The scientific process up close

A tale has recently come to light of a bit of scientific backbiting. John Bargh et al. published a classic piece of research in 1996 showing that behaviour can be affected by trigger words. This is called a priming effect. In the case of Bargh’s research, participants were given a word task. When the word related to being old, the participants left the lab more slowly i.e. the word primed a tendency to a particular response. According to Google Scholar this research has been cited almost 2,000 times – which means it has had a reasonable impact.

In the spirit of good science it is important to confirm such an important finding but no one has done this until just recently. Stephane Doyen and colleagues (2012). They used more participants but failed to find an effect. What they did find is that participants did move more slowly if the experimenter expected them to move more slowly (i.e. had been told the research hypothesis) but no effect was found if the experimenter did not have this expectation. An example of experimenter bias.

This is where the story gets really interesting – Bargh has launched a scathing attack on the new study, including personal attacks on the authors as incompetent. This goes against the principles of science where replication and objective discussions should be welcomed. The scientific community has rallied round and brought the issue out in the open (for example here). This all illustrates the process of science at work.