Analysis of Modal Split of Intra-urban Trips in a Centenary City: A Case Study of Enugu,Nigeria

Abstract

The mode choice of commuters is an important element in the formulation and design of transportation policy. It gives policymakers a better understanding of the dynamics of intra urban trips. However, most of the studies on the modal split of intra urban trips found in the literature focused on Europe, America, and Southeast Asia with little emphasis on Africa. The lacuna this trend creates motivates this study to present evidence from Nigeria using the centenary city of Enugu as a case study. This study is critical for knowledge production that drives explicit outline strategies for the development of a sustainable urban transportation system in Nigeria. Data on modal split and determinant factors were collected from primary source through the use of a questionnaire. Discriminant analysis was used to classify the modal split of intra urban trips in Enugu into two groups’ namely public and private transport. Although public transport modes are the most used for intra urban trips, the mode share indicates that private transport mode (car) is the most used single mode for intra urban trips. More than three-quarters of commuters that use private transport for intra urban trips prefer the mode, whereas a little less than one-fifth of commuters will switch to public transport modes. On the other hand, most of the commuters that use public transport for intra urban trips prefer the modes and very few others will switch to private transport. Car ownership has the strongest influence on the modal split. The study identifies the need for policy interventions that reflects better understating of the dynamics of intra urban trips.

Keywords

modal split intra urban trips public transport private transport discriminant analysis

Introduction

Intra urban trips contribute significantly to promoting the economic growth and prosperity of cities. It is responsible for the exchange of persons, goods and services that takes place between various zones of the city. The mode choice of commuters is a very prominent element in the formulation and design of transportation policy in particular and transportation planning in general. It gives policymakers a better understanding of the dynamics of intra urban trips, which is germane for creating an efficient and effective urban transport system. Most recently, several municipalities have aimed to develop a balanced modal split and sustainable transport modes. Some of the cities have specifically aimed to achieve 30% commuters’ usage of each of the sustainable transport modes (public transport and active transport) for intra urban trips. In particular, 36 cities in advanced economies signed in the Charter of Brussels to accomplish by 2020 a minimum of 15% bicycle preference for intra urban trips, (Mueller, 2018). This goal indicates the policymakers’ aspiration for a modal shift toward sustainable modes. This is probably because of a consensus that private transport does not enhance a sustainable urban transport system (Santos et al., 2013).

A modal split is the proportion of commuters that uses a specific mode type for intra urban trips (car, bus, taxi, cycling, walking, etc). Complex factors such as income, age, car ownership, household structure, gender, among others influence journey makers in choosing from available intra urban trip modes, (Essam & Sadi, 2013). The prediction of the modal split or share of intra urban trips draws from the complex factors. The outcome of such predictions has been applied extensively especially in some advanced economies in transport planning and policy-making (Ahanchian et al., 2019).

For instance, the outcome of a study (Halldorsdottir et al., 2011) in Copenhagen, Denmark, which shows that private transport is the most preferred mode for intra-urban trips accounts for the policy changes that make cycling more attractive to the commuters. Similar studies in Croatia and Slovakia (Drliciak et al., 2020; Novačko et al., 2020), which also indicate the dominance of private transport for intra urban trips, prompts the policy measures that facilitate a modal shift toward public transport. In addition, some studies (Chee & Fernandez, 2013; Loganayegan & Umadevi, 2014) in Southeast Asia accounts for policy changes that promotes public transportation in Malaysia and India. However, in Nigeria, the application of the outcome of the few existing studies, which suggests commuters preference for public transport modes due to transport captivity, on transportation planning and policy-making is not significant despite that the “transportation system is characterized by biased development which favor road transport system over rail transit” (Olatoye & Oyelana, 2019, p. 108).

Most of the plethora of studies on the modal split of intra urban trips found in the literature focused on Europe, America and Southeast Asia with little emphasis on Africa. These studies (such as Bucsky, 2020; Chen et al., 2015; Drliciak et al., 2020; Guerra & Li, 2021; Novačko et al., 2020; Li et al., 2021; Mendiola & Gonzalez, 2021; Shukla, 2019) explains the modal split of intra urban trips and its causal factors. The outcome of most of these studies show private transport as a popular mode used for intra urban trips in their respective cities.

There exists a dearth of studies on the modal split of intra urban trips in Nigeria and other African countries (Olatoye & Oyelana, 2019). The few available studies on Kenya, Ghana, Nigeria, South Africa and other African countries, (Acheampong, 2020; Makajuma & de Langen, 2010; Olatoye & Oyelana, 2019) focus mainly on one aspect of commuters’ trip purpose (either work trips or educational trips). The findings of the studies indicate that public transport modes dominate intra urban trips. However, their outcomes may not be able to produce pragmatic and holistic policy implications that are germane for developing sustainable urban transport systems. The contrary is the case in more developed economies which focus on all commuters’ trip purposes instead of one aspect of urban trip function as found in Africa. This observation motivates the study, which strives to present evidence on the modal split of two-way intra urban trips in Nigeria using the centenary city of Enugu as a case study. The significance of the study drives policies, strategies, and programs for developing a sustainable urban transportation system in Nigeria.

Literature Review

Several studies applied modal split models to evaluate commuters transport behavior regarding choice of mode for intra urban trips. Quite a number of the studies used discrete choice models (multinomial models) to estimate the modal split. In this instance, Chen et al. (2015) used multinomial logit model to examine modal choice of migrant workers in Xian, China. The study found that walking is the most preferred mode for intra urban trips by migrant workers. It follows in descending order by bus, subway and taxi. In addition, age, education level and monthly income have significant effect on the mode choice of the migrant workers. The findings of this study “could be used to support traffic planning and policy strategy which could help the cities with large number of migrant workers” (p. 8).

Similarly, Shukla (2019) used multinomial logit model to analyze the mode choice in intra-urban trips in Ahmedabad, India. The study found that the city’s mode share consist of walking (37%), two-wheeler (26%), bicycle (9%), bus rapid transit (11%), auto-rickshaw (6%), car (4%), Municipal Transport Service (1%), and others (6%). This suggests the dominance of active transport over other intra-urban trip modes. The utility function indicates that as income increases, the use of two-wheeler and car use increases in comparison to public transport bus. Besides, increase in trip length results to decrease in utility of walking trips and bicycle trips. The study avers that the probability of public transportation use is decreasing by increase in income and travel cost.

Meanwhile, the work of Obaid and Hamad (2020) which used multinomial logit model to analyze the mode choice of travelers at Shanjali University City [SUC], United Arab Emirates disagrees with Chen et al. (2015) and Shukla (2019). This is because, unlike the aforementioned studies, which suggest the preference of active transport, the result indicates that car (passive transport) is the most preferred mode for the University City commuters. This follows in descending order by private bus, active transport, public bus and taxi. Besides, the studies found that travel time, travel distance, trip makers’ characteristics and other contributing factors are the main determinants of the mode choice at SUC. The sensitivity analysis indicates that the probability of choosing active transportation decline with increase of distance and duration. Contrastingly, the probability of using the car increases with the increase of distance. The study recommends that “appropriate policy should be formulated to encourage sustainable transportation mode and to reduce daily congestion” (p. 8).

In the same vein, Rasaizadi and Askari (2020) applied multinomial logit model to examine the effect of family structure on modal split in Qazvin-Iran. The study found that individual large families are more likely to choose a bus for their trips. Families in which the age of the eldest child is <6, between 6 and 12 years, and between 12 and 18 years will less likely choose a bus, taxi and other modes respectively. However, other individual families tend to choose motorcycles, whereas individuals of 6 to 30 years old have more tendencies toward using a bus. Besides, age, gender, driving license, trip purpose, family structure, and income significantly influence the modal choice. The study opines that “the use of life cycle variables is useful for studying the effect of family structure on modeling of transportation decisions” (p. 7).

Also, Novačko et al. (2020) applied multinomial logit model to estimate the modal split parameters of intra urban trips in Slavonski Brod, Croatia. They found that the modal share consists of the car (58.8%), walking (26.9%), bicycle (10.9%), and public transport (3.4%). This implies that the car is the most preferred mode for intra urban trips, whereas public transport is the least preferred mode. Besides, the choice of mode depends significantly on the availability of transport modes, cost of transportation and trip purpose. The outcome of this study presents useful information for determining the utility function of each transport mode.

Drliciak et al. (2020) application of multinomial logit model in estimating the modal split of intra urban trips in the city of Zilina Slovakia, which indicate the dominance of car mode, is in conformity with the findings of Novačko et al. (2020). The results show that the modal share consists of car driver (58%), passenger, car (12%), regional bus (23%), bus (1%), bicycle (2%), walking (1%), and train (7%). Also, commuters choice of transport mode depends on safety, travel time and travel cost. Moreover, a 15% increase in modal shift from car to non-motorized vehicles or bus will cause a 50% decrease in emission per work trip. The same effect has a 4.3% shift in modal share from motorcycle to non-motorized vehicles or bus. The outcome of this study prompts the need to promote measures that will facilitate a modal shift toward public transport.

Guerra and Li (2021) used multinomial logit regression to examine the effect of urban form on the mode choice in the United States and Mexican cities. The mean mode share for the US cities shows the dominance of private vehicle (89.9%), followed by transit (7.1%) and non-motorized (3.1%). Contrastingly, the mean mode share for Mexican cities indicates the preponderance of transit (49.3%). This is followed by private transport (27.7%) and non-motorized (22.9%). Besides, the study found that higher income, higher educational attainment and smaller household size tend to highly influence the odds of commuting by private vehicle and lower the probability of commuting by transit or active transport modes in the United States and Mexico’s largest urban areas. The relationship between income and choice of transit, instead of private vehicle, is not significant in the US, whereas the probability of choosing private vehicle shows similar pattern with age in both countries. In addition, men are more likely than women to commute by active modes in the US, whereas it is the opposite in Mexico. However, women appear to have less opportunity to commute by private vehicle than men in Mexico. The study avers that “land use and transportation policies likely have a more substantial role in shaping commute pattern in Mexico than the US” (p. 1).

Moreover, Hu et al. (2022) used nested logit model, a type of discrete choice models, to estimate commuter mode choice in dual-earner household in a small Chinese city. The outcome of this study is not completely consistent with the output from multinomial logit models (Drliciak et al., 2020; Novačko et al., 2020; Obaid & Hamad, 2020; Rasaizadi & Askari, 2020). The study found that the mode choice for female commuters consist of e-bike (63.4%), car (23.2%), walk (4.3%), shuttle bus (3.3%), transit bus (2.9%), motorcycle (1.5%), and bike (1.4%). For male, the most share consist of car (62.4%), E-bike(22.4%),motorcycle(4.8%), shuttle bus (3.8%), foot (3.7%), transit bus (1.7%), and bike (1.5%). This implies that male commuters have higher priority regarding car use than female whereas female commuters show higher propensity to use E-bike than male. In addition, active transport mode and e-bike are popular among older people and low-income households, whereas spouses with high education level will likely use car mode. The outcome of the study suggests that built environment policies aimed at reducing car traffic may need to differentiate between large and small Chinese cities.

In the same vein, Al-Salh and Esztergar-Kiss (2021) applied a combination of multinomial logit model and logit model to estimate the modal split in Budapest, Hungary. The study found that private transport is the most dominant choice for intra-urban trips, and is followed by public transport and walking. In addition, age shows a negative impact for public transport but a positive coefficient for walk and car. Conversely, gender shows a negative effect on walking and traveling by car, whereas it posits a positive impact on public transportation. The findings of this study, “could lead to the development of a micro-simulation-based prototype of mode demand model” (p.18).

Some other studies applied non-discrete choice models to estimate the modal split. To this effect, Maheshwari et al. (2020) used pinch analysis to optimize modal split and its relationship with carbon dioxide emission in Mumbai, India. The study found that the modal split consists of bus (53.8%) railways (16.4%) 2-wheeler (10.5%) cycling (9.4%), walking (7.4%), and 3-wheeler (2.5%). This implies that public motorized transport applies mostly for intra urban trips. However, pollution emission in grams of CO₂ per kilometer by each mode indicates that motorized transport was higher than non-motorized transport, which shows zero emission. The outcome of the study confirms the existence of a strong relationship between modal shift and traffic emission, which are beneficial to future emission reduction strategy.

In another vein, Mendiola and Gonzalez (2021) used spatial econometric model to examine the relationship between urban development characteristics and modal split of intra urban trips in Buenos Aires metropolitan area, Argentina. The modal split indicates that public transport modes apply mostly for intra urban trips, accounting for 45.6% of trips. This follows in descending order by private vehicles and non-motorized modes, which accounts for 35.8% and 18.6% of total trips respectively. In addition, the study found that population density and street intersection density positively influence commuters’ preference for public transport, whereas the modal choice of non-motorized trips depend positively on commuters’ self-containment and street density. For private transport, the modal choice relates significantly with land use characteristics. The implication of this study is that urban development variables are in general highly significant in explaining the modal split of intra-urban trips.

Duleba et al. (2021) introduced a new approach to modal split estimation using the Best-Worst approach in Budapest, Hungary. The study found that public transport is more popular among all groups of commuters, while short- and mid-distance commuters use a car. Car-pooling mobility has the lowest modal share followed by bicycle, walking and home office. In addition, it found that the factors of distance traveled, transport-cost, availability of mode, health effect and travel time influence the modal split of intra urban trips in the city. The study concludes that the result of the Best-Worst approach is competitive in accuracy with mainstream methodologies.

Bucsky (2020) used qualitative approach to examine mode share change due to COVID-19 in the city of Budapest, Hungary. He found that the COVID-19 pandemic has a considerable effect on mode share changes for intra-urban trips. There is a decline in the use of public transport for intra urban trips from 43% to 18% whereas cycling mode share grew from 2% to 4%. However, there is unprecedented growth in the car mode share from 43% to 68%. This is an indication that there is an increase in mode share for private transportation and a reduction in public transport due to the COVID-19 pandemic.

The two main model specifications for modal split found in studies from Africa are multinomial logit models (MNL) and multiple linear regression. Consequently, Salon and Aligula (2012) used multinomial logit model to examine the travel patterns and preferences of the residents of Nairobi, Kenya. They found that 85% and 15% of the residents prefer public transport and car for intra urban trips respectively. This implies that public transport is the dominant mode for intra urban trips among the commuters. Besides, the residents’ trips preferences are influence by infrastructure, income, car ownership, employment and geographic coverage. In addition, the study aver that the residents of Nairobi will choose safer, more comfortable and reliable modes of transport as they become increasingly wealthier. There is the need for the formulation of policies that promotes active and public transport system, and discourages car ownership in Kenya.

Agyemang (2017) used multinomial logit regression to estimate modal choice for long distance trips in Greater Accra metropolitan area, Ghana. The study found that there is a high use of motorized transport (98.5%) compared to non-motorized transport mode which indicate a paltry 1.5%. Moreover, the mode share indicates that bus/trotro is the most preferred mode (50%) for all trips in the city irrespective of the trip purpose. Besides, shared taxis and cars are mostly used transport mode by commuters in the informal and formal sectors respectively. However, age, education, and ownership were significant factors that influenced the modal choice of bus/trotros. Among taxi users, education and ownership of personal car were the two main socio-economic factors that influenced the choice of mode.

However, Vilgoen (2020) used multinomial logit model to investigate the impact of activity behavior on transport mode choice and use in Cape Town, South Africa. The study found the dominance of non-motorized transport (31%) over the city’s intra urban trip modes. This follows in descending order by private car (22%), minibus taxi (17%), traditional public transport [bus and train] (13%), and car taxi (12%). The factors of income, car access, type of activity and distance traveled significantly influence the mode choice and use. Moreover, commuters within the lowest mean income group make use of minibus-taxis, train and walking as trip modes, whereas the high income group prefer the use of private transport. The findings of this study “can help improve the accuracy and behavioral realism of transport decision modeling of the modal choice in the City of Cape Town” (p. 90).

In the same vein, Acheampong (2020) used multinomial logit model to examine the relationship between spatial pattern, intra-urban commuting patterns and mode choice in Kumasi metropolis, Ghana. The study found that mode choice for car increases with increase in income. The high-income household odds of commuting to work in a private car are 29 times higher than low income household. Regarding middle-income household, the odds are relatively smaller, although nearly eight times higher than low-income household. Moreover, workers from low-income households have odds of walking to work five times higher than workers from high-income households. The outcome of this study presents policy implications for sustainable urban transportation and spatial developments outcomes.

Distinctively, Tembe and Matusse (2020) used binomial logistic regression to estimate the modal split in Maputo, Mozambique. The study found, contrary to Salon and Aligula (2012), Agyemang (2017), and Acheampong (2020), that the mode share of urban trips is dominated by walking (45.8%), followed by paratransit chapas (32.9%), private cars (10.2%), buses (9.2%), train (0.6%), and others (1.3%). Income and gender are the two main factors that influence the modal choice. Consequently, the positive relationship between income and choice of bus indicate that as income increases, commuters will likely use buses over paratransit chepas. On the other hand, a negative relationship between choice of buses and gender tend to suggests that women are less likely to use buses for their daily commute. In addition, it is expected that significant proportion of commuters will continue to use walk mode for urban trips due to rapid urbanization. The study predicts that poor quality public transport could result in modal shift from collective to private transport resulting in negative environmental impact.

Olatoye and Oyelana (2019) used multiple linear regression to examine the modal choice of work trips in Lagos, Nigeria. The study found that 32% and 68% of the residents use the private car and public transport for work trips respectively. In addition, transport fare, journey time, driver's skill, safety, comfort and traffic have a significant influence on the residents’ mode choice for work trips. The study recommends the development of water transportation to ease traffic congestion in the city. Another study conducted in Nigeria by Fasina et al. (2020) using multiple linear regression technique to analyze the mode choice of intra-city travel in Lagos shows a consistent pattern with the work of Olatoye and Oyelana (2019). The study found that more than one quarter (26.1%) of commuters use mini bus while about half of the commuters (45%) use private cab(ride-hailing). Also, less than one-eight (11.7%) prefer public bus, whereas 6.1%, 8.3%, and 2.8% of commuters use federal urban mass transit bus, taxi and tricycle for intra-city travel respectively. This is an indication that public transport modes account for the most intra-city travels among the commuters. Besides, transits time, delay, duration, safety, vehicle condition, transit fare and ease of accessibility influence commuter’s choice of intra-city mode. The study opines that there is need for the government to take a concrete step toward integrating the entire public transport mode for better coordination and organization.

It is apparent that most of the available studies in Nigeria used multiple linear regression ( an aggregate model) to estimate the modal split despite the weakness of the model which hinges on its inability to accept disaggregate data and ordinal or nominal discrete variables in the analysis (Ukonze et al., 2020). Consequently, the model does not provide disaggregate information on modal choice at micro (household) level. This weakness makes the use of multiple linear regression least appropriate for estimating the modal split of intra urban trips. This indicates the need for application of a robust model in evaluation of modal split in Nigeria. This challenge provides the ground for this study.

Moreover, discriminant analysis and multinomial models are widely accepted in estimating the modal split, (Okoko, 1998; Ramayah et al., 2010). Nevertheless, evidence from literature review suggests the predominance of multinomial logit models (MNL). None of authors showed evidence on the robustness of discriminant analysis in estimating the modal split. This is in spite of strength of the model which lies in its behavioral structure and enhanced ability in identifying a causal relationship (Ukonze et al., 2020). It is on these bases that this study used discriminant analysis to estimate the modal split of all intra urban trip functions in the centenary city of Enugu, Nigeria. The outcome of this study provides evidence on the utility of discriminant analysis in estimating the modal split. In addition, it will extend the existing literature on modal split models.

Modal Split Theory

Modal split model originated in the United States of America in the 1950s as a response to transportations problems, such as increase in traffic demand and congestion, confronting many American cities. This motivated researches and thinking about transportation as a science, consequently leading to formulation of modal split theory (Ungvarai, 2019). The modal split, also known as modal choice or modal share, is widely used in transportation engineering to evaluate transportation behavior. The modal split theory shows “the percentage of travellers using a particular mode of transport compared to the ratio of all trips mode” (Ungvarai, 2019, p. 5). This involves “calculating the amount of trips between an origin and destination which is broken up into parts belonging to different transport modes” (Ortuzar & Willumsen, 2011). In addition, it estimates how future trips will split between various modes of transport available to an intending journey maker. The outcome explains trip makers’ behavior when they select a travel mode, assuming that various modes of urban transportation are open to the journey maker. Travel modes include bus, car, taxi, bicycle, walking, tricycle, train. Several factors, which may influence the choice for a particular mode include trip type, trip purpose, level of service, cost of transportation, age, sex, income, distance, comfort, availability among others (Okoko, 1998).

The modal split has both advantages and disadvantages especially for transportation analysis. At theoretical level, it has the advantage of being the most developed among the four modeling steps in travel demand. This is because of extensive researches on travel mode choice (Ortuzar & Willumsen, 2011). However, the disadvantage of model especially in transportation analysis lies on the weakness of the model by offering inadequate information on the travel mode of commuters (Ungvarai, 2019). Consequently, the model is solely theoretical and depends on the validity of the existing theory and intuitiveness of guesses (Ortuzar & Willumsen, 2011). The assumption behind the modal split decision is for a two-way trip.

Modal split models are stratified diversion curve models and probabilistic models such as the discriminant analysis and logit analysis models. The diversion curve model is difficult to calibrate when more than two competing travel modes are involved in the analysis. Consequently, the probabilistic models (discriminant analysis) and multinomial models are widely used in analyzing modal split (Okoko, 1998; Ramayah et al., 2010).

The multinomial logit models are types of discrete choice models which models individual choice responses of a trip maker as a function of the characteristics of alternatives available to commuters as well as the socioeconomic attributes of each individual (Khan, 2007). This is a disaggregate approach which recognizes that aggregate behavior is the result of numerous individual decisions and therefore models individual choice responses as a function of the characteristics of the alternative modes available to them as well as the socioeconomic attributes of each individual. The strength of models lies in the ability to predict the consequences of transportation policy measures that affect the choice mode (Ortuzar & Willumsen, 2011). The utility theory of the model indicates that mode choice is usually an indicator of value to an individual. In other words, a decision-maker chooses a single alternative from a set of infinite number of mutually exclusive alternatives where choice set is exhaustive. Consequently, the individual is seen to have selected a mode which maximizes his utility based on the consideration of some factors such as travel time, waiting time, access time transport fares, among others (Khan, 2007).

On the other hand, the discriminant analysis is a multivariate technique that classifies groups of observations into two or more groups based on the K variables measured on each experimental unit (Canizo et al., 2019). It works by finding the linear combinations of the K variables called the discriminant functions. The number of discriminant functions is equal to that of the number of classes minus one. Separations between classes are hyperplanes and the allocation of a given object within one of the classes depends on a maximum likelihood discriminant rule (Abbas et al., 2019). The discriminant function exploits the differentiation between groups. The discriminant coefficient assists in recognizing which variables have more contribution to differentiate about the corresponding dimension. The result of discriminant analysis is a model that predicts group memberships, which enables the transport planners to understand the relationship between the set of selected variables and the observations (Li et al., 2021). The output of discriminant analysis in this study, using Statistical Package for Social Sciences analytical software, yields eigenvalues, wilks lambdas, standardized coefficient (beta), and structure matrix. Consequently, discriminant analysis was used in this study because it enables the inclusion of socioeconomic factors and alternative modes in the analysis by determining which predictor variables are related to the dependent variable, as well as predicting the value of the dependent variable based on the values of the predictor variables. This enables the study to overcome type 1 error.

The Case Study City

Enugu is an inorganic centenary city found in 1909 because of the accidental discovery of coal by a British Mining Engineer Mr. Kitson in the Udi and Okoga areas of the city (Okeke & Ukonze, 2019). Consequently, the city assumed the codename the “Coal City.” As the population grew, the city expanded leading to the development of several layouts. Currently, the city consists of 32 constituent neighborhoods covering an area of 100 square meters (see Figure 1) and a population of over one million people. The city has been an administrative headquarters in Nigeria since 1929, (Jiburum et al., 2021).

Figure 1.

Map of Enugu centenary city.

The predominant land uses in Enugu city are residential, commercial, industrial, institutional, public, and transportation land uses. The public service employs some persons in the active population whereas self-employed, skilled and unskilled manual workers, professionals, and blue-collar workers dominate the private sector.

The trip origin and destination patterns in the city exhibit a dichotomy in distribution among the major trip generation and attractor activities. Residential activities are found predominantly in eastern, southern and northern parts of the city, whereas government, educational and commercial activities occupy the western area of the city. The Central Business District is bound to the west and south by residential layouts. The city portrays a high level of polarization and separation of commercial and institutional land uses, which indicate an explicit and easily identifiable steady journey-to-work trip in the morning and journey-to-home trip in the afternoon. Traffic volume in Enugu has risen approximately 43% over the past three decades with an annual increase of 14%. Private cars constitute over 50% of the vehicle population in Enugu whereas buses, trucks and lorries account for approximately 13.8% of the vehicles. Tricycles make up about 37% of the traffic stream in the city (Clement, 2021).

Methodology

Data used for this research derives from questionnaires administered on household population of Enugu city. Households were the respondents because they use various modes of transportation for intra urban trips. The sample size formula of Mugenda and Mugenda (2009) was employed to determine the sample size of 470 households from the sample frame of 105,762 household population of the nine selected neighborhoods used as the sample for this study. The formula applies because the sample frame is greater than 10,000.

Stratified sampling technique was applied to produce a subpopulation whereby the units of each subpopulation represent each of the three heterogeneous groups (low income, medium income and high come groups) that make up the study population, thus giving each group an equal chance of selection in the sample. Drawing up the sampled 470 household heads (respondents) starts with dividing the 32 constituent neighborhoods of the city into three strata, namely, high density (10 neighborhoods), medium density (13 neighborhoods) and low residential density (9 neighborhoods). Then the selection of three neighborhoods each at random from each of the three strata, resulting in nine neighborhoods selected for studies.

The three selected neighborhoods from low density are Abakpa, Gariki, and, Ogbete, with a respective sample size of 145, 33, and 42 respondents. For medium density, the selected neighborhoods are Uwani, New Haven, and Achara layout with a sample size of, 52, 28, and 80 respondents respectively. For low-density, the selected neighborhoods are Trans Ekulu, Independence layout, and Government Reservation Area. They have a sample size of 19, 38, and 33 respondents respectively. The sample size of each selected neighborhood is proportional to its household population.

Out of 470 copies of questionnaires distributed through home survey, 450 copies representing 95.1% were returned. The questionnaire collected primary data on commuters’ socio-economic characteristics and their modal choice for intra-urban transport in Enugu. The reliability of the instrument test using Cronbach’s Alpha formula yielded a reliability score of 74%, which is adequate for this study. Discriminant analysis was used to classify the modal split of intra-urban trips. The dependent variable is the commuters’ modal choice for intra urban trips and the independent variables are socio-economic characteristics of the commuters. The mathematical expression of discriminant analysis is:

D = v_{1} x_{1} + v_{2} x_{2} + v_{3} x_{3} + \dots + v_{i} x_{i} + a

(1)

Where:

D = discriminate function

V = the discriminant coefficient or weight for that variable

X = respondent’s score for that variable

a = a constant

i = the number of predictor variables

Data Presentation and Analysis

Analysis of Modal Split

Table 1 shows that 45.1% of commuters use private cars for all types of intra-urban trips whereas 44.7%, 7.8%, 2.4%, and 0.7% use buses, tricycles, taxis, and Opay rides respectively. This indicates that public transport modes (buses, tricycles, and taxis) are the most used by commuters (cumulatively 54.9%) for all types of intra urban trips. The bus mode is the most used public transport mode for intra urban trips. Despite this, the mode share of private cars is slightly higher than that of the bus for intra urban, thus indicating the potential for automobile dependency in the city.

Table 1.

Share of Transport Modes for Intra-urban Trips in the Centenary City of Enugu.

Options	Overall intra-urban trips	Educational trips	Religious trips	Recreational trips	Home-to-work trips	Work-to-home trips
Options	Frequency (%)	Frequency (%)	Frequency (%)	Frequency (%)	Frequency (%)	Frequency (%)
Private car	45.1	35.6	48.4	48.9	31.1	33.8
Taxi (Minibus taxi)	2.4	0.2	6.9	4.7	2.4	2.0
Tricycle (Minibus tricycle)	7.8	4.2	25.9	9.3	3.6	2.4
Opay ride (Ride-hailing tricycle	0.7	0.7	2.4	23.8	1.3	2.7
Bus	44.7	59.3	16.4	13.3	61.6	59.1
Total	100	100	100	100	100	100

Source. Field survey.

A breakdown of the modal split for each of the types of intra urban trips in the city shows that the majority of commuters (59.3%) use buses for educational trips. Others in descending are private cars (35.6%) tricycles (4.2%), Opay ride (0.7%) taxis (0.2%). This indicates the dominance of buses and private cars in intra urban school trips. The modal split for religious trips indicates that the majority of commuters (48.4%) use private cars for the trip whereas 25.9%, 16.4%. 6.9%, and 2.4% use tricycles, buses, taxis, and Opay ride respectively. This suggests that private cars were the most preferred mode for faith-based trips in the city. For recreational trips, the majority of commuters (48.9%) use private cars. Others in descending order are Opay ride (23.8%), bus (13.3%), tricycle (9.1%), and taxi (4.7%) This also indicates that the private car was the most preferred mode for recreational trips.

The home-to-work trips show that the bus is the most preferred mode. The majority of the commuters (61.6%) use bus mode whereas, in descending order, 31.1%, 4.9%, 2.4%, and 1.3% of commuters use private cars, tricycles, taxis, and Opay ride respectively. The modal split for work-to-home trips shows a similar pattern to that of home-to-work trips. The majority of commuters (59.1%) use buses for work-to-home trips whereas 33.8%, 2.7%, 2.4%, and 2.0% use private cars, Opay ride, tricycles and taxis respectively. It is apparent from the analysis that buses and cars are the most dominant modes for intra urban trips in the city.

The bus service is commuters’ most preferred mode for educational, home-based-work and work-based-home travels in Enugu. On the other hand, the private car is the most preferred for religious and recreational trips. This finding is consistent with the existing studies (Agyemang, 2017—Ghana; Mendiola & Gonzalez, 2021—Argentinaa; Olatoye & Oyelana, 2019—Nigeria; Rasaizadi & Askari, 2020—Iran; Salon & Aligula, 2012—Kenya) found in developing countries. The afore-mentioned studies indicate that public transport often serves for work, educational and other regular trips, whereas private transport dominates non-essential trips in their respective countries.

Table 2 reveals that users of private transport rate all the performance indices of private transport highly (mean score2 ≥ 3.0). The highest satisfaction is in the area of coverage of service (mean score of 4.9) whereas the least satisfaction is walking distance (mean score of 4.0). This indicates private transport users enjoy a very high level of satisfaction with their intra urban trip mode. Similarly, the users of public transport modes rated some of the performance indices of public transport highly (mean score ≥ 3.0) whereas some performance options were rated low (mean score ≤ 3.0). The highest satisfaction was observed in affordability (mean score of 4.9) and the greatest dissatisfaction was observed in flexibility and noise level (mean score of 2.2 and 1.8 respectively). This is an indication that users of private transport enjoy higher satisfaction levels with the performance of their mode than the public transport users, which exhibited moderate satisfaction levels with their intra urban trips mode.

Table 2.

Commuters Level of Satisfaction With the Performance of Private and Public Transport in the Centenary City of Enugu.

Performance indices	Private transport		Public transport
Performance indices	Mean	Decision	Mean	Decision
Coverage of service	4.9	High	2.7	Low
Reliability	4.6	High	2.9	Low
Safety	4.7	High	3.5	High
Comfortability	4.5	High	3.5	High
Affordability	4.2	High	4.7	High
Speed	4.6	High	2.6	Low
Ease of use	4.5	High	3.4	High
Waiting time	4.6	High	3.2	High
Travel time	4.6	High	3.1	High
Flexibility	4.6	High	1.8	Low
Distance coverable	4.7	High	2.4	Low
Ease of access	4.7	High	3.5	High
Responsiveness to demand	4.6	High	3.4	High
Frequency of service	4.6	High	2.8	Low
Comfort	4.7	High	3.7	High
Walking distance	4.0	High	3.4	High
Regularity of service	4.5	High	3.1	High
Aesthetics	4.2	High	2.9	Low
Waiting area cleanliness	4.6	High	2.5	Low
Availability	4.5	High	3.4	High
Noise level	4.6	High	2.2	Low

Source. Field survey.

The Results of Discriminant Analysis of the Modal Split in the Centenary City of Enugu

The discriminant analysis results in Table 3 classify the modes used for intra-urban trips in Enugu city into two groups namely public transport (54.9%) and private transport (45.1%) modes. The original count shows that 54.9% of commuters use public transport modes for intra urban trips whereas 45.1% use private transport for the same trips in the city. The cross-count validated predicts group membership of the two modes at 90.4% accuracy. This result indicates that public transport (bus, taxi, and tricycle) and private transport (car and opay ride) are the two groups of modes used for intra urban trips in the centenary city of Enugu.

Table 3.

The Modal Split Classification for Intra Urban Trips in the Centenary City of Enugu.

	Mode for intra-urban trips	Predicted group membership		Total
	Mode for intra-urban trips	Private	Public	Total
Original
Count	Private	150	32	182
Count	Public	11	257	268
%	Private	82.4	17.6	100.0
%	Public	4.1	95.9	100.0
Cross-validated
Count	Private	150	32	182
Count	Public	11	257	268
%	Private	82.4	17.6	100.0
%	Public	4.1	95.9	100.0

Source. Discriminant analysis output from Statistical Package for Social Sciences.

Table 4 reveals that one significant discriminant function variate for intra urban trips in Enugu city. It has an eigenvalue of 2.59 and explains 100% of variance that exists in the choice of modes for intra urban trips. This implies that there is an existence of differentiation between public and private transport modes in the model. The canonical correlation suggests that the existence of a significant relationship between commuters’ modal choice for intra urban trips and socioeconomic characteristics of commuters. The wilks lambda result in Table 5 shows that the group difference exhibited by the discriminant function is significant. Consequently, the derived discriminant function is valid.

Table 4.

The Discriminant Function of Modal Split for Intra Urban Trips in the Centenary City of Enugu.

Eigenvalue	% of variance	Cumulative %	Canonical correlation
2.597	100.0	100.0	.850

Source. Discriminant analysis output from Statistical Package for Social Sciences.

Table 5.

Wilks’ Lambda Results of Modal Split for Intra Urban Trips in the Centenary City of Enugu.

Test of function(s)	Wilks’ Lambda	Chi-square	Df	Sig.
1	.278	566.403	11	.000

Source. Discriminant analysis output from Statistical Package for Social Sciences.

Table 6 indicates that the discriminant function’ variate 1’ discriminates users of public transport from users of private transport.

Table 6.

Functions at Group Centroids of Modal Split for Intra Urban Trips in the Centenary City of Enugu.

Mode for all kinds of intra-urban trips	Function
Mode for all kinds of intra-urban trips	1
Private	−1.951
Public	1.325

Source. Discriminant analysis output from Statistical Package for Social Sciences.

Table 7 indicates that private car ownership (0.719), residential status [tenure] (0.524) and income (−0.339) have the highest correlation, in descending order, with the discriminant function, whereas the three lowest factors are educational attainment (−0.172), religion (140), and gender (−0.068). However, neighborhood of the commuters (0.008) did not correlate with the discriminant function.

Table 7.

Structure Matrix of Modal Split for Intra Urban Trips in the Centenary City of Enugu.

	Function
Variables	1
Car ownership	0.719
Residential status of the commuters (tenure)	0.524
Income of the commuters	−0.339
Employment status of the commuters	0.335
Marital status of the commuters	−0.291
Age of the commuters	−0.255
Position of the commuters (household head, wife, son daughter and relation)	0.244
Educational attainment of the commuters	−0.172
The religion of the commuters	0.140
Gender of the commuters	−0.068
Neighborhood of the commuters	0.008

Source. Discriminant analysis output from Statistical Package for Social Sciences.

Table 8 shows the relative contribution of each of the predictor variables in explaining the modal share of intra urban trips in Enugu city. The private car ownership (3.324), religion (1.542), and position of the commuter in the household (0.745), in descending order, made the three highest contributions to the mode share of intra urban trips. In addition, neighborhood of the commuters is not a significant factor that influenced the modal choice.

Table 8.

The Standardized Canonical Discriminant Function Coefficients of Modal Split for Intra Urban Trips in the Centenary City of Enugu.

Variables	Function
Variables	1
Neighborhood of the commuters	0.015
Gender of the commuters	−0.252
Age of the commuters	0.603
Marital status of the commuters	0.222
The religion of the commuters	1.542
Position of the commuters	0.745
Residential status of the commuters	0.576
Educational attainment of the commuters	0.479
Employment status of the commuters	0.115
Income of the commuters	−0.216
Car ownership	3.324
(constant)	−14.391

Source. Discriminant analysis output from Statistical Package for Social Sciences.

Consequently, the discriminant function model for the modal split of intra urban trips in Enugu centenary city is:

\begin{matrix} D = 14.319 - . 603 a g - . 222 ms - 1.542 rg - . 745 ps - . 576 res \\ - . 479 edu - . 115 emp + . 216 in - 3.324 car \end{matrix}

(2)

Where:

ag = age of commuter

ms = marital status of commuter

rg = the religion of commuter

ps = position of commuter

res = residential status of commuter

edu = educational attainment of commuter

emp = employment status of commuter

in = income of the commuter

Table 9 shows a significant difference in modal choice between users of private transport and public transport for intra urban trips accounted for by all factors except for neighborhood of the commuters which is not significant.

Table 9.

Tests of Equality of Group Means of Modal Split for Intra Urban Trips in the Centenary City of Enugu.

	Wilks’ Lambda	F	Df1	Df2	Sig.
Neighborhood of the commuters	1.000	0.075	1	448	.785
Gender of the commuters	.988	5.313	1	448	.022
Age of the commuters	.855	75.671	1	448	.000
Marital status of the commuters	.819	98.813	1	448	.000
The religion of the commuters	.951	22.891	1	448	.000
Position of the commuters	.866	69.500	1	448	.000
Residential status of the commuters	.584	318.984	1	448	.000
Educational attainment of the commuters	.929	34.262	1	448	.000
Employment status of the commuters	.775	130.318	1	448	.000
Income of the commuters	.770	133.561	1	448	.000
Car ownership	.427	600.818	1	448	.000

Source. Discriminant analysis output from Statistical Package for Social Sciences.

Discussion

The discriminant analysis classifies the modes used for intra urban trips in Enugu into two modes namely private transport and public transport (see Table 3). The results show that the majority of commuters use public transport (54.9%) than private transport (45.1%) for intra urban trips. Although public transport modes are the most used for intra urban trips, the mode share (see Table 1) indicates that private transport mode (car) is the most used single mode for intra urban trips. This is an indication that the city is tending toward automobile dependency. The choice of public transport for intra-city trips is significantly higher for males, commuters within the 21 to 30 years’ age cohort, singles, heads of households, and tenants. Others are commuters with lower education attainment, self-employed, low-income earners (<N=30,000) and those who do not have private cars.

Conversely, the choice of private transport was highest for female commuters (predominantly wives) within the 41 to 50 years’ age cohort, married couples, and landlords. Others are commuters with higher education attainments, professionals, high-income earners and car owners. The mode choice of the aforementioned group of commuters relates to their income and job demand.

The findings of this study are in affirmation with the work of Agyemang (2017) who used multinomial logit model specification to estimate the modal split in Greater Accra metropolitan area, Ghana. He found that there is a high dependency on public transport for the city’s intra-urban trips. Besides, three socioeconomic factors namely age, education, and car ownership influence the modal split in Accra, whereas 10 of such factors which include age, marital status, the position of the commuter in the household, level of education, employment status, religion, residential status, gender, car ownership, and income explain the situation for the centenary city of Enugu. This result implies that religion as a factor that influences modal split is peculiar to Enugu city. This result is apparent in the dominant nature of private transport (48.4%) in the city’s religious trips (see Table 1). Consequently, it suggests the existence of automobile dependency (car and opay ride) for faith-based trips. The influence of religion on the modal split is likely because private transport modes provide more efficient services than public transport (see Table 2) for faith-based activities. This account for the reason commuters prefer private transport for religious trips to educational and work trips in the city (also see Table 1). Furthermore, this finding is an indication that some socioeconomic factors that influence modal split can be unique to a city. The uniqueness is a pointer that determinants of the modal split can differ among cities.

The cross-count validation (see also Table 3) of the discriminant analysis predicts the group membership preference of modal split (public transport or private transport) for the intra-urban trips with a 90.4% level of accuracy. The prediction shows that 82.4% of commuters that use private transport for intra urban trips prefer the mode, whereas 17.65% will switch to public transport modes for intra urban trips. This implies that 17.65% of private transport users will switch to public transport modes. This indicates that there is a potential of discouraging automobile dependency and its associated problems through an improved public transport system. On the other hand, 95.9% of commuters that use public transport for intra urban trips prefer their modes whereas 4.1% will switch to private transport. This suggests that a small proportion (4.1%) of public transport users will potentially switch to private transport. The implication is that public transport users have a higher preference for their modes of intra urban trips than is the case of private transport users. In addition, commuters’ switching behavior toward private transport is insignificant. This is expected because most commuters are captive to public transport system (see Table 1). Consequently, it is a reflection that public transport has the potential to dominate intra urban trips in the city. Also, it is an indication that public transport holds great potentials in providing a sustainable transport system in the city. However, the outcome of this result differs from the existing studies (Al-Salh & Esztergas-Kiss, 2021; Drliciak et al., 2020; Guerra & Li, 2021; Novačko et al., 2020; Obaid & Hamad, 2020) found the USA,, Europe, and Middle east which indicate that private transport was the most preferred mode for intra urban trips.

The 10 socioeconomic factors that influenced the commuters’ choice for intra urban trips in the city are as follows in descending order car ownership, religion, the position of the commuters in the household, age, residential status, and educational attainment. Others are marital status, gender, income, and employment status of the commuters. The implication of this result concerning each of the factors is that 1% increase in car ownership, religious trips, the position of commuters in the household, age, and residential status of the commuters will result in 3.344%, 1.542%, 0.745%, 0.603%, and 0.576% increase respectively in the modal split in favor of private transport. Also, a 1% change in commuters’ educational attainment, marital status, and employment status will lead to a respective 0.479%, 0.222%, and 0.115% increase in modal shift toward private transport. On the other hand, income and gender have negative influence on Modal split. This outcome is consistent with the work of Tembe and Matusse (2020) who found a negative relationship between gender and modal choice in Maputo, Mozambique. The implication is that a 1% decrease in the income of commuters will cause −0.216% decrease in modal shift from private transport to public transport. In addition, a 1% decrease in the male genders trips will result to −0.252% decrease in modal shift from public transport to private transport, whereas for female gender, it will also cause −0.252% decrease in modal shift from private to public transport. This implies that male and female commuters are less likely to use public transport and private transport respectively. The neighborhood of the commuters does not have a significant effect on the modal split of intra urban trips in Enugu. The implication is that car ownership had the strongest influence on the modal split of intra urban trips in Enugu. It accounts for 46.6% influence in the modal split. This implies that an increase in car ownership will cause a major increase in the modal shift toward private transport. This finding was corroborated by the works of Novačko et al. (2020), Drliciak et al. (2020), and Al-Salh and Esztergas-Kiss (2021) which suggest that car ownership has the strongest effect on the modal split of intra urban trips in Croatia, Slovakia and Hungary respectively. Consequently, this indicates that formulating policy interventions to promote the use of public transport modes in Nigerian cities, without complementary strategies that will discourage private vehicles and use may not yield the required results about developing a sustainable urban transport system.

A comparison of the outcome of this study with existing studies found in centenary and older cities across the globe show diverse results. The results of the study, which show that public transport modes are the most preferred for intra urban trips, differ from existing studies (Al-Salh & Esztergas-Kiss, 2021; Bucsky, 2020; Drilick et al., 2020; Novačko et al., 2020) found in the centenary cities of advance economies, which indicates a preference for private transport(car). Besides, this study agrees with the existing studies (Agyemang, 2017; Olatoye & Oyelana, 2019; Maheshwari et al., 2020; Salon & Aligula, 2012) in the centenary cities of developing economies, which show the preference for public transport modes for intra urban trips.

This is an indication that the modal split of intra urban trips differ among cities that are 100 years old or older across the globe. The implication is that the age of a city may not likely have a consequential effect on the modal split of intra urban trips. Rather, the uniqueness of the respective economies may have accounted for the variation in the modal split of intra urban trips among the cities. This is because commuters in most of the centenary cities in advanced economies show a consistent pattern of preferring private transport for intra urban trips whereas in developing economies commuters prefer public transport modes. This is a confirmation that modal split can differ among cities because of the peculiarity of determinant factors (such as religion in Enugu city). The outcome of this study is competitive in accuracy with the discrete choice models (Al-Salh & Esztergas-Kiss, 2021; Drliciak et al., 2020; Novačko et al., 2020; Obaid & Hamad, 2020; Rasaizadi & Askari, 2020). This supports the discussions in the theory and methodology sections of this study, which indicate that the discriminant analysis (a probabilistic model) is a robust model for estimating the modal split of intra urban trips.

Policy Implications and Recommendations

There is a high dependence on public transport modes (bus, taxi, and tricycle) for intra urban trips, especially for educational, home-based work, and work-based home trips. This is an indication that public transport has great potentials to provide a sustainable urban transport system. Consequently, policymakers should put measures in place that will improve the level of performance of the public transport system. Such measures should give special attention to the areas considered as a weakness in the performance of public transport. These include coverage of service, reliability, speed, flexibility, distance coverable, frequency of service, aesthetics, waiting area cleanliness and noise level (see Table 2).

Moreover, the bus is the most used public transport mode for intra urban trips, especially by students and workers. Therefore, city managers should formulate policies that will promote the urban mass transit bus system taking into consideration public-private partnerships in the provision of bus transit services that will meet the needs of commuters. The promotion of urban mass transit bus system will likely impact positively on its level of performance which may increase further the preference for public transport modes for intra urban trips, thus resulting in more shifts from private transport to public transport. This will discourage automobile dependency and its associated problems in Nigeria.

The private car is the single most used mode for intra urban trips in Enugu. It experiences a high level of dependence by commuters for faith-based and recreational trips and is the second most important mode for work-based and school-based trips. This indicates a potential automobile-dependent city. There is, therefore, the need for policymakers to develop policies and program that will discourage automobile-dependent cities in Nigeria. The measures should focus on adopting a multi-modal approach in tackling intra urban trips challenges instead of the current approach, which focuses mainly on road transport modes (see mode share in Table 1). Consequently, there is the need to introduce rail and active transport (cycling and walking) options in the modal choice of the intra urban trips in Enugu and other Nigerian cities. This will minimize pressure on the road system and enhance the effectiveness of intra urban public transportation system.

Car ownership is the most important factor that influences the modal split of intra urban trips in Enugu. A unit increase in car ownership will cause a major significant increase and shift in modal split toward private transport in Enugu. This will also lead to automobile dependency and the resultant multiplier effects in the city. Policymakers should therefore formulate policy interventions that will discourage car ownership in Enugu; otherwise implementing policy measures to promote the public transport system in Nigerian cities may not produce the desired goal.

Overall, some socioeconomic factors, such as religion, that influence modal split can be peculiar to a city. This indicates that determinants of the modal split can vary among centenary cities. There is, therefore, the need for policymakers to consider the peculiarity of these factors in their respective cities in making policy decisions regarding the modal split. The policy implications of this study provide the bases for the development of a sustainable urban transport system that is responsive to the needs of commuters.

Conclusion

This study analyzed the modal split of intra urban trips in the centenary city of Enugu. The results of the discriminant analysis classifies the modal split of intra urban trips in the city into two groups, namely public transport (54.9%) and private transport (45.1%). This implies that the majority of commuters use public transport modes (bus, taxi, and tricycle) for intra urban trips. However, the mode share indicates that private transport (car) is the single most used mode for intra urban trips. The bus service is the commuters’ most preferred mode for educational, home-based-work and work-based-home trips whereas the private car is most preferred for religious and recreational trips. Furthermore, religion as a determinant of the modal split is peculiar to Enugu city. This implies that faith-based activities have a strong influence on modal split.

The users of private transport enjoy higher satisfaction levels with the performance of their mode than the public transport users, which exhibits moderate satisfaction level with their intra urban trips modes. The predicted group membership preference shows that 82.4% of commuters that use private transport for intra urban trips prefer the mode, whereas 17.65% will switch to public transport modes for intra urban trips. On the other hand, 95.9% of commuters that use public transport for intra urban trips prefer their modes whereas 4.1% will switch to private transport. This underscores the potentials of public transport in providing an efficient transport system in the city. Car ownership has the strongest influence on modal split and could cause a major significant shift in modal split toward automobile dependency. The outcome of this study is consistent with discrete choice models. Therefore, policymakers should formulate policy interventions that will promote the performance level of the public transport system in the centenary city of Enugu.

Footnotes

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

An Ethics Statement

This is not applicable in this study.

ORCID iDs

Maxwell Umunna Nwachukwu

Uloma Jiburum

Data Accessibility Statement

The data used in this study are already deposited in the Mendeley data repository. The link is .

References

Abbas

F. M.

Wasin

Alqaraghuli

A. A.

(Eds.). (2019). Discriminant analysis and classification, in easy statistics for food science (pp. 161–175). Academic Press. https://doi.org/10.1016/B978-0-12-814262-2.00010-8

Acheampong

R. A.

(2020). Spatial structure, intra-urban commuting patterns and travel mode choice: Analysis of relationships in the Kumasi metropolis, Ghana. Cities, 96, 102432. https://doi.org/10.1016/j.cities.2019.102432

Agyemang

(2017). Mode choice for long distance trip; Evidence from the greater Accra metropolitan area of Ghana. Journal of Transportation Geography, 64, 150–157.

Ahanchian

Gregg

J. S.

Tattini

Karlsson

K. B.

(2019). Analyzing effects of transport policies on travelers’ rational behaviour for a modal shift in Denmark. Case Studies on Transport Policy, 7, 849–861.

Al-Salh

Esztergar-Kiss

(2021). Linking mode choice with travel behaviour using logit model based on utility functions. Sustainability, 13, 4332.

Bucsky

(2020). Modal share changes due to COVID-19: The case of Budapest. Transportation Research Interdisciplinary Perspective, 8, 100141. https://doi.org/10.16/j.trip.2020.100

Canizo

B. V.

Escudero

L. B.

Pellerano

R. G.

Wuilloud

R. G.

(Eds). (2019). Quality monitoring and authenticity assessment of wines: Analytical and chemometric methods. In Grumezescu

A. M.

Holban

A. M.

(Eds.), Quality control in the beverage industry, 17: The science of beverages (pp. 335–384). Academic Press. https://doi.org/10.1016/B978-0-12-816681-9.00010-2

Chee

W. L.

Fernandez

J. L.

(2013). Factors that influence the choice of mode of transport in Penangs; A preliminary analysis. Procedure Social and Behavioral Sciences, 91, 120–127.

Chen

Gan

(2015). Choice model and influencing factors analysis of travel mode for migrant workers: Case study in Xian, China. Discrete Dynamics in Nature and Society, 2015, 236216. https://doi.org/10.1155/2015/236216

10.

Clement

(2021). Modal split for intra-urban movement in Enugu metropolis, Nigeria. [Unpublished MSc dissertation]. University of Nigeria.

11.

Drliciak

Celko

Cingel

Jandacka

(2020). Traffic volumes as a modal split parameter. Sustainability, 12, 10252. https://doi.org/10.3390/su122410252

12.

Duleba

Moslem

Esztergass-kiss

(2021). Estimating commuting modal split by using the Best-Worst approach method. European Transport Research Review, 13, 29.

13.

Essam

Sadi

(2013). Factors affecting mode choice of work trips in developing cities-Gaza as a case study. Journal of Transportation Technologies, 3, 247–259.

14.

Fasina

S. O.

Akanmu

Salisu

U. O.

Okunubi

S. A.

(2020). Intra-city mobility and characterization in a fast growing city of Lagos, Nigeria. Logistics and Sustainable Transport, 11(1), 33–50.

15.

Guerra

(2021). The relationship between urban form and mode choice in US and Mexican cities: A comparative analysis of workers commute. The Journal of Transport and Land use, 14(1), 441–462.

16.

Halldorsdottir

Christenson

Jenson

T. G.

Prato

C. G.

(2011). Modelling mode choice in short trips-shifting from car to bicycle. Retrieved November 1, 2021 from https://trid.trib.org

17.

Sobhani

Ettama

(2022). Exploring commute mode choice in dual earner household in a small Chinese city. Transport Research Part D, 102, 103148.

18.

Jiburum

Nwachukwu

M. U.

Mba

H. C.

Okonkwo

C. N.

Okeke

D. C.

(2021). Determinants of public housing affordability for large income diversity groups in a New-Town Capital City: A case study of Abuja, Nigeria. SAGE Open, 11(3), 1–17. https://doi.org/10.1177/21582440211045

19.

Khan

(2007). Modelling passenger mode choice behaviour using computer aided stated data. [Unpublished Ph.D thesis]. Queens University of Technology.

20.

Wang

Nie

Chen

(2021). Locality adaptive discriminant analysis framework. IEEE Transactions on Cybernetics, 52, 7291–7302. Retrieved November 1, 2021 from https://crabwq.qithub.io

21.

Loganayegan

Umadevi

(2014). Development of modal split modeling for Chennai. International Journal of Modern Trends in Engineering and Science, 1(6), 23–27.

22.

Makajuma

G. A.

de Langen

(2010). Urban transport market determinants in Kenya: Model choice analysis. Journal of Civil Engineering Research and Practice, 7(1), 1–10.

23.

Maheshwari

Jana

Bandyopadhyay

(2020). Optimizing the modal split to reduce carbon dioxide emission for resource-constrained sources. Transportation Research Procedia, 48, 2063–2073.

24.

Mendiola

Gonzalez

(2021). Urban development and sustainable mobility: A spatial analysis in the Buenos Aires metropolitan area. Land, 10, 157. htpps://doi.org/10.3390/land10020157

25.

Mueller

(2018). Health impact assessment of cycling network expansions in European cities. Preventive Medicine, 109, 62–70.

26.

Mugenda

O. M.

Mugenda

A. G.

(2009). Research methods: Qualitative and qualitative approaches. ACT Press.

27.

Novačko

Babojelic

God

Babic

(2020). Estimation of the model split parameters-a case study. International Scientific Journal “Trans and Motauto World”, 5(2), 51–61.

28.

Obaid

Hamad

(2020). Modeling mode choice at Sharjah University city, United Arab Emirates. MATEC Web of Conferences,308, 02004. https://doi.org/10.1051/mateconf/202030802004

29.

Okeke

D. C.

Ukonze

(2019). Conceptualizing urban space (environment) for the delivery of sustainable urban development in Africa; Evidence from Enugu City in Nigeria. Land Use Policy, 87, 104074.

30.

Okoko

(Ed.). (1998). Transportation planning as an integral part of the master planning exercise: The utility of traffic forecasting models in Master planning approach to physical development: The Nigerian Experience. Paraclete Publishers.

31.

Olatoye

T. A.

Oyelana

A. A.

(2019). Analysis of modal choice of residents in Lagos State. The Journal of Human Ecology, 65(1–3), 107–121.

32.

Ortuzar

J. D. D.

Willumsen

L. G.

(2011). Modelling transport. Wiley and Sons.

33.

Ramayah

Ahmad

N. H.

Abdul Halim

Mohamed Zaina

S. R.

(2010). Discriminant analysis: An illustrated example. African Journal of Business Management, 4(9), 1654–1667.

34.

Rasaizadi

Askari

(2020). Effect of family structure on modal split by using the life cycle concept. International Journal Human Capital Urban Management, 5(2), 165–174.

35.

Salon

Aligula

(2012). Urban travel in Nairobi, Kenya: Analysis, insights and opportunities. Journal of Transport Geography, 22, 65–76.

36.

Santos

Maoh

Potoglou

Von Brunn

(2013). Factors influencing modal split of commuting journeys in medium-size European cities. Journal of Transport Geography, 30, 127–137.

37.

Shukla

R. N.

Bhola

A. D.

Gundahya

P. J.

Zala

L. B.

(2019). mode choice analysis of trips of urban residents using multinomial logit model. Journal of Emerging Technologies and innovative Research, 6(5), 268–375.

38.

Tembe

Matusse

(2020). Commuting trips, income and gender in the Metropolitan Maputo, Mozambique. Current Urban Studies, 8, 563–575. https://doi.org/10.4236/cus.2020.84031

39.

Ukonze

F. I.

Nwachukwu

M. U.

Mba

H. C.

Okeke

D. C.

Jiburum

(2020). Determinants of vehicle ownership in Nigeria. SAGE Open, 10(3), 1–13.

40.

Ungvarai

(2019). Modal split: Different approaches to a common term. Materials Science and Engineering, 603, 042091.

41.

Vilgoen

(2020). Investigating the impact of activity behaviour on transport mode choice and use in Cape Town [Unpublished MSc dissertation]. Stellenbosch University.