Your question about relevance is fair. The author of the blog post kindly submitted here complains that she [I wrote "he" before edit prompted by kind reply bel

Your question about relevance is fair. The author of the blog post kindly submitted here complains that she [I wrote "he" before edit prompted by kind reply below] gets math quiz questions when applying for programming jobs. That's a mistake on the part of the hiring companies in two ways. 1) The main thing to look for when hiring programmers is programming work samples. In general, for any kind of work, a work-sample test will tell you more relevant information about which person to hire than any other kind of hiring procedure. (Some other kinds of hiring procedures have incremental validity in addition to work-sample tests, but none have higher validity than work-sample tests.) 2) Insofar as a math quiz question purports to figure out which applicants are smart and which are not, it is legally risky, as are IQ tests and AS ARE EDUCATION CREDENTIAL REQUIREMENTS if the quiz question is not validated as related to bona-fide job requirements. The company that publishes the Wonderlic personnel test hires a lot of psychometricians and lawyers to be able to defend the use of that test for hiring for dozens of kinds of jobs, and I think the balance of the evidence across hundreds of studies on many different job categories in multiple countries shows that general mental ability matters for performing most kinds of jobs well. But in the United States, a smartness test that a hiring manager just makes up is legally risky. It's a good idea to hire smarter rather than dumber job applicants, but it's an especially good idea (for legal protection in the United States) to make sure that the smartness test you use has been validated for your workplace. The research I'm referring to is more generalizable across more occupations than you suggest. In other words, the usual relevance of my FAQ about company hiring procedures in the frequent HN threads about company hiring procedures is that companies almost always go wrong by 1) NOT using work-sample tests, a very good way to find competent workers, and 2) often go wrong by posing trick question tests, presumably to find smart people, without validating those tests and ensuring that those tests don't have disparate impact on applicants. Companies can and should do better, and research tells them how. The people who often complain on HN about company hiring procedures are largely correct, but usually just share anecdotes about what they would do differently if they were hiring, or what they would like hiring managers to do the next time they apply for a job. Research on this topic is abundant, and it applies to people seeking work as programmers as well as to most other occupations.