Quality assurance vs Software development internship as a precursor to Data Science career

Question

Here is my situation,

I'm a bachelors student, and i will soon be enrolling into a data science masters course. I plan on continuing a career in data science.

However, i have some free time right now and i have opportunities to start temporary internships as a developer or in quality assurance as a tester.

Given my inexperience, i was wondering which decision would most benefit me.

I'm leaning towards QA, as i think it would provide me more opportunities to see what decent code actually looks like, which transitions into really any field in comp sci quite nicely. Development seems more impressive on a resume, but i feel like the experiences i would obtain would not transition as well as those in QA.

Forgive my ignorance on the topic, and thank you!

Answer 1

I'm leaning towards QA, as i think it would provide me more opportunities to see what decent code actually looks like, which transitions into really any field in comp sci quite nicely.

If that is your goal, you'll need to dig in on the QA processes in your target companies. Ask specifically their involvement with code.

I worked in QA for 25 years. I've never worked for a company where QA interns saw any significant amounts of code at all. Even the most senior QAers didn't see lots of code. And trust me, the code we did see often wasn't the best.

I often had aspiring developers work for me as QA interns. They learned a lot of valuable things for their future Dev careers - but "seeing what decent code actually looks like" wasn't one of them.

I strongly suspect a Development internship would prove more valuable. You'll obviously get to see at least some code in that role.

Answer 2

Development, almost certainly.

Now, specific jobs vary and a place that has very skilled, empowered SDETs would be the “good” kind of QA experience. But most QA gigs, you won’t even get to see the code! Worst case you’re manually executing test scripts which is work humans shouldn’t be doing; at best you’ll be writing end-to-end tests in selenium or cucumber or something and learning about advanced testing like performance testing, none of which is greatly applicable to data science.

Whereas being able to actually write programs to empower the work you do is never not relevant. Of course, there are churn and burn dev jobs too, where you won’t see much different code or get mentored, but that’s just a bad gig not a bad space.

Plus, to be blunt, you need to be concerned about your resume especially as someone just starting their career, and in most places QA is considered “the place we put people not good enough to develop.” And while that’s not true everywhere, it is true in many places because that’s what they do.

The only reason to start in QA is if you can’t develop (you don’t say what your bachelor’s is in and if you can code or not) and/or if you can’t get a job developing.

Answer 3

Depends on the environment, but QAs usually do not see code.

I suppose it depends on what kind of QA job you get, but QA typically seems to work with the end product, not the code. We developers finish writing and deploying the project, and then hand it to QA to test as though they are a user.

At best, they are working on code to automate the tests, which has few similarities to code for software.

Seeing code is also not that valuable.

Your future data science employer is going to want you to write code. Merely reading it, even if you did that as part of your QA duties, would not be sufficient. I can do a reasonable job of reading code in any generic programming language. I can only write code in a few.