Abstract
M.Ed.
The aim of this study was to apply a Rasch Analysis to the Verbal Intelligence
Quotient Eight (VIQ 8) scale of an isiZulu translation of the Junior South African
Individual Scales (JSAIS) to determine whether it would be a reliable and valid
means of assessing the verbal cognitive functioning and development of learners at a
school in the South African context. The study formed part of the cognitive
assessment leg of a wider research project being undertaken by the University of
Johannesburg’s Institute for Childhood Education (UJICE).
The JSAIS is still a widely used instrument by Psychologists in the South African
context despite the last norming having taken place in 1984 and the norms not being
representative of the South African population. To establish reliability and validity of
formal assessment tools in a multi-cultural and multilingual society such as that of
South Africa, context needs to be taken into account and existing tests need to be
scrutinized carefully for cultural bias.
The Rasch analysis undertaken in this study provides an analysis of item difficulty in
relation to ability of the testees in the sample group, which consisted of 26 learners
in the isiZulu class at a laboratory school in Soweto where research is being
conducted under the auspices of the Soweto Panel Research Programme (SPRP).
This study formed part of the pilot phase of the Programme.
The analysis of the results suggested that on the whole the three sub-tests fitted the
Rasch Model well. The Vocabulary sub-test could be utilized effectively with no
further adaptation, whereas the Picture Riddles sub-test could be adapted by
adjusting the spread of the items. The Word Association sub-test could possibly be
adapted by including a picture stimulus. Apart from these possible adaptations to the
sub-tests as well as omission of items and persons that did not fit the model, it is
recommended that: standardized testing procedures be adhered to, increased sample
sizes be used, testing of the learners take place in the same month and the age of
testees be taken into account in further studies. Other studies into the development of
the traits that the test taps into of learners in this age group, the equivalence of the
isiZulu translation and assumptions regarding “mother-tongue” are also
vi
recommended. Finally it is recommended that testees who performed strongly or
weakly and who did not fit the model due to unwanted sources of variance be
observed further.
The Rasch analysis findings suggest that the JSAIS has some strong psychometric
properties and that adaptation, translation and renorming of the test would be a
worthwhile project to consider in the near future. Further to this, concerning the use
of the test in the longitudinal panel research project, findings suggest that the VIQ 8
scale of the test is likely to be a fairly reliable and valid means of assessing the
cognitive verbal functioning and development of the learners at the school over time,
provided that it forms one part of a systemic multi-method assessment approach
which takes context and development of the learners into account.