A pilot validation study of the OLSET Picture-Language Vocabulary Test.

This study was aimed at establishing the psychometric properties of the OLSET Picture-Language Vocabulary Level 1 Test (OPLVLT) which is a group test developed in South Africa for children in the lower primary school. The test has been used in a number of studies conducted in second language classro...

Full description

Bibliographic Details
Main Author: Malope, Nomxolisi Sindiswa
Format: Others
Language:en
Published: 2011
Online Access:http://hdl.handle.net/10539/9616
Description
Summary:This study was aimed at establishing the psychometric properties of the OLSET Picture-Language Vocabulary Level 1 Test (OPLVLT) which is a group test developed in South Africa for children in the lower primary school. The test has been used in a number of studies conducted in second language classrooms, but little work has been conducted on its standardisation. The standardisation sample was a sample of convenience, comprising a hundred and twenty (120) Grade One Learners drawn from three schools in Johannesburg. Most children were second language speakers of English, and were aged between 6 and 11 years of age. Reliability of the OPLVLT was established by calculating Cronbach alpha which yielded a value of 0.86. Each of the 36 Items had alpha‘s ranging between 0.83 and 0.85 when deleted and thus they were all retained for the test. The construct validity of the OPLVLT was then investigated by correlating scores on the test with scores on an individual test (the Peabody Picture Vocabulary Test PPVT-R) and another group test (the Metropolitan School Language and Listening subtest MRT). These instruments have been widely used internationally, and have well-established reliability and construct validity. The correlation between the OPLVLT and PPVT was 0.17. The correlation between the OPLVLT and MRT was 0.29. Given the low relationship between the tests, an overall reliability estimate for the construct validity analysis was calculated. The Cronbach‘s alpha for this analysis was 0.57, indicating a low level of internal consistency between the different tests chosen for use in the construct validation. Given its high internal reliability as a test, but its low level of correlation with other tests of language ability, the overall conclusion from this pilot study was that the OLSET Picture-Language Vocabulary Level 1 Test was a reliable test which was not highly correlated with either the Peabody Picture Vocabulary subtest or the Metropolitan School Language and Listening subtest in this particular sample. The high internal reliability indices would suggest that the OPLVLT has potential as a psychometric instrument.