Mentoring and Research Misconduct: An Analysis of Research Mentoring in Closed ORI Cases

We are reporting on how involved the mentor was in promoting responsible research in cases of research misconduct. We reviewed the USPHS misconduct files of the Office of Research Integrity. These files are created by Institutions who prosecute a case of possible research misconduct; ORI has oversight review of these investigations. We explored the role of the mentor in the cases of trainee research misconduct on three specific behaviors that we believe mentors should perform with their trainee: (1) review source data, (2) teach specific research standards and (3) minimize stressful work situations. We found that almost three quarters of the mentors had not reviewed the source data and two thirds had not set standards. These two behaviors are positively correlated. We did not see convincing evidence in the records that mentors were causing stress, but it was apparent in the convicted trainees' confessions that over 50% experienced some kind of stress. Secondary data, while not created for this research purpose, allows us to look at concrete research behaviors that are otherwise not very researchable. We believe it is important for mentors and institutions to devote more attention to teaching mentors about the process of education and their responsibilities in educating the next generation of scientists. This becomes a critical issue for large research groups who need to determine who is in charge educating, supervising and assuring data integrity.     

The Prediction of Family Interaction from a Battery of Projective Techniques

This study first explored the adequacy of items on the Rorschach Rating Scale (RRS), which measures Rorschach constructs from a variety of scoring systems. Analyses determined that all items had an adequate capacity to differentiate people and none were clearly redundant. At the same time, the current version of the RRS requires good reading comprehension skills for accurate completion (13th grade level). Next, we developed two sets of RRS scales. Factor analysis of observer ratings (N = 234) indicated the RRS contained six broad, empirically derived factors. In addition, 19 conceptually derived scales were developed from the RRS item pool. Both sets of scales were evaluated for reliability and then compared to the Big Five model (B5M) of personality through a series of factor and regression analyses. Results indicated that RRS scales do not provide adequate definition of all B5M dimensions and B5M scales do not provide adequate definition of all RRS dimensions. We discuss the place of psychotic processes within a comprehensive model of personality and other implications from these findings.