Bias in IT recruitment on the example of Github
27.07.2020
Bias in IT recruitment on the example of Github
The problem of recruiting bias is a fairly common phenomenon, often resulting from the recruiter's lack of experience, but also from the lack of awareness of its existence. Below I have collected some of the most important indicators that are taken into account by recruiters when evaluating a candidate on Github. The same metrics are of course also used on other similar portals.
1. Number of followers
An indicator that is often taken into account is the number of followers of a given developer. This is a very wrong assumption because many times in my career I have encountered profiles of young, less experienced programmers whose number of followers was much higher than that of senior developers.
2. Number of projects
Another indicator is the total number of projects. It is also not authoritative and it is impossible to determine the programmer's experience on its basis. Many developers create empty, often abandoned projects that are never completed or developed further. In addition, a project is never equal to a project - there are programmers who have created one large and advanced project thanks to which we can assess its programming skills, and there are very small projects that usually solve trivial and uncomplicated problems.
3. Number of commits
The total number of commits is very dependent on the project itself. There are projects where commits are less frequent and some projects are frequent. This never shows a candidate's general technical skills and is not a determinant of the right choice.
4. Specific working hours
Many people think that programmers who work at night are better and more talented. However, this is a very bent thesis that has nothing to do with reality because programmers come from different time zones. Even if we consider recruitment only in a specific time zone, e.g. CEST - a given person can still cooperate, e.g. with a client from the USA, which strongly distorts the whole perception. In addition, such an assumption strongly promotes the culture of perceiving programmers as "robots", and everyone has the right to choose the best working time and time of concentration for themselves (at least in the case of flexible working hours in a given organization).
5. Frequency of work on the project
Many outstanding developers lead a normal family life outside of work. So it's hard to assess the programmer's abilities based only on the frequency of work on the project. After all, everyone has the right to choose - there are programmers of the so-called "outsiders", "hobbyists" who devote every moment to coding, but also those who simply have to reconcile their passion for coding with other everyday duties. A seasoned recruiter understands these dependencies.
6. Profile picture rating
An incorrect way of giving opinions is drawing hasty conclusions based on information from the candidate's profile photo - what we look like in the photo is often a humorous issue, a matter of interest in a specific topic and, above all, very individual. Therefore, let's not judge on this basis whether the candidate will be experienced or not, especially since Github is a social networking site that brings together crowds of programmers with different personalities, ways of thinking and interests.
I hope that the above metrics/indicators, but also the problem of bias in recruitment (not only IT) has become more visible to you now and you will know how to use this knowledge when you start recruiting using Github.
Happy hunting headhunters!
Archives
