Appletcollection Vertical Menu java applet, Copyright 2003 GD

Melissa Farrall's WISC-III  v. WISC-IV Comparison

Myth of the WISC-III/WISC-IV Retest: the Apples and Oranges Effect 

According to a study in which 244 children aged 6-16 were administered the WISC-IV and the WISC-III in counterbalanced order, the WISC-IV correlates well with the WISC-III.[1]  The table on page 65 indicates that those taking the WISC-IV would receive scores that are slightly lower than those earned on the WISC-III.

            Direct experience with the WISC-IV has suggested that the pro forma prediction of WISC-IV scores based upon the WISC-III for many children does not work. This difficulty is largely due to what has been coined “the apples and oranges effect.” The WISC-IV and the WISC-III are different in their composition, and they cannot be compared easily on the surface in a meaningful fashion.

These differences between the WISC-IV and the WISC-III become more apparent when assessing the abilities of children with learning disabilities, i.e. those who will, by definition, have greater variation in their skill levels.  Some children, such as those with weaknesses in acquired knowledge (Information and Arithmetic) may earn scores on the WISC-IV that are significantly higher than the WISC-III due to the decreased focus of the WISC-IV on knowledge that is generally acquired in school. On the other hand, given the increased WISC-IV emphasis on the ability to solve novel problems, children with pronounced weaknesses in nonverbal fluid reasoning (Matrix Reasoning and Picture Concepts) may be dismayed to find that their IQ scores have dropped. [additional information to be provided in chart] For the evaluator, these differences present a challenge with respect to how to interpret WISC IQ scores over time. 

            What is an evaluator to do? Certainly, the temptation to compare a WISC-IV Full Scale IQ with that of the WISC-III is overwhelming, and some teams are now in the awkward position of predicting achievement in a child who has now been determined to have less intellectual horsepower than he or she did a few grade levels prior. Conversely, other teams may find that they have to explain why it would be appropriate to revise their earlier estimation and increase expectations and programming for children who have now been declared to be more intelligent.

Clearly, we cannot compare apples and oranges, and a WISC is not a WISC is not a WISC. Please see the chart entitled, “Changes in the Composition of the Wechsler Intelligence Scale for Children, Fourth Edition.” The composition of the WISC-IV Full Scale IQ has changed by fifty percent.   Changes in the composition of WISC-IV Verbal Comprehension, Perceptual Reasoning, and Working Memory Scales vary from a not insignificant twenty-five percent to a wholly unrecognizable seventy-four percent; these figures, of course, presume that we have not engage in “subtest roulette.”[2] Only Processing Speed remains intact. 

For those wishing to examine the question of progress over time, the only option is to examine performance with respect to individual subtest scores (and their confidence bands). When there appears to be a decrease in scores, concerned team members are then faced with the prospect of examining why that change has occurred, and ruling out the possibility that medical concerns or nonacademic factors may be compromising a child’s ability to demonstrate what he or she knows.