5 ms·

I've been involved in interviews for new hires a couple of years now. I'm pretty sure I could have the interviewee talk about cucumbers for 10 minutes and I cou

I've been involved in interviews for new hires a couple of years now. I'm pretty sure I could have the interviewee talk about cucumbers for 10 minutes and I could determine if its a good hire or not. It's all about getting insight about how the person thinks.

ragona 5y ago
To be honest I sometimes get better signal from things like cucumbers than I do with technology. Tech has a real issue with biasing towards people who happen to have worked on a particular topic, and as we all know we get bounced to unfamiliar topics constantly.
jstx1 5y ago
If I'm kind of stiff in conversations with strangers but decent at programming, does that mean that your company doesn't have a place for me?
- halfmatthalfcat 5y ago
  Are you stiff when talking about programming? Regardless what you're talking about, you're going to have to work with a team and communicate the problems at hand. I think that's really the crux of "having a conversational interview". Can you communicate technical things, both broadly and in depth, effectively.
- kube-system 5y ago
  Not the OP, but I have a similar experience. Being “stiff in conversations” is not really enough detail to answer the question. I don’t care about whether people are social butterflies. I care about whether they can accomplish the work as a part of a team. I know that sounds cliche but it’s the truth. I just want people who are accepting to feedback, don’t act unprofessionally, and can effectively work in a team.
  - cema 5y ago
    Accepting feedback, but also providing feedback, also asking for it. So being able to initiate a conversation when needed is an important skill.
- jfengel 5y ago
  The problem with "decent at programming" is that it's often only a fraction of the job. You rarely get formal specifications and clear orders. Nearly everything involves discussion of the user's needs or eliciting the circumstances of a bug. Real programming is very little like they teach in school and even less like coding competition. Being good at programming is great, and mandatory, but if you can't also have a conversation then you can't actually do most jobs.
  - jstx1 5y ago
    In my experience there's a big difference between an interview and talking/communicating/collaborating with coworkers. I don't think that you can use one as a reliable predictor of the other.
    - jfengel 5y ago
      I think that's the point of TFA. You can never predict perfectly but it might be a closer approximation than a traditional interview. At least you're talking about programming, something you should know about. In the actual job you'll have to talk about the subject domain, which you aren't always an expert in.
    - xyzelement 5y ago
      > I don't think that you can use one as a reliable predictor of the other. It's pretty hard to claim that there's no signal from this kind of interview. Candidate A: was able to have a conversation with me, asked good questions, explained their thinking well, was easy to follow, etc. Candidate B: seemed to not understand what I was saying/asking, his answers were rambling and incoherent, and unless I led the conversation he just sat there in awkward silence. You can't make a prediction about which of these is more likely to communicate well at the office?
      - Jensson 5y ago
        All else being equal on paper, the better their social skills the worse all their other skills will be since otherwise the guy with social skills would have had a way stronger career than the guy without and then they wouldn't be equal on paper. The problem is the halo effect here. Candidate A has likely gotten preferential treatment throughout their entire life thanks to his social skills, possibly all the achievements they listed were just because they were good at talking and never had to perform in previous jobs? While if candidate B has similar achievements but acts like that then you know he earned them, since nobody would give such a guy a pass unless they knew their shit. On average a new employer will rate the persons work the same as the persons old employers, since these are the same set. So if both the socially savy person and the socially inept person got the same rating at previous employers, then you should expect to rate these two the same as well. If in these cases your process prefers the first over the second then you will over select for social savviness and under select for technical acumen.
sillysaurusx 5y ago
In all likelihood, you've rejected phenomenal candidates and you didn't even realize it.
- chrisabrams 5y ago
  This is why I have candidates do a take home. We’ll (speaking across entire career not just current employer) have people who do mediocre in the video interviews but then turn in a great take home submission. Not everyone performs their best on the spot. I’ve hired so many talented people just by trying alternative formats to the traditional “today you’re gonna interview with 6 people hope you did your leet code.”
- Aeolun 5y ago
  But did they hire any terrible ones?
- CobrastanJorji 5y ago
  This is the thing I most hate about being an experienced interviewer. I've interviewed hundreds of people. Am I getting better? Who knows! I get no feedback. I can find out whether other interviewers come to the same conclusions that I do, so I can learn to conform, but I will never find out if the person I rejected would've been fine, and I won't even find out if the person we hired did well. How can ANYONE confidently proclaim that they're great at interviewing without some way of measuring the people they reject?
  - Jensson 5y ago
    Google still hires people even if one interviewer rejects them, so at least the people doing statistics there knows the difference between having 1 rejection, no rejections and how well each interviewer performs. But I agree, I interviewed some at Google and we can see the stuff that happened in every other interview. And I was really surprised many times, ultimately I realized that I can't really make good judgements based on an hours worth of data and stopped caring.
wpietri 5y ago
I'm sure you believe that, but I'd bet you don't have data. That's a fantastic way to introduce unconscious bias into the process.

ragona 5y ago

To be honest I sometimes get better signal from things like cucumbers than I do with technology. Tech has a real issue with biasing towards people who happen to have worked on a particular topic, and as we all know we get bounced to unfamiliar topics constantly.

jstx1 5y ago

If I'm kind of stiff in conversations with strangers but decent at programming, does that mean that your company doesn't have a place for me?

halfmatthalfcat 5y ago
Are you stiff when talking about programming? Regardless what you're talking about, you're going to have to work with a team and communicate the problems at hand. I think that's really the crux of "having a conversational interview". Can you communicate technical things, both broadly and in depth, effectively.
kube-system 5y ago
Not the OP, but I have a similar experience. Being “stiff in conversations” is not really enough detail to answer the question. I don’t care about whether people are social butterflies. I care about whether they can accomplish the work as a part of a team. I know that sounds cliche but it’s the truth. I just want people who are accepting to feedback, don’t act unprofessionally, and can effectively work in a team.
- cema 5y ago
  Accepting feedback, but also providing feedback, also asking for it. So being able to initiate a conversation when needed is an important skill.
jfengel 5y ago
The problem with "decent at programming" is that it's often only a fraction of the job. You rarely get formal specifications and clear orders. Nearly everything involves discussion of the user's needs or eliciting the circumstances of a bug. Real programming is very little like they teach in school and even less like coding competition. Being good at programming is great, and mandatory, but if you can't also have a conversation then you can't actually do most jobs.
- jstx1 5y ago
  In my experience there's a big difference between an interview and talking/communicating/collaborating with coworkers. I don't think that you can use one as a reliable predictor of the other.
  - jfengel 5y ago
    I think that's the point of TFA. You can never predict perfectly but it might be a closer approximation than a traditional interview. At least you're talking about programming, something you should know about. In the actual job you'll have to talk about the subject domain, which you aren't always an expert in.
  - xyzelement 5y ago
    > I don't think that you can use one as a reliable predictor of the other. It's pretty hard to claim that there's no signal from this kind of interview. Candidate A: was able to have a conversation with me, asked good questions, explained their thinking well, was easy to follow, etc. Candidate B: seemed to not understand what I was saying/asking, his answers were rambling and incoherent, and unless I led the conversation he just sat there in awkward silence. You can't make a prediction about which of these is more likely to communicate well at the office?
    - Jensson 5y ago
      All else being equal on paper, the better their social skills the worse all their other skills will be since otherwise the guy with social skills would have had a way stronger career than the guy without and then they wouldn't be equal on paper. The problem is the halo effect here. Candidate A has likely gotten preferential treatment throughout their entire life thanks to his social skills, possibly all the achievements they listed were just because they were good at talking and never had to perform in previous jobs? While if candidate B has similar achievements but acts like that then you know he earned them, since nobody would give such a guy a pass unless they knew their shit. On average a new employer will rate the persons work the same as the persons old employers, since these are the same set. So if both the socially savy person and the socially inept person got the same rating at previous employers, then you should expect to rate these two the same as well. If in these cases your process prefers the first over the second then you will over select for social savviness and under select for technical acumen.

sillysaurusx 5y ago

In all likelihood, you've rejected phenomenal candidates and you didn't even realize it.

chrisabrams 5y ago
This is why I have candidates do a take home. We’ll (speaking across entire career not just current employer) have people who do mediocre in the video interviews but then turn in a great take home submission. Not everyone performs their best on the spot. I’ve hired so many talented people just by trying alternative formats to the traditional “today you’re gonna interview with 6 people hope you did your leet code.”
Aeolun 5y ago
But did they hire any terrible ones?
CobrastanJorji 5y ago
This is the thing I most hate about being an experienced interviewer. I've interviewed hundreds of people. Am I getting better? Who knows! I get no feedback. I can find out whether other interviewers come to the same conclusions that I do, so I can learn to conform, but I will never find out if the person I rejected would've been fine, and I won't even find out if the person we hired did well. How can ANYONE confidently proclaim that they're great at interviewing without some way of measuring the people they reject?
- Jensson 5y ago
  Google still hires people even if one interviewer rejects them, so at least the people doing statistics there knows the difference between having 1 rejection, no rejections and how well each interviewer performs. But I agree, I interviewed some at Google and we can see the stuff that happened in every other interview. And I was really surprised many times, ultimately I realized that I can't really make good judgements based on an hours worth of data and stopped caring.

wpietri 5y ago

I'm sure you believe that, but I'd bet you don't have data. That's a fantastic way to introduce unconscious bias into the process.