A consistent trend I have seen is to hire people based on qualities that are not connected to the job. A few of the many possible examples are the following
All of these might be useful in some jobs. If you want to hire a model, for example, attractiveness is quite important. But for a huge number of jobs these traits are useless. Your math ability does not affect most sales jobs, and yet these traits seem to be valued regardless.
To go deeper into an example, I would like to look at intelligence. Most people would agree that it plays some role in almost any job, although considerably less so in some jobs when compared to others. However, people focus on intelligence so much that some jobs have IQ tests built into the interview.
Is this a good use of an organization’s time? Is it a good way to predict job performance? Studies have shown that “the relation of IQ to exceptional performance is rather frail in many arenas. For scientists, engineers, and medical doctors the correlations between ability measures and occupational success are only around 0.2, accounting for only 4% of the variance.”
A better option might be to look at metrics that people tend to undervalue, yet are exceptionally important, such as conscientiousness, value alignment with the organization, and social skills. Indeed, research has found that information about people's personalities gleaned from Facebook profiles is a better predictor of job performance than IQ tests.
An even better alternative would be to replace vague, disconnected metrics with a test that is as similar to the real work that the applicants would do as possible. Such simulation tests have been shown to be much better at predicting someone’s true job performance. Although there will still be some bias, they seem like a much stronger way of evaluation.