'

Conducting Great Engineering Interviews

Понравилась презентация – покажи это...





Слайд 0

Conducting Great Engineering Interviews Reverse Engineering Interviews to Reduce False Negatives Gayle Laakmann McDowell July 2014


Слайд 1

Gayle Laakmann McDowell <dev> </dev> Author Interview Coach Interview Consulting (CS) (MBA) 1


Слайд 2

The Problem(s) Candidates suck at interviewing. You want to hire good employees, not good candidates. Shortage of great engineers. 2 Be a better interviewer. Be a better recruiter. By counteracting…


Слайд 3

What I Coach Download at: CrackingTheCodingInterview.com 3


Слайд 4

Traditional Dev Questions ? ? ? ? ? ? ? Disclaimer: So. Many. Exceptions. 4


Слайд 5

How can you help candidates be their best? 5 Behavioral Problem Solving


Слайд 6

Behavioral Questions: Mistakes Underplaying individual contributions Using “we” not “I” Not elaborating on actions 6 Don’t assume Probe for details


Слайд 7

Generalist vs. Specialist Generalist (algorithms/problem-solving): Hires smart, adaptable people if done well BUT necessitates “onboarding” Specialists (knowledge & experience): Hires immediately useful people BUT is the knowledge actually difficult to acquire? 7


Слайд 8

Generalist vs. Specialist Hires smart, adaptable people BUT requires “onboarding” + Doing it right 8 Hires immediately useful people BUT is it really specialist knowledge?


Слайд 9

Generalist(ish) Hiring So you’ve gone generalist(ish)… 9


Слайд 10

Algorithm Questions “Problem-Solving Questions” = Solving a problem you have NOT seen before. 10


Слайд 11

Problem-Solving Questions few false positives (only good people pass) but… lots of false negatives (many good people fail) 11


Слайд 12

Why Problem-Solving Questions Work Smart people do good work. Adaptable as company changes So why so many false negatives? 12


Слайд 13

Why Smart People Fail Knowledge gaps Intimidation Arbitrary questions Not knowing how to be good These are [partially] solvable problems. 13


Слайд 14

This matters! Goal: good employees, not good candidates. Do you want to eliminate… Everyone who doesn’t know ___? People who are insecure? People who are bad at interviewing? People at random? 14


Слайд 15

1: Knowledge Gaps (Why?) Not knowing fundamental CS knowledge Especially older people! Obscure CS knowledge being tested Making questions “hard” via knowledge. 15


Слайд 16

2: Intimidation (Why?) Low self-confidence. Scary questions. Candidates misunderstanding interview goals. Interviewers being [unintentionally] rude. 16


Слайд 17

3: Arbitrary Questions (Why?) Well-known questions. Easy questions. “A-Ha” moments & singular hurdles. 17


Слайд 18

4: Not Knowing How To Be Good Didn’t use an example. Tried to rush the process. Didn’t walk through problem. 18


Слайд 19

Design Around This Decide what you value. Ask the right questions. Give candidates tools to prepare. Train your interviewers. 19


Слайд 20

1. Decide What You Value Is problem-solving skills a top priority? What sorts of specialist skills do you need? Are fundamentals of CS sufficient? 20


Слайд 21

2. Ask the Right Questions Right topics: Avoid scary topics. Require only basic CS knowledge. Right difficulty: Medium & hard problems. Avoid common problems. Logical path with multiple insights. 21


Слайд 22

3. Give Candidates Tools to Prepare Tell them: What to expect. How they’ll be evaluated. Struggling is normal. Links to preparation resources. 22


Слайд 23

4. Train Your Interviewers How to design good questions. Be extra nice to candidates. Coach candidates to be great 23


Слайд 24

4b. Coach Candidates To Be Great Encourage the right example. Remind them of details. Don’t let them dive into code. Encourage shifting gears. POSITIVITY & SENSITIVITY A good coach can make a HUGE difference! 24


Слайд 25

Need The Right Infrastructure Interview training List of good & bad questions -- and why Standard resource sheets for candidates Feedback from candidates Assigned roles Continuous evaluation & shadowing 25


Слайд 26

It shouldn’t come to this… 26


×

HTML:





Ссылка: