Paper discussing economic rationale of IQ-based hiring practices

TL;DR

I'm looking for a paper that explicitly quantifies the ROI an employer would likely derive if personnel selection involved intelligence testing and only those +2 SD in cognitive ability were hired.

There's a correlation between IQ and job performance, ranging from .20s for low complexity jobs and .50 for high complexity jobs. Hardly unity (r = 1.00) but regardless, it's still considered to be the best predictor of job performance.

A number of years ago as an undergrad I read a paper discussing the economic rationale of an employer administering an IQ test and only selecting candidates who scored +2 SD above the mean. From the pool of candidates remaining, the employer would then interview for the best candidate as per usual. The author's thinking was along the following lines:

• Suppose you had a group of employees who you were paying an annual salary of \$100,000 and their productivity was normally distributed between say, \$100,000 - \$300,000 (mean =$200,000).

• By selecting those who are +2 SD in cognitive ability, the employer would increase the chances that those who are hired will return them closer to $300,000 per year. • Total economic gain = Number of employees x Increased return ($) x Duration of tenure

This topic is relevant to a publication I'm preparing and I've been racking my brain trying to find it. For the life of me I simply cannot find it.

Does this ring a bell for anyone? Ideally, I'd like the specific one that I'm mentioning, but failing that, something in a similar vein. Essentially: suppose you hired only those >130 IQ -- regardless of the rightfulness/wrongfulness of doing so, what I want to know is what the \$ amount benefit to the employer would be.

• Just to clarify, you're looking for a specific paper, or literature in general? For example Linda Gottfredson has written a fair bit on intelligence and work place success. For example, her well known paper "Why g matters". – Eff Jul 16 '18 at 13:39
• Yep, I've got that paper. Ideally, I'm after that actual paper, but if not, a paper that attempts to quantify this personnel selection method e.g. x% more output per employee per year – faustus Jul 16 '18 at 13:45
• Possibly off-topic here, as it pertains more to economy than psychology? – Steven Jeuris Jul 16 '18 at 20:19
• Even if such a paper exists, I have doubts in its real-world relevance. Let's say you only want to hire a carpenter with IQ > 130. How many carpenters would have to reject? Hundreds maybe? What's the cost for interviewing all those and not getting the job done in the meantime? If you can't remember a specific field in which this alleged paper's employees were working, it's probably fair to say the paper is bollocks. – Fizz Jul 17 '18 at 9:05
• @Fizz let me rephrase: rather than IQ >130, you select the top 2% of any applicants. The paper had to do with white collar roles in organisations where I presume they have an HR department. – faustus Jul 17 '18 at 9:10

I believe I know which paper you're referring to. My guess is the widely cited Schmidt & Hunter (1998) paper. The paper summarizes 85 years of practical and theoretical implications of research in personnel selection. To see that they address what you're asking for, here is a quote from the paper

If a superior worker is defined as one whose performance (output) is at the 84th percentile (that is, 1 SD above the mean), then a superior worker in a lower level job produces 19% more output than an average worker, a superior skilled worker produces 32% more output than the average skilled worker, and a superior manager or professional produces output 48% above the average for those jobs. These differences are large and they indicate that the payoff from using valid hiring methods to predict later job performance is quite large.

Then later they state

At one extreme, if an organization must hire all who apply for the job, no hiring procedure has any practical value. At the other extreme, if the organization has the luxury of hiring only the top scoring 1%, the practical value of gains from selection per person hired will be extremely large. But few organizations can afford to reject 99% of all job applicants.

However, a more recent paper on the subject has been published. That's the Schmidt, Oh & Shaffer (2016), which summarizes 100 years of practical and theoretical implications of research in personnel selection. This paper is an update on the former, so you may as well just look only at this latest one.

• perfect. 4536 citations :) – faustus Jul 20 '18 at 22:22
• I take thar paper with a giant boulder of salt skeptics.stackexchange.com/questions/40059/… And for more than one reason faculty.cas.usf.edu/mbrannick/papers/conf/siopk.htm – Fizz Sep 2 '18 at 23:29
• And I managed to re-find the paper that made me really doubt S&H (without asking a ref-request question here, ha!): tandfonline.com/doi/pdf/10.1080/10888691.2014.983635 – Fizz Sep 3 '18 at 0:05
• @Fizz I know of Ken Richardson. However, no, I don't find his arguments convincing. Many of his arguments are against corrections in meta-analyses that attempt to correct for weakly measured data. The problem is that, even if you completely showed that S&H paper was completely untruthworthy for whatever reasons, these criticisms have been dealt with in the wider literature at large. There are many studies that have great data that still finds substantial correlations between intelligence and performance measures. – Eff Sep 3 '18 at 7:42
• @Fizz I generally refrain from attacking researchers personally, and instead discuss their arguments. Unfortunately I feel it's necessary in this case. Ken Richardson and Sarah Norgate are completely outside the mainstream of intelligence research. And why? This is obvious. Sarah's university profile: "Sarah’s research expertise lies mainly in applying developmental psychology to topical societal issues..." and "Sarah is committed to social inclusion and, to this end, her published work has involved critiques of work in behaviour genetics..." She is an activist first, not a scientist first. – Eff Sep 3 '18 at 7:49