1 Introduction

Urban sprawl driven by urbanisation has contributed to a sharp rise in privately owned vehicles and competition for restricted resource space (Yu et al. 2019; Li et al. 2020a). Too many private vehicles on the road network lead to negative externalities, including congestion and emissions. Several of these external effects linked to the heightened utilisation of private vehicles can be addressed by reducing their usage to more viable modes of transport, like public transport (hereinafter, PT). Prior studies have documented the merits of using PT for individuals and cities through the promotion of good health, energy conservation and emission reduction, reduction in vehicular accidents, traffic congestion reduction, and provision of jobs (Nguyen-Phuoc et al. 2021; Cheranchery and Maitra 2018; van Soest et al. 2020; Truong and Currie 2019).

PT has experienced a decrease in usage in recent times (Boisjoly et al. 2018; Miller et al. 2018). This decline was exacerbated when many countries battled the COVID-19 outbreak (Jenelius and Cebecauer 2020; Sogbe 2021; Tirachini and Cats 2020). A decrease in PT usage is caused by many factors, including the built environment (BE), public policy, extent of economic development, and acquisition and utilisation of privately owned vehicles (Zhang et al. 2022). However, de Oña (2021) noted that service quality (hereinafter, SQ), satisfaction and behavioural intention or attitude towards PT are three crucial factors influencing a mode shift. Poor SQ (e.g., reliability, comfort) leads to a satisfaction gap and, subsequently, a negative attitude towards PT (Cheranchery and Maitra 2018). Thus, passengers’ satisfaction with PT and the ability to entice new users through a mode shift hinges on the quality of service (van Lierop and El-Geneidy 2016). This suggests that knowing the dimensions of PT usage and improving on them can mitigate the acquisition and use of private vehicles, which are competitors to PT and threats to the environment and human survival.

Pourbaix (2011) estimated an 80% global increase in daily urban travel using private vehicles by 2025, with the growth occurring in developing countries. Similarly, the World Health Organization (2011) noted that private vehicle usage in emerging countries is expected to supersede that of developed nations by 2030. Recent studies have confirmed these earlier projections (Saxena and Gupta 2023; Soltani 2017; Ntim et al. 2022). Besides, travel requirements in emerging and developed nations may differ due to differences in vehicle ownership, transport systems, infrastructure, and economic status. It is, therefore, necessary to apply successful policies and procedures to increase PT usage. However, this requires a thorough understanding of extant knowledge, clear-cut lacunae, and disagreements that must be resolved. Policymakers and transport authorities may lack the knowledge to make informed judgments without a complete and systematic research evaluation.

As far as we know, no systematic literature review focused on these three factors influencing PT usage. In a prior review, Das and Pandit (2013) reviewed aspects of bus transport service that impact users' perceived quality of service. This contributed significantly to the SQ factors that define the level of service of PT. However, while their study emphasised service factors, it did not provide a clear-cut methodology, and the selected literature is not exhaustive. de Ona and de Ona (2015) presented a review of SQ, providing an extensive review of methodologies deployed in the study of SQ and the merits and demerits of the methods. However, it focused more on the methodologies than the SQ attributes. Besides, most selected papers focused on SQ in the aviation industry. van Lierop et al. (2018) analysed important factors (i.e., satisfaction and loyalty). Their study provided insights into the features that encourage PT users to remain committed to using it. However, the scope of their review was limited by geographical boundaries, with only two papers from Africa and Asia each. Ojo (2019) subsequently provided a literature review of SQ attributes and the methodologies deployed. It highlighted the need to identify important attributes concerning the type of PT, regional context, and culture. However, the study limited the review to 2015; therefore, information about SQ from 2016 to the publication date, 2019 and subsequent years have been excluded.

This study aims to systematically review and examine existing literature on the attributes influencing PT usage, focusing specifically on three important factors: SQ, satisfaction and attitude towards PT usage within the context of developing countries to address the specific needs and concerns of passengers. It identifies and discusses the attributes based on qualitative (subjective) and quantitative (objective) assessments and the impact on the utilisation of PT. It should be mentioned that this research focused on public buses, including bus rapid transit (BRT), as these modes of transport are dominant and synonymous with most developing countries (Nguyen-Phuoc et al. 2021; Cheranchery and Maitra 2018).

The ensuing sections of this paper are structured as follows: Sect. 2, which contains the methodology, provides the scope of the review and presents the strategy for the systematic literature search and selection. It is followed by Sect. 3, which presents the results and the influence of SQ, satisfaction, and attitude towards PT usage. Section 4 offers future research directions and policy implications, including a synopsis of the evidence.

2 Methodology

The study adopted a systematic literature search using the Preferred Reporting Items for Systematic Reviews and Meta-Analysis (PRISMA) guidelines (Moher et al. 2015). The search was done in January 2023 and updated in July 2023. The data source repositories were TRID: TRIS and ITRD, Scopus, Web of Science, Inspec, Compendex and GeoBase. Having identified the keywords following an exploratory review, a search using Boolean operators was carried out:

(“public transport” OR “public transport usage” OR ridership) AND (factors AND “developing country”).

2.1 Inclusion and exclusion criteria

The criteria for including or excluding selected papers in the systematic review were as follows:

  1. 1.

    Papers were included if they focused on research conducted in developing countries.

  2. 2.

    Papers not in English were excluded.

  3. 3.

    Papers related to COVID-19 and PT were excluded. The study recognises the broader context of declining PT usage and the concurrent rise in private car usage in developing countries, as highlighted in the literature even before the pandemic outbreak. Moreover, it is essential to emphasise that while COVID-19 has undoubtedly impacted PT usage patterns in the short term, it is transient, and its long-term implications are uncertain. It is possible that some changes observed during the pandemic, such as increased remote work and a preference for private transportation to maintain social distancing, may persist to some extent even after the pandemic has subsided. However, historical trends suggest that PT remains a vital component of urban transportation systems, and factors such as population growth, urbanisation, environmental concerns, and congestion will continue to drive demand for efficient PT solutions. Therefore, while COVID-19 may have influenced PT usage patterns in the short term, its long-term impact remains to be seen, and other factors, such as SQ, satisfaction, and attitude towards PT, are most likely to play a significant role in shaping future trends.

  4. 4.

    Papers that combined SQ and satisfaction of various modes, including paratransit, were excluded.

  5. 5.

    Only peer-reviewed research or journal articles were included.

  6. 6.

    Papers on SQ, satisfaction, and attitudes towards public bus transport, including BRT in developing countries, were included. This is because bus transport is a widespread means of transportation synonymous with many developing countries.

  7. 7.

    Papers published between 2000 and 2023 were considered.

The search in the repositories mentioned above yielded the following results: TRID: TRIS and ITRD (15,000), Scopus (32,010), Web of Science (3752), Inspec (1603), Compendex (2012) and GeoBase (1558). Further details of the search process and results are provided in Table 1 and Fig. 1. This resulted in a total of 55,935 articles gathered, with subsequent field restrictions applied, leading to 2483 records (refer to Table 1). This number was imported into Rayyan (Ouzzani et al. 2016), and 962 duplicates were removed, leaving 1521 records. The authors then screened titles and abstracts to determine whether they focus on (i) SQ and satisfaction with public bus transport and (ii) attitudes towards PT. The inclusion and exclusion criteria are listed above. This was done to narrow the selection to articles aligned with the topic, resulting in 335 chosen for full-text review. Of these, 13 articles were unavailable, leaving 322 for full-text review. In the full-text review, 92 records were excluded for lacking relevance to the research theme, 67 for combining SQ and satisfaction of different modes, 35 for focusing on paratransit and related modes, and 30 for concentrating on the built environment as a factor. The snowballing method identified 10 additional studies from reference lists, of which 6 were retained following the same criteria. The systematic search produced an extensive body of literature on the subject, with a final selection of 104 papers for the review. The PRISMA flow diagram in Fig. 1 depicts the number of records at each stage and the reasons for exclusion and inclusion.

Table 1 Summary of search results
Full size table
Fig. 1
figure 1

PRISMA flow diagram

Full size image

Various pieces of information were documented from each of the records. Initially, general details, including author(s) and publication year, were extracted to provide the research’s context and scope. Subsequently, we scrutinised the methodologies employed for data analysis (such as SEM), the data source (primary or secondary), approaches for collecting data (such as questionnaires or interviews), and the sample size. Finally, data were compiled on the constructs that were assessed.

3 Results

3.1 Patterns in literature

China leads in publications, accounting for nearly a quarter of all studies (23), followed by India with under one-fifth (18), Brazil (7), Ghana (6), Vietnam (5), Bangladesh (6), and Ethiopia (6). Other countries collectively share the remaining research. Azerbaijan is transcontinental, with 1 study and 1 review article focusing on developing countries. It is evident that research predominantly centres on Asian countries, and there is a dearth of research in developing nations beyond Asia. These countries can potentially introduce fresh perspectives to the research on the subject. Figure 2 illustrates the publication count by country, while Fig. 3 shows the geographical distribution.

Fig. 2
figure 2

Distribution of publications by country (by July 2023)

Full size image
Fig. 3
figure 3

Distribution of publications by continent (by July 2023)

Full size image

3.2 Methodological approaches and empirical findings in PT research

The articles in this literature review utilised qualitative and quantitative data and data analysis methods. Quantitative data involves collecting data through surveys, experiments, or statistical analysis. On the other hand, qualitative data provides insights into attitudes, opinions, behaviours, or motivations and is often gathered through interviews or observations. The plurality of studies analysed in this review relied solely on quantitative data (98). In two cases, qualitative data was used. Mogaji and Nguyen (2021) utilised semi-structured interviews to investigate the level of contentment of passengers with disabilities. In addition, Simangunsong et al. (2023) used interviews to assess the perception of BRT users. In four cases, qualitative data was integrated with quantitative data. Thai and Quan (2023) combined interviews and surveys to analyse the determinants influencing urban individuals' inclination to utilise buses. Kacharo et al. (2022) studied the safety and security of female passengers using focus group discussions and questionnaires. Araya et al. (2022) combined a questionnaire survey, ethnographic observations and document reviews to evaluate the safety and security of women on public buses. Finally, Busco et al. (2023) examined the connection between bus transport and social exclusion using surveys and focus group discussions. The majority of studies relied on primary data sources (94); however, a few utilised secondary sources, including data gathered by transit agencies (7) or initially gathered for different research projects (3). The primary method of data collection in most cases was through questionnaire surveys.

There are vast differences in sample sizes, with the least being 29 respondents from a questionnaire survey (Ganji et al. 2021) and the highest being 588,000 from a transit agency entry data (Guzman et al. 2019). Common analysis methods deployed include the service quality (SERVQUAL) model and service performance (SERVPERF) model, which have been widely used, structural equation modelling (SEM), importance performance analysis, different kinds of regression analysis and machine learning algorithms such as random forest (RF) and gradient boost decision trees (GBDT).

While most studies leaned towards quantitative analysis, a few incorporated qualitative data, demonstrating a balanced exploration of the subject matter. Moreover, the range of sample sizes across studies underscores the diversity in research approaches, from smaller-scale surveys to extensive transit agency data analysis. This variation reflects the complexity and breadth of the topic under investigation, accommodating different methodological requirements and research objectives. A spectrum of techniques was employed in terms of analysis methods. The diversity of research methodologies and analytical approaches showcased in the literature contributes to a comprehensive understanding of PT usage dynamics and SQ assessment. Researchers can glean valuable insights into the complex interplay between SQ, passenger satisfaction, and mode choice behaviour by exploring qualitative and quantitative findings.

The literature debates the best possible methodology for examining SQ dimensions and satisfaction. The SERVQUAL model established by Parasuraman et al. (1985) measures commuter differences in appreciation and expectation of SQ dimensions. Nonetheless, Cronin and Taylor (1992), as cited by Esmailpour et al. (2020), highlighted that the measure of commuters' expectations is not of the essence and, in lieu, put forward the SERVPERF model, which utilises performance perception to measure SQ. Matzler et al. (2004), cited by Zhang et al. (2019a), highlighted that, albeit salient, IPA has a linear and symmetric association between satisfaction as a dimension and satisfaction in general, which they claimed is a distorted supposition. The authors made a case for the three-factor theory as a better alternative to eliminate distortions.

Similarly, Tuan et al. (2022) indicated that the IPA has crucial conceptual and methodological constraints, favouring the three-factor theory. Zheng et al. (2022) made a case for the three-factor theory, arguing that the other methods, including linear regression, are flawed because they presume linearity and symmetricity for SQ dimensions. Fang et al. (2021) shared the same viewpoint concerning IPA; however, by contrast, they added that the three-factor theory also has the same weakness as IPA in determining the hierarchy of significance of commuter satisfaction and, therefore, proposed GBDT and Impact Asymmetry Analysis (IAA).

3.3 Identifying qualitative and quantitative factors that impact perceptions of SQ and satisfaction

Over time, researchers have employed a broad array of factors to examine the utilisation of PT. Among these, SQ and satisfaction encompass a vast number of studies. In this section, we present a summary of these factors, which were informed by qualitative and quantitative assessments, and their impact on utilising bus transport.

Nathanail (2008) posited that the SQ of PT is assessed through two leading indicators. The objective indicator, provided by operators, is a quantitative metric for comparison with standards or past performance. The subjective indicator, derived from surveys, offers qualitative measures, capturing users' perceptions by evaluating the gap between actual and ideal service levels. Quantitative factors are typically measured using numerical data. For example, frequency—quantifying the regularity and frequency of PT services, indicating how often vehicles are available, travel time—assessing the duration it takes PT vehicles to travel between stops or destinations, etc. and qualitative factors involve subjective assessments of availability or passenger comfort. Although these two measures exist, a plurality of the studies employed qualitative factors. On the other hand, very few researchers have focused on measuring these factors quantitatively. The SQ and satisfaction factors that significantly impact PT usage overlap. Nonetheless, we discuss them separately for clarity.

Cronin and Taylor (1992) defined SQ as “an attitude that reflects a comprehensive, long-term evaluation.” Parasuraman et al. (1985) provided a historical definition of SQ: “SQ perceptions result from comparing consumer expectations with actual service performance.” This has led to the widely embraced perspective that excellent SQ entails delivering performance that either meets or surpasses customers' expectations of the service. Commuter perception constitutes a crucial aspect of SQ. In our analysis, 47 papers assessed commuters’ perception of SQ. Most papers examined SQ with various factors, such as availability, reliability, safety, security, cleanliness, etc. Table 2 summarises the literature on SQ, which is incorporated in this review.

Table 2 Summary of selected literature on service quality
Full size table

Two factors that featured prominently in most papers in the review are the perception of passengers regarding their safety (30) and security (19) during transit, with one paper each considering quantitative insights. The authors asserted that safety and security are the most influential determinants of public bus usage and are strongly associated with gender (Ikhlaq et al. 2017; Alomari et al. 2023; Suman et al. 2016; Deb and Ali Ahmed 2018; Houria and Farès, 2019; Adom-Asamoah et al. 2021; Nguyen-Phuoc et al. 2021; Busco et al. 2022, 2023; Araya et al. 2022; Kacharo et al. 2022; Verma et al. 2020). Safety and security have an overarching impact on other latent attributes (Wang and Zhu 2014), and this is reflected while on-board the bus (Hu et al. 2015; Orozco-Fontalvo et al. 2019; Yarmen and Sumaedi 2016) or while waiting at the bus stop or station, especially at night (Busco et al. 2022; Esmailpour et al. 2020; Suman et al. 2016; Noor and Iamtrakul 2023, 2024). Addressing safety and security concerns is vital for building public trust, encouraging ridership, and creating a positive experience for passengers using PT.

Another significant qualitative factor in this review is service reliability (20), with 19 papers measuring qualitative aspects. Chakrabarti and Giuliano (2015) defined reliability “as one which consistently operates according to its schedule or plan.” The varying aspects of this factor complicate it. For example, the arrival and departure of PT may be timely, but disparities in the dwell time over the period may lead to delays and impact schedule adherence, potentially affecting the reliability of the service about meeting predetermined schedules. While timely performance significantly influences passengers’ use of bus services (Deepa et al. 2022; Cheranchery and Maitra 2018; Randheer et al. 2011; Sam et al. 2018; Das and Pandit 2015), in-vehicle time poses a significant obstacle to the use of bus services (Maitra et al. 2015; Thai and Quan 2023; Shen et al. 2016). In-vehicle time refers to the duration passengers spend travelling inside the bus or other PT vehicles. This time spent in transit can significantly influence passengers' decisions regarding bus services. Factors such as longer travel times or perceived inefficiencies during the journey can pose obstacles or challenges that may deter individuals from utilising bus services. Therefore, understanding and addressing issues related to in-vehicle time is crucial for improving the attractiveness and efficiency of bus services. In addition, waiting time indirectly impacts passengers’ perception of the reliability of PT (Hu et al. 2015; Chaudhary 2020; Fu et al. 2018). As demonstrated through the various studies, a reliable PT system contributes to increased ridership and positively impacts overall transportation behaviour. The ability of PT to adhere to schedules, provide consistent service, and minimize disruptions plays a pivotal role in shaping users' perceptions and preferences. Addressing and improving reliability issues can be crucial for promoting PT's widespread and sustainable use.

Affordability as a qualitative factor involves the assessment of user perceptions of the cost of PT services. In our review, 16 papers examined affordability, with twelve considering qualitative insights. The authors opted for varied terminologies to label this attribute. Most authors measured affordability as fare (6), cost or coined travel cost (7), ticket price (2), pricing or price (2) and affordability (1). Many PT users view affordability as a deciding factor. A principal reason for using PT is low bus fares (Kacharo et al. 2022; Birago et al. 2017; Alomari et al. 2023; Sam et al. 2014) and an increment in transport fares engender a decrease in PT usage (Maitra et al. 2015; Joewono et al. 2016; Adom-Asamoah et al. 2021) and low-income earners are the hardest hit by the increases (Toro-González et al. 2020). This indicates that for most passengers, the use of PT is contingent on affordability.

In 14 papers, the authors asserted that commuters’ perceived comfort is a significant driver of PT usage. Some of these papers examined the overall influence of comfort on PT usage and its association with in-vehicle crowding (Saleem et al. 2023; Sam et al. 2014; Deepa et al. 2022; Thai and Quan 2023; Atombo and Wemegah 2021; Birago et al. 2017; Houria and Farès, 2019; Adom-Asamoah et al. 2021), in-vehicle comfort, precisely seating comfort, the temperature inside the bus and facilities at the station (Deb and Ali Ahmed 2018; Shen et al. 2016; Alomari et al. 2023; Esmailpour et al. 2020; Hu et al. 2015). In our review, eight papers evaluated crowdedness in the bus or at stations (Shen et al. 2016; Deb and Ali Ahmed 2018; Maitra et al. 2015; Chaudhary 2020). Shen et al. (2016) highlighted that crowdedness substantially influences commuters’ perception of comfort and has a more pronounced effect on commuters standing than those seated.

In our review, 16 papers measured accessibility, with nearly all (13) considering qualitative insights. Aspects considered include the ease with which passengers can navigate PT systems, such as information availability or ease of use (Busco et al. 2023; Houria and Farès, 2019; Ikhlaq et al. 2017). Other qualitative factors include cleanliness inside the bus and seats (8) (Deepa et al. 2022; Chaudhary 2020; Saleem et al. 2023; Hu et al. 2015; Birago et al. 2017), availability (9) (Adom-Asamoah et al. 2021; Hu et al. 2015; Atombo and Wemegah 2021), driver behaviour, speed and customer care.

Regarding quantitative factors, three papers measured accessibility by considering aspects such as service coverage (Insani et al. 2021) and the number of vehicles in service (Toro-González et al. 2020; Rabay et al. 2021). Challenges with transfers, access and egress modes can deter commuters from using PT. Insani et al. (2021) noted that switching between buses contributes to extended travel time, posing a distinct drawback for PT users, and integration with other modes of PT is recommended. Toro-González et al. (2020) concluded that increasing the number of vehicles in operation results in a rise in PT demand. Nonetheless, this depends on the speed of the vehicles. Four papers measured affordability (Guzman et al. 2019; Rabay et al. 2021; Toro-González et al. 2020; Insani et al. 2021). One paper each measured indirect aspects of reliability and frequency of PT; Insani et al. (2021) measured waiting and travel time, while Toro-González et al. (2020) measured the frequency of arrival and departure.

Other researchers advocated and used a combination of qualitative and quantitative assessments. Combining these measures could offer a more practical and dependable tool for assessing transit SQ (de Ona and de Ona 2015). In our review, two papers used both measures to assess commuters’ perception of SQ and satisfaction. Rong et al. (2022) measured perceived and actual time-related factors such as stopping, dwell, waiting, and travel times. In addition to time-related factors, Suman et al. (2016) measured connectivity, cost, security, punctuality, accessibility, comfort and safety.

3.3.1 Satisfaction

While SQ involves a cognitive assessment of the variance between initial expectations and perceived performance, satisfaction is its emotional equivalent. It pertains to the customer's contentment or dissatisfaction with the service (Carvalho Dos Reis Silveira et al. 2022). Table 3 shows a summary of the literature on satisfaction, which is incorporated in this review. In our review, 52 papers examined commuters' satisfaction with PT usage. Virtually, all the papers used a range of service factors, such as accessibility, reliability, and cleanliness, to mention a few. In some instances, we observed that alternative factors like passenger expectation (2) and perceived value (2) were considered satisfaction factors. Also, mood, emotions and cognitive activities (1) were used as satisfaction attributes. In urban environments characterised by high population density and extensive PT networks, factors influencing commuters' satisfaction with PT services are of paramount importance. Geographical nuances, social dynamics, and the nature of PT offerings are pivotal in shaping passengers' experiences and perceptions.

Table 3 Summary of selected literature on satisfaction
Full size table

Commuters face many challenges when utilising bus transport across diverse urban landscapes, from bustling metropolises to smaller urban centres. Accessibility is critical, particularly for individuals with disabilities or mobility limitations. Ramps for wheelchairs (Ali and Abdullah 2023; Ji and Gao 2010), accessible stations (Jayakumar et al. 2023; Chaisomboon et al. 2020; Girma 2022), information accessibility regarding signage about routes, schedules, and facilities (Nguyen and Pojani 2023; Han et al. 2022; Girma et al. 2022), information regarding announcements to inform passengers about stops, transfers, and other relevant information (Fang et al. 2021; Lan et al. 2022; Fu and Juan 2017) and technological accessibility regarding real-time information through apps or displays, assisting passengers in planning their journeys (Mendez et al. 2019; Bose and Pandit 2020; Li et al. 2020b). These are essential elements ensuring inclusivity and ease of use for all passengers.

In the rapidly urbanising environments of developing countries, where PT infrastructure may be limited and urban congestion is common, the comfort of passengers during their transit journeys becomes paramount. Factors such as comfort while waiting and riding (Githui et al. 2009; Zhang et al. 2019b; Nwachukwu et al. 2019; Tuan et al. 2022), the temperature in the bus (Wu et al. 2016; Freitas 2013), seating comfort (Andaleeb et al. 2007; Githui et al. 2009) and ventilation (Esmailpour et al. 2020), and the condition of vehicles (Umme et al. 2022; Singh 2016) significantly influence passenger satisfaction. The qualitative insights gleaned from 32 examined papers shed light on how these comfort-related factors impact passenger perceptions and contribute to their overall satisfaction with bus transport services. Studies conducted by Andaleeb et al. (2007), Allen et al. (2019), Nwachukwu et al. (2019), and others highlighted the tangible effect of improved comfort measures on passenger satisfaction, emphasising the importance of addressing these factors within the unique context of developing country urban settings.

Furthermore, safety and security concerns loom large in commuters' minds, influencing their overall satisfaction with bus transport. Whether waiting at a bus stop (Sun et al. 2020; Zhang et al. 2019a) or navigating through crowded vehicles (Ali and Abdullah 2023), passengers prioritise their well-being and peace of mind. Enhancing safety measures and providing a secure environment can significantly improve the attractiveness of bus transport services.

Affordability is another pressing issue, especially in urban areas where the cost of living may be high. Passengers evaluate the financial accessibility of bus transport services, weighing fares against their budgetary constraints. As such, pricing strategies and fare structures are pivotal in shaping commuters' perceptions of affordability and satisfaction.

In addition to these qualitative factors, quantitative metrics such as reliability, comfort, and availability also come into play. On-time performance, seating comfort, and frequency of services directly impact passengers' experiences and satisfaction levels. PT operators must maintain high standards across these dimensions to meet urban commuters' diverse needs and expectations. Table 4 shows the categorisation of qualitative and quantitative factors influencing satisfaction.

Table 4 Quantitative and qualitative factors influencing commuter satisfaction with PT
Full size table

3.4 Attitude towards PT

Private car ownership and usage are often considered deterrents to PT utilisation. Reza Jalilvand et al. (2012) highlighted that a person's attitude influences their actions, and the theory of planned behaviour (TPB) suggests attitudes impact behavioural intentions (Ajzen 1991). Scholars have thus explored the attitudes and intentions of non-PT users, particularly private car users. Social-symbolic, affective and instrumental aspects influence the acquisition and use of cars (Gatersleben 2007). In our review, the papers on attitudes towards PT usage are scanty, and most travel behaviour research to date has primarily centred on instrumental aspects. Table 5 shows a summary of the literature on attitudes towards bus transport usage, which is incorporated in this review.

Table 5 Summary of selected literature on attitudes toward public transport
Full size table

Four papers focused on instrumental aspects, while one paper focused on social-symbolic aspects. The instrumental aspects assessed include cost, flexibility, safety, reliability, travel time, direct lines, comfort, speed and convenience (Tao et al. 2019; Al-Ayyash and Abou-Zeid 2019; de Magalhães and Rivera-Gonzalez 2021; Li et al. 2019b) and social-symbolic aspects such as success and status (Li et al. 2019a). The intention of car users to use PT is considerably influenced by cost, reliability, comfort, travel time, and connectivity and enhancing these factors will boost the intention to use PT (Li et al. 2019b; de Magalhães and Rivera-Gonzalez 2021).

In addition to understanding the factors influencing attitudes towards PT usage, it is crucial to consider the implications of these attitudes for PT policy and planning. For instance, emphasising instrumental aspects such as cost, reliability, and comfort highlights the importance of improving SQ to encourage PT adoption among private car users. Similarly, the recognition of social-symbolic aspects, such as perceptions of success and status associated with car ownership, underscores the need for targeted marketing and messaging campaigns to challenge prevailing cultural norms and promote the social desirability of PT usage. By addressing instrumental and social-symbolic factors, policymakers and urban planners can develop more effective strategies to incentivise PT usage and reduce reliance on private cars.

3.5 Evolution and trends in PT SQ and satisfaction research

Until the last two decades, few studies on PT SQ and satisfaction were conducted in emerging countries. The rapid growth of urban populations and increased income levels resulting from economic improvement generated a significant demand for individual mobility, hence the increased need for assessing this subject.

According to our review, the service factors influencing PT usage investigated over the decades are no different from those assessed today except for a few variations and additions. Before 2020, the investigation predominantly involved qualitative assessments of SQ and satisfaction factors. Since 2020, there has been a shift towards quantitative assessments, including accessibility, reliability, affordability, and transfers.

Since time immemorial, SQ factors like cleanliness, accessibility, reliability, comfort, affordability, etc., have been used in assessing SQ. However, in the last couple of years, researchers introduced factors which impact service reliability into the analysis, such as frequency, waiting time, turning frequency, stopping frequency, travel time, arrival interval, on-time performance, punctuality, information and transfer convenience.

The methodology for data collection and data analysis on the subject has also evolved from SERVQUAL and SERVPERF models and descriptive statistics (Aniley and Negi 2010; Randheer et al. 2011) to more robust methods such as SEM and ML algorithms like GBDT and RF (Nguyen-Phuoc et al. 2021; Lan et al. 2022; Fu 2022). With technological advancements, different data collection methods have been adopted, such as using social media, ticketing data, PT data, and automatic vehicle location (AVL) data (Rong et al. 2022; Suman et al. 2016; Tavares et al. 2021; Zheng et al. 2021; Rabay et al. 2021).

An emerging trend involves examining the preferences of different commuter groups. The extant literature revealed disparities in the desired quality and satisfaction of SQ dimensions among PT users. This diversity among commuters has led researchers and practitioners to emphasise heterogeneity in the PT market. Segmenting commuter groups can help transport authorities provide tailored services to meet the satisfaction of specific user groups (Allen et al. 2019) and allocate resources effectively for targeted interventions (Fu 2022). Our review revealed that such disaggregation has been chiefly conducted for youth, students, tourists, groups considered vulnerable (such as older people, women, and people with disabilities, PWD’s) and more recently, for captive and choice riders.

Based on the scarce literature available regarding segmentation according to elderly passengers, it can be inferred that elderly passengers’ satisfaction is greatly influenced by station broadcast, driver’s habit, punctuality, awnings, complaint handling, safety and security, driver courtesy, information services, waiting time with station broadcast being the most important (Yuan et al. 2019; Chaisomboon et al. 2020; Lan et al. 2022; Busco et al. 2023). Moreover, older public bus users encounter deficiencies in bus stop infrastructure, accessibility and information (Busco et al. 2023).

Researchers have explored segmentation based on the frequency of PT usage, dividing commuters into frequent and occasional riders. Fang et al. (2021), Tuan et al. (2022), Cheranchery and Maitra (2018) and Maitra et al. (2015) investigated the relationship between bus service dimensions and overall satisfaction for both rider groups. Aside from finding notable variance in how these groups perceived SQ, they found dimensions both groups prioritise in common: availability and accessibility (Tuan et al. 2022), waiting area, driver's behaviour, complaint handling, stop announcements (Fang et al. 2021).

Concerning students, diverse factors influence their satisfaction with PT. Agyeman and Cheng (2020) found that frequent breakdowns and extended travel times hinder school bus transport, affecting pupils' learning environment, while for university students, reliability and punctuality are of concern when using public buses (Javid and Al-Kasbi 2021). Furthermore, fare, comfort, reliability, and safety are vital attributes that impact the choice of PT (Sam et al. 2014) and fear of sexual harassment caused female students to avoid public buses (Nguyen and Pojani 2023). Jomnonkwao et al. (2022) switched the scope slightly by focusing on parents' perceptions of school bus SQ, highlighting that infrastructure, information, and safety dimensions influence their views.

Regarding youth commuters, there is proof that enhancing factors like performance, comfort, and assurance would boost satisfaction among youth PT passengers. What’s more, Busco et al. (2023) concluded that young PT users lack safety and SQ and face issues of harassment on PT systems. Nwachukwu et al. (2019) are the sole researchers who examined tourists' satisfaction. According to their results, tourists are discontent with PT usage, citing several SQ factors, particularly accessibility.

In our review, only two papers investigated the travel satisfaction of disabled travellers. Mogaji and Nguyen (2021) found gender-based disparities in travel satisfaction, noting that disabled women's satisfaction was more influenced by security, vulnerability, and the need for assistance compared to disabled men, while high floors in PT vehicles, crowded conditions, and untrained staff are factors affecting PT usage for people with disabilities (Ali and Abdullah 2023). Moreover, they also emphasised the lack of sufficient space for mobility aids in PT as a prominent challenge.

The segmentation also considered the viewpoint of female commuters. From the literature, safety and security are major quality issues and a significant barrier to public bus usage, given that a substantial portion of women experiences sexual molestation and psychological and physical violence often perpetrated by conductors and male co-passengers on buses and at bus stops (Busco et al. 2023; Saigal et al. 2021; Kacharo et al. 2022; Verma et al. 2017; Orozco-Fontalvo et al. 2019; Araya et al. 2022; Umme et al. 2022). In a recent review, Noor and Iamtrakul (2023) highlighted the prevalence of harassment on PT and the lack of responsive policy support to address this issue in developing countries. Figure 4 shows a schematic representation of the evolution and trends in PT SQ and satisfaction research in developing countries.

Fig. 4
figure 4

Evolution and trends in PT SQ and satisfaction research in developing countries

Full size image

4 Future research directions and policy implications

4.1 Synopsis of the evidence

The results in Sect. 3 reveal that using PT for commuting is affected by various SQ factors, satisfaction-related aspects, and attitudes. These effects span pre-commute, commute, and post-commute phases. We highlight the attributes which have been discussed in this literature review.

Safety and security stand out as one of the crucial dimensions influencing users' perception of SQ and satisfaction. This finding contrasts with van Lierop et al. (2018) conclusions, who concluded that safety is not such a strong indicator of satisfaction in developed countries like those in Europe.

Commuters’ perception of reliability also correlates with usage and satisfaction. Overall, this aspect wields a substantial impact on satisfaction. This observation is consistent with the findings of van Lierop et al. (2018), who highlighted that enhancing the onboard experience is only advantageous if passengers are content with the service's reliability. On the other hand, comfort is a crucial factor during both the pre-commute and commute phases.

The convenience of individuals accessing and using PT services is crucial in determining PT systems' overall effectiveness and inclusivity. Accessibility, according to the review, is among the most critical attributes that impact satisfaction. Improving accessibility in PT is essential for creating a system that is convenient, equitable, and capable of meeting the population's diverse needs.

Affordability is another important factor. Many PT users are motivated by its affordability, and any rise in cost disproportionately affects individuals in the lower income brackets (Adom-Asamoah et al. 2021).

Availability is a significant service factor for frequent and occasional users, especially in developing countries (Tuan et al. 2022). Passengers anticipate a PT system that offers easy access without prolonged waiting times or difficulties reaching their desired destinations.

Another notable factor is crowdedness, which illustrates how a single SQ attribute can affect others. Crowdedness affects both comfort and safety on buses and can contribute to incidents of harassment against female passengers (Orozco-Fontalvo et al. 2019).

Challenges with first-mile and last-mile (hereinafter, FM/LM) connectivity are apparent, yet they are underexplored in the context of developing countries, necessitating further investigation. Other underexplored factors include connectivity (how well different modes of transportation are connected), transfers (the process of switching between various modes of transport), waiting environment (the conditions while waiting for transport) and waiting facilities (amenities provided for passengers while waiting). Climate change and heat stress can significantly impact the comfort and accessibility of PT in various ways. As temperatures rise due to climate change, extreme heat events become more frequent, posing challenges for passengers and the transport infrastructure. Rising temperatures due to climate change can lead to increased heat stress, especially during heatwaves, making waiting at outdoor transit stops uncomfortable and prolonged exposure to high temperatures while waiting for transit or during the journey can result in heat-related illnesses, affecting the well-being of passengers. Fraser and Chester (2017) posited that the design of transit systems, including the placement of stops and schedules, contributes to environmental exposure and may pose potential health risks, particularly during periods of extreme heat due to access and waiting. Improving the aesthetics of bus stops and incorporating diverse design elements can positively impact the thermal comfort of bus riders by enhancing their perception of beauty and pleasantness (Dzyuban et al. 2022). Addressing the effect of climate change on PT comfort and accessibility requires a comprehensive approach that involves collaboration between transit agencies, urban planners and policymakers. Sustainable and resilient solutions can increase the overall PT SQ in the face of changing climate conditions.

Until 2000, limited studies investigated SQ and commuters' satisfaction; since then, researchers focused primarily on qualitative assessments of the factors. The turning point for quantitative and qualitative assessments began in 2020 due to technological advancements that aided data collection procedures. As mentioned in Sect. 3, there is a difference in how frequent and occasional users perceive SQ and their satisfaction. Further understanding of occasional users and improvements in these aspects can entice them to increase usage. In addition, the attitudes of private car users can provide valuable insights into their preferences, perceptions, and behaviours related to using PT systems.

4.2 Future research directions

Historically, transport researchers have explored the factors that influence PT usage. While our review uncovers valuable information, it also highlights significant areas that require further investigation. The first of these areas concerns exploring a less studied population group: private car users. Given the rise of private car acquisition and usage in developing countries, as mentioned in the introduction, shifting attention to this group is crucial. Recent attempts (Li et al. 2019a, b; Al-Ayyash and Abou-Zeid 2019; Tao et al. 2019; De Magalhães and Rivera-Gonzalez 2021) have begun to address this, yet their approaches have been somewhat limited. For example, the target population for Al-Ayyash and Abou-Zeid (2019) included only university students and staff, which may not represent generalisability. Particular focus is required on SQ, satisfaction, and attitude towards PT. These areas are crucial for transport authorities and policymakers to intervene effectively.

The second area highlighted in the literature is problems with connectivity—FM/LM, but there is a paucity of literature. PT involves multiple journey stages, including waiting, walking, and switching between different modes, causing inconvenience (Park et al. 2021). The literature on transfers and satisfaction is scanty and conflicting. Das and Pandit (2015) noted more than two, Andaleeb et al. (2007) highlighted no transfers, and Li et al. (2020b) indicated transfer convenience. Future studies must further probe transfers and their influence on PT usage and satisfaction. In addition, research should explore how transit authorities can adopt adaptive measures and sustainable practices to ensure that PT remains a reliable and comfortable option for all, even in the face of changing climate conditions.

Additional research avenue pertains to data collection methodologies. Qualitative research, such as interviewing individuals central to the subject, yields insights into PT usage factors. While most studies used questionnaires, there is potential for qualitative approaches. This approach aids in comprehending underlying influences by delving into the motivations behind mode shifts. For example, Mogaji and Nguyen (2021) investigated disabled commuters' satisfaction with PT using semi-structured interviews. Mystery shopping (MS) can also be employed for data collection. While MS in PT is known and implemented in practice, its exploration in academic literature remains limited, with only a few papers currently available (Voß et al. 2020). de Ona and de Ona (2015) and Voß et al. (2020) argued that using questionnaires in academic settings frequently lacks in-depth justification.

Finally, an area that may be explored is Mobility-as-a-Service (MaaS), which is gaining attention in developed countries and emphasising car access over ownership. Integrating bus transit within MaaS initiatives holds significant potential for enhancing urban mobility by providing users with seamless, multimodal transportation options. Several key factors influence the successful integration of bus transit into MaaS ecosystems, each playing a crucial role in shaping the accessibility, efficiency, and sustainability of PT networks.

Technology integration: One of the fundamental aspects of integrating bus transit into MaaS platforms is the seamless integration of technology. Users can conveniently access, plan, and pay for their bus journeys by leveraging mobile applications, contactless payment systems, and real-time tracking (Lyons et al. 2020). This technological integration enhances the overall user experience and encourages the adoption of bus transit as part of MaaS solutions.

Interoperability: Another critical factor is interoperability, which ensures the smooth integration of various transport modes within MaaS platforms. By enabling users to seamlessly switch between buses, trains, ridesharing, and micro-mobility services, interoperability enhances the flexibility and convenience of multimodal travel (Bushell et al. 2022). This interoperability requires collaboration between different transport providers and the development of common standards for data exchange and service integration.

Data sharing: Collaboration and data sharing among stakeholders are essential for optimizing bus transit services within MaaS ecosystems. By sharing data on routes, schedules, fares, and occupancy levels, transit agencies and technology providers can improve the efficiency and reliability of bus services (Chen and Acheampong 2023). Therefore, service integration is somewhat linked to the integration of data, which is indispensable for enabling service providers to make informed operational decisions (Kamargianni and Goulding 2018). This data-driven approach enables better planning, real-time monitoring, and predictive analytics, leading to more responsive and customer-centric transit operations.

Physical integration: This involves the physical infrastructure and connectivity between different modes of transport within a MaaS system, such as buses, trains, bikes, and ride-sharing services. Physical integration ensures seamless transfers and convenient access for passengers, enhancing the overall efficiency and attractiveness of the transit network (Saliara 2014).

Fare integration: It involves integrating fare payment systems across various modes of transport within a MaaS ecosystem. This allows passengers to use a single ticket or payment method to access different transit modes, including buses, trains, and ride-sharing services (Audouin 2019). Fare integration simplifies the passenger payment process, reduces transactional barriers, and encourages multimodal usage, ultimately enhancing public transport systems' overall efficiency and user experience.

Policy and regulation: Supportive policies and regulations play a crucial role in enabling the integration of bus transit into MaaS initiatives. Policies that promote open data standards, fare integration schemes, and public–private partnerships create an enabling environment for innovation and collaboration. Clear regulatory frameworks and support from central governments help address legal and governance issues and facilitate the implementation of MaaS solutions (Smith et al. 2018).

Partnerships and collaboration: Collaboration among public and private sector stakeholders is essential for driving innovation and fostering collective action towards integrating bus transit with MaaS initiatives. Partnerships between transit agencies, technology providers, urban planners, and community organisations enable the exchange of information, the consolidation of resources, and collaborative decision making (Merkert et al. 2020). Collaborative efforts contribute to developing integrated, user-centric MaaS solutions that address the diverse needs of urban populations.

4.3 Policy implications

The review offers a valuable understanding of the contemporary knowledge regarding the determinants influencing PT usage and the approach used to research them in developing countries. In this concluding section, we recommend enhancing the commuting experience and promoting greater utilisation of PT.

Identifying distinct SQ and satisfaction concerns while investigating ways to curb the rise in private vehicle usage can raise awareness among policymakers, transport operators, and transport authorities. Emphasising the alignment of SQ improvements with passengers' needs and expectations is crucial. This user-centric approach can foster more inclusive and customer-focused PT systems, as commuting experiences affect travellers' subjective well-being (De Vos et al. 2013). There are several instances where SQ improvements have an impact on ridership (with experiences also related to developed countries). For example, enhancements in comfort-related features of the Chicago transit system led to a notable 5% rise in the number of riders, equivalent to 15 million additional trips per year over 5 years (Foote 2004), enhancing the speed and frequency of buses in Ireland led to a decrease in car usage, reducing it from 34 to 22% (Redman et al. 2013), the introduction of an Integrated Ticketing System in cities such as London, New York, Haifa, Washington, Madrid, Chicago, various locations in Italy, San Francisco, Quebec, Montreal, and Seoul resulted in an increase in weekday patronage for both subway and bus services (Suman and Bolia 2019). SQ improvements are needed, and discourse on this subject requires attention.

In numerous developing countries, entrepreneurs primarily lead transport services, receiving limited governmental backing (Yeboah and Asibey 2019). Investments in transport infrastructure are necessary to curtail growing private car use for commuting. PT authorities should allocate resources towards sustainable modes of transport. For example, substantial investments in Australia's bus systems in recent decades have resulted in significant increases in ridership, primarily attributed to extensive service reorganisation and frequency improvements. Among these investments, BRT systems have demonstrated notable impacts at the corridor level, contributing to a 40% increase in new passengers from car drivers, a 17% increase from car passengers, and a 27% growth in new trips (Currie and Wallis 2008).