Hacker News new | past | comments | ask | show | jobs | submit login
Interviewing for Evidence (dannorth.net)
23 points by BerislavLopac 81 days ago | hide | past | favorite | 11 comments



That's the right mindset - when interviewing, we commonly look at their accomplishments in the past, but what we're really looking for is evidence that the candidate will be able to perform well at the new position.

So we need to collect evidence by asking about specific situations which we assume has been similar to what the candidate will need to do on the new position (Experiential Evidence). This is stronger than the other types of evidence, but especially when the candidate has no direct work experience we'd need to defer to hypothetical evidence where we try to set up a scenario close to what the candidate will likely experience on the new job (hypothetical evidence).

Opinion Evidence is at times useful as well, but most of the time I think that candidates do not voice their opinion - they try to guess, and then articulate, _my_ opinion on the matter. Which makes a lot of sense.

The article describes credential evidence (qualifications) as the most straightforward type of evidence. I, however, have found that this is the most useless type of evidence given how the vast majority of "qualifications" mostly prove that you had the money to pay for the certificate.

It needs to be said that candidates do lie and exaggerate, so collecting evidence is not enough - it's the big picture that emerges based on all this evidence. That's where the good old judgement call comes in. But if you have multiple interviewers involved, it is much easier to argue about this judgement call on the basis of evidence than on the basis of gut feeling.


It seems like a trend among bloggers to try and coin their own phrases or terms, but they often don't quite fit. Or you have to do some mental gymnastics to understand them.

> Experiential evidence

Well, they can lie, this isn't really 'evidence.'

> Opinion evidence

Sounds like an oxymoron


Dan North lives from speaking at conferences, corporate events and tutoring classes, so coming up with novel, memorable terms is his trade mark. I attended such a class ("software, faster" [2] or something like that) around 2016. It was very entertaining, a much needed escape from corporate reality and full of great insights and "aha" moments, but none of that had a lasting impact on me. Dan's way is to reduce complex topics into simple categories - like in the OP post, he speaks about the "Dreyfus model of clinically acquired skills" [1], or pulls out a complexity-impact diagram etc. This reduction of complex topics to simple categories greatly impresses upper management with simple messages and gave us, the foot folk, hopes that we finally found the lever with which we'd make things better. But the category choices and reduction procedure seem arbitrary and I've found the messages imparted of little value in work life and corporate reality.

[1] https://dannorth.net/tags/dreyfus/

[2] https://dannorth.net/courses/software-faster/


Well put. I think this kind of writing is entertaining, but at the end agree that I'm usually not left with any big takeaways from it.


While I think the authors quadrant and rankings of the quadrants are good, I don’t see how this saves judging the candidates on metrics such as “they really impressed me” or “they knew what they were talking about.


My campaign for office might rely on opinion evidence, but the campaign of my opponent relies on hypothetical evidence.


I expected to be given some evidence proving that this is indeed the right way to hire.


sounds like CYA approach wrapped in smart sounding "evidence" word combinations. Like some MBA class on defensive hiring.

>Adjacent to the topic of evidence-based interviewing, I am a fan of interviewing in pairs.

Que the "Office Space" scene. I mean what can be better evidence of the stuff being an MBA style BS than it is being pictured in "Office Space"?!


Interviewing is subjective. Tell yourself that it is not. Problem solved.


Terrible, terrible article. In summary "where do you see yourself in 5 years" is now labelled as "evidence". : facepalm:


From the title, I expected something about how to maximize the reliability of witness testimony and avoid planting false suggestions.




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: