Gifted Issues Discussion homepage Forums Identification, Testing & Assessment High Working Memory on WISC

My DS7 scored in the "superior" range for processing and the mildly gifted level for working memory. Good scores, but not outstanding. Broad math on WJ III achievement was 178.

I wish I had known about the optional arithmetic subtest on WISC in the working memory section. It's supposed to demonstrated the highest correlation with giftedness (even higher than vci subtests). I would have loved to have known this in advance and opted for the extra subtest.

I suspect DD would have done better on "arithmetic," too, but can they give optional subtests without a specific reason?

The psych that tested DD did it and just indicated on the report that she substituted. I think the NACG suggests that gifted kids tend to do better on arithmetic than digit span because arithmetic is more interesting/meaningful. It would be interesting to see how DS would do on arithmetic. It could be why DD's working memory score is so much better than DS's although if the psych had done digit span on her instead of arithmetic, I think she would have done fine. She can repeat about 5 numbers backwards. DS seems to have more trouble with it.

So I've looked up the criteria for scatter, and found that most definitely, the VCI should be interpreted with caution. That level of scatter (range of 6 scaled score points) is both statistically significant (p<.05) and rare (base rate of 5.5%). The Comprehension score differing from the average VCI score by that much occurs at an even smaller base rate of only 2% in the norm population.

In short, you should definitely discuss additional testing regarding language-based intelligence.

The perceptual subtests are not that divergent. I should point out that Picture Concepts is a little bit different from the other PRI subtests, in that it is 1) more amenable to verbal mediation, and, 2) not particularly spatial.

On another note: if you want to know more about visual memory span, the Wechsler Nonverbal, WRAML-2, and the UNIT all have some measure of visual/spatial memory span.

The psychologist who administered DS's WISC said that the PRI is an area that tends to improve with age. I'm not sure what she is basing this off of, but she has 20+ years experience with gifted kids.

He was 6 1/2 when he took the WISC. Still scored "gifted" in PRI but the score was 10 points lower than verbal.

For what it's worth, when my DS took his private school entrance exam (before K) they said he scored in the 4th percentile for comprehension. I think they were standing at the ready to accommodate a learning disability.

Every subsequent test- WISC, Gates, Iowa, and WJ has measured his comprehension between the 95th and 99th percentile.

My point is that if your child experiences any anxiety, has a tester who doesn't know their stuff, or has never taken any kind of test before, it can seriously affect the score.

That is a sufficient reason. I usually substitute when there is a significant difference between Digit Span and Letter-Number Sequencing, and Arithmetic lines up better with one of them, and then discuss the profile in the narrative. Or if none of them line up, I use the standard two subtests (DS & LNS), and include cautions about the WMI. BTW, typically, the default substitution is Arithmetic for LNS, not for DS. I do know examiners who routinely make this substitution, for all examinees, regardless of suspected giftedness. It can also be interesting to compare the process score for DS forwards and backwards with the rest of the WM subtests (one can obtain scaled scores for forwards/backwards separately).

Do you mean you would do all three tests and use the two scores of the three that are closest to each other? With DD, she decided ahead of time that she would do arithmetic instead of digit span. Possibly because DD told her she likes math in school, or because of what the NAGC says. I thought that testers had to decide ahead of time which tests would be used in scoring.

Do you have any thoughts as to why there would be a big gap between Digit Span and Letter-Number Sequence for DS? It seems like practically the same test. I think after the skull fractures/brain injury his scores were 10 and 14, and then 6 months later 11 and 15 (or something along those lines), so there was the same gap both times he did it.

Not exactly. (You are correct that examiners are supposed to select subtests before administration, based on perceived need. Sometimes you can't, though, such as when one is spoiled during testing.) I would do the original two: ds and lns, and then if they were widely divergent, I would add ar, and then use the two that were most consistent for the Index score, and also include a discussion of the diversity (and thus lowered reliability of Index score) among the subtest scores, with my hypotheses about the explanation. I justify the selection after testing with the same reasoning that says that excessive scatter among the subtests in a cluster makes the Index score invalid. I suppose this is my way of minimizing the impact of a questionable Index score on the IQ, and of creating a better fit between the numbers people see and the clinical interpretation.

WRT your DS, was there any mention of differences between digit forwards and digits reversed in the Digit Span subtest? Forwards is more closely related to short-term rote memory and attention, while Backwards usually tracks better with LNS, as the mental manipulation aspect of it is more similar.

Also, six months out from a TBI, the brain is still healing. Actually, we continue to see improvements in cognitive functions for a couple of years after the trauma.

Thank you. There was nothing the report about forwards vs. backwards. I should ask him to try some numbers myself and see if I notice if he's struggling with one or the other.

One of the cranial nerves in his brain was paralyzed due to blunt force trauma and concussion, and it took about 6 months for the nerve to heal enough for his eyes to become completely aligned. So that sounds about right for recovery from a brain injury. During the first testing (about 4 months after the brain injury), the psych did a whole bunch of other tests and DS had absolutely wretched scores for fine motor ability (Grooved Pegboard), like below the .01 percentile. So we suspect that the brain injury made his preexisting condition (DCD) even worse. Strangley he scored very high on subtests like matrix reasoning even though he probably had some double vision. And block design wasn't bad either but quite a bit lower than the other PR subtests.

Well, the matrix reasoning score makes sense, as the task is more about abstract reasoning and pattern recognition than it is about straight visual skills, although, of course, you have to be able to see the designs to solve them. Likewise, block design has significant motor and motor speed elements, which were probably impacted, and relies on vision to complete, but it also contains spatial reasoning and nonverbal problem solving components that would not necessarily be as compromised by vision issues. The blocks are big enough, and the color patterns distinct enough, that he was probably able to solve the items accurately, with just some points knocked off for speed. Any comments on slow-but-accurate performance, loss of bonus points for speed, or overtime responses? Those are usually tipoffs that the real issue with a relatively lower block design score is motor speed. Sometimes there is also an alternate score reported (it can be generated for any WISC-IV, but not everybody reports it) for block design no bonus points, which is a less-timed version of scoring. There are still time limits for completing each item, but speed is not factored into the scaled score. If you compare the standard bd score and the bd no bonus points score, and they are noticeably different (there are tables for statistical significance), that suggests that motor speed is compromising performance, rather than actual perceptual reasoning. Typically, you do not substitute the alternative block design scoring into the Index scores, but it is valuable clinical information.