speed for all public transport trip legs\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eAerial speed is calculated by dividing the aerial distance between start and end point of a public transport trip leg as extracted from GPS traces; the average aerial speed is obtained by averaging the values across all public transport trip legs conducted during the 24-hour reporting period.\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003ePeople who with high aerial speeds on the public transport leg of a trip are \u003cb\u003emore likely\u003c/b\u003e to enjoy ease of movement.\u003c/p\u003e \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\" morerows=\"2\" rowspan=\"3\"\u003e \u003cp\u003eDistance\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eTotal aerial distance traveled\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eTotal aerial distance traveled across all trips, defined as summation of the aerial distance between origins and destinations as extracted from GPS traces for each trip.\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003ePeople who travel over longer (aerial) trip distance overall are \u003cb\u003emore likely\u003c/b\u003e to enjoy ease of movement.\u003c/p\u003e \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eDirectness of travel\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eAverage ratio between road distance and air (= Euclidian) distance across all motorized trip legs.\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003ePeople with a lower ratio (i.e., people who can travel on more direct routes for each motorized trip leg) are \u003cb\u003emore likely\u003c/b\u003e to enjoy ease of movement.\u003c/p\u003e \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eLong walking trips\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eNumber of walking trips and walking trip legs over 2 kilometers in length.\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003ePeople who make less long walking trips are \u003cb\u003emore likely\u003c/b\u003e to enjoy ease of movement.\u003c/p\u003e \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eEffort\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eTrip leg ratio\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eRatio between the total number of trip legs a person conducts and the total number of out-of-home activities conducted by the person (as specified above).\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003ePeople with a lower trip leg ratio (i.e., less trip legs per conducted activity) are \u003cb\u003emore likely\u003c/b\u003e to enjoy ease of movement.\u003c/p\u003e \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eComposite mobility score\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e\u003cb\u003eComposite measure of all OeM parameters\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e\u003cb\u003eAverage respondent z-score derived from z-scores for OeM parameters available for the respondent\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e\u003cb\u003ePeople with a higher mobility score are more likely to enjoy ease of movement.\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e\u003c/tr\u003e\u003c/tbody\u003e\u003c/table\u003e\u003c/div\u003e "},{"header":"Methods","content":"\u003ch2\u003eData\u003c/h2\u003e\u003cp\u003eWe rely on GPS-based travel behavior surveys conducted in the four metropolitan areas of Israel: Tel Aviv, Jerusalem, Haifa and Be'er Sheva (Table\u0026nbsp;\u003cspan refid=\"Tab2\" class=\"InternalRef\"\u003e2\u003c/span\u003e). It includes a total of six distinct surveys conducted between 2010 and 2018. While the surveys were conducted in different contexts across space and time, with different transport, land use, population and economic conditions, these differences are relatively modest so as not to prevent their combined use for our analyses.\u003c/p\u003e\u003cp\u003eThe travel behavior surveys combined GPS location tracking enriched with information provided by the respondents through a dedicated app. Additionally, they included a questionnaire to obtain key socio-economic characteristics of respondents and their household. The time period for which respondents reported on their trips varied between a ‘day’ (starting at home in the morning and ending at home in the evening) and up to 48 hours. We acknowledge that this short observation period is a major limitation for obtaining reliable estimates of a person’s ease of movement. Yet, since we will conduct our analyses at the level of population segments rather than individuals, we argue that the data are adequate to obtain proof-of-concept of the proposed approach.\u003c/p\u003e\u003cp\u003eJointly, the six surveys contain data for 106,049 respondents who made 699,785 one-directional trips during the days of observation. For the purpose of the study, the data set was cleaned to eliminate respondents for which essential socio-economic or trip data was missing. Moreover, professional drivers and respondents younger than 18 years old were excluded from the sample, as their travel behavior may be expected to be distinctly different from the ‘regular’ adult population and may thus distort the calculation of EoM parameters. The resulting cleaned data set consists of 68,424 respondents (65%) who made 620,973 trips (89%) during the days of observation.\u003c/p\u003e\u003cp\u003eGiven that the reporting period varied between respondents, the EoM parameters have been calculated for each respondent for a 24-hour period. In case a respondent reported on more than 24 hours, we used only the first 24-hour period to calculate the parameters, to avoid problematic transposing of observations across a longer time period towards a 24-hour period. Respondents reporting for less than 24 hours but still for an entire day (i.e., who explicitly marked their first activity and last activity of the day in the app) were fully included in the sample. In contrast, respondents who reported for less than 24 hours on their travel patterns and did not explicitly mark their first activity and last activity through the app, were excluded from the sample.\u003c/p\u003e\u003cdiv class=\"gridtable\"\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c7\" colnum=\"7\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c8\" colnum=\"8\"\u003e\u003c/div\u003e\u003ctable float=\"Yes\" id=\"Tab2\" border=\"1\"\u003e\u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 2\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eOverview of raw and cleaned data obtained from GPS-based travel behavior surveys (see also Appendix A).\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e\u003ccolgroup cols=\"8\"\u003e\u003c/colgroup\u003e\u003cthead\u003e\u003ctr\u003e\u003cth align=\"left\" colname=\"c1\" morerows=\"1\" rowspan=\"2\"\u003e \u003cp\u003eMetropolitan area\u003c/p\u003e \u003c/th\u003e\u003cth align=\"left\" colname=\"c2\" morerows=\"1\" rowspan=\"2\"\u003e \u003cp\u003eData collection\u003c/p\u003e \u003cp\u003eperiod\u003c/p\u003e \u003c/th\u003e\u003cth align=\"left\" colspan=\"2\" nameend=\"c4\" namest=\"c3\"\u003e \u003cp\u003eNumber of households\u003c/p\u003e \u003c/th\u003e\u003cth align=\"left\" colspan=\"2\" nameend=\"c6\" namest=\"c5\"\u003e \u003cp\u003eNumber of respondents\u003c/p\u003e \u003c/th\u003e\u003cth align=\"left\" colspan=\"2\" nameend=\"c8\" namest=\"c7\"\u003e \u003cp\u003eNumber of trips\u003c/p\u003e \u003c/th\u003e\u003c/tr\u003e\u003ctr\u003e\u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eRaw\u003c/p\u003e \u003c/th\u003e\u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003eCleaned\u003c/p\u003e \u003c/th\u003e\u003cth align=\"left\" colname=\"c5\"\u003e \u003cp\u003eRaw\u003c/p\u003e \u003c/th\u003e\u003cth align=\"left\" colname=\"c6\"\u003e \u003cp\u003eCleaned\u003c/p\u003e \u003c/th\u003e\u003cth align=\"left\" colname=\"c7\"\u003e \u003cp\u003eRaw\u003c/p\u003e \u003c/th\u003e\u003cth align=\"left\" colname=\"c8\"\u003e \u003cp\u003eCleaned\u003c/p\u003e \u003c/th\u003e\u003c/tr\u003e\u003c/thead\u003e\u003ctbody\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\" morerows=\"1\" rowspan=\"2\"\u003e \u003cp\u003e\u003cb\u003eJerusalem\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e\u003cb\u003e2010–2011\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e3,632\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e3,174\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e16,961\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e11,139\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e113,294\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e98,337\u003c/p\u003e \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e\u003cb\u003e2014–2017\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e4,193\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e3,387\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e17,918\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e11,087\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e113,394\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e96,748\u003c/p\u003e \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\" morerows=\"1\" rowspan=\"2\"\u003e \u003cp\u003e\u003cb\u003eTel Aviv\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e\u003cb\u003e2014\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e4,586\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e4,287\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e16,149\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e11,287\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e112,698\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e97,859\u003c/p\u003e \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e\u003cb\u003e2016–2017\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e4,239\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e4,145\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e16,937\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e10,247\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e108,447\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e98,398\u003c/p\u003e \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eHaifa\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e\u003cb\u003e2016–2017\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e5,108\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e4,772\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e18,547\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e12,186\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e125,476\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e113,976\u003c/p\u003e \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eBe'er Sheva\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e\u003cb\u003e2014–2015\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e6,309\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e5,967\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e19,537\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e12,478\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e126,476\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e115,655\u003c/p\u003e \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c2\" namest=\"c1\"\u003e \u003cp\u003e\u003cb\u003eTotal\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e\u003cb\u003e28,067\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e\u003cb\u003e25,732\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e\u003cb\u003e106,049\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e\u003cb\u003e68,424\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e\u003cb\u003e699,785\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e\u003cb\u003e620,973\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e\u003c/tr\u003e\u003c/tbody\u003e\u003c/table\u003e\u003c/div\u003e\u003ch2\u003eCalculating ease of movement parameters and mobility risk score\u003c/h2\u003e\u003cp\u003eThe underlying assumption of the approach taken in this paper is that ease of movement is a relative phenomenon. That is, someone can move with ease in comparison to what is common in a particular society. In line with this understanding and given the fact that the units and values vary widely between the 17 EoM parameters, we made use of z-scores for each parameter. A z-score describes the position of a raw score in terms of its distance from the mean in standard deviation units. The z-score is positive if the value lies above the mean and negative if it lies below the mean. For each travel parameter, we thus calculate the mean for the entire sample of respondents and subsequently ascribe a z-score to each respondent based on the observed value, using the following widely-known equation:\u003c/p\u003e\u003cp\u003e\u003cimg width=\"92\" src=\"https://myfiles.space/user_files/83062_751fab6dfaef2446/83062_custom_files/img1721821587.png\" alt=\"A math equations with black textDescription automatically generated with medium confidence\" height=\"37\"\u003e\u003c/p\u003e\u003cp\u003ewhere x is the observed value for a respondent, μ is the mean for the sample, and σ is the standard deviation for the sample.\u003c/p\u003e\u003cp\u003eFor ease of interpretation, all 17 ease of movement parameters have been transposed, so that higher positive z-scores always imply a higher ease of movement, while negative z-scores imply relative difficulty in movement. By using z-scores in this way, it becomes straightforward to compare across respondents and EoM parameters. Moreover, it makes it possible to calculate the composite mobility score with ease. This score is calculated by first taking, for each respondent, the average value of the z-scores for the relevant EoM parameters. The resulting average values are then again normalized for the entire sample using z-scores, so that the average composite mobility score for the entire sample will be zero. Lacking theoretical arguments for ascribing weights to particular EoM parameters, no weighing is applied when calculating a respondent’s composite mobility score.\u003c/p\u003e\u003cp\u003eNote that the use of z-scores does \u003cem\u003enot\u003c/em\u003e imply that the approach will always identify people at risk of mobility problems, even when none exist. The size of the differences in z-scores matters. In cases where differences in ease of movement between people are small, we would expect small differences in z-scores and weak relationships between z-scores and respondents’ socio-economic profiles. In contrast, in contexts with substantial differences in ease of movement, we would expect large differences in z-scores, as well as systematic patterns in z-scores between distinct population segments. Thus, while the approach relies on a comparison across people and is thus relative in nature, it does not inevitably imply that substantial risks of mobility problems will be identified even if none exist.\u003c/p\u003e\u003ch2\u003ePre-assessment of ease of movement parameters\u003c/h2\u003e\u003cp\u003eBefore conducting the analyses, we conducted a range of tests to determine which ease of movement parameters and which respondents to include in the analyses. These tests are described in detail in Appendix B. Based on the tests, we decided to exclude respondents with zero trips during the observation period (N = 5,443; 8%), while including all ease of movement parameters in the analyses. Obviously, making no trips at all may be a strong indicator of difficulties in movement, which is indeed confirmed in part by the socio-economic profile of the relevant respondents (Table B.5 in Appendix B). Yet, given the short observation period and the limited number of ease of movement parameters that can be calculated for this sub-segment of the sample, it was decided to exclude them from the analysis.\u003c/p\u003e\u003cp\u003eThe number of ease of movement parameters that can be calculated for the respondents included in the analyses (N = 62,981; 92%) depends on their observed travel behavior. Since all included respondents made at least one trip during the observation period, it was possible to calculate at least 12 parameters for each respondent. This minimum was calculated for only 15% of respondents (N = 9,525), while for 48% of the respondents it was possible to calculate 15 or more parameters (N = 30,088) (Appendix C).\u003c/p\u003e"},{"header":"Results","content":"\u003cp\u003eThe purpose of the current paper is to provide a first proof-of-concept. Hence, we have conducted four known-group analyses to determine whether the proposed approach can deliver the expected results: by level of access to a private motorized vehicle; by age; by gender; and by disability. For the latter three groups, we conduct the analyses for the entire sample and by vehicle access level.\u003c/p\u003e \u003cp\u003eThe first known-group analysis distinguishes between respondents based on their access to private motorized vehicles (cars or motorbikes). We distinguish between four distinct groups: (1) respondents who do not have a private motorized vehicle in the household (N\u0026thinsp;=\u0026thinsp;15,653, 24.9%); (2) respondents who have a private motorized vehicle in the household and share it with more than one other adult (N\u0026thinsp;=\u0026thinsp;15,272, 24.3%); (3) respondents who have a private motorized vehicle in the household and share it with only one other adult (N\u0026thinsp;=\u0026thinsp;16,047, 25.5%); and (4) respondents who are the sole user of their own private motorized vehicle (N\u0026thinsp;=\u0026thinsp;15,919, 25.3%).\u003c/p\u003e \u003cp\u003eFigure \u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e provides the average scores for all 17 EoM parameters as well as for the composite mobility score. The results are nearly perfectly in line with expectations: as access to a motor vehicle goes up, z-scores on all but one parameter and on the composite mobility score go up (the only exception is \u0026lsquo;number of trips in night hours\u0026rsquo;, where respondents who share a motor vehicle with one adults have a higher z-score than respondents with their own vehicle). Among others, respondents without private vehicle access conduct much more long walking trips, less overall trips and less trips in evening hours, more trips as a passenger, and travel at lower speeds when using public transport. Particularly striking are the differences in number of independent motorized trip legs and average aerial speed across all independent motorized trips.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eThe composite mobility score for each individual respondent separately is also surprisingly in line with expectations, in spite of the short observation period (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e). The vast majority of respondents without access to a motor vehicle show composite mobility scores well below the average, with 65.4% of respondents having a z-score below \u0026minus;\u0026thinsp;2. Moreover, only 3.1% of this population segment has an above-average composite mobility score. The exact opposite holds for respondents who have their own motor vehicle: 61.7% has a composite mobility score above +\u0026thinsp;2, while only 5.7% has a below-average score. The situation for the other two segments of respondents falls between these extremes and exactly in line with expectations: respondents who share a motor vehicle with two or more adults show lower composite mobility scores than respondents who share their vehicle with only one other adult. Interestingly, the range of observed composite mobility score for these two segments is substantially smaller than that for the respondents without or with their own motor vehicle.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eThe second known-group analysis analyzes the composite mobility score by age, for the entire sample and for the four vehicle access levels. The literature shows that both younger and older people are more likely to experience mobility problems. For the younger segment is related in part to relatively low income levels and limited vehicle access (e.g., Ralph \u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e2015\u003c/span\u003e; Klein and Smart \u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e2017\u003c/span\u003e; Ralph \u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e2017\u003c/span\u003e), while for the older segment mobility problems are partly the result of increasing travel-related impairments and related (gradual) driving cessation, lower income levels, as well as concerns over social safety (e.g., P\u0026aacute;ez, Scott et al. \u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e2007\u003c/span\u003e; Luiu, Tight et al. \u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e2017\u003c/span\u003e; Corran, Steinbach et al. \u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e2018\u003c/span\u003e; Palm, Nakshi et al. \u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). Our results are perfectly in line with these results, with below-average composite mobility scores for respondents up to age\u0026thinsp;~\u0026thinsp;33 and from age\u0026thinsp;~\u0026thinsp;64 onwards. The pattern is remarkably similar for each of the four segments by vehicle access, even though the age range within which the average respondent experiences an above-average ease of movement varies. This range is largest for respondents who have their own motor vehicle (ranging from ages\u0026thinsp;~\u0026thinsp;24 to ~\u0026thinsp;70), and the smallest for respondents without access to a motor vehicle (ranging from ages\u0026thinsp;~\u0026thinsp;33 to ~\u0026thinsp;54). For all motor vehicle segments, average composite mobility score drops sharply at older age, again in line with expectations (e.g., Siren and Hakamies-Blomqvist \u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e2004\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eThe fact that respondents without access to a motor vehicle have above-average composite mobility scores between ages\u0026thinsp;~\u0026thinsp;33 and ~\u0026thinsp;54, in combination with the declining scores for all population segments irrespective of vehicle access at older age, raises the question whether the composite mobility score is not shaped too strongly by parameters related to trip frequency. The above-average composite mobility scores occur especially in the age range when respondents may be both employed and are taking care of children, which may result in relatively high trip rates, positively affecting the composite mobility score even if respondents may experience substantial mobility problems. Hence, we conducted an additional analysis and calculated the composite mobility score based on 12 EoM parameters only, excluding the five trip frequency parameters (number of trips per day; number of trips in evening hours; number of trips in night hours; number of activities per day; and number of distinct activity types visited). This analysis results in virtually identical patterns for all population segments (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e). Only for respondents with their own motor vehicle does the elimination of frequency-related EoM parameters result in relatively higher composite mobility scores for respondents from age\u0026thinsp;~\u0026thinsp;36 onwards, suggesting that their relatively high ease of movement only partially translates into a higher trip frequency in comparison to people with lower levels of access to a motor vehicle. For all other vehicle segments, the analysis shows that the observed lower composite mobility scores at younger and older ages are hardly shaped by trip frequency parameters, suggesting that the composite mobility score indeed captures ease of movement.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eThe third known-group analysis explored the impact of gender. Extensive research shows that women have distinct travel patterns from men and many studies also show that women experience more mobility problems, in part due to limited access to a motor vehicle, lower incomes, and concerns over social safety (e.g., Law \u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e1999\u003c/span\u003e; Siren and Hakamies-Blomqvist \u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e2006\u003c/span\u003e; Uteng and Cresswell \u003cspan citationid=\"CR43\" class=\"CitationRef\"\u003e2008\u003c/span\u003e; Priya Uteng \u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e2021\u003c/span\u003e; Zhang, Zhao et al. \u003cspan citationid=\"CR46\" class=\"CitationRef\"\u003e2022\u003c/span\u003e). Our results are in line with the literature: women have a somewhat lower, statistically significant, composite mobility score than men (-0.31 versus +\u0026thinsp;0.22; p\u0026thinsp;\u0026lt;\u0026thinsp;0.05). Comparable modest differences in composite mobility score are found for all four population segments by vehicle access, with the differences being significant for all four segments except respondents who share a vehicle with three or more adults (Table\u0026nbsp;\u003cspan refid=\"Tab3\" class=\"InternalRef\"\u003e3\u003c/span\u003e). These results underscore that women\u0026rsquo;s ease of movement is not only hindered because of lower vehicle access, but by other factors as well, as suggested by the literature.\u003c/p\u003e \u003cp\u003eA further analysis shows that women tend to show lower scores on most EoM parameters (Appendix D). The situation is particularly striking for women with access to a car, who show significantly lower z-scores on all parameters, except for number of activities per day and number of long walking trips, where not significant differences between women and men are observed (Appendix D). The pattern is more diverse for respondents without a car. In line with the literature, women without a car show much lower z-scores for number of trips in the evening and night hours (p\u0026thinsp;\u0026lt;\u0026thinsp;0.001). Surprisingly, they also perform a smaller diversity of activities during the day than men without a motor vehicle. Also somewhat at odds with the literature is that women show higher z-scores for a number of speed parameters and number of long walk trips, suggesting that they travel on average at higher speeds and make less long walking trips than their male counterparts.\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab3\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 3\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eComposite mobility score by gender for each vehicle segment separately, for the entire sample (N\u0026thinsp;=\u0026thinsp;62,981).\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"11\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c7\" colnum=\"7\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c8\" colnum=\"8\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c9\" colnum=\"9\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c10\" colnum=\"10\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c11\" colnum=\"11\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\" morerows=\"2\" rowspan=\"3\"\u003e \u003cp\u003ePopulation segment\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colspan=\"4\" nameend=\"c5\" namest=\"c2\"\u003e \u003cp\u003eNumber of respondents\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colspan=\"6\" nameend=\"c11\" namest=\"c6\"\u003e \u003cp\u003eComposite mobility score\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003ctr\u003e \u003cth align=\"left\" colspan=\"2\" nameend=\"c3\" namest=\"c2\"\u003e \u003cp\u003eWomen\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colspan=\"2\" nameend=\"c5\" namest=\"c4\"\u003e \u003cp\u003eMen\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colspan=\"2\" nameend=\"c7\" namest=\"c6\"\u003e \u003cp\u003eWomen\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colspan=\"2\" nameend=\"c9\" namest=\"c8\"\u003e \u003cp\u003eMen\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colspan=\"2\" nameend=\"c11\" namest=\"c10\"\u003e \u003cp\u003eStatistical test\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003en\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003e%\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003en\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c5\"\u003e \u003cp\u003e%\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c6\"\u003e \u003cp\u003eMean\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c7\"\u003e \u003cp\u003eSD\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c8\"\u003e \u003cp\u003eMean\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c9\"\u003e \u003cp\u003eSD\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c10\"\u003e \u003cp\u003ep-value\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c11\"\u003e \u003cp\u003et value\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eFull sample\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e33,842\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e100%\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e29,139\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e100%\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e-0.31\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e1.85\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e0.22\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e1.72\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e \u003cp\u003e2.58*\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c11\"\u003e \u003cp\u003e0.04\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eNo motor vehicle\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e8,938\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e26.4%\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e6,715\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e23.0%\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e-2.71\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e2.41\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e-2.42\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e1.18\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e \u003cp\u003e2.58*\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c11\"\u003e \u003cp\u003e0.03\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eVehicle shared with multiple adults\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e8,582\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e25.4%\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e6,780\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e23.2%\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e-0.84\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e2.36\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e-0.78\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e1.31\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e \u003cp\u003e0.16\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c11\"\u003e \u003cp\u003e1.07\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eVehicle shared with one adult\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e8,564\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e25.3%\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e7,483\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e25.6%\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.75\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e2.75\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e0.87\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e1.63\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e \u003cp\u003e2.38*\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c11\"\u003e \u003cp\u003e0.04\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSole user of vehicle\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e7,758\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e22.9%\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e8,161\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e27.9%\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e2.46\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e2.42\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e2.71\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e2.31\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e \u003cp\u003e2.69*\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c11\"\u003e \u003cp\u003e0.01\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003ctfoot\u003e \u003ctr\u003e\u003ctd colspan=\"11\"\u003e*p\u0026thinsp;\u0026lt;\u0026thinsp;.05\u003c/td\u003e\u003c/tr\u003e \u003c/tfoot\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003eThe final known-group analysis compares ease of movement for people with and without impairments. The literature shows that people with physical, sensory, or cognitive impairments experience a range of restrictions on movement and are less mobile as a result (e.g., Imrie \u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e1996\u003c/span\u003e; Shoval, Wahl et al. \u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e2011\u003c/span\u003e; Sammer, Uhlmann et al. \u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e2012\u003c/span\u003e; Ratering, Van der Heijden et al. \u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). People with impairments not only experience more mobility problems than others due to a mismatch between their abilities and the specifics of the transport system, but also because disability affects economic opportunity, resulting in lower incomes and thus in lower vehicle access, both of which affect ease of movement.\u003c/p\u003e \u003cp\u003eSince only the travel behavior surveys conducted in the Jerusalem metropolitan area included a question about respondents\u0026rsquo; disability status, the known-group analysis is limited to this sub-sample (N\u0026thinsp;=\u0026thinsp;20,986; 33% of the sample for all four metropolitan areas). From this group, 37% (7,764 respondents) indicated that they are disabled. This share is much higher than commonly seen in the literature, which typically reports shares ranging between 6% and 10% (Martens \u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e2018\u003c/span\u003e). Hence, this set of respondents probably also includes people with only modest impairments, so that the results of our analysis should be treated with some caution (see Appendix E).\u003c/p\u003e \u003cp\u003eYet, in spite of these concerns, the results are in line with expectations. While the Jerusalem sub-sample without impairments enjoys a slightly higher ease of movement than the entire set of respondents (which, of course, also includes people with impairments in the other metropolitan areas), the sub-sample with impairments has a composite mobility score well below the average: +0.12 versus \u0026minus;\u0026thinsp;0.48. The difference between the two sub-segments is statistically significant (p\u0026thinsp;\u0026lt;\u0026thinsp;0.01). Significant differences comparable in size can also be observed when conducting the analysis by vehicle access level. In all cases, the composite mobility score for people without impairments is higher than for people with impairments (Table\u0026nbsp;\u003cspan refid=\"Tab4\" class=\"InternalRef\"\u003e4\u003c/span\u003e). Moreover, in line with expectations, for both respondents with and without impairments the composite mobility score increases substantially as the level of vehicle access goes up. Interestingly, the differences by vehicle access are much larger than the differences between respondents with and without impairments. For instance, while people with impairments with their own private vehicle have a substantially lower composite mobility score than their non-impaired counterparts (+\u0026thinsp;1.87 versus +\u0026thinsp;2.69; p\u0026thinsp;\u0026lt;\u0026thinsp;0.05), they still have a much higher score than non-impaired respondents without vehicle access (+\u0026thinsp;1.87 versus \u0026minus;\u0026thinsp;2.66).\u003c/p\u003e \u003cp\u003eThe large differences across level of car access in combination with the relatively modest differences related to impairment, suggest that vehicle access level is a much stronger predictor of ease of movement than impairment itself. To determine whether this holds, we conducted a stepwise regression with backward elimination. The initial model included a broad range of socio-economic and land use variables, among which a disability variable (see explanation below). The final model retained six significant predictors: age, household size, disability, car ownership, population density, and employment opportunity. These six factors collectively explained 46.4% of the variance in the composite mobility score, with a statistically significant model fit (F (6, 20979)\u0026thinsp;=\u0026thinsp;58.64, p\u0026thinsp;\u0026lt;\u0026thinsp;.001). The analysis confirms the dominant importance of car ownership level in explaining the composite mobility score, even if disability was also found to have a significant impact (Table\u0026nbsp;\u003cspan refid=\"Tab4\" class=\"InternalRef\"\u003e4\u003c/span\u003e).\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab4\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 4\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eStepwise regression for composite mobility score for respondents in the Jerusalem region (N\u0026thinsp;=\u0026thinsp;20,986).\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"5\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eVariable\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eβ\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eSEB\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003et-value\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c5\"\u003e \u003cp\u003ep-value\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAge\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e-0.26\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.17\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e-2.51**\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e.006\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eHousehold size\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e-0.29\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.18\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e-3.29***\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eDisability\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e-0.27\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.16\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e2.27**\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.008\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eCar ownership level\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.57\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.29\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e8.62***\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003ePopulation Density\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.28\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.19\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e2.62**\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e.004\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eEmployment opportunity\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.25\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.35\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e2.41**\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e.005\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003e*p\u0026thinsp;\u0026lt;\u0026thinsp;.05, **p\u0026thinsp;\u0026lt;\u0026thinsp;.01, ***p\u0026thinsp;\u0026lt;\u0026thinsp;.001.\u003c/p\u003e \u003cp\u003eFinally, we employed another stepwise regression with backward elimination for the entire sample, to identify the factors that most significantly predict a respondent\u0026rsquo;s composite mobility score. The included factors in the initial model are age, gender, years of education, household size, number of children in the household, household type, religion, driving license (binary: yes/no), motor vehicle access level (four distinct categories, as specified above), population density in zone of residence, and employment opportunity in zone of residence. Table\u0026nbsp;\u003cspan refid=\"Tab5\" class=\"InternalRef\"\u003e5\u003c/span\u003e presents the final model. Five significant variables were retained: age, household size, motor vehicle access level, population density, and employment opportunity. These five factors collectively explained 44.6% of the variance in mobility score, with a statistically significant model fit (F (5, 62976)\u0026thinsp;=\u0026thinsp;56.73, p\u0026thinsp;\u0026lt;\u0026thinsp;.001). Most importantly, all retained variables show the expected sign, supporting the claim that the composite mobility score indeed captures ease of movement. As shown in the table, motor vehicle access level (β\u0026thinsp;=\u0026thinsp;0.53, p\u0026thinsp;\u0026lt;\u0026thinsp;.001) and household size (β = -0.38, p\u0026thinsp;\u0026lt;\u0026thinsp;.001) emerged as the strongest predictors of composite mobility score. Age also emerged as a relevant predictor, confirming the bivariate analyses presented above showing that especially at older age composite mobility score drops substantially. Gender was not retained in the final model. While at odds with the literature, this result is in line with the finding above that women and men show relatively modest differences in composite mobility score.\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab5\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 5\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eStepwise regression for composite mobility score for the entire sample (N\u0026thinsp;=\u0026thinsp;62,981).\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"5\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eVariable\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eβ\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eSEB\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003et-value\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c5\"\u003e \u003cp\u003ep-value\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAge\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e-0.28\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.16\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e-2.44**\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e.007\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eHousehold size\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e-0.38\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.14\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e-4.07***\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eCar ownership level\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.53\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.31\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e8.43***\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003ePopulation density\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.29\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.17\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e2.81**\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e.003\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eEmployment opportunity\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.26\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.37\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e2.83**\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e.002\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003e*p\u0026thinsp;\u0026lt;\u0026thinsp;.05, **p\u0026thinsp;\u0026lt;\u0026thinsp;.01, ***p\u0026thinsp;\u0026lt;\u0026thinsp;.001.\u003c/p\u003e"},{"header":"Conclusion","content":"\u003cp\u003eThis paper presented an approach to use data on observed travel behavior to determine who is being served well and who is being served poorly by the transport system. Drawing on the extensive literature analyzing and explaining travel behavior, we identified 17 parameters assumed to capture a person\u0026rsquo;s relative ease of movement. We included parameters regarding trip frequency, speed, travel distance, and effort. We acknowledge that none of these parameters by themselves is sufficient to determine whether someone is served well or poorly by the transport system, as travel behavior may be the result of choice as well as constraint or, perhaps more precisely, choices made within constraints, with the strength of the latter depending on a person\u0026rsquo;s specific circumstances. However, we argue that jointly the parameters are likely to differentiate between well-served and poorly-served people.\u003c/p\u003e \u003cp\u003eWe applied our approach to data from six GPS-based travel behavior surveys conducted in Israel\u0026rsquo;s four main metropolitan areas (N\u0026thinsp;=\u0026thinsp;62,981 valid respondents). Given the variety in measurement units and values, we relied on z-scores for all ease of movement parameters (EoM parameters), with negative values suggesting low levels of ease of movement and thus mobility problems, and positive values relative ease of movement, both in comparison to the entire sample. In addition to the 17 ease of movement parameters, we calculated a composite mobility score for each respondent based on the respondent\u0026rsquo;s z-scores for all 17 EoM parameters.\u003c/p\u003e \u003cp\u003eWe subsequently conducted four known-group analyses to verify the validity of the proposed approach. Results from these analyses are in line with expectations and thus confirm the potential of the approach. The first analysis, comparing four population segments differing in their level of access to private motorized vehicles (car or motorbike), showed that z-scores systematically increase for all EoM parameters as access to private motorized vehicles improves. Respondents without a motor vehicle scored poorest on all parameters on average. Among others, they conduct much more long walking trips, less overall trips and less trips at night, more trips as a passenger, and travel at lower speeds when using public transport.\u003c/p\u003e \u003cp\u003eThe second known-group analysis examined the composite mobility score across age. In line with the literature, we found that both younger and older population segments experience lower ease of movement, irrespective of whether the composite mobility score was based on the complete set or a reduced set of EoM parameters. Results also show a sharp decline in ease of movement as people age for all car access segments. The third known-group analysis by gender shows a somewhat lower ease of movement for women, irrespective of vehicle access, again in line with the literature. Finally, a comparison between people with and without impairments for the Jerusalem sub-sample (N\u0026thinsp;=\u0026thinsp;20,986) also showed differences in line with expectations: people with impairments have a significantly lower ease of movement than people without impairments. Regression analyses for the entire sample and the Jerusalem sub-sample confirmed that vehicle access, age, and disability shape ease of movement, but did not corroborate the relevance of gender.\u003c/p\u003e \u003cp\u003eTaken together, these results thus provide first evidence in support of the fundamental claim of this paper: an estimate of a person\u0026rsquo;s ease of movement can be derived from a systematic analysis of their observed travel behavior. This outcome is particularly striking, taking into account the short observation period per respondent. Clearly, more research is needed to refine and test the approach. Several directions for further research can be highlighted. First, the approach should be tested on datasets that include observations on people\u0026rsquo;s travel behavior across multiple days, as studies have shown that people\u0026rsquo;s travel behavior fluctuates substantially over time (Schlich and Axhausen \u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e2003\u003c/span\u003e; Raux, Ma et al. \u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e2016\u003c/span\u003e; Deschaintres, Morency et al. \u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e2022\u003c/span\u003e). Second, a systematic assessment of the relevance of different ease of movement parameters can strengthen the approach. The current set is suitable for contexts in which car-based travel provides superior ease of movement in most cases. In the ideal case, only \u0026lsquo;mode-agnostic\u0026rsquo; ease of movement parameters would be included, so that the approach can be applied across a range of contexts, including metropolitan areas with outstanding public transport service. Such a systematic assessment of potential EoM parameters should also take into account the data that are typically collected in GPS-based travel behavior surveys across the world, so that an approach can be developed that would allow for comparative analyses across geography. Finally, and perhaps most essential, the approach should be scrutinized for external validity, by collecting, among the same population sample, both quantitative data on respondents\u0026rsquo; travel behavior and qualitative data on their travel experiences and challenges, either through bespoke surveys or in-depth interviews (Murphy, Gould-Werth et al. \u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e2021\u003c/span\u003e; Singer and Martens \u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e2023\u003c/span\u003e), and subsequently assessing whether ease of movement estimates derived from the former indeed correlate systematically with the latter.\u003c/p\u003e"},{"header":"Declarations","content":"\u003ch2\u003eAuthor Contribution\u003c/h2\u003e\u003cp\u003eDiana Saadi: Data cleaning and preparation; Data analysis; Methodology; Software; Validation; Visualization.Karel Martens: Conceptualization; Funding acquisition; Methodology; Project administration; Supervision; Validation; Writing - original draft; Writing - review \u0026amp; editing. All authors reviewed the manuscript.\u003c/p\u003e\u003ch2\u003eAcknowledgement\u003c/h2\u003e\u003cp\u003eThe authors would like to thank Sami Adri of Netivei Ayalon Highways Company for generously sharing the processed travel behavior survey data.\u003c/p\u003e\u003ch2\u003eData Availability\u003c/h2\u003e\u003cp\u003eData used in the paper are managed by the Netivei Ayalon Highways Company, a government-owned company. Data may be requested from the company.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n\u003cli\u003eCarroll, P., R. Benevenuto and B. Caulfield (2020). \u0026quot;Identifying Hotspots of Transport Disadvantage and Car Dependency in Rural Ireland.\u0026quot; \u003cu\u003eTransport Policy\u003c/u\u003e.\u003c/li\u003e\n\u003cli\u003eCasas, I. (2007). \u0026quot;Social Exclusion and the Disabled: An Accessibility Approach.\u0026quot; \u003cu\u003eThe Professional Geographer\u003c/u\u003e \u003cstrong\u003e59\u003c/strong\u003e(4): 463-477.\u003c/li\u003e\n\u003cli\u003eChapin, F. S. (1974). \u003cu\u003eHuman activity patterns in the city: things people do in time and in space\u003c/u\u003e. New York, Wiley.\u003c/li\u003e\n\u003cli\u003eChen, N. and G. Akar (2016). \u0026quot;Effects of neighborhood types \u0026amp; socio-demographics on activity space.\u0026quot; \u003cu\u003eJournal of Transport Geography\u003c/u\u003e \u003cstrong\u003e54\u003c/strong\u003e: 112-121.\u003c/li\u003e\n\u003cli\u003eCorran, P., R. Steinbach, L. Saunders and J. Green (2018). \u0026quot;Age, disability and everyday mobility in London: An analysis of the correlates of \u0026lsquo;non-travel\u0026rsquo; in travel diary data.\u0026quot; \u003cu\u003eJournal of Transport \u0026amp; Health\u003c/u\u003e \u003cstrong\u003e8\u003c/strong\u003e: 129-136.\u003c/li\u003e\n\u003cli\u003ede Dios Ortuzar, J. and L. G. Willumsen (2011). \u003cu\u003eModelling Transport\u003c/u\u003e. Chichester, John Wiley \u0026amp; Sons.\u003c/li\u003e\n\u003cli\u003eDelbosc, A. and G. Currie (2011). \u0026quot;Transport problems that matter - social and psychological links to transport disadvantage.\u0026quot; \u003cu\u003eJournal of Transport Geography\u003c/u\u003e \u003cstrong\u003e19\u003c/strong\u003e(1): 170-178.\u003c/li\u003e\n\u003cli\u003eDeschaintres, E., C. Morency and M. Tr\u0026eacute;panier (2022). \u0026quot;Cross-analysis of the variability of travel behaviors using one-day trip diaries and longitudinal data.\u0026quot; \u003cu\u003eTransportation Research Part A: Policy and Practice\u003c/u\u003e \u003cstrong\u003e163\u003c/strong\u003e: 228-246.\u003c/li\u003e\n\u003cli\u003eGiuliano, G. (2004). \u0026quot;Land use and travel patterns among the elderly.\u0026quot; \u003cu\u003eTransportation in an Aging Society\u003c/u\u003e \u003cstrong\u003e27\u003c/strong\u003e: 192-210.\u003c/li\u003e\n\u003cli\u003eHidayati, I., W. Tan and C. Yamu (2021). \u0026quot;Conceptualizing Mobility Inequality: Mobility and Accessibility for the Marginalized.\u0026quot; \u003cu\u003eJournal of Planning Literature\u003c/u\u003e \u003cstrong\u003e36\u003c/strong\u003e(4): 492-507.\u003c/li\u003e\n\u003cli\u003eHine, J. P. and M. S. Grieco (2003). \u0026quot;Scatters and clusters in time and space: implications for delivering integrated and inclusive transport.\u0026quot; \u003cu\u003eTransport Policy\u003c/u\u003e \u003cstrong\u003e10\u003c/strong\u003e(4): 299-306.\u003c/li\u003e\n\u003cli\u003eImrie, R. (1996). \u0026quot;Disability and discourses of mobility and movement.\u0026quot; \u003cu\u003eEnvironment and Planning A\u003c/u\u003e \u003cstrong\u003e32\u003c/strong\u003e: 1641-1656.\u003c/li\u003e\n\u003cli\u003eJeekel, H. and K. Martens (2017). \u0026quot;Equity in transport: Learning from the policy domains of housing, health care and education.\u0026quot; \u003cu\u003eEuropean Transport Research Review\u003c/u\u003e \u003cstrong\u003e9\u003c/strong\u003e(53).\u003c/li\u003e\n\u003cli\u003eKim, S. and G. F. Ulfarsson (2021). \u0026quot;Staying home or going places: Mobility factors of older minority women\u0026apos;s daily trip making in the United States.\u0026quot; \u003cu\u003eJournal of Transport \u0026amp; Health\u003c/u\u003e \u003cstrong\u003e21\u003c/strong\u003e: 101031.\u003c/li\u003e\n\u003cli\u003eKlein, N. J. and M. J. Smart (2017). \u0026quot;Millennials and car ownership: Less money, fewer cars.\u0026quot; \u003cu\u003eTransport Policy\u003c/u\u003e \u003cstrong\u003e53\u003c/strong\u003e: 20-29.\u003c/li\u003e\n\u003cli\u003eKolodinsky, J. M., T. P. DeSisto, D. Propen, M. E. Putnam, E. Roche and W. R. Sawyer (2013). \u0026quot;It is not how far you go, it is whether you can get there: Modeling the effects of mobility on quality of life in rural New England.\u0026quot; \u003cu\u003eJournal of Transport Geography\u003c/u\u003e \u003cstrong\u003e31\u003c/strong\u003e: 113-122.\u003c/li\u003e\n\u003cli\u003eLaw, R. (1999). \u0026quot;Beyond \u0026apos;women and transport\u0026apos;: towards new geographies of gender and daily mobility.\u0026quot; \u003cu\u003eProgress in Human Geography\u003c/u\u003e \u003cstrong\u003e23\u003c/strong\u003e(4): 567-588.\u003c/li\u003e\n\u003cli\u003eLevinson, D. M. and K. J. Krizek, Eds. (2005). \u003cu\u003eAccess to destinations\u003c/u\u003e.\u003c/li\u003e\n\u003cli\u003eLucas, K. (2012). \u0026quot;Transport and social exclusion: Where are we now?\u0026quot; \u003cu\u003eTransport Policy\u003c/u\u003e \u003cstrong\u003e20\u003c/strong\u003e: 105-113.\u003c/li\u003e\n\u003cli\u003eLucas, K., J. Bates, J. Moore and J. A. Carrasco (2016). \u0026quot;Modelling the relationship between travel behaviours and social disadvantage.\u0026quot; \u003cu\u003eTransportation Research Part A: Policy and Practice\u003c/u\u003e \u003cstrong\u003e85\u003c/strong\u003e: 157-173.\u003c/li\u003e\n\u003cli\u003eLuiu, C., M. Tight and M. Burrow (2017). \u0026quot;The unmet travel needs of the older population: a review of the literature.\u0026quot; \u003cu\u003eTransport Reviews\u003c/u\u003e \u003cstrong\u003e37\u003c/strong\u003e(4): 488-506.\u003c/li\u003e\n\u003cli\u003eMartens, K. (2018). \u0026quot;Ageing, impairments and travel: Priority setting for an inclusive transport system.\u0026quot; \u003cu\u003eTransport Policy\u003c/u\u003e \u003cstrong\u003e63\u003c/strong\u003e: 122-130.\u003c/li\u003e\n\u003cli\u003eMurphy, A. K., A. Gould-Werth and J. Griffin (2021). \u0026quot;Validating the Sixteen-Item Transportation Security Index in a Nationally Representative Sample: A Confirmatory Factor Analysis.\u0026quot; \u003cu\u003eSurvey Practice\u003c/u\u003e \u003cstrong\u003e14\u003c/strong\u003e(1): 27185.\u003c/li\u003e\n\u003cli\u003eNordbakke, S. and T. Schwanen (2015). \u0026quot;Transport, unmet activity needs and wellbeing in later life: exploring the links.\u0026quot; \u003cu\u003eTransportation\u003c/u\u003e \u003cstrong\u003e42\u003c/strong\u003e(6): 1129-1151.\u003c/li\u003e\n\u003cli\u003eNutley, S. (2005). \u0026quot;Monitoring rural travel behaviour: a longitudinal study in Northern Ireland 1979-2001.\u0026quot; \u003cu\u003eJournal of Transport Geography\u003c/u\u003e \u003cstrong\u003e13\u003c/strong\u003e(3): 247-263.\u003c/li\u003e\n\u003cli\u003eP\u0026aacute;ez, A., D. Scott, D. Potoglou, P. Kanaroglou and K. B. Newbold (2007). \u0026quot;Elderly mobility: demographic and spatial analysis of trip making in the Hamilton CMA, Canada.\u0026quot; \u003cu\u003eUrban Studies\u003c/u\u003e \u003cstrong\u003e44\u003c/strong\u003e(1): 123-146.\u003c/li\u003e\n\u003cli\u003ePalm, M., P. Nakshi, E. Yousefzadeh Barri, S. Farber and M. Widener (2024). \u0026quot;Uncovering suppressed travel: A scoping review of surveys measuring unmet transportation need.\u0026quot; \u003cu\u003eTravel Behaviour and Society\u003c/u\u003e \u003cstrong\u003e36\u003c/strong\u003e: 100784.\u003c/li\u003e\n\u003cli\u003ePriya Uteng, T. (2021). Chapter Two - Gender gaps in urban mobility and transport planning. \u003cu\u003eAdvances in Transport Policy and Planning\u003c/u\u003e. R. H. M. Pereira and G. Boisjoly, Academic Press. \u003cstrong\u003e8: \u003c/strong\u003e33-69.\u003c/li\u003e\n\u003cli\u003ePyrialakou, V. D., K. Gkritza and J. D. Fricker (2016). \u0026quot;Accessibility, mobility, and realized travel behavior: Assessing transport disadvantage from a policy perspective.\u0026quot; \u003cu\u003eJournal of Transport Geography\u003c/u\u003e \u003cstrong\u003e51\u003c/strong\u003e: 252-269.\u003c/li\u003e\n\u003cli\u003eRalph, K. M. (2015). \u003cu\u003eStalled on the road to adulthood? Analyzing the nature of recent travel changes for young adults in America, 1995 to 2009\u003c/u\u003e, University of California, Los Angeles.\u003c/li\u003e\n\u003cli\u003eRalph, K. M. (2017). \u0026quot;Multimodal Millennials? The Four Traveler Types of Young People in the United States in 2009.\u0026quot; \u003cu\u003eJournal of Planning Education and Research\u003c/u\u003e \u003cstrong\u003e37\u003c/strong\u003e(2): 150-163.\u003c/li\u003e\n\u003cli\u003eRatering, C., R. Van der Heijden and K. Martens (2024). \u0026quot;Moving around with Anxiety Disorder.\u0026quot; \u003cu\u003eTransportation Research Part F: Psychology and Behaviour\u003c/u\u003e \u003cstrong\u003e100\u003c/strong\u003e: 493-506.\u003c/li\u003e\n\u003cli\u003eRaux, C., T.-Y. Ma and E. Cornelis (2016). \u0026quot;Variability in daily activity-travel patterns: the case of a one-week travel diary.\u0026quot; \u003cu\u003eEuropean Transport Research Review\u003c/u\u003e \u003cstrong\u003e8\u003c/strong\u003e(4): 26.\u003c/li\u003e\n\u003cli\u003eSammer, G., T. Uhlmann, W. Unbehaun, A. Millonig, B. Mandl, J. Dangschat and R. Mayr (2012). \u0026quot;Identification of mobility-impaired persons and analysis of their travel behavior and needs.\u0026quot; \u003cu\u003eTransportation Research Record\u003c/u\u003e \u003cstrong\u003e2320\u003c/strong\u003e(1): 46-54.\u003c/li\u003e\n\u003cli\u003eSchlich, R. and K. W. Axhausen (2003). \u0026quot;Habitual travel behaviour: Evidence from a six-week travel diary.\u0026quot; \u003cu\u003eTransportation\u003c/u\u003e \u003cstrong\u003e30\u003c/strong\u003e(1): 13-36.\u003c/li\u003e\n\u003cli\u003eScott, D. M. and M. W. Horner (2008). \u0026quot;The role of urban form in shaping access to opportunities: An exploratory spatial data analysis.\u0026quot; \u003cu\u003eJournal of transport and land use\u003c/u\u003e \u003cstrong\u003e1\u003c/strong\u003e(2): 89-119.\u003c/li\u003e\n\u003cli\u003eShay, E., T. S. Combs, D. Findley, C. Kolosna, M. Madeley and D. Salvesen (2016). \u0026quot;Identifying transportation disadvantage: Mixed-methods analysis combining GIS mapping with qualitative data.\u0026quot; \u003cu\u003eTransport Policy\u003c/u\u003e \u003cstrong\u003e48\u003c/strong\u003e: 129-138.\u003c/li\u003e\n\u003cli\u003eShoval, N., H.-W. Wahl, G. Auslander, M. Isaacson, F. Oswald, T. Edry, R. Landau and J. Heinik (2011). \u0026quot;Use of the global positioning system to measure the out-of-home mobility of older adults with differing cognitive functioning.\u0026quot; \u003cu\u003eAgeing \u0026amp; Society\u003c/u\u003e \u003cstrong\u003e31\u003c/strong\u003e(5): 849-869.\u003c/li\u003e\n\u003cli\u003eSinger, M. E. and K. Martens (2023). \u0026quot;Measuring travel problems: Testing a novel survey tool in a natural experiment.\u0026quot; \u003cu\u003eTransportation Research Part D: Transport and Environment\u003c/u\u003e \u003cstrong\u003e121\u003c/strong\u003e: 103834.\u003c/li\u003e\n\u003cli\u003eSiren, A. and L. Hakamies-Blomqvist (2004). \u0026quot;Private car as the grand equaliser? Demographic factors and mobility in Finnish men and women aged 65+.\u0026quot; \u003cu\u003eTransportation Research Part F: Traffic Psychology and Behaviour\u003c/u\u003e \u003cstrong\u003e7\u003c/strong\u003e(2): 107-118.\u003c/li\u003e\n\u003cli\u003eSiren, A. and L. Hakamies-Blomqvist (2006). \u0026quot;Does gendered driving create gendered mobility? Community-related mobility in Finnish women and men aged 65+.\u0026quot; \u003cu\u003eTransportation Research Part F: Traffic Psychology and Behaviour. Special Issue on Older drivers\u0026apos; safety and mobility: Current and future issues\u003c/u\u003e \u003cstrong\u003e9\u003c/strong\u003e(5): 374-382.\u003c/li\u003e\n\u003cli\u003eTao, S., S. Y. He, M.-P. Kwan and S. Luo (2020). \u0026quot;Does low income translate into lower mobility? An investigation of activity space in Hong Kong between 2002 and 2011.\u0026quot; \u003cu\u003eJournal of Transport Geography\u003c/u\u003e \u003cstrong\u003e82\u003c/strong\u003e: 102583.\u003c/li\u003e\n\u003cli\u003eUteng, T. P. and T. Cresswell, Eds. (2008). \u003cu\u003eGendered mobilities\u003c/u\u003e. Transport and Society. Aldershot, Ashgate.\u003c/li\u003e\n\u003cli\u003eUteng, T. P. and T. Cresswell, Eds. (2016). \u003cu\u003eGendered mobilities\u003c/u\u003e. London/New York, Routledge.\u003c/li\u003e\n\u003cli\u003evan D\u0026uuml;lmen, C., M. \u0026Scaron;imon and A. Kl\u0026auml;rner (2022). \u0026quot;Transport poverty meets car dependency: A GPS tracking study of socially disadvantaged groups in European rural peripheries.\u0026quot; \u003cu\u003eJournal of Transport Geography\u003c/u\u003e \u003cstrong\u003e101\u003c/strong\u003e: 103351.\u003c/li\u003e\n\u003cli\u003eZhang, M., P. Zhao and X. Tong (2022). \u0026quot;Constructing women\u0026apos;s immobility: Fear of violence and Women\u0026rsquo;s constricted nocturnal travel behaviour.\u0026quot; \u003cu\u003eTravel Behaviour and Society\u003c/u\u003e \u003cstrong\u003e26\u003c/strong\u003e: 178-192.\u003c/li\u003e\n\u003cli\u003eZmud, J., M. Lee-Gosselin, M. Munizaga and J. A. Carrasco, Eds. (2013). \u003cu\u003eTransport Survey Methods: Best Practice for Decision Making\u003c/u\u003e. Bingley, Emerald.\u003c/li\u003e\n\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":true,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"transportation","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"port","sideBox":"Learn more about [Transportation](http://link.springer.com/journal/11116)","snPcode":"11116","submissionUrl":"https://submission.nature.com/new-submission/11116/3","title":"Transportation","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"em","reportingPortfolio":"Springer Hybrid","inReviewEnabled":true,"inReviewRevisionsEnabled":false},"keywords":"ease of movement, travel behavior, travel behavior surveys, mobility problems, transport disadvantage, GPS","lastPublishedDoi":"10.21203/rs.3.rs-4450289/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-4450289/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eThis paper presents an approach to use GPS-based travel behavior surveys to determine who is being served well and who is being served poorly by the transport system. We draw on the extensive literature on travel behavior, which has shown that people\u0026rsquo;s travel behavior is at least in part shaped by the travel barriers they experience. Starting from this basic insight, we define 17 parameters that may provide insight into a person\u0026rsquo;s relative ease of movement. These ease of movement parameters cover dimensions related to trip frequency (e.g., overall and in evening hours), transport mode use (e.g., as driver or passenger), travel speed (e.g., for public transport legs), distance (e.g., trip detour ratio), and effort (e.g., ratio between trip legs and out-of-home activities). None of these parameters by themselves is sufficient to determine whether someone is served well or poorly by the transport system, as behaviors may be the result of choice as well as constraint. However, we argue that jointly the parameters are likely to differentiate well-served from poorly-served people. We apply our approach to data from six GPS-based travel behavior surveys conducted in Israel\u0026rsquo;s four main metropolitan areas (N\u0026thinsp;=\u0026thinsp;62,981). We calculate z-scores for all ease of movement parameters, with negative values suggesting mobility problems and positive values relative ease of movement compared to the entire sample. We conduct four known-group analysis, comparing mean z-scores by level of access to a private motorized vehicle, age, gender, and disability. Results are systematically in line with expectations: population segments identified in the literature as experiencing (more severe levels of) transport disadvantage show systematically lower composite mobility scores. These outcomes are particularly striking, taking into account the short observation period of only one day per respondent. Taken together, these findings provide a first indication that revealed travel behavior patterns can be used to identify population segments poorly served by the transport system and thus to determine both success and failure of the existing transport system. While more research is needed, the approach holds promise to determine the impacts of transport investments on people\u0026rsquo;s ease of movement.\u003c/p\u003e","manuscriptTitle":"Determining people’s ease and difficulty of movement based on observed travel behavior","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2024-07-24 11:52:07","doi":"10.21203/rs.3.rs-4450289/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"decision","content":"Revision requested","date":"2024-12-05T01:34:07+00:00","index":"","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2024-11-13T12:59:07+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2024-11-04T03:36:13+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"144845626031415730135914728619872725456","date":"2024-10-14T00:33:05+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"111314552361099831865689304603794596134","date":"2024-10-12T09:21:08+00:00","index":"hide","fulltext":""},{"type":"reviewersInvited","content":"","date":"2024-10-10T07:47:05+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2024-07-01T19:21:08+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2024-05-21T06:44:30+00:00","index":"","fulltext":""},{"type":"submitted","content":"Transportation","date":"2024-05-20T16:26:13+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"transportation","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"port","sideBox":"Learn more about [Transportation](http://link.springer.com/journal/11116)","snPcode":"11116","submissionUrl":"https://submission.nature.com/new-submission/11116/3","title":"Transportation","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"em","reportingPortfolio":"Springer Hybrid","inReviewEnabled":true,"inReviewRevisionsEnabled":false}}],"origin":"","ownerIdentity":"edb0e80d-0cef-410b-a35e-2da19c3da55f","owner":[],"postedDate":"July 24th, 2024","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"published-in-journal","subjectAreas":[],"tags":[],"updatedAt":"2025-04-21T16:09:59+00:00","versionOfRecord":{"articleIdentity":"rs-4450289","link":"https://doi.org/10.1007/s11116-025-10604-x","journal":{"identity":"transportation","isVorOnly":false,"title":"Transportation"},"publishedOn":"2025-04-17 15:58:02","publishedOnDateReadable":"April 17th, 2025"},"versionCreatedAt":"2024-07-24 11:52:07","video":"","vorDoi":"10.1007/s11116-025-10604-x","vorDoiUrl":"https://doi.org/10.1007/s11116-025-10604-x","workflowStages":[]},"version":"v1","identity":"rs-4450289","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-4450289","identity":"rs-4450289","version":["v1"]},"buildId":"qtupq5eGEP_6zYnWcrvyt","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}