Search

The differences in raw scores between two versions of the Standard Progressive Matrices Test (SPM) and differences in available norms for one version and the currently studied population were investigated. The SPM was originally designed to measure the general factor of intelligence unrestrained by language. Scores for the untimed SPM version published in Britain in 1983 (SPM British) and the timed SPM version published by the Australian Council for Educational Research (SPM-ACER) were compared for 107 male corrective trainees (Maori, European, and Pacific Islanders) in New Zealand ranging in age from 16 years to 20 years 4 months. Norms for the SPM-ACER for Australian prison inmates and norms for the SPM British for children in New Zealand aged 8 to 16 years were compared for use for the New Zealand offenders. Both forms of the SPM were found equivalent in relation to use by the population sampled in this study, but the two sets of norms were not equivalent. It is suggested that the SPM-ACER should be abandoned for the study sample because the norms appeared to grossly inflate the performance levels of these corrective trainees. The use of the SPM British and New Zealand norms was recommended. Instructions for the comparative testing are included. (SLD)

A New Zealand pilot study examined Tests of Reading Comprehension (TORCH) scores compared to PAT: Reading Comprehension scores and compared with teacher ratings. TORCH is a reading test package published in 1987 by the Australian Council for Educational Research. It consists of 14 untimed passages intended to assess the extent to which readers in the Standard 2 to Form 5 age range are able to obtain meaning from text. It is based on a modified cloze format in which students are presented with a passage of text, and a retelling of the passage in different words. Approximately 350 pupils from Standard 3, Form 1, and Form 3 were tested on a TORCH passage and the appropriate level of the PAT. Teachers also rated their pupils' reading comprehension on a 1-5 scale. The study looked at overall correlations between the tests, and between teacher ratings and each of the tests, as well as correlations by gender, socioeconomic status, and ethnic group. Overall, results indicated that pupils' scores on TORCH correlated well with PAT and with teacher ratings. Since teacher ratings are based on standard classroom practice, these results present an encouraging picture of the validity of TORCH for the New Zealand classroom. For Maori and Pacific Island groups, correlations with both teacher ratings and the PAT were generally the same or higher than correlations for the European group. (Eleven tables of data are included.) (SR)