PURPOSES : This study aimed to explore crowding impedance for high-risk travelers on various modes of public transit during the COVID-19 pandemic and develop a transport policy to encourage the proper use of public transport.
METHODS : A stated preference survey was conducted to investigate the behaviors of travelers on various modes of public transit, with special emphasis on crowding inside vehicles. Multinomial logit-based modeling was used to estimate the explanatory variables identified as parameters based on the surveyed data. A crowding multiplier was adopted to represent the behavioral differences for the high-risk travelers on various modes of public transit.
RESULTS : The established model was solved using the ‘mlogit’ R package program to estimate the identified parameters. The results demonstrated significant behavioral difference for the high-risk travelers on public transit during the COVID-19 pandemic. The proposed crowding multiplier successfully captured the reduced likelihood of high-risk travelers to be sensitive to crowding on the subway; furthermore, it revealed that non-crowding travelers on the subway are less sensitive to crowding than bus travelers.
CONCLUSIONS : This study estimated crowding impedance for high-risk travelers on various modes of public transit during the COVID-19 pandemic and suggested an appropriate transport policy for those travelers.

ABSTRACT
1. 서론
2. 선행연구
    2.1. 잠재선호 조사 방법론 연구
    2.2. 대중교통 혼잡도 추정 연구
    2.3. 연구의 기여도
3. 연구 방법
    3.1. 연구의 분석과정
    3.2. 로짓 모형
4. 설문조사 구성 및 데이터 분석
    4.1. 설문조사 구성
    4.2. 데이터 분석
5. 결과 및 분석
    5.1. 모형의 분석 결과
    5.2. 고위험군의 혼잡저항 분석
    5.3. 고위험군을 위한 대중교통 정책 방향
6. 결론 및 한계점
REFERENCES

• 조신형(조지아 공과대학교 박사후연구원) | Cho Shin-Hyung
• 윤혁렬(서울연구원 교통연구실 선임연구위원) | Yun Hyuk-Ryul
• 김동성(서울연구원 공공투자관리센터 연구위원) | Kim Dong-Sung Corresponding author