Browsing by Subject "TRIALS"
Now showing 1 - 2 of 2
- Results Per Page
- Sort Options
Item Open Access Can Results-Free Review Reduce Publication Bias? The Results and Implications of a Pilot Study(Comparative Political Studies, 2016-11) Findley, MG; Jensen, NM; Malesky, EJ; Pepinsky, TB© 2016, © The Author(s) 2016. In 2015, Comparative Political Studies embarked on a landmark pilot study in research transparency in the social sciences. The editors issued an open call for submissions of manuscripts that contained no mention of their actual results, incentivizing reviewers to evaluate manuscripts based on their theoretical contributions, research designs, and analysis plans. The three papers in this special issue are the result of this process that began with 19 submissions. In this article, we describe the rationale for this pilot, expressly articulating the practices of preregistration and results-free review. We document the process of carrying out the special issue with a discussion of the three accepted papers, and critically evaluate the role of both preregistration and results-free review. Our main conclusions are that results-free review encourages much greater attention to theory and research design, but that it raises thorny problems about how to anticipate and interpret null findings. We also observe that as currently practiced, results-free review has a particular affinity with experimental and cross-case methodologies. Our lack of submissions from scholars using qualitative or interpretivist research suggests limitations to the widespread use of results-free review.Item Open Access Classification of crystallization outcomes using deep convolutional neural networks.(PloS one, 2018-01) Bruno, Andrew E; Charbonneau, Patrick; Newman, Janet; Snell, Edward H; So, David R; Vanhoucke, Vincent; Watkins, Christopher J; Williams, Shawn; Wilson, JulieThe Machine Recognition of Crystallization Outcomes (MARCO) initiative has assembled roughly half a million annotated images of macromolecular crystallization experiments from various sources and setups. Here, state-of-the-art machine learning algorithms are trained and tested on different parts of this data set. We find that more than 94% of the test images can be correctly labeled, irrespective of their experimental origin. Because crystal recognition is key to high-density screening and the systematic analysis of crystallization experiments, this approach opens the door to both industrial and fundamental research applications.