Enclothed cognition brushes up well

The idea that our clothing affects how we think and feel about ourselves is hugely popular. The right clothes for the right occasion can even make a big difference to how we act.

But clothes can have further reaching effects than just making us feel good. Over a decade ago, a pair of US-based researchers first described 'enclothed cognition', reporting that people performed better on a test of attention when wearing a lab coat that was described as a doctor's coat versus a lab coat described as a painter's coat. That 2012 study has been cited more than 600 times and covered by more than 160 news outlets, note the authors of a new paper in Personality and Social Psychology Bulletin. 

Since that initial publication, dozens of other studies have reported that the symbolic meaning of clothing can influence cognition. For example, one study found that women feel more powerful when wearing heeled rather than flat shoes, while another claimed that wearing a police uniform causes people to pay more attention to people of a low socio-economic status. 

However, in 2019, a high-powered direct replication attempt of the original 2012 paper did not find any differences between the 'painter' and the 'doctor' conditions, casting wider doubt on the topic. Prominent findings in the field of embodied cognition and social priming more broadly have also recently famously failed to replicate. 

In their response to the 2019 replication failure, the authors of the original study argued that while the replication attempt did call their specific finding into question, the "sum total" of the available data suggested that the core idea of embodied clothing is "generally valid." Ascertaining whether or not this is the case is clearly important for the research in this area. 

In order to get to the bottom of this, Dr C. Blaine Horton Jr of Columbia University and colleagues set out to conduct a thorough analysis of every experimental, peer-reviewed study on enclothed cognition published in English that they could find. In total, they identified 40 such studies, from 24 articles. 

The researchers then used a statistical tool called a z-curve analysis. This can reveal the presence of publication bias — that is, whether negative findings have been under-published. It can also provide other important insights, such as whether there is evidence for 'p-hacking' — when data is manipulated so that a statistically significant finding can be reported. 

Overall, Horton Jr and their colleagues found that for the papers published before 2016 (the year after the Open Science Collaboration drew attention to the replication crisis in psychology, and after which research practices were improved), their analysis could not rule out the possibility that questionable research practices, or an over-preponderance of false positives, played a role in findings in support of the idea of enclothed cognition. 

However, all was not lost. Their analysis also suggested that after 2016, there was no publication bias and that questionable research practices were unlikely to have played a role in the described effects. The analysis also suggested that a majority of the significant effects from the papers published since 2016 were likely to replicate under the same conditions. 

The team was also able to estimate that the size of the effect clothing has on thoughts, feelings, and behaviours is small to moderate. The effect sizes were consistent across studies looking at different types of clothing and various potential outcomes, the researchers note. 

Sample size was highlighted as a limitation of many of the examined studied. The authors argue that most included too few participants, with the average overall sample size in the studies was 103, when at least 150 would have been ideal. 

Overall, though, this new work suggests that the core idea of enclothed cognition is generally valid. It also demonstrates an exciting meta-psychology approach that might be used in other fields with high profile papers that have failed to replicate, but ample papers that support the core idea. (The concept of the growth mindset, for example, may be one possible target.) Just as one positive finding cannot be viewed as the final word on any subject, neither can one failed replication attempt. 

Read the paper in full: https://doi.org/10.1177/01461672231182478