Single Review on past events

There is wide agreement that complete reporting of experimental outcomes is required for the overall evaluation of the validity of a model or hypothesis. However, at the same time, there is a problem with “negative results”, which often go unpublished. Outlets for communicating these research outcomes are lacking, and current editorial trends prevent researchers from publishing this information. Prof. Jutta Schickore from the Department of History and Philosophy of Science at Indiana University Bloomington is interested in this problem. In her Jour Fixe presentation she addressed the topic “Null and void and unpublishable? Scientists' perceptions of negative results, non-significant research outcomes and replications.”

Schickore wanted to find out how working scientists themselves interpret the notions of “negative result” and “replication”. She tried to explore what significance (in the everyday sense of the term) they attach to these kinds of research outcomes, whether they think that these kinds of research outcomes are worthy of publication, and what experience they have (if any) trying to publish these research outcomes.

Together with two colleagues, she conducted a web-based survey of 2,535 individuals from 49 departments (sciences and medicine) at Indiana University as well as 40 in-depth interviews with a sample of the survey respondents. The interviews focused on conceptualizing types of research outcomes that were perceived as hard to publish or unpublishable, experience in publishing these types of papers, and the perceived value of such work.

They found that depending on the discipline the term “negative result” is used in various different ways. There is no consensus across or within disciplines about what constitutes a “negative result” and what constitutes a “replication”. Also, there are (at least) two aspects of the problem: the problem of non-reporting negative results and the problem of non-reporting replications. Outcomes of replications seem to be a large and important component of unreported research outcomes. Therefore it would be misleading to equate “unpublishable result” with “negative result”. It appears that very few scientists are used to talking about methodological issues. Only very few researchers (all clinicians) couched their concerns about negative results, replications, and publication barriers in methodological terms. By far the most frequent concern was about conservatism in the scientific community.

“Our research suggests that there should be more emphasis in science education on the importance of full reporting. In this context, appreciation for replications should be a goal” she concluded. “Different types of replications need to be distinguished and their different methodological purposes and values need to be highlighted.  Overall, it seems that researchers lack the conceptual tools to reflect on methodological issues. While this may not be a problem in the day-to-day business of research, it becomes a problem in contexts in which scientists are required to express the reasoning behind their methodological moves – for instance, in controversies, in cross-disciplinary research, and in exchanges with wider publics.”

Jutta Schickore received her PhD from the University of Hamburg, Germany, in 1996. Her research interests include historical and philosophical aspects of microscopy, the problem of error in science, studies of the eye and vision, the history of philosophy of science especially from the 19th century, and the relation between history and philosophy of science.