Gifted Issues Discussion homepage Forums Identification, Testing & Assessment .

.

What is the average picture concepts score of gifted kids?

I sort of made my own schema of intelligence with two binary options for two vectors, creating four different categories of higher level reasoning:

Verbal vs Nonverbal
This does not refer to auditory vs visual stimulus but rather whether or not crystalized intelligence (even to a basic degree) is a requirement for succeeding at the task. Picture concepts is a verbal task with a visual stimulus because it relies on simple general knowledge to express categorical reasoning. VFI/VR is fluid reasoning with a crystalized knowledge threshold for expressing the former.

Reasoning vs Memory:
Verbal knowledge and visuospatial skills are vectors of memory (contextual verbal long-term retrieval and decontextual figural working memory, respectively) that correlate much more with verbal and nonverbal reasoning than contextual verbal short-term memory with a visual stimulus (picture span) and decontextual nonverbal short-term memory and working memory with an auditory stimulus (digit span and letter-number sequencing).

Nonverbal reasoning is assessed by the WISC-V FRI in the forms of matrix reasoning (untimed fluid reasoning) and figure weights (fluid reasoning with a short-term memory and quantitative reasoning threshold).

I'm wondering if older WISC-V examinees (or just kids who know algebra) use equations to solve figure weights problems. I for one have used verbal mediation on visuospatial tasks since I was 6/7 and modified my strategies throughout the years.

I'm wondering if some very verbally gifted kids tend to show sometimes even AWMI/WISC-IV WMI performance higher than even fluid reasoning tasks of the PRI (including the verbally loaded picture concepts) when they are young (like ages 6/7) because they don't understand time limits very well or don't use verbal mediation yet. Their FRIs might rise and WMIs fall as the latter becomes more about memory than understanding relatively complex (for the 6/7 y olds) verbal procedures and verbal mediation is used more on the former. Visuospatial skills tend to stay the same. This is all anecdotal, though.

Also, primarily verbally gifted kids with also high FRIs tend to score higher on arithmetic than figure weights, possibly because the former uses an auditory stimulus.

You have a lot of thoughtful questions here, some of which I may inadvertently not comment on or not get to this round...remind me if something important is missed. Quantitative notes below are all regarding the most current revisions (WISC-V, WAIS-IV).

With regard to one of your initial questions, the correlation coefficient between Si and PC is .39, which is among the lowest correlations for reasoning subtests (as distinct from cognitive proficiency subtests). PC also has the lowest correlation to any of the reasoning Index scores of any reasoning subtest (between .23 and .48, vs mostly .5ish to .8ish for the other subtests). That, of course, is why it is not a core subtest (particularly because of its weaker correlation to the FSIQ). It has the weakest correlation of any non-processing speed subtest.

The NAGC data on performance in GT learners found that the mean PC score was 13.6. That was the lowest mean score of all of the reasoning subtests. FWIW, my experience has been that GT examinees sometimes experience some artifactual penalties on PC because they find other relationships among the stimulus items that are legitimate and logically consistent, but not normative. There seems to be an element of reading "what would typical learners say" to scoring well on it. I've even gone back on testing of limits (clinical inquiry) and prefaced the second exposure with statements of that nature.

( Table 1, using extended norms).

There are various positions on whether Co is verbal knowledge or verbal reasoning. One can make an argument for either. Its correlations are nearly identical to Si (.59) and Vc (.60), which are generally considered the less ambiguous representatives of verbal reasoning and verbal knowledge. Its correlations on the WAIS-IV are slightly more tilted toward Vc, but still not by a ton. Co is also heavily culture-bound, which I would postulate may be part of its uncertain assignment to knowledge or reasoning. If a learner is fluent in the culture, I would suggest that it might be a better measure of contextualized social reasoning (as it is often interpreted to be), but if an examinee is not, then clearly acquired knowledge will significantly impact performance, in which case one might better describe it as an indicator of knowledge of social conventions. (I have used either or both of these interpretive lenses, depending on the larger context of the student.)

Notably, Si and Co were the two highest mean subtest scores in the NAGC's GT sample. (Both 18.1).

I'm wondering if more time spent on item design could fix the artificial attenuation of gifted kids' picture concepts scores.

When I make analogies or matrices or number series I try to eliminate ambiguity, but official test developers have other constraints (like biases) and finite time.

It would be helpful if the "what would typical learners say" was viewed as standard administration, because it's not a hint or change in difficulty.

Your knowledge on the comprehension subtest is interesting.

Thank you for answering my questions.

These next three things are unrelated:

If I were to make a GAI-type index it would have information, vocabulary, similarities, matrix reasoning, figure weights, picture concepts, and visual puzzles.

I wonder why visual puzzles is not in the GAI, EGAI, or FSIQ like block design.

I might get an opportunity to take the WISC-V soon and want to ask some questions on which subtests are spoiled or inflated and of the unspoiled ones which ones would be relevant for me to take.

Perhaps PC could be improved for GT learners (it has the lowest reliability scores of any reasoning subtest for GT, and GT is also the subgroup with the lowest reliability for PC), but since it didn't turn out to be a preferred subtest for the core, it probably wasn't worthwhile for Pearson to invest more time/money into tweaking it.

The simplest answer to why the EGAI has BD and not VP is that it was derived from the original standardization sample for the GAI, plus the two extra verbal comprehension subtests. It's not entirely clear why BD is in the GAI/FSIQ rather than VP, as their psychometrics are similar, but I suspect it is because BD allows the clinician to observe the externalized process of solution development, and it historically is a good rapport-building subtest with which to open testing.

I would not describe any of the subtests as spoiled or inflated. It's more a question of how one interprets the scores than of tossing some of them out (except in the case of those where standardized conditions were broken during administration). Also, the highest-reliability index scores do not allow substitutions (one is allowed for the FSIQ, but that's it), so picking and choosing subtests would result in less--and less reliable--data, rather than more. The more relevant factors for an imminent WISC in your case would be:

1. to make sure that the examiner knows beforehand that you and your parents are interested in the Extended Norms, if they are applicable (so that the examiner does not prematurely discontinue any subtests after reaching a max scaled score, and so that they are more likely to calculate ExN scores if needed);

2. to discuss whether the EGAI will have value to you; and

3. to include your specific concerns about your executive function development and possible ADHD in the intake, so that they can select appropriate measures.

I'm interested in extended norms if they apply.

I'll ask about the EGAI.

Is the Wisconsin Card Sorting Test possibly accentuated because I've taken an free online version (not from a psychologist)?

I wouldn't be overly concerned about inflated scores from free online tests, since they typically do not have the actual test items (which are copyrighted), but the most straightforward way to manage those types of validity questions is to tell your examiner upfront about all of the unofficial and official testing you've done in the past.

Most (legitimate) tests have a valid retest interval (meaning how long you should wait between retests of the identical measure) of at least six months (for academic-type tests), 12 months (many neuropsych tests) or 24 months (for cognitive tests).

The main thing you can do at this point is to stop taking online tests, and to be honest and thorough with your examiner.

I haven't taken online tests in more than six months and planned on telling the examiner during the rapport-establishment stage in our first meeting.