Abstract: [Purpose / significance] This paper is written to address the reproducibility crisis and improve the truth of scientific research. [Method / process] First the definition of reproducibility is introduced. And then selective reporting of results is analyzed to be the main incentive, including HARKing, cognitive biases and p-hacking. Following it, the TOP guidelines for promoting transparency and open science are presented in detail. [Results / conclusion] On the basis of the TOP guidelines, four practices are deeply discussed. They are registered reports , open science, preprints and TOP factor. These practices are significantly effective and should be for wider adoption, evaluation and improvement.

摘要： 【目的/意义】应对可重复性危机，提升科学研究的可靠性。【方法/过程】从可重复性定义出发，分析了研究不可重复的主要诱因是对结果的选择性报道，包括HARKing、认知偏误和p值篡改，并详细介绍了促进透明和开放科学的TOP准则。【结果/结论】以TOP准则为依托，综合了四种适合广泛采用、评价和改进的应对措施，如注册报告、开放科学、预印本和TOP因子。