A dissertation submitted to
The Department of Speech Pathology and Audiology
School of Human and Community Development
Faculty of Humanities
University of the Witwatersrand
In fulfilment of the requirements of the degree
Master of Arts in Speech-Pathology
March, 2017 / Purpose: There is a need to develop and refine assessment measures on bilingual
children, since language measures used on monolingual individuals cannot and should
not be directly applied to the bilingual population (Hoff et al., 2012; O’Brien, 2015).
The occurrence of Afrikaans-English bilinguals in South Africa provides a rewarding
area of investigation for the Speech-Language Therapist (SLT) (Penn & Jordaan,
2016), as the Afrikaans language is well-researched and many individuals from this
population are considered to be more balanced bilinguals than other bilingual groups
(Coetzee-Van Rooyen, 2013).The assessment of vocabulary in bilingual children has
received particular attention because limited vocabulary is one of the first signs of
language impairment (Ellis & Thal, 2008). This research aimed to determine how
Grade 1 Afrikaans-English bilingual children perform on a bilingual vocabulary
assessment.
Design: A quantitative, descriptive, cross-sectional and comparative design was used
in this study.
Method: The Expressive One-Word Picture Vocabulary Test 4 (EOWPVT-4) (Martin
& Brownell, 2011a) and the Receptive One-Word Picture Vocabulary Test 4
(ROWPVT-4) (Martin & Brownell, 2011b) were used to assess 30 grade 1 Englishspeaking
monolinguals. In addition an adapted Afrikaans expressive one word
vocabulary test based on the EOWPVT-4 and an adapted Afrikaans receptive one
word vocabulary test based on the ROWPVT-4 were used to assess 30 grade 1
Afrikaans-English bilinguals. Permission from the schools involved, informed consent
from the parent/s or guardian/s as well as child assent were obtained. The data
gathered from testing was tabulated, interpreted with the use of mean scores and
standard deviations (SD) and analysed using within- and between -group statistical
comparisons. Mean raw scores were converted to percentages for ease of comparison
between receptive and expressive scores.
Results: Within-language comparisons revealed that on the English test, receptive
and expressive scores within both the English monolingual and bilingual groups were
significantly correlated. Expressive scores could therefore be predicted from receptive
scores or vice versa in both the English monolingual and bilingual groups. However,
the receptive and expressive score on the Afrikaans tests were not significantly
correlated. In the bilingual group, the receptive score in Afrikaans was significantly
higher than the expressive score suggesting that although the bilingual participants
had good knowledge of Afrikaans vocabulary they could not always express this in a
naming test. They frequently used the English word. Afrikaans is possibly being used
less in the home and school environments so that the English words are more familiar.
Nonetheless, both the monolingual and bilingual participants had significantly higher
scores on the receptive vocabulary assessment than on the expressive vocabulary
assessments in both English and Afrikaans.
Between-group comparison revealed that the differences between the scores of the
English monolingual and Afrikaans-English bilingual learners were not significant on
either the receptive or expressive vocabulary measure in English. The bilingual group
performed as well as the English participants on the English tests, suggesting that they
are not disadvantaged in the language of instruction. The norms used in the EOWPVT
and the ROWPVT were applicable to both the monolingual and bilingual groups’
scores for the age range of the participants and highlighted that these tests were
suitable in assessing an English monolingual and Afrikaans-English bilingual child in
South Africa. When composite scoring was used the bilinguals scored significantly
better than their monolingual peers on both the receptive and expressive measures,
which confirmed the premise behind this study- that composite scoring should be used
to gain an accurate assessment of a bilingual child’s vocabulary.
Adaptation of the English tests into Afrikaans, as opposed to O’Brien’s study (2015),
which adapted English tests into isiZulu, may have positively affected the results as
all English words had direct translation equivalents in Afrikaans, which was not the
case in isiZulu. The comparison between simultaneous and sequential bilinguals
within the bilingual group demonstrated that the simultaneous bilinguals’ mean
receptive and expressive scores surpassed those obtained by the sequential bilingual
participants. A significant difference was identified between simultaneous and
sequential bilinguals’ composite receptive scores and Afrikaans expressive scores.
Finally, only one monolingual participant scored below the peer group mean on both
the receptive and expressive vocabulary tests, indicating low proficiency in English
and risk of language impairment; however no bilingual participants were found to be
language impaired when composite scoring was used.
Conclusion: Bilingualism remains a rewarding area of investigation in South Africa.
Afrikaans-English bilingual children performed significantly better than O’Brien’s
(2015) isiZulu-English participants on a translated, originally English vocabulary test.
Throughout this study the refinement of valid assessment tools for accurate
description of bilingual children’s vocabulary was highlighted. The well-researched
technique of composite scoring has proven to be valuable in avoiding overdiagnosis in
South African bilingual children. / MT2018
| Identifer | oai:union.ndltd.org:netd.ac.za/oai:union.ndltd.org:wits/oai:wiredspace.wits.ac.za:10539/23821 |
| Date | January 2017 |
| Creators | Van Zyl, Ashleigh |
| Source Sets | South African National ETD Portal |
| Language | English |
| Detected Language | English |
| Type | Thesis |
| Format | Online resource (xiii, 111 leaves), application/pdf, application/pdf |
Page generated in 0.0029 seconds