COMPLEX SENTENCE COMPREHENSION WITHIN A SOUTH AFRICAN ADOLESCENT POPULATION

[摘要] South African Speech-Language Pathologists are assessing and treating many adolescentswith varying speech and language difficulties. This task is complicated by the fact that itis currently largely unknown what the language abilities of mainstream adolescents are.Some research has suggested declining language abilities, as well as significant effects ofgrade, gender and language background on cognitive academic language skills. Asindividuals grow older, the demands of the classroom environment become greater,forcing adolescents to use complex language skills in order to learn. This research paperaimed to assess the complex sentence comprehension abilities of a sample of SouthAfrican adolescents, through the use of the Grammar/Listening subtest of the Test ofAdolescent Language, developed by Hammill, Brown, Larsen, and Wiederholt (1980).Additionally, the effects of grade, gender, language, time spent reading for school andrecreationally each week, number of television programmes watched per week, length oftime spent playing computer or video games per day, preferred learning styles(group/alone and auditory/visual/both), most recent English and school report marks, anduse of cellular telephones, were used to determine related factors and educationalvariables which might be linked to each other. Additionally, a working memory measurewas included, in order to ascertain that this factor was not having a negative effect uponthe comprehension scores. It was found that gender was an over-riding factor throughoutthe study. Females tended to do better on the complex sentence comprehension test, andalso spent more time reading for school each week, obtained better English and schoolreport marks, preferred to learn alone, and sent more sms’s. Males were only inclined toplay more computer or video games per week. In opposition to the initial hypothesis thatthese learners would perform poorly on the TOAL subtest, it was found that participantsgenerally performed within the average limits of the test. Correlations, one-wayANOVA’s, chi squared analyses and t-tests were performed for the secondary aims, inorder to determine any relationships between the variables. Overall, though, gender wasthe key variable in the study, which is in line with other literature in the field. Homelanguage and educational level had minimal effects. Further research has beenrecommended.

[发布机构] University of the Witwatersrand