Blogger Razib Khan at the Gene Expression blog claimed that bias is "rife in any science which utilizes statistics." The GiveWell blog has suggested researchers publish their questions, theories, and planned methods of data collection before their begin their work, so that if the results come out differently from what the researchers expected, then this cannot be hidden. S. Pelech argues that one of the major reasons why bias can have such a profound impact on the outcome of a scientific study is that there is so much emphasis on conducting hypothesis-driven research. In conducting system-wide, unbaised research to seek what is really going on inside of organisms and their cells, half the battle is to collect sufficient data and the other half is to look at the results with fresh eyes and let the data reveal to the observer what is happening. Read More...

Tags: Research Bias

Daniele Fanelli at the University of Edinburgh examined more than 4,600 scientific papers published between 1990 and 2007, and found "a steady decline in studies in which the findings contradicted scientific hypotheses." During those 17 years, positive results increased from around 70 percent in 1990 to about 86 percent in 2007, and she speculates that the growing pressure to report only positive results may lead to a "decline" in scientific research around the world. S. Pelech observes that with the rapid progress made in the biological sciences with improved tools and techniques over the last few decades, it is not surprising that our knowledge about the world is becoming increasingly extensive. Nevertheless, it is striking that with around 23,000 proteins encoded by the human genome, it appears that over 95% of the biochemistry/molecular biology publications arise from less than 5% of these proteins. It would seem likely then that there is in fact a high degree of redundancy in scientific publications, which contributes to a very high rate of positive results. Read More...

Tags: Publication, Research Bias

John Tierney of the New York Times noted that having a corporate connection does not necessarily bias a researcher’s work, but many journals insist on declarations of corporate sponsorship from industry. S. Pelech argues that research biases exist just as much in publications arising from universities and research institutes as from industrial labs. Read More...

Tags: Conflict-of-interest, Research Bias