Do Users Prefer Particular Public Bus Transport Operators in Bhopal, India?

1. QUESTIONS

In Bhopal, three different public bus transport operators (PBTOs) run along various routes, sometimes covering the full or partial length of the same routes. These PBTOs operate in parallel across the city. Among them, My Bus is a Special Purpose Vehicle (SPV) managed by Bhopal City Link Limited (BCLL) under the Bhopal Municipal Corporation, while the minibuses and suburban midi buses are run by private operators. The Regional Transport Office (RTO) oversees these PBTOs, assigning routes and fleets, which are detailed in Table 1. Though ridership across these PBTOs was heavily impacted during the COVID-19 pandemic, it gradually returned to normal levels afterward.

The existence of three PBTOs operating in Bhopal, either along full or partial stretches of the same routes, raises the question of whether users prefer one operator over another, especially given that each operator must reach a certain ridership to break even. This prompts two key research questions: (i) Do users in Bhopal differentiate when choosing a PBTO? (ii) What travel behavior attributes influence the choice of one PBTO over another?

2. METHODS

The study employs both qualitative and quantitative methods. Due to the dynamic nature of the research, some steps and procedures evolved as the study progressed. A quantitative approach was used to measure target users’ responses to a set of questions, enabling data comparison and statistical analysis. The qualitative approach provided insights into the decision-making processes related to travel across the three PBTOs in Bhopal. The overall methodology is outlined in Figure 1.

Figure 1.Methodology of the Study

Source: Developed by the Authors

Identification of travel behavior attributes: A conceptual framework for understanding the mode based paradigm of defining travel behaviour is described in Figure 2.

Figure 2.Conceptual Framework for the Identification of Attributes of Travel Behaviour

Source: Developed by the Authors

The mode-based paradigm of travel behavior suggests that the choice of a particular transport mode is influenced by the following factors (Khisty and Lall 2003):

• The characteristics of the transport mode such as occupancy, ride time, frequency, and other service-related factors.

• The characteristics of the trip including trip distance, time of travel, and similar trip-specific elements.

• The characteristics of the trip maker defined by socio-economic, behavioral, and psychological factors, such as family income, vehicle ownership, and family size.

These factors form the basis for this study to identify the travel behavior attributes that help determine whether users have a preference for a specific PBTO. From a literature review, 29 relevant attributes were identified and validated through an expert opinion survey, which included 22 experts from academia, industry, and government. The experts ranked each attribute on a 3-point Likert scale: 3 for Agree, 2 for Neutral, and 1 for Disagree (see Table 1 in Appendix I). Attributes were retained based on the modal values derived from these rankings. These 29 attributes, listed in Appendix I, were used in the primary survey (conducted through four sets of questionnaires) and in statistical analysis to identify any preference differences.

The survey was conducted after finalizing the travel behavior attributes, following these steps:

• Complete Enumeration: The survey included all users of the three PBTOs, meaning no specific sample size was set. A total of 767 responses were collected: 337 (44%) from My Bus, 207 (27%) from Midi Bus, and 223 (29%) from Mini Bus.

• User Classification: Based on the study’s hypothesis, users were classified into three categories: (i) Daily, (ii) Weekly, and (iii) Occasional users. Further, users were categorized as either dedicated or non-dedicated to a specific PBTO. A user was considered dedicated if they consistently chose the same PBTO, regardless of how frequently they traveled. Most respondents (44%) were daily dedicated users. The majority of respondents (78%) were male.

• Preliminary Field Study: A reconnaissance survey was conducted to define the study area in Bhopal, mapping the routes where all three PBTOs operate using GIS-based mapping.

• Survey Procedure: Two volunteers, along with the principal investigator, conducted the survey on both weekdays and weekends during peak hours (9 AM - 11 AM and 5 PM - 7 PM) and one non-peak hour (12 PM - 2 PM). They boarded all three PBTOs from a common bus stop at various times and interviewed users as they boarded the buses at shared stops.

• Data Collection Tool: The survey was carried out using the open-source software KoBo Toolbox. The data collected was stored in Excel and coded for further analysis.

The data sets from the primary survey were stored in Excel format using the KoBo Toolbox, categorized for each of the PBTOs. Users were classified as either daily-dedicated or daily non-dedicated. Most responses to the survey questions were recorded as dichotomous (yes/no), with ‘yes’ coded as 1 and ‘no’ coded as 0, except for age, which was recorded as specific age groups. These codes facilitated the statistical analysis of user preferences (see Table 2).

To assess the reliability of the survey responses, Cronbach’s Alpha was calculated, since the responses were recorded in binary form. The formula for Cronbach’s Alpha is:

$$\mathbf{\alpha = (N \cdot \overline{c})/(\overline{v} + (N - 1). \overline{c})\ }$$

Where:

• N is the number of test items,

• $\overline{c}$ is the mean covariance between pairs of test items,

• $\overline{v}$ is the average variance.

The Cronbach’s Alpha values for the survey responses were as follows:

• My Bus: 0.712

• Mini Bus: 0.933

• Midi Bus: 0.923

These values meet the threshold for response consistency, as outlined by Tavakol and Dennick (2011), indicating that the responses from users of each PBTO were reliably consistent.

To assess differences in user preferences for specific PBTOs, a Kruskal-Wallis One-way ANOVA by Rank test was conducted. The test checks if the variance between operators is greater than the variance within operators to determine if there is a significant difference in user choices. The Chi-square value (7.36) exceeded the critical value (5.99) for 2 degrees of freedom at a 5% significance level, and the p-value was below 5%, leading to the rejection of the null hypothesis. A Post Hoc test further identified significant differences in user preferences between Mini-Midi, Mini-My Bus, and Midi-My Bus, with a p-value of 10%, confirming the rejection of the null hypothesis.

Linear Probability Modelling (LPM) was used to explore the relationship between user preferences for a PBTO and 29 identified predictor variables (from Table 1). The LPM model takes the form:

$$Y_i=\beta_i +\beta_z X_i+u_i$$

The response variable (Y), representing the probability of choosing a particular PBTO, is binary: 1 if the user chooses a PBTO and is a daily dedicated user, and 0 if they do not.

3. FINDINGS

The results of the association analysis reveal a significant relationship between the travel behavior attributes and the users’ preference for one PBTO over another. Linear Probability Model (LPM) results are summarized in Table 3.

ACKNOWLEDGMENT

The authors acknowledge the support from the officials of My Bus, Bhopal City Link Limited (BCLL) and private operators of midi and mini buses. The authors acknowledge the hard work and support in conducting the primary survey by the volunteers. The authors are deeply grateful to the experts who participated in the expert opinion survey for short listing the attributes of travel behaviour.

This study was undertaken as part of Doctoral Research by the Authors under Department of Urban and Regional Planning, School of Planning and Architecture, Bhopal, India.