Make science PhDs more than just a training path for academia

I really like referencing the illustrative guide to a PhD when I think of a PhD [1]. The whole point of a PhD is to create that dent in the picture. I feel like the training path for academia is just an effect from pushing the boundary since you need to adhere to the scientific method. It's the same argument that some people make about engineer degree in college. I did not expect my university to teach me all practical skills (like how to weld, or machine things) because that is not the point of the education. The education is meant to gear you towards making principle based designs rather than history based design and in order to that, which requires a lot of theory.

[1] http://matt.might.net/articles/phd-school-in-pictures/

It ought to be a training path for academia too!

It’s always struck me as strange that the skills needed to get a faculty job (essentially, the ability to produce interesting data) are mostly disjoint from those needed to keep the same job (management, grant writing).

This topic always makes me a little suspicious. Are we trying to meet an industry demand, or are we looking for a relief valve for an excess of PhDs?

Doctoral students are such a good deal for universities, I can see why they’d want lots of them. Yes professional students and MS students pay tuition and sometimes TA a bit, but for high skilled low wage labor that lands big grants, there’s nothing like a few doctoral students to write the code or do the lab work.

I’m not convinced that industry bound students need to lean these skills at low wages in a grad program with long completion times and high attrition rates. Elite JD students are completed in a consistent 3 years and have an attrition rate typically below one half of one percent, and they don’t graduate ready to practice law. They learn this working brutal hours under stressful conditions at $180,000+ a year. Elite PhD programs take over twice as long and have attrition rates of about 35-50% depending on the discipline.

A shorter, more consistent practice oriented doctoral degree makes sense compared to a PhD but... is there a market for it? I’m not just asking if they’d get hired, I’m asking if they’d be hired at sufficiently higher rates than MS or BS holders to justify the expense and loss of earnings. Would it be competitive with those degrees + an MBA?

There are still reasons to do it. The nature of the job is also a factor. Senior CRUD bug fixer (or overstressed lawyer) might pay a little (or a lot) better that Senior analyser of fascinating data sets, and lave lower entry costs. But I can see why someone would pay a bit more to earn a bit less to get a more interesting job. There are limits though. Is a PhD or even one if these alternate doctorates really essential? Or even worth the opportunity cost?

I'd rather ensure research was better funded and thus a more viable career path (with more research institutes separate from colleges to level out the pyramid structure)

Otherwise it just seems like turning a PhD into job training - something which is already happening to many undergraduate degrees as the BSc becomes the new High School Diploma.

One thing that is nice about computer science, perhaps as compared the natural sciences, is that a Ph.D. is absolutely a valuable industry degree. Just look at some of the best examples: Jeff Dean, Sanjay Ghemawat to name two. Many modern companies in the computer industry - though not all - value Ph.D.s and actively try to hire them.

In The Netherlands we have a specific degree on the PhD level but focused on practical research, called PDEng (Professional Doctorate in Engineering).