Tuesday, April 05, 2005

WISC-III/WJ III CHC study: Shameless plug

Yes.... a shameless (but, non-commercial) plug. Be sure to read my stated conflict of interest statements on this blog's home page (WJ III coauther)

The following Wechsler/WJ3 cross-battery confirmatory factor analysis article was just published in the most recent edition of School Psychology Quarterly. Contact the journal to subscribe and get your copy now.....before supplies are depleted. Don't be the last one on your block to read it!

Phelps, L., McGrew, K. S., Knopik, S. N., & Ford, L. (2005). The general (g), broad, and narrow CHC stratum characteristics of the WJ III and WISC-III tests: A confirmatory cross-battery investigation. School Psychology Quarterly, 20(1), 66-88.

Abstract: One hundred, forty-eight randomly selected children (grades three-five) were administered the WISC-III, WJ III Tests of Cognitive Abilities, WJ III Tests of Achievement, and seven research tests selected from the WJ III Diagnostic Supplement. The validity of the existing WISC-III and WJ III broad Cattell-Horn-Carroll (CHC) test classifications was investigated via the application of CHC-organized, broad-factor, cross-battery confirmatory factor analyses (CFA). Likewise, the validity of the WISC-III and WJ III narrow CHC ability classifications was investigated via the evaluation of a three-stratum hierarchical (narrow+broad+g) CHC CFA cross-battery model. The Tucker-Lewis Index, the Comparison Fit Index, and the Root Mean Square Error of Approximation evaluated the fit for the resulting models. All statistical values indicated good to excellent fit.

Select study highlights
  • This study represented the first-ever three-stratum, CFA CHC-based analyses of a jointWechsler/Woodcock data set.
  • At the broad CHC factor level, prior WISC-III/WJ III test classifcations where supported.At the broad factor level, the WISC-III included a greater proportion of Gv tests (Picture Completion, Picture Arrangement, Block Design, Object Assembly) than the WJ III (Spatial Relations and Block Rotation).
  • WISC-III Block Design test was the strongest single indicator of Gv (and an integrated narrow spatial relations/visualization [SR/Vz] ability factor). Consistent with prior CHC-based Wechsler studies, the results continued to NOT support the interpretation of Wechsler Block Design as a measure “reasoning” (Gf). WISC-III Object Assembly also appeared to be a strong indicator of Gv, although its interpretation at the narrow ability level was indeterminate.
  • Relatively low Gv loadings for WISC-III Picture Completion and Picture Arrangement, plus an additional secondary Gc loading for Picture Completion, reinforced prior research that, in the context of CHC-defined assessments, the use of these two WISC-III tests is discouraged as scores on these two tests may confound ability profile interpretation of Gv composites.
  • The two WJ III tests (Visual Matching, Decision Speed) and two WISC-III speeded tests (Coding, Symbol Search) were all identified as strong indicators of broad Gs, but differentiation at the narrow Gs ability level was not supported.
  • At variance from prior suggested interpretations of WISC-III Arithmetic as an indicator of quantitative reasoning (under Gf), when included together with select WJ III math achievement measures, WISC-III Arithmetic was found to be a mixed measure of Gq (quantitative knowledge) and Gs.
  • At the narrow ability level, support was found for interpreation of the WJ III Numbers Reversed and Auditory Working Memory Tests as measures of working memory (MW), whereas WJ III Memory for Words and Memory for Sentences are best interpreted as measures of memory span (MS). In contrast to apriori hypothesis, the WISC-III was found to be a measure of MS and not MS and MW.
  • The WJ III Planning test was found to be less of an indicator of Gv as reported in previous studies, and, instead, loaded primarily on Gf. The possible involvement of working memory, an ability linked to Gf and g, was hypothesized as a possible reason.
  • Consistent with the extant Wechsler CHC CFA research, the WJ III was found to provide valid measures of three broad CHC domains (Gf, Glr, Ga) not measued by the WISC-III.

1 comment:

Anonymous said...

Definitely an article I would enjoy reading. Okay question:
Arithmetic a mixed measure of Gq AND Gs? Gs, not Gsm? Please advise if that's a typo or if it somehow loaded on processing speed?

Planning loading on Gf makes intuitive sense I think (no pun intended.) There seems to be a link-executive processes related to BOTH concept formation and planning. Although planning is not concept formation-it does seem to respond to basic heuristic use when applied well.