A Setswana translation and adaptation of the Stark Griffin™ Dyslexia screening test

Abstract

South African learners, particularly in Grade 4, demonstrate significant challenges in reading comprehension. The 2021 PIRLS study found that 81% of these learners could not read with comprehension. Notably, even Setswana-speaking learners, who receive initial instruction in their home language, struggle. It seems that factors beyond the medium of instruction, such as the development of indigenous languages and equitable resource distribution, are crucial. The shift to English-medium instruction in the intermediate phase further exacerbates the problem, making it challenging to diagnose reading difficulties, especially in second language learners, accurately. Linguistic disparities, encompassing  phonetic, phonological, morphological, and orthographic differences and challenges in phonological awareness, contribute to these difficulties and hinder the identification of specific learning disorders, such as developmental dyslexia. Critically, South African standardised screening tools for developmental dyslexia (DD) are only available in English, leaving English second-language speakers, such as Setswana learners, without adequate diagnostic support. To address this gap, this study focused on translating and adapting an existing English DD screening tool into Setswana for clinical use. The translation and adaptation processes, and the initial pilot study results, are discussed. An exploratory sequential research design with both quantitative and qualitative components was used to compare the outcomes of the translated Stark Griffin™ Dyslexia Screening Test (SGDST) set with those of the source test (English SGDST). This research study consisted of two phases: translation and adaptation (Phase 1) and pilot test (Phase 2). Following expert panel adaptations, including simplified terminology, a pilot study of the SGDST in Setswana revealed administration errors and significant differences in (1) eidetic encoding for primary school groups, (2) decoding between primary and high school groups, and (3) English and Setswana versions of the test, particularly in primary schools. The SGDST Setswana version received positive ratings, despite one report of unspecified grammatical errors, which could not be triangulated. Errors, including incorrect basal/ceiling levels and encoding stimuli, were attributed to translation quality, misinterpreted instructions, and human error. Despite these errors, decoding accuracy between control and experimental groups showed no significant differences, except for eidetic encoding in primary schools, where a significant difference was found. Significant differences emerged when comparing primary and high school performance. Decoding accuracy was significantly higher in primary schools for both control group and experimental group, reflecting expected developmental trends. The experimental group’s Setswana decoding performance also showed a significant difference between the primary and high schools, with the primary school achieving 0% accuracy. The ITC checklist showed positive feedback, but reported grammatical errors, suggesting a need for refinement. The study highlights the challenges of adapting and administering standardised tests in African languages, emphasising the need for improved training for administrators and further validation.