Psigometriese eienskappe van die ESSI-lees- en speltoets

In drawing up the ESSI reading and spelling test, four phases were followed. In an earlier article, the development work that took place during the first three phases was discussed. This article focuses on the last phase during which the psychometric characteristics of the ESSI reading and spelling test were investigated. Firstly, the skewedness of the test scores on all the tests (vocabulary lists) was investigated. The results showed that there was a reasonable measure of symmetrical distribution of scores on all the Afrikaans and the English tests. The kurtosis of the scores was also investigated, and with some exceptions, it was clear that the scores had a normal distribution. Attention was also devoted to the standard error of measurement for the tests. Not one of the standard errors of measurement values for the tests was larger than 2. Information on the reliabilities of the tests is also included. Various reliability indices were calculated and all of them exhibited high levels of reliability. Concerning the validity of the tests, attention was devoted to content and criterion-related validity of the tests. It was possible to ensure that the requirements of content validity were met by requesting experienced persons to identify the words for the test. Information on the predictive validity of the tests indicated that all the tests have sound predictive validity. (South African Journal of Education: 2002 22(2): 149-153)