Walkability in European Cities from a Tourist Perspective: A Comparative Study of Thessaloniki and Lyon using Choice-based Conjoint Analysis

1. INTRODUCTION

1.5. The Most Important Attributes of Walkability

Walkability, defined as a space’s ability to allow and encourage pedestrian use, depends on many factors. These vary in influence depending on the users and the walk’s purpose: an older adult shopping will have different needs than a teenager walking alone at night or a worker commuting to work daily.

1.5.1. Accessibility and Connectivity

Pedestrian accessibility refers to the ease of access for pedestrians in a place, neighborhood, or area. This differs from connectivity, which is the ability to reach that place by any transport, whether by car, public transport, train, or bicycle. These two points highlight the connection between an area and its surrounding spatial environment. The more accessible a place is, the more likely it is to be visited and explored. A difficult-to-access place is often overlooked by tourists, especially those without a private vehicle.

For private vehicle users, mainly cars, scooters, and bicycles, safe and easily accessible parking spaces should be considered. For public transport users, it is essential to ensure they arrive at the target location efficiently, with timetables that facilitate their arrival and return from the site.

The streets must be designed for everyone to access the space they want. People with Disabilities (PwD) (e.g., wheelchair users) need a wide enough path to move freely. It is also important to be able to avoid stairs. In addition, a blind person needs an environment with fixed markers, such as defined pavements and pedestrian crossings, rather than a meeting zone, where respondents are presented with a series of product profiles, which may invite them to travel long distances to the place to be discovered.

1.5.2. Tediousness of Walking

Walking may be considered the easiest way to get around town, but it is not always the fastest or least tiring. The difficulty of a path is a crucial factor, as it affects a pedestrian’s psychological and physical fatigue. Individuals are accustomed to different urban forms, and for some population groups, a straight path is unpleasant because it does not allow quick linking between two amenities. A different plan would be more easily disorienting and misleading for someone used to a Hippodamian city. As Hall argues, people interpret and navigate space through culturally conditioned spatial conventions [11]. This helps explain why unfamiliar urban environments may be more difficult for tourists to read and navigate. It can, therefore, be very challenging for a person to navigate an unfamiliar city.

In cities, many spatial breaks divert pedestrians from their direct path. These deviations can be natural, such as landforms or waterways. On hills, in most cities, streets are designed to have the smallest possible slope, even if it means extending the road at the top. These roads prioritize cars, but the routes for pedestrians become longer. Their route is far from straight if they do not find a staircase. In addition to the longer journey due to the constant bends, the hills are tiring to climb. Thessaloniki and Lyon are hilly cities, where the difficulty of walking uphill is evident among individuals who are not accustomed to physical exercise. This discontinuity can also be attributed to the planning choices of the location, including buildings, roads, railways, or facilities. Barriers can hinder movement, forcing pedestrians to take a detour. For example, this is the case when entering most of Lyon’s parks: Gerland, the Tête d’Or, the Parc des Hauteurs, and many others. To enter these parks, one must access them through the gates during opening hours. This sometimes means making detours to enter or leave the park. In contrast, most parks in Thessaloniki are open at all hours and rarely have designated entry points.

More subtly, the pathways of a square can be contrary to the natural route users wish to take. Initially designed for their aesthetic qualities or uses, squares and their paths can be repurposed. When the designed path does not align with users’ expectations, lines of desire can emerge. Figure 1 depicts a “desire path” in a Thessaloniki park, illustrating how pedestrians often create informal shortcuts that differ from the designed circulation. Such behavior reflects a mismatch between planned and actual pedestrian movement patterns, an important element for assessing walkability. These improvised paths appear on a lawn that is too heavily trodden by pedestrians seeking to save time and who do not follow designated routes.

Moreover, the route’s difficulty can be similar to the difficulty of a short discontinuity in the itinerary. Deviating from or interrupting one’s trajectory to avoid obstacles is often difficult. Obstacles such as parked scooters, bins, or uneven paving can reduce pedestrian comfort and accessibility, especially for PwD or tourists unfamiliar with the urban layout. Figure 2, taken in Thessaloniki, exemplifies how such hindrances force detours and slow movement.

1.5.3. Sensory Experience

During our research in Thessaloniki, we noticed that some streets were not suitable for walking in heavy rain. Pedestrians took refuge in the halls of buildings and the entrances of restaurants and shops until the rain subsided. The heat and sunshine can discourage tourists and pedestrians alike from going out at certain times. As seen in many Mediterranean countries, businesses and residents are less active during the hottest hours of the day in the summer months. Weather constraints such as rain, wind, snow, and low or high temperatures are not real problems inside a car, train, bus, tram, or metro. In a car, a simple action activates the windscreen wipers, closes the windows, and starts the air conditioning, thereby resolving the issue. On foot, however, and for two-wheeled vehicles as well, these problems persist regardless of the actions taken. The surrounding environment can provide protection, and some cities are designed to accommodate this more effectively than others. Elements such as trees, buildings, roofs, canopies, and indoor spaces can shelter pedestrians from weather conditions. A city without pedestrian protection from sun, rain, and snow prevents people from walking freely and comfortably year-round.

1.5.4. Green Elements in the Street and Places

Trees and shrubs protect against the weather, cool the area, block the wind, and make the space more attractive in the summer and throughout the year. In hot weather and direct sunlight, artificial equipment can provide shade and reduce exposure. However, the plants in the city help lower the temperature during the day and at night, making the surrounding areas cooler. The control of this high heat is all the more pertinent with the advent of frequent and intense heatwaves associated with global warming. Similarly, plants and green spaces help absorb rainwater, which can hinder movement and pose a risk to the population in extreme cases. Additionally, vegetated areas offer a pleasant and peaceful ambiance. As witnessed after the COVID-19 lockdowns, people gathered in parks not only for the fresh air but also for the calm and aesthetic quality, in contrast to the confinement of living indoors.

1.5.5. Aesthetic Quality of the Environment

The aesthetics of cities attract tourists. However, people have only recently been attracted to urban landscapes. It was in the 19^th century that the concept of tourist contemplation and aesthetic pleasure began to emerge among the wealthy majority. However, since then, the city has attracted people not only for its beauty. The aesthetics of a city are also shaped by the activities taking place within it, which can make the urban environment itself more attractive to visitors. It should be noted that these activities are not necessarily artistic; they may include traditions, ways of life, culinary habits, and many other elements, all of which may seem banal to a local but can be striking to a tourist.

Tourists can be drawn to places that offer unique points of interest, such as museums and monuments, or to observe stunning landscapes. In contrast to traditional sightseeing, activities such as shopping, sports, and relaxation attract people. If these activities are spatially concentrated enough, they move from one place to another on foot. Tourists do not always go out with a specific purpose; sometimes they go out to “see the city center” or “see the view.” A district with many monuments attracts tourists to walk around the city, while a residential or industrial district is generally less attractive.

In addition to not attracting pedestrians in general, some less lively areas can make people feel unsafe, which repels pedestrians, especially tourists unfamiliar with the area, who are reluctant to cross through locations that seem dangerous.

1.5.6. Security and Safety

Tourists and residents want to move around peacefully, without risk, as pedestrians, whether due to incivility by various people or to the ever-increasing traffic. The insecurity is fueled by concerns about marginal or infrequent behavior, such as hit-and-run drivers on the roads or the use of electric scooters, which is frightening due to their novelty. The presence of pedestrian crossings and lighting can mitigate this fear.

These crossing facilities make it easy and safe to get around during rush hour and when fewer pedestrians are on the streets, such as at night. Sometimes there are no pedestrian crossings within a reasonable distance, and pedestrians must cross directly into the lane, which is dangerous or can lead to long delays to find a safe opportunity. Moreover, in Thessaloniki, as in many cities, pedestrian crossings are not always clearly marked, which may lead tourists to believe that they do not have priority to cross at a particular location. At the same time, the fear can arise from another pedestrian, who is also a human being and may have malicious intent. Public lighting plays an important role in this area by enhancing perceived safety and visibility; in the event of an incident, it may facilitate timely assistance or intervention. In addition to lighting, the activity and crowd density of the place play major roles in an individual’s perceived perception of safety; proximity to others can enhance perceived security.

Although previous studies have examined walkability determinants, pedestrian route choice, and tourism-related preferences, limited evidence exists on how tourists explicitly trade off street-level walkability attributes under different urban and climatic conditions using a unified experimental framework. This lack of comparative, attribute-level evidence constrains planners’ ability to derive context-sensitive yet transferable pedestrian design priorities. Addressing this gap requires a methodologically consistent, field-based approach that can isolate marginal effects and interactions among key walkability attributes.

2. MATERIALS AND METHODS

The CBC experiment operationalizes the literature-derived attributes and tests their trade-offs within and between the two cities, enabling a comparative assessment.

We use CBC to isolate each attribute’s marginal contribution to route choice and to quantify trade-offs, reporting relative importances (which attributes matter most) and level utilities (direction/magnitude for each level).

Building on a user-centric urban-design perspective (safety, comfort, legibility), accessibility theory (impedance, reachability within time budgets), and an affordances lens (reliance on immediately observable cues by short-term visitors), we operationalize four (4) design-actionable constructs with observable attributes: equipment for crossing roads and safety (perceived safety/legibility), proximity of the public transport stop (accessibility), presence of green/shade (thermal comfort), and obstacles on the sidewalk (micro-impedance), adding also weather for reference. A randomized CBC experiment estimates part-worth utilities and relative importances by city to identify within-city priorities (RQ1), tests between-city differences (RQ2), and examines compensations/trade-offs among attributes (RQ3). The comparative pairing of Thessaloniki and Lyon provides a structured basis to discuss regional transferability to similar European cities (RQ4).

To frame the study, there is a need to define the study variables: location, date, and study population.

As explained above, the study’s population is tourists. The study defines a tourist as any person who travels to discover a place for pleasure, culture, enrichment, and relaxation. Tourists have different requirements and needs from residents and therefore require special attention.

Moreover, being in an unfamiliar place and culture makes the tourist feel out of place, including the city’s planning, habits, customs, pedestrians’ walking styles, and the distribution of pavement between users and facilities. This may arouse curiosity in some people while confusing, disturbing, or disorienting others.

The city’s walkability will encourage them to go out and enjoy it. If the city’s walkability is too low, the tourist may choose to stay at the hotel rather than visit the city, which would be a failure for the city and its market.

As explained above, the study areas are Thessaloniki, Greece, and Lyon, France. We selected these two cities to capture diverse urban and tourist profiles and explore how perceptions of walkability vary between distinct European urban contexts. Thessaloniki and Lyon were chosen for their distinct yet similar traits. Both cities are the second-largest in their countries and draw many international tourists. They stand out due to differences in key urban and mobility aspects, including transportation infrastructure, street layout, and microclimatic conditions. This makes them ideal for a comparative study. Thessaloniki is a coastal Mediterranean city with limited green space and a developing public transport system, which contrasts sharply with Lyon. Lyon has a river and a well-established metro and tram system, along with better urban greening. This variety offers valuable insights into how different urban environments influence tourists’ perceptions of walkability.

Thessaloniki is the center of Greece’s second-largest urban area, with a population exceeding 1,000,000 and an area of approximately 1,200 km². Thessaloniki, like many coastal cities, has been developed around the sea. Nowadays, the Region of Central Macedonia welcomes approximately four million tourists per year (before the COVID-19 crisis) and around one and a half million tourists in Thessaloniki. The transport network in Thessaloniki is centered around private cars and buses. The Thessaloniki Metro, which has been under construction since 2006 and delayed by significant archaeological finds, was not yet operational at the time of the survey.

Lyon is also the center of France’s second-largest urban area, with 2,2 million inhabitants and an area of over 530 km². Between the Alps and the Massif Central, the Roman city was built in the 1st century BC at the confluence of the Rhône and Saône rivers. Like Thessaloniki, Lyon has hills, Fourvière and the Croix-Rousse. Lyon attracts nearly five million tourists each year (before the COVID-19 crisis). Unlike Thessaloniki, Lyon has metros and tramways, but private cars are also used in the city, including in the city center.

We chose to compare these two cities not only because they are among the largest metropolitan urban areas in Greece and France, respectively, but also because they are readily accessible to both tourist populations and because of their contrasting urban, climatic, and transportation characteristics, which enrich the comparative analysis. They are both cities steeped in history, founded many centuries ago, and having undergone many successive developments.

To define the most important attributes (i.e., characteristics) of the study, as it is impossible to study all the aforementioned parameters in depth, we conducted qualitative surveys among tourists in Thessaloniki at various locations and times, involving people with different profiles. These surveys were conducted through discussions with people of all ages and genders who walked for pleasure or with a specific goal in mind. The exchange was mainly centered on a simple question: “In your opinion, what parameter is the most important in walking?” and “Are there things that encourage you to walk or not?” Among the study criteria discussed with the interviewees, several were mentioned frequently: accessibility, green spaces, security, weather protection, and sidewalk obstacles.

We then sought to determine the importance of the walkability parameters to the tourists, so we conducted a Conjoint Analysis. Conjoint Analysis is a statistical technique often used in marketing to determine customers’ preferences regarding the attributes of a product (good or service). The CBC technique presents respondents with a series of sets of product profiles, in which they must select one profile from two or three options. These profiles are generated randomly so that each respondent has a variety of combinations of attribute levels, and the results show which attributes are most important and which levels have the greatest utility [49, 50]. CBC analysis has also been applied in passenger transport service design contexts, including driverless bus acceptance and flexible transport services [49, 50].

In this study, the Discover platform of Sawtooth Software was utilized, an application designed for CBC analysis. For this second study, in-person tablet surveys were conducted in the streets of Thessaloniki and Lyon. Apart from the CBC questions, there were also demographic questions. A street-intercept, time–location sampling approach was employed at tourist activity nodes. The street-intercept surveys were conducted in central, high-tourist areas in both cities. In Thessaloniki, interviews were conducted primarily along the New Waterfront, at Aristotelous Square, and in adjacent pedestrian corridors. In Lyon, surveys were conducted in the Presqu’île district, near Bellecour Square, and along central pedestrianized streets with significant tourist activity. These locations were selected for their high visitor concentration and accessibility. While the immediate environment may influence respondents’ perceptions, the CBC experiment presented randomized hypothetical scenarios, reducing direct contextual bias from the interview location. Enumerators approached passersby across multiple days and dayparts during summer 2021 for three months, screened them for tourist status (non-residents aged 18 and above), and administered the questionnaire on-site. This approach captures on-trip decisions in the environments of interest when a full sampling frame is unavailable. The final analytic sample consisted of 163 respondents, with 78 surveys conducted in Thessaloniki and 85 in Lyon.

The total sample size (N = 163; Thessaloniki: 78; Lyon: 85) falls within the range commonly used in stated-preference and CBC studies examining urban mobility and tourism behavior. In CBC designs, statistical efficiency depends more strongly on the number of choice tasks per respondent and the experimental design structure than on absolute sample size alone. Given the balanced design, the randomized presentation of profiles, and the aggregation of multiple-choice observations per respondent, the effective number of choice observations exceeds the raw respondent count. This ensures stable estimation of part-worth utilities and relative importance scores for the selected attributes. Nevertheless, subgroup comparisons should be interpreted with appropriate caution. According to established CBC design guidelines, minimum sample sizes of 50–100 respondents per segment are generally sufficient for stable utility estimation in moderate-complexity multi-attribute choice experiments.

The surveys were conducted during Summer 2021, a period when both Greece and France had largely lifted strict lockdown measures, allowing domestic and international tourism subject to vaccination or testing requirements. Public spaces, outdoor attractions, and transport systems were operational, although certain health protocols (e.g., mask mandates in indoor settings and transport) remained in place.

The characteristics of the attributes were chosen according to the rules of the CBC analysis (measurable, orderable, and not only felt by the respondent). In addition to the CBC requirements, candidate attributes were screened against operational criteria relevant to urban design practice: (i) the attribute had to be actionable through policy or design interventions; (ii) it had to be directly observable by tourists during typical visits; and (iii) it had to be comparable across Thessaloniki and Lyon. These criteria ensured clarity for respondents and feasibility for planners (Table 2).

The CBC model estimates (Table 2) will address RQ1–RQ4 by quantifying within-city priorities, testing between-city differences, and evaluating attribute trade-offs.

The Sawtooth Software Discover platform recommended 12 scenarios per profile for each respondent to ensure sufficient responses.

3. RESULTS

3.1. Demographics

In total, 163 tourists responded, of whom 78 were in Thessaloniki and 85 in Lyon.

In Thessaloniki, 54 men, 22 women, and two (2) people did not want to specify their gender.

In Lyon, there were 50 men, 30 women, and five (5) people who did not want to specify their gender.

In the pre-survey questionnaire, respondents’ nationalities were asked: in Thessaloniki, 26 nationalities were reported, and in Lyon, 41. For both surveys, a diverse range of nationalities is represented. This will allow us to analyze the parameters that favor walking, thereby obtaining the broadest possible view.

In Fig. (3), we observe an overrepresentation of people under 34 years old. Specifically, the age distribution of respondents reveals a clear predominance of younger tourists in both cities. The 18–24 age group accounts for 42% of respondents in Thessaloniki and 41% in Lyon, while the 25–34 age group accounts for 16% and 17%, respectively. Participation drops sharply among older cohorts: only 4% in Thessaloniki and 6% in Lyon are aged 35–44, 9% and 11% are aged 45–54, 6% in each city are 55–64 years old, and just 1% in Thessaloniki and 4% in Lyon are aged 65 or older. This concentration of younger participants may influence perceptions of walkability, as younger visitors often display greater tolerance for longer walking distances and are more inclined toward active mobility. This may be explained by young people’s tendency to travel despite the COVID-19 pandemic. To put this into context, the summer when the survey was conducted, we were in the process of a gradual exit from confinement, and many countries were reopening, with Greece and France slowly resuming tourism and economic activity.

Figures 4 and 5 show that most tourists reported feeling safe in both cities.

Figure 4 indicates that the majority of tourists in Thessaloniki reported feeling safe or totally safe during their stay. Specifically, 35.9% of the total sample were male respondents who reported feeling “safe,” while 10.3% were female respondents reporting the same perception. Similarly, 25.6% of the total sample were men who felt “totally safe,” compared to 9.0% women. Neutral responses represented 5.1% of the total sample for each gender, while perceptions of being unsafe were relatively limited (3.9% men and 2.6% women). No respondents reported feeling “totally unsafe.” Overall, 83% of surveyed tourists in Thessaloniki reported feeling safe or totally safe during their stay.

Figure 5 indicates that perceptions of safety in Lyon were generally lower than in Thessaloniki. Specifically, 17.7% of the total Lyon sample were male respondents who reported feeling “safe,” while 12.9% were female respondents reporting the same perception. Similarly, 27.1% of the total sample were men who felt “totally safe,” compared to 9.4% women. Neutral responses were more frequent than in Thessaloniki (7.1% men and 4.7% women). Reports of feeling “unsafe” accounted for 5.9% of men and 2.4% of women, while 3.5% of both genders selected “totally unsafe.” Overall, 16% of respondents in Lyon expressed some degree of insecurity.

Through the following figure, it is shown that tourists in Lyon and Thessaloniki use many different means of transport (Fig. 6):

Figure 6 shows that in Thessaloniki, the most frequently used transport modes among tourists are buses (64.1%), cars (60.3%), and taxis (55.1%), with no reported use of the metro or tramway. Bicycle use is 20.5% and e-scooter use is 19.2%, while boat use is limited to 15.4%. In Lyon, the tramway/metro (75.3%), buses (61.2%), and e-scooters (54.1%) dominate, followed by bicycles (40.0%), cars (29.4%), and taxis (28.2%), with no reported boat use. The high car usage in Thessaloniki is surprising for tourists but may be explained by travel beyond the city to the wider Region of Central Macedonia or across Greece. Affordable bus tickets and low taxi/Véhicule de Tourisme avec Chauffeur (VTC) prices, compared to those in many European countries, also make road-based modes of transportation attractive. In contrast, Lyon’s dense metro and tramway network, combined with efficient public transport and bike- and e-scooter-sharing systems like Vélo’v, reduces dependence on cars despite higher taxi and VTC costs.

3.2. CBC Results

Using the CBC analysis questions, the utility of the levels and the importance of the attributes could be determined.

Utility is the parameter that enables the ranking of levels of a given attribute. The higher the utility, the more preferred the level is. Conversely, the lower the utility, the less preferred the level is. However, a negative level does not mean it was disliked; it simply means it was less preferred than the others. The same is true for a positive value, which could be judged as less unfavorable than the other levels of each attribute.

The primary information of interest is the importance of each attribute. Attribute importance represents the impact that the attribute had on the respondent’s choices. This data allows for a direct comparison of the attributes.

The importance of the attributes in Thessaloniki is presented in Fig. (7) [RQ1]:

Figure 7 shows that in Thessaloniki, “green elements in the street and places” is the most important attribute for tourists, with an importance of 31.15%, far surpassing the other factors. This prominence is likely due to the city’s hot summer climate, with midday temperatures averaging 35 °C, making shaded and vegetated areas particularly valued. The next most important attributes are proximity of the public transport stop (18.56%) and equipment for crossing roads and safety (17.86%), followed closely by the number of obstacles on the sidewalk (16.61%) and weather protection (15.81%). The close values among these four suggest that, while greenery is a clear top priority, tourists also place nearly equal importance on safe, accessible, and obstacle-free walking.

The importance of the attributes in Lyon is presented in Fig. (8) [RQ1]:

Figure 8 shows that in Lyon, the five walkability attributes are rated more evenly than in Thessaloniki, with values ranging from 17.38% to 22.22%. The most important factors are equipment for crossing roads and safety (22.22%), followed closely by the proximity of the public transport stop (21.18%) and green elements in streets and public places (20.65%). Weather protection accounts for 18.57%, and the number of sidewalk obstacles ranks last at 17.38%. This balanced distribution indicates that tourists in Lyon consider various elements of walkability nearly equal, with a marginal inclination towards secure road crossings and accessible public transport.

4. DISCUSSION

Interpreting the results through the framework, we address [RQ1–RQ3] (within-city priorities, between-city differences, and trade-offs) and then discuss [RQ4] (transferability). The study’s findings align with accessibility and walkability theory: crossing quality and proximity to public transport influence route choice, while green spaces and shade serve as climate-sensitive comfort enhancers; between-city differences reflect contextual constraints.

Addressing RQ1–RQ3, we find that the greatest difference lies in the importance of green spaces in Thessaloniki, whereas this attribute ranks third in Lyon [RQ2]. This can be seen in the fact that Thessaloniki is a much warmer city, and, as mentioned earlier, green spaces create coolness within the urban fabric.

The city of Thessaloniki has little green space in its city center (Fig. 9), which is also near its center of activity attracting tourists, while in Lyon, there is not much green space in the city center either, but the rivers that run through the city help to cool the city down and provide plenty of space close to the city center where one cannot feel the heavy heat (Fig. 10).

It should be noted that the surveys in Lyon were conducted two weeks later than in Thessaloniki, in mid-August, when the weather was cooler in both cities. In addition, the summer of 2021 was particularly cool in France and in Lyon compared to the heat waves of previous years. This may have led people in Lyon to feel the heat less than usual.

The pronounced importance of green elements in Thessaloniki (31.15%) is consistent with research emphasizing the role of vegetation and microclimatic moderation in shaping pedestrian behavior under high-temperature urban conditions. Behavioral thermal regulation has been shown to significantly influence pedestrian path choices in hot environments, as individuals actively seek shaded and vegetated routes to minimize thermal discomfort [39]. Similarly, studies examining preferences for urban green infrastructure have demonstrated that tree canopy and natural elements substantially enhance perceived environmental quality and walking attractiveness [45]. Broader built-environment analyses further confirm that greenery and environmental comfort are central components of perceived walkability [38]. The study’s findings extend this literature by quantifying the magnitude of these preferences using CBC trade-offs, demonstrating that under Mediterranean summer conditions, greenery outweighs even accessibility and crossing safety in shaping tourist route choice.

The second most noticeable difference in the results is the importance tourists in Lyon attribute to traffic-related equipment, which received higher importance than all other attributes [RQ2]. This result was unexpected because pedestrian crossings are always marked on the ground in Lyon, unlike in Thessaloniki. Similarly, there are many traffic lights for safe road crossings. It is therefore worth considering why this result was obtained:

• It may be due to people’s behavior behind the wheel, driving faster or more aggressively, whether in cars, bikes, scooters, or electric scooters. So, tourists in Lyon would feel less safe.
• Conversely, it is possible that tourists in Lyon feel safer in traffic than in their place of origin. As a result, they would pay greater attention to unfamiliar situations and to those for which they have not developed appropriate habits.

The more balanced distribution of attribute importance observed in Lyon, with crossing safety ranking highest (22.22%), aligns with existing studies highlighting the critical role of intersection design, signalization, and perceived safety in pedestrian decision-making. Conjoint and modeling approaches have consistently shown that safety-related infrastructure and accessibility features significantly influence route preferences in urban contexts [36, 37]. Built-environment research further identifies crossing facilities and connectivity as foundational determinants of walkability in compact European cities [38]. From a tourism perspective, accessibility to destinations and integration with transport networks are key components of visitor mobility behavior [42]. The Lyon results, therefore, reinforce prior evidence suggesting that in temperate, well-connected cities, tourists evaluate walkability as a multi-dimensional construct in which safety and accessibility carry weight comparable to environmental comfort.

In both cities, the presence of pavement obstacles and the availability of weather shelter were ranked last and second-to-last [RQ3]. These results, therefore, indicate a lack of interest, or at least less interest, in these attributes compared to the others. Without a doubt, tourists and others often perceive these attributes as secondary or do not consider the possibility of a city where streets are blocked, and shelters are nonexistent. However, it is possible that some areas, such as Vieux Lyon, have no shelter from which to take refuge during a rainstorm. The same is true for pavements with obstacles.

Apart from the need for green spaces in Thessaloniki, we can see that tourists in both cities have similar needs, such as being close to public transport stops and having facilities for safe road crossing [RQ3].

The findings are most comprehensible when considered in relation to the conceptual frameworks outlined in the Introduction. Walking is the oldest and most accessible mode of transportation, and it is significantly influenced by urban design, local climate, and cultural perspectives. In this context, visitors in Thessaloniki exhibit a pronounced inclination towards green places. This corroborates prior research underscoring the importance of urban greenery for climate, environmental health, and psychological well-being. This is particularly evident in Mediterranean climates, where compact structures and insufficient shade exacerbate urban heat. The World Health Organization (2022) [47] underscores the importance of green, pedestrian-friendly landscapes for health and sustainability. The result confirms that green space is not merely a passive setting; it significantly influences mobility behavior, particularly in hot urban areas such as Thessaloniki. Such pedestrian-oriented urban interventions signify a shift toward integrating greener and more walkability-oriented approaches into public space design. Nonetheless, their efficacy for travelers depends on their accessibility and proximity to major tourist routes, as this study demonstrates.

In Lyon, the dominance of traffic safety infrastructure illustrates the relationship between perceived and actual safety and the urban experience. Despite Lyon’s superior pedestrian crossings and traffic signals compared to Thessaloniki, the persistent focus of tourists on this feature indicates that design alone does not ensure a perception of safety, particularly when vehicle conduct or the unfamiliarity of micromobility devices introduces an element of unpredictability. This supports Hall’s (1966) [11] cultural-anthropological argument that space is not neutral but is interpreted differently by users based on prior experience and cultural conditioning. What appears to be a well-designed crossing to a resident may still seem unsafe to a visitor from a quieter or more regulated pedestrian environment.

The findings further substantiate the notion presented in the Introduction that walkability is intricately connected to intermodality. Walking is an essential part of every trip, and efficient public transportation links enhance its appeal. The ranking of proximity to transport stops as second in both cities indicates that tourists appreciate the convenience and reassurance that alternative transportation offers while navigating new urban environments. This underscores the necessity for transport design that is accessible, coherent, and seamlessly integrated, as articulated in the “Accessibility and connectivity” section of the Introduction.

The lower rankings for weather protection and sidewalk obstacles, although uniform across cities, need attention. As explained in the Introduction, individuals generally observe these attributes just when they are deficient or malfunctioning. In cities with intermittent rainfall or harsh temperatures, the absence of shelter may seem less urgent under temperate conditions. Nevertheless, the comfort of tourists may be compromised under adverse conditions, so urban resilience planning must incorporate this consideration through meticulous design rather than merely reacting to criticism.

From a methodological standpoint, our study contributes to tourism and walkability research by operationalizing micro-scale street-level attributes within a comparative CBC framework. While tourism studies have widely applied choice experiments to examine residential redevelopment preferences, destination attributes, or niche tourism products [40, 41, 43], relatively few have modeled pedestrian route-level walkability determinants through trade-off-based experimental design. Furthermore, research linking tourism and walkability has often relied on aggregate indices such as Walk Score rather than behavioral preference modeling [44]. By applying identical randomized CBC experiments across two European cities, this study bridges transport behavior analysis and tourism mobility research, providing quantifiable and context-sensitive insights into how visitors prioritize safety, accessibility, greenery, and micro-impedances.

This paper further supports the human-centric social perspective presented in the Introduction. Hall (1966) [11] asserts that architects and urban planners will avoid disaster only if they regard humans as interlocutors with their environment. Tourists, as transient occupants of space, provide a distinctive viewpoint on a city’s walkability, free from habitual influences. In contrast to residents, they assess a location based on immediacy, clarity, comfort, and sensory perceptions. This provides essential insights into the factors that render a space welcoming or difficult to navigate.

From a broader planning perspective, these findings indicate a shift towards user-centered, evidence-based urban planning. They also validate the efficacy of Conjoint Analysis in uncovering latent preferences that may not be discernible through observation or hierarchical planning. This study prioritizes tourist perceptions, contrasting with earlier top-down pedestrian planning initiatives that often failed to sufficiently incorporate user perspectives.

Finally, it is important to note that although Thessaloniki and Lyon vary considerably in size, climate, and infrastructure, the analogous rankings of certain attributes (proximity of public transport stops and safety) suggest a common cross-cultural rationale for tourists’ preferences in walkable environments. Nonetheless, local particularities, such as temperature in Thessaloniki or traffic patterns in Lyon, might alter the relative significance of many characteristics. Urban planners should therefore seek to blend universal design principles with localized adaptation, especially when designing for short-term users, such as tourists [RQ4]. The comparative evidence supports RQ4 by distinguishing between priorities that are likely transferable (such as crossing quality and the proximity of the public transport stop) and those that are context-sensitive (such as green/shade intensity and obstacle mitigation).

Several limitations should be noted. First, the study’s findings are based on a relatively small, uneven sample of 163 respondents, limiting the extent to which the results can be applied broadly. Nevertheless, it is worth noting that in Conjoint Analysis, we do not expect a sample to be statistically significant in the same sense as in common polling exercises, as respondents provide multiple answers to the same CBC problem. Moreover, the tourist population is significantly lower compared to the resident population. Second, all data were gathered in the summer of 2021, which may not account for seasonal variations or post-pandemic changes in tourist behavior. Third, although Conjoint Analysis offers valuable insights into perceived preferences, it fails to assess real walking behavior or variables such as heat exposure and noise in urban environments. Ultimately, because the poll was administered exclusively to tourists, the findings cannot be extrapolated to the local population.

It is also important to acknowledge the temporal context of data collection. The surveys were administered during Summer 2021, when international tourism was gradually recovering from pandemic-related disruptions. Although major lockdowns had been lifted in both countries, residual health awareness and behavioral adaptations may have influenced preferences. For instance, heightened sensitivity to crowding and enclosed environments could partially explain the strong emphasis on green space and outdoor comfort, particularly in Thessaloniki. At the same time, the prominence of accessibility and structured infrastructure in Lyon may reflect a preference for predictability and organized public space during a period of gradual normalization of mobility. While it is difficult to isolate pandemic effects from structural urban characteristics, the temporal context likely interacted with environmental and climatic factors in shaping perceived walkability.

Future research could increase the sample size by including residents for comparison with tourists and integrate stated-preference methodologies with behavioral data, such as GPS tracking or ethnographic observation. While the present study adopts a multidisciplinary approach that integrates urban planning and behavioral insights, future research could be strengthened by involving a sociologist or social geographer with expertise in tourism. This would provide a more profound examination of tourists’ socio-spatial behaviors and cultural anticipations, therefore enhancing the understanding of walkability assessments and their contextual variability.

CONCLUSION

In this study, five attributes were selected as characteristics, and a quantitative research study, in the form of Conjoint Analysis, was conducted in both Thessaloniki and Lyon. The five characteristics that were chosen as most important were the following: 1) proximity of public transport stop: (i) < 5 min, (ii) 5-15 min, (iii) > 15 min; 2) equipment for crossing roads and safety: (i) traffic lights, (ii) pedestrian crossing, (iii) no traffic lights or pedestrian crossing; 3) weather protection: (i) totally covered, (ii) partially covered, (iii) uncovered; 4) green elements in the street and places: (i) vegetated, (ii) few plants, (iii) no green elements; 5) number of obstacles on the sidewalk: (i) no obstacles, (ii) few obstacles, (iii) obstacles.

Tourists in the two cities completed 163 questionnaires, with 78 in Thessaloniki and 85 in Lyon.

In Thessaloniki, the order of the attributes’ importance was found to be as follows: 1) green elements in the street and places: 31.15%, 2) proximity of public transport stop: 18.56%, 3) equipment for crossing roads and safety: 17.86%, 4) number of obstacles on the sidewalk: 16.61%, and 5) weather protection: 15.81% [RQ1].

In Lyon, the order of the attributes’ importance was found to be as follows: 1) equipment for crossing roads and safety: 22.22%, 2) proximity of public transport stop: 21.18%, 3) green elements in the street and places: 20.65%, 4) weather protection: 18.57%, and 5) number of obstacles on the sidewalk: 17.38% [RQ1].

Many interesting results were derived from comparing tourists’ perceptions of walkability in the two cities, and even more from comparing the utility of each level for each attribute.

The population has neglected walking for many years in favor of ever-faster, simpler modes of transportation. However, as cities’ collective environmental awareness has increased, actions have been taken to promote walking once again. However, for this to happen, it is necessary to consider the needs and expectations of pedestrians and potential pedestrians who will use the public space facilities.

The walkability of a place depends on many factors, and population surveys help us learn about people’s priorities. As a result, urban and transportation planners can consider various parameters, such as safety, sensory well-being, tree cover, attention to the aesthetics of the surroundings, and ease of movement, at different scales.

Through this study, we observed trends among tourists, finding similar expectations in the cities of Thessaloniki and Lyon, and noting a significant attachment to green spaces in the Greek city. We can assume that this attachment is due to the city’s climate and the development of green areas.

In another study, one might expect the results for the inhabitants of these cities to vary across cities. Tourists’ expectations are linked to the habits they have acquired in their home countries. Tourists are a population that visits the city for a relatively short period, walks around, and have cultural needs different from those of those who live in the city.

Tourist cities, such as Thessaloniki and Lyon, must ensure their spaces are well walkable for both travelers and residents. To achieve this, they must reconcile the expectations of each group without abandoning any population [RQ4].

Given the prominence of green elements in Thessaloniki (31.15% relative importance), urban interventions should prioritize microclimatic enhancements along pedestrian routes. This includes increasing tree canopy coverage along key tourist corridors, installing lightweight shading structures in exposed segments of the waterfront and central streets, and integrating small-scale green infrastructure (e.g., pocket parks, planted corridors). Given the Mediterranean summer context, such interventions may yield immediate improvements in pedestrian comfort and tourist satisfaction. Additionally, maintaining clear, obstacle-free sidewalks in high-tourism zones should remain a secondary yet consistent priority, ensuring accessibility for all user groups.

In Lyon, where crossing safety emerged as the most important attribute (22.22%), policy efforts should focus on intersection-level improvements. This includes enhanced pedestrian signal timing, raised crossings in mixed-traffic areas, improved visibility at junctions, and traffic-calming measures in tourist-dense corridors. Given the balanced importance of attributes, integrated pedestrian planning, combining safe crossings, multimodal connectivity, and clear wayfinding, appears more appropriate than single-attribute interventions.

This paper presents an applied, comparative CBC of tourists’ walkability preferences across two European cities that differ in climate, multimodal network maturity, and baseline green infrastructure. By estimating relative attribute importances and level utilities, we quantify trade-offs among four design-actionable, observable attributes, i.e., crossing equipment/markings, proximity of the public transport stop, green/shade, and obstacles, plus weather protection. Two findings are transferable to similar European urban contexts: (i) upgrading crossing quality/legibility and (ii) ensuring nearby public transport access consistently emerge as top priorities.

Context-sensitive levers should then be calibrated to local conditions: in warmer climates and sun-exposed corridors, green/shade and weather protection become more salient; where micromobility/traffic complexity is higher, crossing treatments predominate. In practice, cities can phase investments by first addressing crossings and public transport proximity along priority tourist corridors, and then tuning shade/obstacle mitigation to their climatic and street-morphology profiles.

5. STUDY LIMITATIONS

Several limitations should be acknowledged when interpreting the findings of this study. First, the sample size (163 respondents) is relatively modest and unevenly distributed between the two cities. Although this is acceptable for CBC analysis, where each respondent provides multiple-choice observations, it still limits statistical generalizability beyond comparable tourist contexts.

Second, data collection took place exclusively during the summer of 2021. As a result, the findings primarily reflect warm-season conditions. They may overemphasize attributes related to thermal comfort (e.g., green elements and shade), while underrepresenting priorities that could emerge in colder or wetter seasons. Pandemic-related travel patterns at that time may have also influenced the demographic profile of respondents, notably the overrepresentation of younger tourists.

Third, the study relies on stated preferences rather than observed behavior. While CBC is well-suited to identifying trade-offs and relative attribute importance, it does not capture actual walking routes, exposure to environmental conditions, or real-time constraints. Consequently, the results reflect perceived rather than revealed behavior.

Fourth, the focus on tourists means that the findings cannot be extrapolated to residents, whose walking behavior is shaped by routine travel, local knowledge, and long-term exposure to the urban environment. Tourists evaluate walkability in terms of immediacy, clarity, and comfort, which differ fundamentally from those of commuters or residents.

Finally, although the selected attributes were intentionally limited to observable, design-actionable, and comparable factors, this necessarily excludes other potentially influential dimensions of walkability, such as noise, crowding, air quality, slope intensity, or social interactions. These omissions may partly explain the relatively balanced importance scores observed in Lyon.

Future research should expand the sample size, incorporate multiple seasons, include resident populations, and combine stated-preference methods with behavioral data (e.g., GPS tracking or systematic observation) to strengthen external validity and deepen understanding of tourists’ walkability experiences.

REFERENCES