Psychometric evaluation of the Kaufman Assessment Battery for Children, Second Edition (KABC-II) in rural South Africa

Mitchell, Joanie Marie (2015-12)

Thesis (MA)--Stellenbosch University, 2015

Thesis

ENGLISH ABSTRACT : Millions of children living in low- and middle-income countries (LMICs) are not fulfilling their cognitive potential. Early interventions may aid in alleviating the effects of risk factors and thus promote cognitive development. To determine the effectiveness of these interventions, accurate measures of cognitive development are necessary. Measuring the intelligence of children in LMICs has often been contentious, as traditional intelligence quotient (IQ) assessments have largely been used. Traditional IQ assessments measure acquired knowledge that is known to be biased towards school exposure. Recently, given improvements in our understanding of the human brain, there has been a shift towards using assessments that measure cognitive processing, i.e. the skill-set necessary to solve tasks. Given their ability to measure cognitive potential to learn, rather than acquired learning, these assessments are deemed more culture fair. The KABC-II is an example of such an assessment and is increasingly used in LMICs. The KABC-II consists of four scale indexes evaluating performance (Sequential, Learning, Simultaneous and Planning) and has often been translated and adapted for these contexts. However, studies examining the psychometric properties of these adapted assessments are sparse. The primary aim of this research was to evaluate the psychometric properties of a translated and adapted version of the KABC-II in a sample of healthy, isiZulu-speaking children of primary school-age in rural South Africa. The evaluation was achieved through two objectives; firstly by evaluating performance on the adapted KABC-II and secondly through examining whether the underlying structure of the KABC-II was maintained. Using data from a sub-sample of 382 children (part of a larger project), performance on the KABC-II was evaluated using descriptive statistics. The underlying factor structure of the battery was examined using confirmatory and exploratory factor analyses. The confirmatory factor analysis showed two models representing plausible structures of the KABC-II which were good fits to the data. One model included an overall factor (as in the original KABC-II) and the other excluded an overall factor. Within the model including the overall factor, correlations between two of the four scale indexes (Simultaneous and Planning) were shown to be high. This strong association was confirmed in the exploratory factor analysis, where the subtests of these scale indexes emerged as one factor. Performances on these scale indexes were the weakest, with the Planning Scale subtests having the lowest means. The low means may indicate that the Planning Scale might not be optimally culturally appropriate in this context, but could alternatively suggest a deficit in executive functioning skills required to complete these tasks. A lack of executive functioning skills may be due to the presence of biological and psychosocial risk factors, including maternal education. In conclusion, the KABC-II was an appropriate assessment to use in this rural Zulu context, as it was appropriately translated, adapted and piloted. Future research should aim to establish the strength of the overall factor and the appropriateness of the Planning Scale subtests in LMICs. Further research should be done to investigate the effects of biological and psychosocial risk factors on cognitive development. Such studies may inform interventions which could lead to improvements in the cognitive development of children in rural South Africa and other LMICs.

AFRIKAANSE OPSOMMING : Miljoene kinders in lae- en middelinkomste lande (LMICs) bereik nie hulle kognitiewe potensiaal nie. Vroeë intervensies kan moontlik die gevolge van risikofaktore verlig en derhalwe kognitiewe ontwikkeling bevorder. Om die doeltreffendheid van sulke intervensies te bepaal, word akkurate metings van kognitiewe ontwikkeling benodig. Meting van intelligensie van kinders in LMICs is dikwels omstrede, omrede tradisionele intelligensie kwosient (I.K.) assesserings grotendeels gebruik word. Tradisionele I.K. assesserings meet kennis wat bevooroordeeld is tot skoolblootstelling. Onlangs, gegee verbeteringe in ons begrip van die menslike brein, het daar ‘n verskuiwing plaasgevind na die gebruik van assesserings wat kognitiewe verwerking meet, m.a.w. meting van die vaardighede benodig om take op te los. Gegewe hul vermoë kognitiewe potensiaal om te leer, eerder as verkry leer te meet, is hierdie assesserings word as meer kultuurbillik beskou. Die KABC-II is ‘n voorbeeld van hierdie tipe assessering en word al hoe meer in LMICs gebruik. Dit bestaan uit vier skaalindekse vir prestasie evaluering (naamlik Sequential, Learning, Simltaneous en Planning) en word dikwels vertaal en aangepas vir gebuik in hierdie konteks. Studies wat die psigometriese eienskappe van hierdie assesserings ondersoek, is egter skaars. Die primêre doel van hierdie navorsing was evaluering van die psigometriese eienskappe van ‘n vertaalde en aangepaste weergawe van die KABC-II onder ‘n groep gesonde, Zulusprekende kinders van laerskoolouderdom in landelike Suid-Afrika. Die evaluering is bereik deur twee doelwitte; eerstens deur die evaluering van prestasie op die aangepaste KABC-II en tweedens deur te bepaal of die onderliggende strukture van die KABC-II gehandhaaf is. Data van ‘n sub-groep van 382 kinders (deel van ‘n groter projek) is gebruik om prestasie op die KABC-II te evalueer met behulp van beskrywende statistiek. Die onderliggende faktorstruktuur van die toetse is ondersoek deur bevestigende en ondersoekende faktorontledings. Die bevestigende faktorontledings het twee geloofwaardige modelstrukture van die KABC-II getoon, wat goed by die data gepas het. Een model het ‘n algehele faktor ingesluit (soos in die oorspronklike KABC-II), terwyl die ander een ‘n algehele faktor uitgesluit het. Die model wat die algehele faktor ingesluit het, het hoë korrelasies getoon tussen twee van die vier skaalindekse, naamlik Simultaneous en Planning. Hierdie sterk assosiasie is in die ondersoekende faktorontleding bevestig, waar die subtoetse van hiedie skale as een faktor na vore gekom het. Prestasie op hierdie skaalindekse was die swakste, met die laagste gemiddeld gevind by die Planning Skaal subtoetse. Die lae gemiddeld mag moontlik daarop dui dat die Planning Skaal in hierdie konteks nie optimaal kultureel gepas is nie, maar alternatiewelik mag dit ook dui op ‘n gebrek aan die nodige uitvoerende funksionering om hierdie take te voltooi. ‘n Gebrek aan uitvoerende funksionering mag te wyte wees aan die teenwoordigheid van biologiese en psigososiale risikofaktore, insluitend moeder-onderwys. Die slotsom is bereik dat die KABC-II was‘n gepaste assessering vir gebruik in hierdie landelike Zulu konteks, soos dit was gepas vertaal, aangepas en geloods word. Toekomstige navorsing moet poog om die sterkte van die algehele faktor en die toepaslikheid van die Planning Skaal subtoetse in LMICs te bepaal. Verdere navorsing word ook benodig om die gevolge van biologiese en psigososiale risikofaktore op kognitiewe ontwikkeling te ondersoek. Sulke studies mag intervensies toelig wat kan lei tot verbeterde kognitiewe ontwikkeling van kinders in landelike Suid-Afrika en ander LMICs. Die bevestigende faktorontledings het twee geloofwaardige modelstrukture van die KABC-II getoon, wat goed by die data gepas het. Een model het ‘n algehele faktor ingesluit (soos in die oorspronklike KABC-II), terwyl die ander een ‘n algehele faktor uitgesluit het. Die model wat die algehele faktor ingesluit het, het hoë korrelasies getoon tussen twee van die vier skaalindekse, naamlik Simultaneous en Planning. Hierdie sterk assosiasie is in die ondersoekende faktorontleding bevestig, waar die subtoetse van hiedie skale as een faktor na vore gekom het. Prestasie op hierdie skaalindekse was die swakste, met die laagste gemiddeld gevind by die Planning Skaal subtoetse. Die lae gemiddeld mag moontlik daarop dui dat die Planning Skaal in hierdie konteks nie optimaal kultureel gepas is nie, maar alternatiewelik mag dit ook dui op ‘n gebrek aan die nodige uitvoerende funksionering om hierdie take te voltooi. ‘n Gebrek aan uitvoerende funksionering mag te wyte wees aan die teenwoordigheid van biologiese en psigososiale risikofaktore, insluitend moeder-onderwys. Die slotsom is bereik dat die KABC-II was‘n gepaste assessering vir gebruik in hierdie landelike Zulu konteks, soos dit was gepas vertaal, aangepas en geloods word. Toekomstige navorsing moet poog om die sterkte van die algehele faktor en die toepaslikheid van die Planning Skaal subtoetse in LMICs te bepaal. Verdere navorsing word ook benodig om die gevolge van biologiese en psigososiale risikofaktore op kognitiewe ontwikkeling te ondersoek. Sulke studies mag intervensies toelig wat kan lei tot verbeterde kognitiewe ontwikkeling van kinders in landelike Suid-Afrika en ander LMICs.

Please refer to this item in SUNScholar by using the following persistent URL: http://hdl.handle.net/10019.1/98043
This item appears in the following collections: