(Other than that: yes, I agree, except that actually conducting hiring interviews that way would probably actually lose more signal than it eliminated noise, at least in the fields I’m familiar with interviewing in. Alas.)
Software development, engineering, mathematics (in industry rather than academia).
The loss of signal could probably be eliminated in all of these, with some effort. The sort of thing I’m thinking of where signal would be lost by default is where you ask the interview candidate to design something, write a bit of code, sketch a system they worked on in the past, etc., on paper or whiteboard. If the candidate has to be behind a curtain, that’s difficult to do and probably involves irksome extra latency (e.g., a system where they write or sketch whatever they want to and then step aside, and only then does the interviewer get to see what they did).
You could work around this with computerized whiteboards—the candidate sits in one room and the interviewers in another, both rooms have electronic whiteboards, and they are coupled so that anything written on one shows up on the other too.
(Or by using something other than whiteboards that’s easier to decouple in this way. For instance, for a coding task some kind of collaborative text editor may do better.)
I see. (I had guessed you were talking of people who have to directly¹ interact with perspective customers, so you have to know what they look and sound like in order to know what first impression perspective customers might get.)
Of course what “directly” means depends on where you are; I hear there’s a country where people will boycott a Web browser solely because of the political stance of the CEO of the company making it on a topic with hardly anything to do with software. ;-)
Yes, that would be another example. But, I think, a different sort of example. Let’s suppose that candidate A comes across better than candidate B in interview simply because of widely-shared prejudices affecting the interviewers. For the kind of job you describe, that (rather horribly) means that candidate A probably is better able to do the job than candidate B.
(It might well be that the best thing overall is to try to stop people in that situation favouring A over B on account of prejudice anyway, in the hope that over time this reduces the overall level of prejudice and everyone is better off.)
Just curious… What fields are those?
Software development, engineering, mathematics (in industry rather than academia).
The loss of signal could probably be eliminated in all of these, with some effort. The sort of thing I’m thinking of where signal would be lost by default is where you ask the interview candidate to design something, write a bit of code, sketch a system they worked on in the past, etc., on paper or whiteboard. If the candidate has to be behind a curtain, that’s difficult to do and probably involves irksome extra latency (e.g., a system where they write or sketch whatever they want to and then step aside, and only then does the interviewer get to see what they did).
You could work around this with computerized whiteboards—the candidate sits in one room and the interviewers in another, both rooms have electronic whiteboards, and they are coupled so that anything written on one shows up on the other too.
(Or by using something other than whiteboards that’s easier to decouple in this way. For instance, for a coding task some kind of collaborative text editor may do better.)
I see. (I had guessed you were talking of people who have to directly¹ interact with perspective customers, so you have to know what they look and sound like in order to know what first impression perspective customers might get.)
Of course what “directly” means depends on where you are; I hear there’s a country where people will boycott a Web browser solely because of the political stance of the CEO of the company making it on a topic with hardly anything to do with software. ;-)
Yes, that would be another example. But, I think, a different sort of example. Let’s suppose that candidate A comes across better than candidate B in interview simply because of widely-shared prejudices affecting the interviewers. For the kind of job you describe, that (rather horribly) means that candidate A probably is better able to do the job than candidate B.
(It might well be that the best thing overall is to try to stop people in that situation favouring A over B on account of prejudice anyway, in the hope that over time this reduces the overall level of prejudice and everyone is better off.)