Dataset For Discrete Participation Patterns in Demand-Responsive Transit Choice: Insights from a Latent Class Analysis

Dataset Description This dataset contains stated-preference survey data collected from 500 respondents who evaluated six hypothetical demand-responsive transit (DRT) choice situations. Each choice situation included two DRT service alternatives (A and B) and an explicit reject-service alternative, resulting in 3,000 choice observations. Files user_history.csvContains 3,000 choice observations (500 respondents × 6 choice situations). Each row represents the alternative selected by a respondent in a stated-choice task. id: Respondent identifier (1–500) situation: Choice situation (0–5) choice: Chosen alternative (0 = A, 1 = B, 2 = reject service) features.csvContains 9,000 alternative-level observations (3,000 choice situations × 3 alternatives). Each choice situation contains three rows corresponding to alternatives A, B, and reject service. id: Respondent identifier situation: Choice situation alternative: Alternative identifier (0 = A, 1 = B, 2 = reject) access: Walking time from origin to boarding location wait: Waiting time ivt: In-vehicle travel time egress: Time from alighting to the final destination linc: Household income indicator (1 = monthly income < 3,000,000 KRW) license: Driving license ownership (1 = has a driving license) constant: Reject-service indicator (1 = reject alternative) Variable Definitions access: Walking time from origin to the boarding location. wait: Waiting time, calculated as vehicle access time minus walking access time. ivt: In-vehicle travel time. egress: Time from alighting to the final destination. linc: Household income indicator. license: Driving license ownership. constant: Indicator for the reject-service alternative. All time variables are integer-coded in minutes according to the stated-preference survey design. Data Structure Each respondent completed six hypothetical choice situations. Each situation consisted of two DRT service alternatives (A and B) and one reject-service alternative. The user_history.csv file records the selected alternative, while features.csv provides the corresponding alternative-specific attributes. The reject alternative is represented by alternative = 2 and constant = 1. Its service-related time attributes are coded as zero. Intended Use The dataset can be used for research involving stated-preference analysis, discrete choice modeling, demand-responsive transit choice, machine learning, contextual decision-making, and related transportation research. Data Privacy The dataset has been prepared without directly identifying information about individual respondents.