Blind academic CVs

[lekkerkroketje]

This is translated TAAT, but I hope you'll humour me!

There is a thread somewhere about 'anit-racist' actions. One of the point suggested was using blind CVs for recruitment with just qualifications, job history and a personal statement. Academia can be pretty discriminatory against a whole variety of people, being in places a rather closed white-male club. I know that that's partly due to the nature of the job and partly due to problems getting people into the earlier parts of the system among many other forms of discrimination and I wasn't really looking for a discussion on that. But all of this is exacerbated by some recruiters refusing to advertise other than by word of mouth or choice mailing lists (I know that's not technically allowed in the UK, but it certainly happens elsewhere!) or binning CVs from institutions they're not familiar with. So maybe some sort of blind CVs would have an advantage.

The obvious question is how on earth would you do it? Removing undergrad institute should be relatively painless, because a decent PhD should cancel out any disadvantage there. But what about PhD institute? And how to hide the exact identifying subject matter? And post-doc positions? Where, and what worked on, and for how long? Publications and citations depend on funding, institute, the give aways about origins and sex in author names which affect citations, collaboration opportunities, time out for whatever reason so H number is going to bias towards rich-country white males. If it's a small field, these things are already identifying. So then does the pool of applicants shrink down even further to the two identifiable candidates vs the risky unknowns? I guess that's part of the point, to get the risky unknowns to interview.

I'm not currently hiring, but was thinking about it for the future, along with the recurring moan about the closed incestuous nature of the business. Do any of your institutions attempt blind recruiting? Does it actually work? What would you do differently?

[GCAcademic]

I can’t see how this could possibly work in academia. Publications are such a huge part of the application /CV. And are likely to remain so as long as the REF exists. Many departments will read and hypothetically score publications before even coming up with a long list for interview. Good luck changing that practice when the REF stakes are so high.

[lekkerkroketje]

I don't much fancy my chances of changing the system, especially given my lack of proficiency in the legal language where I am. In the current climate, I should probably go and flog myself for not caring enough either to learn the language or to change the world!

We don't have anything quite as awful as the REF, but I guess that's just another example of how officially self reinforcing the system is. I suppose I was more thinking for grad student and post doc recruitment where there's a bit more flexibility. The ivory towers are pretty unassailable, but at lower levels personal decisions and biases determine who gets a chance.

[parietal]

It is not feasible to blind an academic CV. As soon as you include the titles of publications, then someone can type that into Google Scholar and find the authors. But publications are one of the key things that are used to evaluate CVs.

But other things can help - a lot of work has been done by ATHENA SWAN etc to reduce bias against women, and the same principles can apply

• CVs should include only the top 4 or 5 papers to avoid bias towards the people with big collaboration networks.
• making hiring committees genuinely aware of their own bias and how to counteract it.

and there are probably others, but these are a start.

[MedSchoolRat]

How do you know what the 'top' 4 or 5 papers are? I've got like, 100 items on my pubs list going back almost 30 years. How would I choose?

We do that unconscious bias training, annually. I don't believe in it, tbh. After you've done the course a few yrs it's very automatic to renew but nothing sinks in. A lot like the manual lifting & fire safety etc. training.

So back to OP...
You could have a blind recruitment round that was done on skillset, experience, areas of expertise and even something like H-index, rather than pubs list. To get to a short list of 6 people, maybe. However, none of the jobs I've been involved in (that I applied or I helped recruit for) had more than 4 applicants in recent years. So... a blind round would have been waste of time.

Nobody looked at funders when I was involved, but I have relatively subordinate roles, not looking at grants won or institutes led, etc.

[worstofbothworlds]

We get 30+ or more applicants for any jobs, maybe a blind round would be good for us but everyone would be guessing anyway!

How do you know what the 'top' 4 or 5 papers are?
This I can do quite easily. They are the ones with the best data, most comprehensive studies, best journal etc. I have a paper from 10 years ago that gets cited loads and one from 20 years ago ditto. A couple more recently that are the result of a really big study in both cases, one in a very big journal.

I had to write an article blind recently (come to think of it, I need to find another destination for it as it was rejected after a ridiculous number of revisions). It was crazy - this was a review article so I was at least not saying "there's this dataset from Smith and we followed up on it but we have nothing to do with Smith" when it's obvious you do as you went to the same area of the world/asked the same people questions/used the same rather niche organism. But I could choose between saying "we found, Author (2010)" or Smith (2010) found and referring to myself in the 3rd person all the time.

[murmuration]

Ah, see, I wonder about that "top" papers. As some of my top papers are a decade old or more, and then I worry about not looking current. I wonder if one could ask for "top" and "most recent"? Although I'm also sore from being downgraded from "rising star" to "hasn't done much lately" on exactly the same project with no difference other than the invited resubmission occurred after maternity leave... So maybe "top" with an emphasis that they don't have to be recent could be more levelling.

[worstofbothworlds]

Yes, usually it will say "most recent/most relevant/most representative".
I will sometimes put in the one that's a decade old (because I'm just following that one up) and then a couple more recent that aren't as earth shattering.

[geekaMaxima]

Yes, I've seen specs that ask for top 5 papers of the last 10 years, though not blinded. It attracted people with anything between 8-20 years experience in academic posts, and I thought it was a really good leveller - maternity leave became (almost) invisible, it put the focus on quality over quantity, and it kept it recent enough that people's expertise wasn't out of date (relevant in my STEM field).

As it wasn't blinded, it had the interesting effect of making some known names (to me) look less impressive when compared against relative unknowns in concrete terms. I liked that. The post went to an lesser-spotted known candidate in the end, and I think the top-5-papers approach played a distinct role.

[dwnldft]

But don't panels just look up the full publication lists anyhow? They certainly do in my field.

[worstofbothworlds]

Not for all applications, if you have 30+

[dwnldft]

I guess different fields have different cultures. We get way more than 30 applications but would still automatically look up all the publication lists when reviewing them.

[CaraDune]

Like others I don't see how you could do this for CVs, just from a practical point of view. About the only way you could institute it, I think, would be to reduce the publications section of the initial CV to "this applicant has X number of papers in journals with the following research impacts, which have been cited Y number of times"). You could, I suppose, use this at the paper sift stage, but it certainly wouldn't fly at the shortlist stage.

On change I would like to see would be for all peer-reviewed journals to introduce double-blind reviewing, so that the referees didn't get to see the authors' names on the manuscript (referees are anon in my field - though we always have a good game of "guess the reviewers" when the comments come in).

[dwnldft]

"this applicant has X number of papers in journals with the following research impacts, which have been cited Y number of times")

But this would violate DORA agreements. We shouldn't be judging quality of papers by the impact factors of the journals in which they are published. Citation metrics are also recognised to have many limitations and should be used carefully and within context.

[geekaMaxima]

But don't panels just look up the full publication lists anyhow? They certainly do in my field.

I have been in shortlisting meetings where the shortlist was based on the top-5 papers and the full publication list (that was included in some CVs anyway) was expressly discounted. Anyone who referenced the full pub list was just told "we're not considering that" by the chair.

I couldn't make the final panel in the end so I'm not sure if the interview stage was run as tightly 🤷‍♀️

[MedSchoolRat]

I guess I'll never need to know this, but I still don't know how I'd choose my 'top' articles. The one I worked hardest on & maybe learnt most from has only been cited once. The ones cited most I'm not a lead author on or they are ancient; or authors are in alphabetical order. The one that got most publicity hasn't even been published (& won't get much publicity after published, it was the preprint that got attention but things moved on fast).

Best journal... well, I actually had something to out to review with The Lancet recently but of course it was rejected. I still think it should go on my cv that it actually got to review. The published article arguably in the 'best journal' was rewritten by another coauthor so much that it doesn't feel like mine any more (I did sweat some blood on that one, tbf).

[OhTheRoses]

3 and 4 publications that can be included in the prevailing REF cycle. If you had 6 publications in 2014 and nothing above 4m2 it doesn't exactly indicate a rising trajectory.

[OhTheRoses]

6 4 publications in 2004 and nothing above 2 since.