Application form advice
OK, so I’ve finished interviewing lots of data scientists over the last few weeks. I’m sorry that there were so many applications for the data scientist (hundreds) that I really can’t give people feedback to individuals whose application was not selected for interview. However, I thought what I could do is give some advice for people who were not successful in getting to interview, and quite honestly I would say the same to a lot of you anyway so I think it’s almost as good. I’m going to make two observations first, and then I’ll give some advice.
The first observation I will make is that there were lots of really great candidates, at application and interview. It’s great to see how many people want to work as data scientists in the NHS, and I think my team’s commitment to open source really resonated with people, which is lovely.
The second observation I will make is I spoke to loads of data scientists who have all the right ideas and are trying to do the right thing, but they are being held back either by a lack of interest from their organisation, or, worse, by their organisations actively blocking them. I’m quite sad about that. People tell me they know about GitHub, and they use it when they get home for personal projects, but they’re not allowed to use it at work, or they’re not allowed to use Python, or they’re forced to do things in PowerPoint because that’s the file format people are used to, that kind of thing. When I’m interviewing people I find it a bit sad, I wonder what they would be doing at work and what they would be saying in the interview if they worked somewhere that was developing them and allowing them to work in a way that would help them get the best results out of what they were doing.
Anyway, rant over. Here’s what I’m looking for in an application form. If you look at the essential criteria you’ll see three basic things that run through what I want in an application form, what I want in an interview, and what I want in a data scientist. Can you write code, can you manage your work and show initiative, and can you work in a team.
A really good application form will address very clearly all of the essential criteria, with examples of specific stuff that you’ve done. I’m sure you all know that. To improve some of the applications I received though I would suggest the following.
Write enough. Honestly, some of the forms were kind of short. It’s not enough to say “I’m good at something, I did it such and such a time”. Flesh it out a bit. You used GitHub. When? How? Whom with? An application that very clearly, with examples, and fleshed out examples, shows experience and knowledge of the essential criteria will be a good application.
For a truly great application, and one which is more likely to be shortlisted, show some insight into your work that illustrates the elements of the essential criteria. That’s what I really want to hear. You used GitHub, at a specific time and place, for a specific purpose, with a specific set of people. What insights did you gain? For example, on Friday the team and I were having a look at GitHub projects, we already use Kanban to manage our sprints but we were seeing what GitHub can add to this process. There’s a clear trade off between complexity and power when you’re using GitHub to manage projects. You may not have done anything fancy like that. Who merges pull requests? The reviewer of the owner? How do you let people know there are pull requests? What is your git workflow? How big should a commit be? How old/ behind main is it OK for a branch to be?
Talk about the way you code. Sometimes when I’m coding, especially dashboards, I find it useful to move away from the screen and use a pen and paper for a bit. Do you refactor early, late? Does your work environment make it hard to get feedback so you end up making three of everything? Bring out the learning you have from your work around each of the essential criteria, with clear and detailed examples, that’s the application that brings you to life as a data scientist and will maximise your chances of getting an interview.