At Applied we are proud that hiring managers love our platform. We are equally glad that candidates also love the application process. We design the hiring process to remove bias. Does this mean that groups who could be more prone to suffering bias from others in traditional hiring processes end up rating our process more highly?
After completing a job application through our platform we invite candidates to submit their feedback about the application process. They feedback on their rating of the process on a scale of 1-10 with 10 being the best. They can also add comments if they wish.
Most candidates give our application process really high ratings. What we found out that was new is that:
- Non-white candidates did rate the process slightly higher than white candidates.
- Both groups tended to rate our application process very highly and therefore it is works well for both majority and non-majority ethnic groups
- White candidates were slightly more likely to rate our application process as 7 or 8 out of 10 than non-white candidates
- Non-white candidates were more likely than white candidates to rate the process as 9 or 10 out of 10
- Overall 65% of candidates gave us at least 9 out of 10.
- Overall the average rating is 8.57
After looking at the ratings we decided to have a look at the text of what was written. The three comments below were received on the day this post is being written:
“This anonymous application framework is great! This should be how people are hired.”
“Great system and easy to use!”
Whenever we look at the data we seem to find recent comments along the same lines. Here are some other comments we have received in 2019:
“This was a very unusual and thought-provoking first step in an application. It was great to be treated like an individual and human being from the outset rather than as a numbered candidate whose significance only features in the final application stages so thank you.”
“It was the first time an application required of me to write for specific situations and show how i deal with things in real life situations! Even if I don't get selected, it was a great experience and i would love to do that again!”
“I think this is a much better way of assessment than other applications I have completed as it was really tailored towards the specific skills the candidate can bring to the job.”
But we decided to take a different perspective on this by using word clouds to see which words occurred frequently across the data. There are nearly 8000 pieces of written feedback so it would take quite a while to read them all! First, in our analysis, we excluded words that were bound to come up a lot due to the context e.g. application, questions.
For the 65% of candidates who give us scores of 9 or 10 out of 10 the 2 most commonly occurring words for white candidates were:
And for non-white candidates it was similar:
We are very happy with these results and proud to be providing a platform that enables hiring managers to find great recruits in an unbiased way. We are also happy that candidates generally find using the platform a very positive experience.
Even if we expand the word clouds to look at all feedback not just those from candidates who rated us highly, and we expand the number of words we still see a very positive picture. For non-white candidates we see this for the top 4 words:
Meanwhile, for white candidates we see this:
The feedback for both groups is very similar with overlap in three of the four words. The only difference is that white candidates seem to focus more on the cv. This may reflect the very bias we are trying to address. Traditional selection processes using a cv are prone to favour white candidates as numerous experiments have shown. In the past white candidates may have seen greater success using cv’s and may therefore be more inclined to see the cv as a useful part of the hiring process. Non-white candidates on the other hand more commonly thank us for our innovative approach. So in summary, the great news is that we have found a way to level the playing field whilst also developing a process that candidates rate highly.