China initiated the "Five No's for Publication" in December 2015 as a response to rising incidents of retraction. Use the number of retracted publications and their original publication time as proxies to investigate the effect of the Five No's policy on academic misconduct. We searched the Retraction Watch Database for research articles published by Chinese scholars from 1 March 2010 to 29 February 2020. The short- and long-term trends of the number of publications were presented by conducting an interrupted time series analysis in quarterly time units. Of 4,215 retracted papers with Chinese authors, 2,881 involving academic misconduct were identified. In the first quarter (12.01.2015-02.29.2016) after the implementation of the Five No's, an average reduction of 55.80 (p < 0.001) publications that involve academic misconduct was observed, although there was an increase in the trend of publications of 3.34 per quarter (p < 0.01) in the long run (12.01.2015-02.29.2020), relative to the pre-intervention period (03.01.2010-11.30.2015). The validity of these results was further supported by three different robustness checks. China's government should strengthen enforcement, promote education, and improve the scientific evaluation system to consolidate the influence of the Five No's policy and foster an ethical research environment.
Keywords: Academic misconduct; Retraction Watch; interrupted time series analysis; policy evaluation; scientific integrity.