Comparison of scores from the Wechsler Adult Intelligence Scale–IV and the Woodcock Johnson III in an Australian university sample

Objective: This study investigates, in an Australian university student sample, the score equivalence of the Wechsler Adult Intelligence Scale-IV (WAIS-IV) norms for which are based on the US standardization sample versus the Woodcock Johnson III (WJ III) cognitive test battery, normed for the Australian population based on a standardisation sample aged between four and twenty-one. Mean IQ scores in the two instruments were compared in sample scores of all and those restricted to 21 years or younger. Additionally, corresponding broad ability scores were compared. Method: Tests were assigned in blocked random order to 164 (141 aged 21 years or younger) participants (M = 20.38, Mdn = 19, SD = 5.19). Thirty-eight (32 aged 21 years or younger) undertook both tests. Scores from all, and the subset aged 21 years or less were analysed. Results: The mean WAIS-IV FSIQ were significantly higher than mean WJ III GIA in both independent and repeated measures analyses. Effect sizes, measured with population SD of 15, were d = 0.34 and d = 0.53 in the independent and repeated measures study. Conclusions: Considering the potential impact of the observed effect sizes in high-stakes decision for the Australian population, we appeal for further research on local norming. KEY POINTS What is already known on this topic: When used in Australia, the WAIS-IV scores are based on the US standardization data, while the WJ III has been normed for the Australian population. Local and current norming of tests is essential for valid and fair use of a cognitive abilities test. WAIS-IV scores are routinely used to make high-stakes decisions in Australia. What this paper adds: Empirical evidence that the current WAIS-IV scores and WJ III scores in Australia are not equivalent in a sample of university students. If the observed effect sizes were to hold in the critical cut-off regions used for making high-stakes decision, a substantial number of eligible beneficiaries will be affected. Though the samples in this study are restricted, the partial random sampling, consistency, and effect size warrant more comprehensive study.