Abstract
Background: Many young children in South Africa are enrolled in early learning programmes (ELPs) that use a different language of learning and teaching (LOLT) from their mother tongue. In which language should they be assessed?
Aim: To investigate the effect of the language of testing on Early Learning Outcomes Measure 4&5 (ELOM 4&5) performance.
Setting: Preschool Programmes.
Methods: The nationally standardised ELOM 4&5 was administered to isiXhosa-speaking children (39 boys and 46 girls) attending English LOLT programmes (mean age = 62.12 months; standard deviation [sd] = 4.36). The language of assessment order was counterbalanced (English first and isiXhosa second or vice versa). A mixed linear model with fixed and random effects was fitted with ELOM 4&5 Total score at assessment time 2 in language 2 (English or isiXhosa) as the dependent variable. The model included the following predictors: ELOM 4&5 Total score in the language assessed at time 1, days between assessments, sex, age in months and ratings of Task Orientation.
Results: Children performed better in isiXhosa regardless of the language of administration order. Children assessed in English first performed better when tested in isiXhosa second.
Conclusion: Early Learning Outcomes Measure 4&5 test scores of English LOLT isiXhosa-speaking children in ELPs are likely to be more valid indicators of their ability when children are tested in their mother tongue language.
Contribution: This first South African study to investigate the effects of language of test administration on ELOM 4&5 performance in children attending English LOLT programmes indicates that isiXhosa speakers should be assessed in their mother tongue as required by the national home language assessment policy.
Keywords: Early Learning Outcomes Measure 4&5; language of assessment; early childhood; early learning programme; language of learning and teaching; isiXhosa; English.
Introduction
The Tashkent Declaration and Commitment to Action for Transforming Early Childhood Care and Education (ECCE) (UNESCO 2022) emphasises the importance of learning approaches that affirm multilingual education and the promotion of mother tongue language of instruction. The 1997 South African Language in Education Policy (LiEP) follows an additive multilingual approach that requires first (home) language instruction in the early years (Department of Basic Education [DBE] 1997). The underlying principle of the policy is to maintain the use of home language as the Language of Learning and Teaching (LOLT), especially in the early years of learning, while providing access to additional languages. The recent DBE Mother Tongue Based Bilingual Education (MTbBE) initiative addresses the transition from ‘mother tongue’ in Grade 4, promoting the use of bilingualism and African languages (DBE 2024). A mother tongue:
[i]s the language most familiar to the child when they begin formal schooling. But children could have more than one mother tongue. Typically, the mother tongue is the language the child uses regularly for communication in their home and/or community. It is usually the language which the child learns at Home Language level in school. The notion of a ‘mother tongue’ is also known by other terms such as most familiar language(s), primary language(s), first language(s), or home language(s). (Bua-lit, 2024, p. 1)
Following South African policy and the relevant evidence base, the Early Learning Outcomes Measure (ELOM 4&5), which has been standardised for use in 11 South African languages, has consistently been administered in the child’s ‘home language’ (Dawes et al. 2025; Giese et al. 2023). However, it is our experience and confirmed in studies that because of diverse language enrolments in early learning programmes (ELPs) (particularly, but not only, in urban areas), as well as parental preferences, many children who have African home languages attend facilities where the language of learning, teaching and assessment (LOLTA) is English (Mabiletja 2018). Furthermore, many children grow up exposed to multiple languages at home and in their communities, and there are also many attending ELPs who are exposed to other regional languages at home; so determining their ‘home’ language is not a simple matter.
In this contribution, we use the term ELP inclusively to refer to all South African programme types, including centre-based Early Child Development Centres (ECDs) and non-centre-based services such as playgroups and childminder services.
According to the 2016 Annual School Survey, only 6.6% of children in Grade R public schools are English ‘home language’ speakers (DBE 2023), and the 2021 ECD Census (DBE 2022) found that 36% of ELPs are unilingual. English is the language used in almost half of these (49%). A significant proportion of children who have a language other than English as their home language are therefore likely to attend an English LOLT ELP.
English is widely perceived to be the language of upward mobility in a world of work where English dominates (Evans & Cleghorn 2014; Msila 2009; Trudell et al. 2016; Venketsamy & Miller 2021). This can lead to a preference for children’s instruction in English among African language parents from early on (Mabiletja 2018). These authors note that a factor influencing the choice of preschool and primary school is access to English as the medium of instruction in preparation for the world of work, where English dominates. It is also our experience that, as preparation for English LOLT primary schools, ECD Resource and Training Organisations often report that parents choose to send their children to English LOLT preschools. A small study of parent perceptions of the value of English at preschool for isiZulu children confirmed the high value placed on English as a universal language, which is important for further education (Saneka & De Witt 2019). This is despite evidence that children learn best when the first language of instruction in an ELP or school is their home language (e.g. Benson 2005; Bühmann & Trudell 2007; Pinnock 2009; Taylor & Von Fintel 2016). In addition, even where the LOLT is an African language, English and Afrikaans terms may be used for teaching numeracy and mathematics (counting, shape names, etc.), and many learning support materials, including alphabet charts, are in English (Wildsmith-Cromarty, Dyer & Modipa 2023).
Results of studies on school-age African language children in South Africa have shown that when the child’s home language and the LOLT differ, children obtain lower scores when assessed in a language other than that used at home (Van Staden, Bosker & Bergbauer 2016; Vorster, Mayet & Taylor 2012). Analyses of SACMEQ III data from Grade 6 learners collected in 15 Southern African countries in 2010 showed that speaking the language of instruction was a statistically significant pupil-level predictor of both reading and mathematics (Hungi 2011). Multilingual children have also been studied in other African countries. For example, Knauer et al. (2019) tested rural Kenyan children aged 2–6 years on receptive vocabulary tests in three languages: ‘Luo (the local language), Swahili and English (official languages) at two time points, 5–6 weeks apart’ (p. 1). ‘Baseline receptive vocabulary scores in Luo and Swahili were strongly associated with receptive vocabulary scores in English at follow-up’ (p. 1), indicating the importance of multilingual testing if one is to understand the cognitive growth of multilingual children.
Recent South African early-grade reading studies by Mohohlwane et al. (2024) indicate that children perform best when provided with home language instruction. There is also no evidence that an early exit from home language to English as the LOLT improves reading performance. They conclude that while children in Grade 1 can learn English, they have insufficient mastery of this as a LOLT. For pre–Grade R children and Grade R children, this is clearly even more likely to affect performance.
Ensuring valid estimates of language, cognitive and other abilities in young children who use one or more of our African languages at home, but who attend an English LOLT educational environment, poses questions for educators and researchers alike. To our knowledge, no research on the effects of language of testing among children prior to schooling has been conducted in South Africa.
The problem
Language of assessment is a crucial issue for the functional equivalence of ELOM 4&5 test administration for children exposed to multiple languages in the early years. When tests have functional equivalence, the instructions are understood the same way in adapted languages of testing and do not therefore advantage children assessed in the original language from which the test was adapted and translated (Pena 2007; Van de Vijver & Tanzer 2004). This issue was addressed in the standardisation of the ELOM 4&5 (see Method section).
The ELOM 4&5 has been used to assess more than 20 000 children in a variety of settings in all official languages except South African Sign Language (SASL) (Giese et al. 2023). Practice to date is that children are assessed in their home language/mother tongue unless they indicate otherwise at the start of the assessment. But is this necessarily the best practice when home language and LOLT differ? In which language should a child whose language is not English but who attends an English LOLT ELP be assessed on the ELOM 4&5 to obtain the fairest measure of their ability? Our study seeks to address this question. The findings may allow guidance on an appropriate approach to the language of assessment on the ELOM 4&5.
Research methods and design
Design
The study entails a comparison of the performance on the ELOM 4&5 of isiXhosa language children attending ELPs when administered in that language and in English, the ELP LOLT. As we were unable to ascertain the possible variety of languages spoken in participants’ homes, we refer to isiXhosa as the child’s ‘home language’ based on information provided by the child’s ELP and by the child during the assessment.
To test the effects of the language of test administration on test performance, each child was assessed a number of days apart. For statistical analyses, the required sample size was computed with G*Power 3 software (Faul et al. 2007), setting power at 0.8, effect size at 0.20 (Cohen’s d), alpha at 0.05 and assuming an F test in a multiple regression (the procedure most similar to a mixed linear model). The required sample size computed in this way was 39.
Measures
Early Learning Outcomes Measure 4&5
The ELOM 4&5 is a 23-item standardised test that provides a reliable and fair assessment of children regardless of their socio-economic and ethnolinguistic background and is available in all the official languages of South Africa except SASL (a version for SASL is in development). Content, construct, age and concurrent validity (with the WPPSI-IV), as well as test–retest reliability, have been established (Anderson 2021; Dawes et al. 2020; Snelling et al. 2019). To date, the ELOM 4&5 has been used to assess some 20 000 4- and 5-year-old children in South Africa for various purposes, including investigations of the effectiveness of ELPs (Dawes et al. 2023; Giese et al. 2023) and in the Thrive by Five national survey of the development of 5-year-olds (Tredoux et al. 2023). The ELOM 4&5 assesses children’s development in the following domains:
- Gross Motor Development (GMD) (4 items)
- Fine Motor Development and Visual-Motor Integration (FMC&VMI) (4 items)
- Emergent Numeracy and Mathematics (ENM) (5 items)
- Cognition and Executive Functioning (CEF) (4 items)
- Emergent Literacy and Language (ELL) (6 items)
Early Learning Outcomes Measure 4&5 administration takes about 45 min, and items are designed to be enjoyable and are aligned with activities commonly found in preschool curricula.
Procedure
Assessment
Children were individually assessed by two trained, accredited assessors who are fluent in the child’s language. Assessors were tested for the reliability of their scoring using a video of a child being tested and scored by a highly experienced occupational therapist (OT). The scoring of the two assessors was compared with that of the OT. One assessor demonstrated 89% agreement with the expert scoring, while the other obtained 96% agreement. All data were captured on password-protected tablets using the standard assessment kit. Once the assessment had been completed, the data were uploaded to a secure server and anonymised.
Participants
A convenience sample of isiXhosa-speaking children in their final year of an English LOLT ELP (prior to entering Grade R) was enrolled from ELPs that agreed to their children being assessed.
All children passed the ELOM 4&5 disability screen. As explained in the ELOM 4&5 Technical Manual (Dawes et al. 2025), four questions drawn from the World Health Organization 10-point Disability Screen are answered by the child’s teacher or caregiver. They are designed to detect children who are likely to have disabilities or developmental problems in hearing, sight, mobility or intellectual functioning that are likely to negatively affect their ability to perform the items and to not provide a fair assessment of their abilities. Children who have one or more problems are excluded from the ELOM 4&5 administration. In this case, the child’s teacher was requested to answer the disability screen questions.
The study sample included 39 boys and 46 girls (N = 85) (mean age: 62.12; standard deviation [SD] = 4.36 months at first assessment). All children were in programmes charging in the region of R480 per month. As there was minimal variation in fees, the effects of variations in fee level could not be tested.
Children were randomly assigned to one of two counterbalanced assessment conditions: ELOM 4&5 assessment in either English or isiXhosa first and the other language second, as shown in Table 1.
| TABLE 1: Early Learning Outcomes Measure 4&5 samples by test language order. |
Eighty-four children’s scores were valid (all data available for both administrations). Their scores were used in the analyses that follow.
Ethical considerations
Ethical clearance to conduct this study was obtained from the University of Cape Town, Ethics Review Committee of the Faculty of Humanities on 04 June 2024. The ethical approval number is PSY2024-014. Primary caregivers were sent informed consent forms and requested to consent to their child’s assessment. Only children whose caregivers gave consent were enrolled. Children were requested to assent to testing but could withdraw at any time with no consequences to them.
Results
Statistical analyses were undertaken using SPSS 30.0 (IBM Corp. 2023). Correlations between languages of administration for ELOM 4&5 Total and all domains are presented in Table 2. Statistically significant and high correlations are evident for all pairs, with those for GMD and FMC being lower than other domains and ELOM 4&5 Total score.
| TABLE 2: Early Learning Outcomes Measure 4&5 paired samples correlations. |
The first substantive question to be explored was whether or not there was a difference in the child’s performance when assessed in English (the ELP LOLT) or isiXhosa (the child’s home language). A dependent (paired samples) t-test was used to test the null hypothesis of no difference.
The results provided in Table 3 show statistically significant differences favouring children tested in isiXhosa for the ELOM 4&5 Total and three of the five domains: GMD, ENM and ELL. It is evident (following Cohen’s 1988 conventions) that, apart from GMD, effect sizes for statistically significant comparisons are moderate. Overall, children did better when assessed in their home language rather than their LOLT (English).
| TABLE 3: Early Learning Outcomes Measure 4&5 paired samples t-tests across language administrations*. |
Testing predictors of performance in the second language of assessment
The original design required the assessments in English and isiXhosa to be undertaken 10 days apart. Challenges in the field occasioned by holidays, festival days and child absences, among other factors, resulted in a broader range in the time between the two assessments. The average number of days between assessments was 22.61 (SD = 8.74; mode = 14 days; range = 40 days). We were concerned that the time between assessments might affect the relationship between performances in the two languages, and we therefore used simple linear regression to assess the strength of the relationship between ELOM 4&5 Total scores in the two languages, with the interval of time as a predictor. For this analysis, we estimated a required sample size of 39 with G*Power 3 (power = 0.8; ES = 0.20; alpha = 0.05).
We treated language as a dichotomous variable and tested for differences in Total ELOM 4&5 score across the language of administration. Reframing the analysis as a linear regression allowed us to enter control variables. Predictors of the ELOM 4&5 Total score obtained in the language tested in the second assessment (either English or isiXhosa) were:
- Early Learning Outcomes Measure 4&5 Total score on the first language of assessment
- Sex
- Age at first assessment
- Days between the first and second assessments
- Task orientation score: a Total score derived from the assessor’s rating of the child’s attentiveness, concentration, diligence and interest shown while being tested.
We had hoped to include the effect of the amount of time the child had spent in an English Language ELP as a covariate. However, this was not possible because this information was not available. The dependent variable was the second language of assessment ELOM 4&5 Total score.
As reported in Table 4, we found that the Total ELOM 4&5 score on the first assessment in isiXhosa and the assessor’s observation of the child’s Task Orientation score on the second language of testing (English) both significantly predicted performance on the second language of testing (English). Child age and time between assessments were also significant but weak predictors of ELOM 4&5 Total when administered in English second. It is unclear why this is the case.
| TABLE 4: Predictors of English Early Learning Outcomes Measure 4&5 Total score when administered second. |
As reported in Table 5, the only significant predictor of ELOM 4&5 Total score when children were tested in the second assessment in isiXhosa was the ELOM 4&5 Total score when assessed in English on the first occasion.
| TABLE 5: Predictors of isiXhosa Early Learning Outcomes Measure 4&5 Total score when administered second. |
It is of interest that children’s Task Orientation (concentration, interest and attention while being tested) only predicted ELOM 4&5 Total scores when tested in English. A plausible explanation is that being tested in their less familiar language placed a higher cognitive demand on the child and required greater attention and concentration.
To test for an interaction between the language in which the ELOM 4&5 was administered and the order in which it was administered (i.e. before or after the English version) on ELOM 4&5 Total scores, we conducted a mixed linear model, taking ELOM Total scores as the outcome variable, the language in which the assessment was conducted, age at first assessment, days between assessments, the order in which the test was conducted (i.e. English then isiXhosa or the inverse) and the interaction of home language and the order in which the tests were conducted, as predictors.
Given that there were repeated assessments, children were nested within assessments, and a random effect for child was used in the model. Statistically significant effects were found for language of assessment (B = 15.52, SE = 1.55, df = 1, 83, p < 0.001), the child’s age (B = 1.27, SE = 0.40, df = 1, 81, p < 0.001) and the interaction between language test order and language of administration (B = -18.53, SE = 2.32, df = 1, 83, p < 0.001).
Pairwise comparisons were undertaken to further understand the interaction effect between test language order and the language in which ELOM 4&5 was administered. There was a significant difference between test administration in English first vs administration in isiXhosa first (t. ratio = –7.98; df = 83, p < 0.001). Results are displayed in Figure 1.
 |
FIGURE 1: Interaction effect of test language order on Early Learning Outcomes Measure 4&5 Total scores in isiXhosa and English. I bars are 95% confidence intervals. |
|
The results indicate that the English LOLT preschool children performed better on the ELOM 4&5 in both isiXhosa (their home language) and in English when first assessed in English.
In sum, a ‘practice’ test administration in English preceding administration in isiXhosa contributed more to their performance when tested in their home language than a prior (practice) administration in their home language. Practice in the ‘less familiar’ preschool LOLT (English) advantaged their test performance in their home language (isiXhosa) more than practice in the latter. That said, isiXhosa home language test performance remained superior to performance when tested in English (the preschool LOLT).
Limitations
We did not have access to reliable information on the participants’ exposure to English LOLT classes prior to the year of the study. As a result, we were not able to explore the effects of English LOLT ‘dose’ on performance in both languages. This weakness should be addressed in future research. It would also be desirable to have information on the languages to which children are exposed at home and an indication of which is the most commonly used language. This would be useful in determining their most familiar language.
Discussion and conclusion
With the exception of two domains (FMC&VMI and CEF) in which there was no difference in performance, children who had attended an English LOLT ELP performed better on the ELOM 4&5 when tested in their home language. The time between assessments in English and isiXhosa did not predict scores at the second administration. Regression analyses showed that when tested in English, the degree of task orientation was significantly associated with ELOM 4&5 Total scores. This may indicate that children were more focused on the test when assessed in English. Task orientation did not contribute to ELOM 4&5 Total scores when children were tested in isiXhosa. Analysis of the interaction between language of first administration (English or isiXhosa) and order of administration indicated that English LOLT preschool children performed better on the ELOM 4&5 in both isiXhosa and in English when first assessed in English but still obtained higher scores overall when tested in isiXhosa, indicating that home language speakers with at least 10 months’ exposure to an English LOLT ELP should be assessed in their home language on the ELOM 4&5. That higher task orientation scores were associated with better ELOM 4&5 results when isiXhosa speakers were tested in English suggests that children found the English administration more challenging, and that having been in an English LOLT class for 9 months had likely not reduced the effort required during the test.
Our study of children attending an English LOLT ELP affirms the South African LiEP and provides similar findings to a recent Kenyan study of language of testing in young children (Knauer et al. 2019). Administration of the ELOM 4&5 in children’s isiXhosa mother tongue/home language produces consistently better performance regardless of whether the English or isiXhosa version was administered prior to the other version. The main implication of our finding, therefore, is that for children in English LOLT ELPs prior to Grade R, the child’s home language (their mother tongue) should be the language in which the ELOM 4&5 is administered.
It remains a challenge, however, to decide on the most appropriate language of assessment for children living in households and communities where several languages are used. As we noted at the outset of this contribution, urban ELP classes in particular may accommodate children from a variety of African language backgrounds, making the implementation of home language LOLT policies well-nigh impossible in these multilingual classrooms. In such situations, the default LOLT often becomes English. We also noted that children may be exposed to several languages at home and in their communities. As a result, a proportion of children in ELPs and in the Foundation Phase of school are likely to have multilingual language repertoires, within which some linguistic skills are likely to be more dominant than others as a function of exposure to usage in their homes and communities (Coetzee-Van Rooy 2018).
So, what language should be used when assessing the development of these children on tests such as ELOM 4&5? Clearly, the decision will influence the extent to which the assessment provides a fair indication of their abilities. Further research is clearly necessary if we are to improve our practices with children with multilingual language repertoires.
For now (and based on our experience), if possible, we suggest asking the child’s main caregiver (e.g. mother) which languages are spoken at home with the child, and if more than one, asking which is spoken most often on an everyday basis. That (dominant language) would likely indicate the child’s most familiar language and hence the most appropriate assessment language. In the absence of the caregiver or another member of the household, the child’s educator would be the key informant on the child’s ‘home language’. At the point of test administration, children should be the final arbiters. They should be asked which language they would prefer to use during the assessment and be told that they can switch to another language should they wish to do so.
Acknowledgements
We are grateful to Elizabeth Girdwood for overseeing the fieldwork for both studies, to Tawanda Madamombe for supporting assessor’s fieldwork queries and to Tyryn Carnegie for preparing data files for analysis. We thank early learning programme principals and educators for accommodating the study and the learners for their participation.
Competing interests
The author reported that they received funding from DataDrive2030, which may be affected by the research reported in the enclosed publication. The author has disclosed those interests fully and has implemented an approved plan for managing any potential conflicts arising from their involvement. The terms of these funding arrangements have been reviewed and approved by the affiliated university in accordance with its policy on objectivity in research.
CRediT authorship contribution
Andrew Dawes: Conceptualisation, Methodology, Formal analysis, Investigation, Writing – original draft, Funding acquisition. Linda Biersteker: Conceptualisation, Methodology, Investigation, Writing – original draft, Project administration, Writing – review & editing. Colin Tredoux: Methodology, Formal analysis, Writing – original draft.
Funding information
The study was funded by DataDrive2030 (https://datadrive2030.co.za), an NPO that supports research and data collection on early learning in South Africa. https://datadrive2030.co.za
Data availability
The data are available on open access at Data First at the University of Cape Town (https://www.datafirst.uct.ac.za).
Disclaimer
The views and opinions expressed in this article are those of the authors and are the product of professional research. They do not necessarily reflect the official policy or position of any affiliated institution, funder, agency or that of the publisher. The authors are responsible for this article’s results, findings and content.
References
Anderson, K.J., Henning, T.J., Moonsamy, J.R., Scott, M., Du Plooy, C. & Dawes, A.R.L., 2021, ‘Test–retest reliability and concurrent validity of the South African Early Learning Outcomes Measure (ELOM)’, South African Journal of Childhood Education 11(1), a881. https://doi.org/10.4102/sajce.v11i1.881
Benson, C., 2005, The importance of mother tongue-based schooling for educational quality, Commissioned Study for EFA Global Monitoring Report, 24, UNESCO, Paris, viewed from https://unesdoc.unesco.org/ark:/48223/pf0000146632.
Bua-lit, 2024, What is mother tongue-based Bi/multilingual education? viewed 10 December 2025, from https://bua-lit.org.za/wp-content/uploads/2025/03/PAMPHLET_2_English.pdf.
Bühmann, D. & Trudell, B., 2007, Mother tongue matters: Local language as a key to effective learning, UNESCO, Paris, viewed from https://unesdoc.unesco.org/ark:/48223/pf0000161121.
Coetzee-Van Rooy, S., 2018, ‘Dominant language constellations in multilingual repertoires: Implications for language-in-education policy and practices in South Africa’, Language Matters 49(3), 19–46. viewed 10 December 2025, from https://doi.org/10.1080/10228195.2018.1493613
Cohen, J., 1988, Statistical power analysis for the behavioral sciences, 2nd edn., Lawrence Erlbaum Associates, Hillside, NJ.
Dawes, A., Biersteker, L., Girdwood, E., Snelling, M.J.T.L., Tredoux, C.G. & Kafaar, Z., 2025, Early learning outcomes measure: Technical manual, 4th edn., Datadrive2030, Westlake, Cape Town, viewed 10 December 2025, from https://datadrive2030.co.za/wp-content/uploads/2025/10/Early-Learning-Outcomes-Measure-45-Years-Assessment-Tool-Technical-Manual-4th-Edition-1.pdf.
Dawes, A., Biersteker, L., Snelling, M., Horler, J. & Girdwood, E., 2023, ‘To what extent can community-based playgroup programmes targeting low-income children improve learning outcomes prior to entering the Reception Year in South Africa? A quasi-experimental field study’, Early Education and Development 34(1), 256–273. https://doi.org/10.1080/10409289.2021.2005748
Department of Basic Education (DBE), 1997, Language in education policy, Department of Basic Education, Pretoria, viewed from https://www.education.gov.za/Portals/0/Documents/Policies/GET/LanguageEducationPolicy1997.pdf.
Department of Basic Education (DBE), 2022, ECD census 2021: Report, Department of Basic Education, Pretoria, viewed 10 December 2025, from https://datadrive2030.co.za/wp-content/uploads/2022/09/ecdc-2021-report.pdf.
Department of Basic Education (DBE), 2023, South African national literacy strategy and plan: 2024–2030, Department of Basic Education, Pretoria, viewed from https://static.pmg.org.za/250422DBE_-_Strategic_Plan_2025_-_2030.pdf.
Department of Basic Education (DBE), 2024, What is MTBB? Public education resource, viewed 10 December 2025, from https://wcedeportal.co.za/eresource/260996.
Evans, R. & Cleghorn, A., 2014, ‘Parental perceptions: A case study of school choice amidst language waves’, South African Journal of Education 24(2), 1–17. viewed 10 December 2025, from https://doi.org/10.15700/201412071203
Faul, F., Erdfelder, E., Lang, A.-G. & Buchner, A., 2007, ‘G*Power 3: A flexible statistical power analysis program for the social, behavioral, and biomedical sciences’, Behavior Research Methods 39, 175–191. https://doi.org/10.3758/BF03193146
Giese, S., Dawes, A., Biersteker, L., Girdwood, E. & Henry, J., 2023, ‘Using data tools and systems to drive change in early childhood education for disadvantaged children in South Africa’, Children 10, 1470. https://doi.org/10.3390/children10091470
Hungi, N., 2011, ‘Characteristics of Grade 6 pupils, their homes and learning environments’, SACMEQ Working Paper 1, September 2011, viewed 10 December 2025, from https://www.researchgate.net/publication/315105499_Characteristics_of_Grade_6_Pupils_their_Homes_and_Learning_Environments#fullTextFileContent
IBM Corp., 2023, IBM SPSS statistics for windows, Version 30.0 [Software], IBM Corp, New York.
Knauer, H.A., Kariger, P., Jakiela, P., Ozier, O. & Fernald, L.C., 2019, ‘Multilingual assessment of early child development: Analyses from repeated observations of children in Kenya’, Developmental Science 22(5), e12875. https://doi.org/10.1111/desc.12875
Mabiletja, M.M., 2018, ‘Parents’ direct or indirect involvement in the choice of language of learning and teaching’, South African Journal of African Languages 38(3), 355–362. https://doi.org/10.1080/02572117.2018.1518051
Mohohlwane, N., Taylor, S., Cilliers, J., & Fleisch, B., 2024, ‘Reading Skills Transfer Best from Home Language to a Second Language: Policy Lessons from Two Field Experiments in South Africa’, Journal of Research on Educational Effectiveness 17(4), 687–710. https://doi.org/10.1080/19345747.2023.2279123
Msila, V., 2009, ‘School and intra-township migration: Black parents scrambling for quality education in South Africa’, Journal of Education 46, 82–98.
Pena, E.D., 2007, ‘Lost in translation: Methodological considerations in cross-cultural research’, Child Development 78, 1255–1264. https://doi.org/10.1111/j.1467-8624.2007.01064.x
Pinnock, H., 2009, Steps towards learning: A guide to overcoming language barriers in children’s education, Save the Children UK, London.
Saneka, N.E. & De Witt, M., 2019, ‘Barriers and bridges between mother tongue and English as a second language in young children’, South African Journal of Childhood Education 9(1), a516. https://doi.org/10.4102/sajce.v9i1.516
Snelling, M., Dawes, A., Biersteker, L., Girdwood, E. & Tredoux, C., 2019, ‘The development of a South African early learning outcomes measure: A South African instrument for measuring early learning programme outcomes’, Child: Care, Health and Development 45, 257–270. https://doi.org/10.1111/cch.12641
Taylor, S. & Von Fintel, M., 2016, ‘Estimating the impact of language of instruction in South African primary schools: A fixed effects approach’, Economics of Education Review 50, 75–89. https://doi.org/10.1016/j.econedurev.2016.01.003
Tredoux, C., Dawes, A., Mattes, F., Schenk, J.-C., Giese, S., Leach, G. et al., 2023, ‘Are South African children on track for early learning? Findings from the South African Thrive by Five Index 2021 Survey’, Child Indicators Research 17, 601–636. https://doi.org/10.1007/s12187-023-10093-3
Trudell, B., 2016, ‘Language choice and education quality in Eastern and Southern Africa: A review’, Comparative Education 52(3), 281–293. https://doi.org/10.1080/03050068.2016.1185252
UNESCO, 2022, Tashkent declaration and commitments to action for transforming early childhood care and education, viewed from https://www.unesco.org/sites/default/files/medias/fichiers/2022/11/tashkent-declaration-ecce-2022.pdf.
Van de Vijver, F. & Tanzer, N.K., 2004, ‘Bias and equivalence in cross-cultural assessment: An overview’, European Review of Applied Psychology 54(2), 119–135. viewed 10 December 2025, from https://doi.org/10.1016/j.erap.2003.12.004
Van Staden, S., Bosker, R. & Bergbauer, A., 2016, ‘Differences in achievement between home language and language of learning in South Africa: Evidence from prePIRLS 2011’, South African Journal of Childhood Education 6(1), a441. https://doi.org/10.4102/sajce.v6i1.441
Venketsamy, R. & Miller, D., 2021, ‘Factors affecting parents’ choice of schools for Grade 1 learners’, South African Journal of Childhood Education 11(1), a913. https://doi.org/10.4102/sajce.v11i1.913
Vorster, C., Mayet, A. & Taylor, S., 2012, ‘Language and learning’, in N. Taylor, S. van den Berg & T. Mabogoane (eds.), What makes schools effective? Report of South Africa’s National school effectiveness study, Pearson Education.
Wildsmith-Cromarty, R., Dyer, C. & Modipa, T., 2023, ‘Enhancing visibility of local African languages in South Africa through learning to read’, Journal of Multilingual and Multicultural Development 44(9), 860–876. Pearson Education, Cape Town. https://doi.org/10.1080/01434632.2022.2078832
|
Hide
Show all
Email this article (Login required)
