To our knowledge, this study is the first survey on residents’ knowledge, attitudes and practices towards research misconduct and factors associated with research misconduct in China. The questionnaire applied has good reliability and validity with the reference to previous studies. The results may help policy-makers and hospital managers identify key populations among residents for research misconduct and design the content of residency training. Our study suggests that limited residents served as a primary investigator for a research project (37.6%) and published a paper as the first author or corresponding author (38.2%). 53.7% admitted to have committed research misconduct. The average total score of self-reported knowledge among residents was 39.44 ± 14.46 (range from 12 to 60). The average total score of perceived consequences for research misconduct among residents was 25.16 ± 8.47 (range from 7 to 35). Previous studies have already shown the worrisome prevalence of self-reported research misconduct among medical faculty members (4, 11, 17). Our study revealed the worse results among residents in hospitals, which demonstrated the insufficiency of research integrity management in hospitals. However, most residents (88.5%) attended a course on research integrity, which was surprisingly contrary to the high prevalence of research misconduct and attending a course on research integrity contributed to research misconduct in the study, indicating the shortcomings in the current research integrity courses and urgency to update and implement the content of courses.
In our study, multiple submissions (50.6%) was the most frequent form of research misconduct, and 12.2% of the residents conducted ≥ 4 times multiple submissions. Since multiple submissions was only perceived as a severe deviance by scientific journals, some researchers regarded it as little apparent harm (18). The Ministry of Education of the People’s Republic of China has not emphasized the severity of multiple submissions (15), and it was not even considered as research misconduct by some researchers (19). Falsifying research data, materials, literature or annotations, or fabricating research results was the second most common form of research misconduct with an alarming self-reported rate of 49.0%. Bjørn et al. reported that 10.0% of participants believed the common incidence of falsification, fabrication, and plagiarism (FFP) and some respondents were willing to conduct these practices based on their perceived true conclusions (20). Regarding funding applications, FFPs were more acceptable and regarded as not important as in publications, so the actual prevalence of FFPs may be worse than expected.
Lower self-reported knowledge was associated with higher research misconduct prevalence. We also collected the self-reported reasons for research misconduct from residents. The top 3 reasons are “researchers lack research ability”, “researchers are influenced by academic environment” and “researchers deviate in personal value”. These results indicate that a lack of personal research ability and knowledge about research misconduct lead to residents’ involvement in research misconduct. The reputation and income of medical staff are closely associated with the professional title, and research achievements play an important role in the professional title promotion system and the evaluation system in Chinese tertiary hospitals, such as publication of papers or grant application. The contradiction of research ability and promotion pressure would contribute to the incidence of research misconduct. Previous studies also suggest that promotion pressure and individual morality are the main perceived reasons for research misconduct (1, 4). Consistent with earlier surveys, personal morality was the main influencing factor, suggesting its importance in the enhancement of personal morality (18). Lower perceived consequences for research misconduct were significantly correlated with research misconduct with the enormous OR, and this may provide clues for the design of training. Limited researchers considered that education on research misconduct has an effect on reducing the incidence of research misconduct (4). Therefore, more courses focusing on consequences for research misconduct should be conducted for residents to reduce the incidence of research misconduct.
Residents with postgraduate degree or above may be more likely to conduct research misconduct. Oren et al. (21) also reported that PhD nurses tend to fabricate, select or omit data to enhance publication. Majid et al. (22) suggested that postgraduate students have a higher estimation of research misconduct than undergraduate students. More statistical skills in postgraduate students may make it easier for them to fabricate, select or omit data easier, and the desire to be successful drive them to conduct research misconduct (18). Besides, the restricted time and overloaded work are the barriers to research, and may lead to the incidence of research misconduct. Those Serving as a primary investigator for a research project have a lower inclination to conduct research misconduct. This may be related to the experience of researchers. Primary investigators usually have relatively rich research experience and knowledge, and junior researchers usually have poor knowledge of research misconduct on the contrary (19).
Our results reflect the weakness of research misconduct knowledge, the importance of perceived consequences and practices for research misconduct, and the factors linked with research misconduct among residents. Those who were postgraduate or above, had lower scores of research misconduct knowledge and perceived consequences, did not serve as a primary investigator for a research project, published papers as the first author or corresponding author, and attended a course on research integrity tended to conduct research misconduct. This revealed a terrible phenomenon in which residents with research experience almost have a tendency to conduct research misconduct despite being trained, suggesting the necessity for the reformation of residents’ education. More attention should be paid to residents’ education by hospital managers and policy-makers, and we propose several recommendations to improve research integrity. First, research integrity training with an emphasis on research skills and consequences for research misconduct could be included in the compulsory curriculum, and more training forms should be conducted to provide more convenient access to research integrity knowledge, such as seminars and online training, etc. second, a scientific evaluation system may be established that focuses more on the quality of publications than quantity to reduce the incidence of research misconduct. Third, auditing and surveillance system can be implemented in hospitals, and the department in charge of research integrity should be set up and maintain its authority and independence.
The study has a few limitations. First, convenience sampling was applied due to the sensitivity of research misconduct, which may affect the results. Furthermore, the questionnaire was derived from self-report, and bias may be involved in the process despite assurances of anonymity. Finally, although based on other studies, the questionnaire was self-designed and measurements may differ from the study objectives.