[Scale Development on Perception of Academic Harassment among Medical University Faculties]

In academia, harassment may often occur and remain unrevealed in Japan, which discourages young researchers from pursuing their career. It is necessary to estimate and improve the perception of "academic harassment" among university faculties. Therefore, in this study, we aim to develop a scale of perception of academic harassment.

Prior to a quantitative survey, a task team consisting of medical doctors, researchers, nurses, hospital workers, and managers in general affairs division identified 36 items related to academic harassment. In February 2016, we sent a self-administered questionnaire to 1,126 academic faculty members who worked in a medical university located in Tokyo, Japan. We instructed them to score the extent to which they consider each item as related to academic harassment based on a Likert scale. We carried out maximum likelihood factor analyses with promax rotation and computed Cronbach's alpha to develop a scale and investigate the reliability of the scale.

In total, 377 returned the questionnaires (response rate, 33.5%; male, 73.8%). In factor analyses, we removed 17 items owing to low factor loadings, and four factors were eventually extracted. The first factor was termed "Harassment in organization (7 items)" because it included conditions of forcing a particular person to work on chores or lectures for students that may prevent one's academic research outputs. The second factor was termed "Violence and denying personal character (4 items)". The third factor was termed "Research misconduct (5 items)" including conditions of excluding a particular person from the coauthor list of research outputs or pressuring a person to fabricate, falsify, or plagiarize research outputs. The fourth factor was termed "Research interference (3 items)" including a condition of interference with conference attendance. Cronbach's alpha values of these four factors ranged from 0.83 to 0.91, suggesting that the scale had high reliability. The means of these factors did not differ according to gender but were higher in participants aged 50 or older than in younger participants.

The results suggest that the scale of perception on academic harassment consisting of four factors with 19 items is valid and reliable to some extent.