Problems and complaints resulting from harassment on railways have been highlighted as a social problem. Actual cases of such problems were investigated, and countermeasures were reviewed and evaluated by conducting questionnaire surveys. Approximately 3,500 passengers were surveyed to ascertain the circumstances of harassment and demands for countermeasures, and a monitor survey of about 80 passengers was carried out to evaluate the effects of such countermeasures.

      In the monitor survey, evaluation of the effects of countermeasures was conducted by focusing on the information content of educational etiquette posters as such a countermeasure. The questionnaire results were used to formulate seven common types of on-board annoyance for poster data (violence, littering, using cellphones, standing near the door, cutting in line, applying make-up, and sitting on the train floor.) The evaluation results verified the validity of posters for obviously antisocial behavior such as crime and violence, and also for behaviour where the numeric data shown on the posters was surprising (Fig.1). The validity of a method showing a ranking of annoyances as a way of presenting data was also verified.

      The following five countermeasures were also presented and individually evaluated from five view points: (1) posters showing annoyance data; (2) campaign notices inside the train or on the car body; (3) use of the opinions of harassment sufferers (such as using a PA system to give voice to sufferers); (4) explanation using posters, for example, of why such behavior causes annoyance, and (5) humorous Japanese poems about train etiquette. These five evaluation results showed that using the opinions of harassment sufferers in all five points had a positive effect, raising expectations of the effectiveness of such countermeasures (Fig.2).

Degree of annoyance caused by obvious violence Degree of annoyance caused by make-up application - comparison of the numeric data between men and women is surprising
Fig.1 Examples of posters with data whose validity was verified in the monitor evaluation	Fig.2 Evaluation results of the five-point countermeasures