Favorable research environment is a key determinant of research integrity according to a ten-country survey across Central and Eastern Europe

Our study was designed to investigate research integrity among scientists actively working in biomedicine. Using the unique opportunity of the Alliance for Life Sciences, a networking initiative of leading research institutions and universities in Central and Eastern Europe, we organized our own in-depth survey on research integrity in the countries involved. We employed a standardized questionnaire consisting of 19 closed questions with simple, multiple-choice, or scaled responses to explore the situation regarding research integrity at Alliance for Life Sciences member institutions. The questionnaire was in English language and was administered through the Qualtrics platform anonymously. Finally, 10 institutions from 10 different countries participated in the study and the aggregated group of respondents consisted of 752 scientists. First, the analysis of the obtained data included descriptive frequencies of the responses to all types of questions. Second, the construction and analysis of the model of latent variables was included to verify assumptions about individual aspects of the observed behavior and their interrelationships. Confirmatory factor analysis was applied to verify the domain structure, followed by multivariate analysis of variance to assess the effects of institutional affiliation, gender, seniority, and ethics training. Our data provide the first systematic description of research integrity at the biomedical research institutions in 10 Central and Eastern European countries. Because the majority of our respondents were experienced researchers, our study is really valuable for mapping the state of research integrity in participating countries. The descriptive part of the results brings a detailed insight into the research environment in relation to research integrity, as well as recent and former personal experiences with scientific misconduct. It also covers the personal acceptability of various forms and consequences of scientific misconduct and personal estimations of scientific misconduct. The unique findings of our study came from the model showing the correlations between latent and higher-order variables, which reflect relevant domains of the questionnaire. Moreover, seniority and ethics training were identified as predictors of adherence to the principles of research integrity. These results emphasized the importance of the favorable research environment regarding scientific integrity, which is apparently interconnected with the incidence of various forms of misconduct as experienced recently or in the past. Our study using a model of latent variables yielded unique results that clearly showed the importance of fostering a favorable research environment, especially through systematic education in research integrity and available institutional policies that both reduce the risk of scientific misconduct.