I use a coding test not only to see if someone can write basic code, but also to assess general intelligence and see/hear their problem solving process. A resume definitely doesn't tell you all you need to know and experience does not necessarily mean effectiveness.
I actually do understand why people administer these tests in our field. There's so little to guide you otherwise.
The problem is that it's kind of awful, and I personally really do believe that it's driving people away from the field.
Here's my take (I've posted this on HN a few times). Many people find in person, at the whiteboard exams quite stressful. This is fairly common, many people describe exams to be among the more stressful events in their lives.
Over time, I believe that institutions that conduct exams evolved a bill or rights between institution and applicant. Exams are used, but they are administered consistently and fairly, the topic and subject matter will not be a huge surprise (questions are kept secret of course, but their nature should be predictable). People who pass or fail will receive an answer, and often get a score and feedback to know how they did. The people who grade the exams are provided training to ensure they evaluate people fairly, consistently, and without bias, and are expert in their field.
Now, I know that universities, medical boards, bar exam reviewers, don't always live up to this, but it is the intended goal. I believe that our field, software, subjects people to exams repeatedly, but without any of the bill or rights I descried above.
For example, how do you grade? Are you qualified to assess general intelligence? Can you understand that someone might not be eager to submit to your general assessment of their intelligence?
Now, if you want to say nobody is entitled to a job, I certainly agree. But the industry seems to think it's entitled to workers, and they just seem utterly blind to how off putting their practices are. And in an industry with concerns about gender, racial, and age discrimination, double secret assessments of people's general intelligence by interviewers of unknown qualifications is about the worst thing you can do.
This is fairly common, many people describe exams to be among the more stressful events in their lives.
Interviews are always exams, whether it's writing code on a whiteboard or trying to figure out what the interviewer wants to hear about where you see yourself in 5 years.
In a way, yes. But by blurring this distinction, you make it impossible to distinguish between an hour at the whiteboard solving a data structures problem, and an hour answering questions like "tell me about a time you overcame a challenge and what you learned".
Because people outside our field experience interviews in the second way more often, I think they don't fully understand what goes on in google style interviews. This is why I would call what we go through "Exam-style interviews", drawing a distinction between the two - even though I agree with you that they have elements in common.
