TE
科技回声
首页24小时热榜最新最佳问答展示工作
GitHubTwitter
首页

科技回声

基于 Next.js 构建的科技新闻平台,提供全球科技新闻和讨论内容。

GitHubTwitter

首页

首页最新最佳问答展示工作

资源链接

HackerNews API原版 HackerNewsNext.js

© 2025 科技回声. 版权所有。

People are still bad at gauging their own interview performance

158 点作者 nns超过 8 年前

19 条评论

fecak超过 8 年前
&gt;some portion of interviewees are losing interest in joining your company just because they didn’t think they did well, despite the fact that they actually did.<p>This is the key takeaway for me. This is IMO and experience probably based on a &quot;sour grapes&quot; type defense mechanism, where candidates will internally try to talk themselves out of wanting a job that they felt they didn&#x27;t get. They walk out of the interview feeling they didn&#x27;t perform, and instead of regretting the lost opportunity they start to focus on even the most minute &quot;negatives&quot; about the job&#x2F;company.<p>The more time you give a candidate to stew about those reasons they don&#x27;t want the job, the more time they&#x27;ll have to find reasons that may not even exist.<p>Positive feedback should be given almost instantly. Some companies and candidates are reluctant to give positive feedback quickly because they feel it may hurt their negotiation leverage.<p>I don&#x27;t have data on this, but anecdotally (over almost 20 years in recruiting) I&#x27;ve seen this countless times while collecting post-interview feedback from candidates. The ones that feel they did poorly are likely to mention a negative about the interviewer or company, which becomes somewhat awkward when we come back with a job offer.
评论 #12456697 未加载
评论 #12456619 未加载
评论 #12457023 未加载
评论 #12460362 未加载
评论 #12457605 未加载
评论 #12457881 未加载
评论 #12458976 未加载
RyanCavanaugh超过 8 年前
I would draw the opposite conclusion from that data. The mode difference was 0, and 91% of people were within 1 star of the interviewer.<p>Without knowing what the interviewer&#x27;s bar is for any particular star level, I don&#x27;t see how the interviewees could do any better. Rounding to 1-star increments amplifies relatively small changes -- if the interviewer rounds up 3.6 to 4 and you round 3.4 down to 3.
评论 #12457231 未加载
评论 #12456331 未加载
评论 #12456550 未加载
评论 #12458117 未加载
Sleaker超过 8 年前
&gt;some portion of interviewees are losing interest in joining your company just because they didn’t think they did well, despite the fact that they actually did.<p>Maybe this is because the interviewee feels partly abused, as if communication isn&#x27;t a two-way street. I think interviews would be less terrible if interviewers actually told you at the end that you did above&#x2F;aswell as&#x2F;below expectations, and&#x2F;or why instead of just waiting 1-2 weeks before getting back to you and then telling you that you are&#x2F;aren&#x27;t moving on to the next step. A huge part of the interview process produces emotional ups&#x2F;downs that no one wants to deal with and which ostracize potential employees, but the interviewers don&#x27;t really do much to help with this.<p>I also think that having a 1-4 range is not a good guide for gathering sweeping datasets. Adding any amount of numbers will help to show actually how much. Or actually using human language to define them would be better. 1) Terrible&#x2F;Unhirable. 2) Unhirable&#x2F;Not Passing 3) Passing 4) Exemplar would actually be beneficial right now arbitrary numbering doesn&#x27;t help to give people much context as to what they should choose, and the numbers are so close that you&#x27;re going to get outliers where maybe some people think that a 2 is passable but not great, and interviewer thinks 3 is passable but not great, or vice-versa.
评论 #12456822 未加载
sushid超过 8 年前
&gt; Therefore, every bar above 0 is impostor syndrome country, and every bar below zero belongs to its foulsome, overconfident cousin, the Dunning-Kruger effect.<p>Dunning-Kruger effect is attributed to &quot;a metacognitive inability of the unskilled to recognize their ineptitude.&quot; [1] Seems like a harsh accusation for being off by a single star.<p>And the author makes a generalization that the imposter effect is &quot;better&quot; than Dunning-Kruger effect. I don&#x27;t clearly see why this would be the case for the interviewee.
评论 #12457946 未加载
p4wnc6超过 8 年前
Maybe I&#x27;m missing something, but why isn&#x27;t the implication going the other direction?<p>In my experience, the problem is that interviewers have no idea how to correctly value a candidate&#x27;s performance. Maybe the candidates are closer to being well-calibrated, but their self-assessments don&#x27;t match up with the interviewers&#x27; because the interviewers don&#x27;t know how to gauge what they are looking for?<p>Making the assumption that an interviewer knows how to measure the response of a candidate, even in cases of extremely quantitative questions with well-defined answers, is highly suspect to me. I think virtually no one knows how to do that effectively.
评论 #12456376 未加载
评论 #12456692 未加载
评论 #12456723 未加载
Kephael超过 8 年前
These interview training programs seems to be really reminiscent of coding bootcamps and appear to have grown rapidly in a very short period of time. I think I&#x27;d almost prefer interviewees being judged based on the schools they attended and companies they worked at. I&#x27;m not sure that I find the ability to cram for data structures and algorithms questions to be indicative of intelligence and problem solving skills. Schools attended and companies worked at could be a very strong potential signal of applicant performance and not involve several rounds of &quot;implement that algorithm from memory&quot;. Are interviews becoming more difficult as a result of the proliferation of all the interview test prep material?
评论 #12456908 未加载
评论 #12457592 未加载
评论 #12456751 未加载
评论 #12457440 未加载
评论 #12458512 未加载
GoToRO超过 8 年前
I often wonder how would these people hire soldiers? &quot;So I see you have no previous experience in killing people... We are looking for people that have at least 5 years of experience in killing people.&quot;<p>My point is you need smart enough people for your job and nothing more. You do an IQ test, a personality one and that&#x27;s all. The rest can be learned.
评论 #12456613 未加载
评论 #12456489 未加载
评论 #12456654 未加载
评论 #12456924 未加载
评论 #12456422 未加载
delphinius81超过 8 年前
Once you have an engineering job, a typical project tasking involves: some team design discussion, some engineering research, some work, some review, more work. The requirements are either known upfront, or determined as a team.<p>I do not understand why interviews are not conducted in this manner. Pick some arbitrary idea and have the candidate work with the interviewer to design something, then let the candidate do a little implementation research&#x2F;design, review with the next interviewer, wipeboard code and review some part of the project with next reviewer. Have a coffee break at some point to get to know the person. Treat the interview process like a normal day at work.
评论 #12459091 未加载
评论 #12458696 未加载
Ologn超过 8 年前
After having interviewed dozens of people, the idea occurred to me that my ratings for interviewees were a Gaussian curve with a normal distribution. Most people blended into each other - they knew the topics the same amount. Enough to be employed in their current job. Then one in six was a standard deviation below or worse, and one in six was a standard deviation above or better.<p>That might be one reason for the impostor syndrome. Because if we interviewed six people, one would be unqualified, four would be OK but not spectacular, and one would know the subjects better than anyone. So in an interview of six people the best qualified one of the six would have the job. The average four were not truly impostors, but would not get the job unless they had an inside recommendation or if we were really needed someone and there was a shortage of candidates coming in.
Apocryphon超过 8 年前
I understand companies are unwilling to give negative feedback to applicants for fear of legal repercussion. But often not getting any form of feedback is highly detrimental to an applicant who&#x27;s invested days if not weeks of going through the process, only to receive a boilerplate &quot;thank you for applying, please try again&quot; response.<p>One wonders if it&#x27;s possible to devise some sort of neutral, yet helpful, assessment that could be given to applicants without fear of legal retaliation. Or, if eventually an applicant will sue a company for <i>not</i> providing feedback after they were rejected.
andrewstuart2超过 8 年前
In related news, &quot;People are still bad at knowing the unknown&quot; and &quot;Everybody thinks they&#x27;re average; perhaps because their sample size can only be 1.&quot;
评论 #12461360 未加载
Fricken超过 8 年前
I&#x27;ve never interviewed for a job I didn&#x27;t get. Possibly because I aim too low. But if I were to present myself honestly as a potential employee, I would enter the interview without greeting or making eye contact, then I would sit as far away from the interviewer as possible and doodle in a notebook for 15 minutes before raising my head and explaining everything I think is wrong about the company.
EdiX超过 8 年前
Mmmhhh I would have drawn the opposite conclusion, most people estimate ±1 star away from the real result, it&#x27;s not bad considering the actual performance is not entirely objective.<p>Glad to see that once again Dunning-Krüger proved wrong.
评论 #12457566 未加载
collyw超过 8 年前
Surely a lot depends timing and circumstances. One interviewer said they gave me the job based on asking if the guys accent was Dutch (he was Norwegian and found that hilarious). It was a maintenance programming job and had a high turnover of staff. Other interviews you may have done well but had some serious competition.
nfriedly超过 8 年前
I used to do a lot of interviews for front-end engineers. The process I settled on was to give them the following setup:<p>- some mock-ups from when we first designed our product gallery<p>- a json file that listed the products names, pictures, description, etc.<p>- a blank css file<p>- a html file that loaded the css and jQuery, and then made an ajax request to load the json file. (This was back when jQuery was pretty much state of the art for front-end.)<p>With all of that in place, I&#x27;d explain that I&#x27;d like them to work on building a product gallery that looked similar to the mockups. I would further explain that, since we only had an hour, I didn&#x27;t expect a polished or even fully functional product, but I&#x27;d just like them to dive in and see how far they got. They could use their editor of choice and ask any questions they wanted.<p>I felt like this gave better insight to candidates skill and workflow. My colleagues all gave more traditional interviews, so I also felt like there was a good counterbalance if someone didn&#x27;t do well with this kind of interview.
评论 #12463355 未加载
merb超过 8 年前
That there still needs to be discussions about that topic. On a workplace people should be treated equally. No matter how many topics will be brought up. A world with different genders, stereotypes, whatever can only work if <i>both</i> sides won&#x27;t interpret too much into it and just enjoy their work with their colleagues.<p>Actually we don&#x27;t employ a woman (I&#x27;m working at a 3 person company). And we actually hire, but the only woman will probably decline on their side, she actually has experience in things we actually not need yet, so I guess we can&#x27;t provide her a new challenge. way probably the best application.
mark-r超过 8 年前
Don&#x27;t miss the footnote: &quot;I’m always terrified of misspelling “Dunning-Kruger” and not double-checking it because of overconfidence in my own spelling abilities.&quot;
kafkaesq超过 8 年前
What a hatchet job.<p>Unfortunately we don&#x27;t have the raw data available; but all we can infer from these heatmaps is that yes (and exactly as one would expect) there&#x27;s a bit of variance between between the two sets of ratings (in both directions). Maybe a bit more than can be explained by the sample size, or from the simple fact people (on both sides of the process) are forced to discretize their assessments (so even if both parties were to objectively agree that the &quot;real&quot; performance was in the 2.4-2.7 range, which it is a good chunk of the time.... inevitably you&#x27;ll find them be off by +&#x2F;- 1 in their quantile assignments a good chunk of the time -- which is exactly what the data seem to show).<p>In particular: independent of how we explain the variance (whether due to sampling effects, or bona fide DK&#x2F;IS) note that the vast majority of the variance is in the +&#x2F;- 1 range. And on the alleged DK side (-2 variance) the measure is apparently quite miniscule.<p>However, that&#x27;s not the original D-K effect. What D-K specifically predicted (and were able to show with much better data and methodology -- in their very specific sample population, at least) is that, <i>among the general population</i> there will be a striking degree of -2 variance, specifically in the bottom quartile. Going by the oft-quoted abstract from their original paper:<p><i>Across 4 studies, the authors found that participants scoring in the bottom quartile on tests of humor, grammar, and logic grossly overestimated their test performance and ability. Although their test scores put them in the 12th percentile, they estimated themselves to be in the 62nd.</i><p>In other words: &quot;Performers in the bottom quartile estimated themselves to be performing, on average, at mean performance in the third quartile.&quot;<p>So if were to look for something analogous to a D-K effect, we&#x27;d expect to find a &quot;hot spot&quot; (or at least the strongest amount of red ink, for that quartile) in the upper left of the first chart.<p>Instead we find most performers in the bottom quartile (correctly) identifying themselves as in the bottom quartile; with a smaller portion bleeding up into the second quarter; and an an almost negligible portion placing themselves in the D-K zone -- namely, the third quarter.<p>If anything what the data show is that (even assuming there&#x27;s nothing to contend with in how the actual data were taken), while inevitably there&#x27;s some variance (both ways) in the two sets of rankings, and probably some of due to distorted self-perception -- on the whole there&#x27;s actually <i>much less</i> of a D-K effect among engineers than among the general population (or rather, &quot;Cornell undergraduates taking psychology classes&quot; -- which, lest we forget, were the actual subject class used in the original D-K study).<p>So basically the opposite of what the author of the original article is claiming to infer from this data.
matmann2001超过 8 年前
Probably because interviewing companies have stopped giving performance feedback.