Scientific research suffers due to pressure for positive results

Scientific papersThe scientific community can be a dog-eat-dog world, with many scientists vying for the same jobs and grants. The researcher going home with the price is often the one with the most publications in high-ranking.

A study by the University of Edinburgh shows that this competition may hurt the quality of scientific research, since many scientists aim for positive research results, as those are most likely to be published.

Negative results are as valuable but not as interesting

Papers reporting negative or null results however can be just as valuable as those reporting positives ones. But since they are less likely to be published and a scientists’ renown is often measured in number of publications and citations, negative or null results are usually just not as interesting to scientists.

The research published in Scientometrics encompasses over 4,600 papers from all scientific disciplines published. Looking at papers in which a hypothesis had been tested, the researchers found a rise of more than 22% in frequency of hypothesis supporting results between 1990 and 2007.

Differences between disciplines and countries

Furthermore the results showed significant differences between disciplines and countries. With the social and biomedical sciences being the most positive disciplines and Japan the most positive country. The growth was particularly low in the UK which along with the other European countries showed a lower increase than the US.

Is it the journals or the scientists?

According to the researchers changes in methodology cannot explain away the patterns. This led them to hypothesize that the decrease in negative outcomes has to lie either with journals simply not publishing them or with scientist pursuing predictable results or producing positive outcomes by re-interpreting, selecting or even manipulating data.

Whatever the cause, the decrease in published papers with negative outcomes may result in a waste of resources through repetition of research or a misinterpretation of other research because not all data is available or complete.

© Jorn van Dooren |

Comments are closed.