ACN - 101321555 Australasian Human Research Ethics Consultancy Services Pty Ltd (AHRECS)

Resource Library

Research Ethics MonthlyAbout Us

ResourcesResearch IntegrityReply to de Winter and Dodou (2014): Growing bias and the hierarchy are actually supported, despite different design, errors, and disconfirmation-biases (Papers: Daniele Fanelli | 2014)

Australasian Human Research Ethics Consultancy Services Pty Ltd (AHRECS)

Reply to de Winter and Dodou (2014): Growing bias and the hierarchy are actually supported, despite different design, errors, and disconfirmation-biases (Papers: Daniele Fanelli | 2014)

 


View full details | Go to resource


I appreciate the efforts that de Winter and Dodou (2014) have put into replicating and challenging claims made by Fanelli (2010, 2012), as well as those of Pautasso (2010). This is how all sciences should make progress, and it is therefore both a duty and an honour to respond to this challenge.

The results presented are largely in agreement with claims by Fanelli (2012 and 2010), but this fact is obfuscated by a somewhat selective interpretation of findings, reinforced by differences in study design, and major flaws in the sampling and analytical design.

FLAWS IN INTERPRETATION:

1) Fanelli (2012) claimed that negative results are disappearing in percentage, which is exactly what is found here. Even de Winter and Dodou (2014) quote Fanelli (2012) as using percentage figures, so I am quite baffled as to why they consider their results at odds with mine. For the record, the absolute number of negative results in Fanelli (2012) did not show a decline, and it was never claimed in the paper that it did

Fanelli D (2014) Reply to de Winter and Dodou (2014): Growing bias and the hierarchy are actually supported, despite different design, errors, and disconfirmation-biases. Peer J – commentary. (non peer-reviewed) This is the original pre-print the comment replies to.
Access this preprint



Resources Menu

Research Integrity


Human Research Ethics

0