Also how did you prevent people finding both postings? When searching for jobs if identical they will appear in the same list. Also the smart ones applying will apply on the one without because of all the various biases that the poster has and they will play some reverse psychology.
yes, this is a major flaw in their analysis. there's no way to draw a reliable conclusion because the experiment is uncontrolled in this way, among other factors (like the subjective evaluation of candidate quality which was also uncontrolled for).
Regarding performance of responses I am referring to count of unique respondents meeting minimum requirements.
Regarding 'better candidates', this is based upon the count of candidates which made it through team interviews and coding challenges.
Our process requires multiple manager/leader approvals at each step of the process which is intended to reduce bias.
It is possible the soft attributes of personality and communication are reflected better in one group than the other. I do not have evidence but is a possible source of unintended bias.
RandomWorker|3 years ago
clairity|3 years ago
kokanator|3 years ago
Regarding performance of responses I am referring to count of unique respondents meeting minimum requirements.
Regarding 'better candidates', this is based upon the count of candidates which made it through team interviews and coding challenges.
Our process requires multiple manager/leader approvals at each step of the process which is intended to reduce bias.
It is possible the soft attributes of personality and communication are reflected better in one group than the other. I do not have evidence but is a possible source of unintended bias.