Roundtable – Open Science
About this Session
Time
Thu. 11.04. 16:45
Room
Plenary Hall (R1)
Speaker
Speakers – Mario Gollwitzer, Brian Nosek, Abel Brodeur, Thomas Hinz.
Moderated by Malika Ihle.
Scientists, politicians, and the general public need published research they can trust. However, many factors potentially affect the reliability of results, such as confirmation or hindsight biases, conflict of interest, irreproducible workflows, and a sub-optimal peer-review process which often does not detect these shortcomings. All empirical research relies on manifold decisions of researchers (on e.g. designs, samples, case selection, model selection). These decisions often impact results in an opaque way. However, research findings that can be reproduced by independent scholars and that stand extensions in the form of robustness checks might be considered “credible”. Open research practices, i.e. providing accessible study plan, protocol, dataset, data preparation and data analysis code, and publication, allows to reproduce analyses and explore variation, to detect error, and to provide enough precision to allow replication, therefore enhancing the credibility of the research. Opening research resources also allow for their reuse, increasing the research’s impact. Over the last decade, many efforts have been undertaken to promote open science – including in the social sciences. The panel will introduce three major projects: SCORE (Systematizing Confidence in Open Research and Evidence), I4R (Institute for Replication), META-REP (Meta-scientific program to analyse and optimise replicability in the behavioral, social, and cognitive sciences). With main proponents of these initiatives, we will discuss (1) how successful these projects have been so far in reanalyzing and replicating social science research, (2) what would lead to greater adoption of open research practices, and (3) what are the roles of each stakeholder (e.g. funders, publishers, institutions, researchers) in promoting this change.
Moderated by Malika Ihle.
Scientists, politicians, and the general public need published research they can trust. However, many factors potentially affect the reliability of results, such as confirmation or hindsight biases, conflict of interest, irreproducible workflows, and a sub-optimal peer-review process which often does not detect these shortcomings. All empirical research relies on manifold decisions of researchers (on e.g. designs, samples, case selection, model selection). These decisions often impact results in an opaque way. However, research findings that can be reproduced by independent scholars and that stand extensions in the form of robustness checks might be considered “credible”. Open research practices, i.e. providing accessible study plan, protocol, dataset, data preparation and data analysis code, and publication, allows to reproduce analyses and explore variation, to detect error, and to provide enough precision to allow replication, therefore enhancing the credibility of the research. Opening research resources also allow for their reuse, increasing the research’s impact. Over the last decade, many efforts have been undertaken to promote open science – including in the social sciences. The panel will introduce three major projects: SCORE (Systematizing Confidence in Open Research and Evidence), I4R (Institute for Replication), META-REP (Meta-scientific program to analyse and optimise replicability in the behavioral, social, and cognitive sciences). With main proponents of these initiatives, we will discuss (1) how successful these projects have been so far in reanalyzing and replicating social science research, (2) what would lead to greater adoption of open research practices, and (3) what are the roles of each stakeholder (e.g. funders, publishers, institutions, researchers) in promoting this change.