The use of neuropsychological tests to assess intelligence

Abstract

Objective: We sought to derive a ‘neuropsychological intelligence quotient’ (NIQ) to replace IQ testing in some routine assessments. Method: We administered neuropsychological testing and a seven-subtest short form of the Wechsler Adult Intelligence Scale to a community sample of 394 adults aged 18–96 years. We regressed Wechsler Full Scale IQs (W-FSIQ) on 23 neuropsychological scores and derived an NIQ from 9 measures that explained significant variance in W-FSIQ. We then compared subgroups of 284 healthy and 108 unhealthy participants in NIQ and W-FSIQ to assess criterion validity, correlated NIQ and W-FSIQ scores with education level and independence for activities of daily living to assess convergent validity, and compared validity coefficients for the NIQ with those of ‘hold’ and ‘no-hold’ indices. Results: By design, NIQ and W-FSIQ scores correlated highly (r = .84), and both were higher in healthy participants. The difference was larger for NIQ, which accounted for more variability in activities of daily living. The NIQ and ‘no-hold’ index were better predicted by health status and less predicted by educational status than the ‘hold’ index. Conclusions: We constructed an NIQ that correlates highly with Wechsler FSIQ. Tests required to obtain NIQ are commonly used and can be administered in about 45 min. Validity properties of NIQ and W-FSIQ are similar. The NIQ bore greater resemblance to a ‘no-hold’ than ‘hold’ index. One can obtain a reasonably accurate estimate of current Full Scale IQ without formal intelligence testing from a brief neuropsychological battery.

Keywords:

• Intelligence
• neuropsychology
• biological intelligence
• cognitive function

Notes

1. This was done by subtracting 50 from each T-score, multiplying the difference by 1.5, and adding the product to 100. The mean of these standard scores (not the formula shown in Table 6) is used to compute NIQ.