Evidence for Language Policy in Government Pre-Primary Schools in Nigeria: Cross-Language Transfer and Interdependence

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

 

This paper presents the findings from a study focused on Hausa and English sound recognition and decoding for 4-year-olds in Nigeria. The study presents correlational analyses between the four measures, as well as a path-analysis suggesting oral language skills predict decoding within language, and decoding in one language predicts decoding in the other language.

Overall, I think this paper would make an important contribution to the literature on multilingual development in Nigeria with some revisions. My most important comment is the following:

• The overall study design is cross sectional, and the four measures are highly correlated. These two facts necessitate the authors to be very clear about the limitations of findings when it comes to cross-linguistic transfer. Transfer implies skills developed in one language at one time transfer to another language at another time. Very few studies have the luxury of these longitudinal data, which is the case here. This does not mean the study is not important, or does not “suggest” transfer, but you need a strong limitations section. Alternatively, you could test other pathways (e.g. going from decoding toward sound) and show that these do not converge in the way that theory suggests. Given the high levels of correlation between all 4 measures, I would be surprised.

 

Below are some more specific comments.

 

Introduction

The introduction and literature review are concise and well organized. Only a few small suggestions.  

• The first paragraph on linguistic diversity in Nigeria is very important. It would be helpful to note a bit more about the languages. Specifically, since your study is located in Hausa-speaking communities, it would be helpful to say more about the language, how it differs from the other large groups you mention (Igbo and Yoruba), and how the three relate (or not) to English (e.g. script, sounds, etc.)
• The second paragraph starts with a human rights framework, which is important, but perhaps you could give a nod to the pedagogical reasons to support L1 instruction (which you detail later)
• Section 1.1 is important to note the context. However, most of the projects/initiatives you mention focus on elementary schools. Could you find examples or discuss the ECE context, since this is the focus of your study
• Section 1.2 is very important theoretically. You do a good job of identifying some studies in Africa. Another recent cross-linguistic study in Uganda between two Ugandan languages and English could help as another example, but also to build on the issue of transfer in small problem spaces (letter sounds) vs large problem spaces (comprehension) discussed by Snow and Kim (2007): https://www.tandfonline.com/doi/full/10.1080/01434632.2020.1788038
• Section 1.2 provides a great amount of detail on a few studies. It would be more important to give a bit less detail, and highlight the parts theoretically important to cross-linguistic transfer. For example, studies focused on younger children show transfer in small problem spaces. We have very limited evidence of transfer in larger problem spaces (Kim and Piper is an exception). If you make arguments about Grade 5 and 6, you need to point out that none of the studies with African languages have evidence here.

 

Materials and Methods

This is very clear. You could mention the extent to which the instrument you used (Phonics Screening Check) has been validated in the UK, and how it was adapted in Hausa by academics at the Federal University of Gusa, and whether there was a validity study done.

 

Results

 

The results of the first three research questions are very clear. However, the biggest takeaway is that all 4 measures are very strongly correlated. This makes sense, but it begs the question “are all these measures pointing to the same overall construct for these 4-year-olds?” In other words, can you correlate these to other measures (e.g. SES measures, age, etc.).

For RQ4, you developed an SEM model that allows English sounds to predict English letters, and similar in Hausa. Did you test cross paths as well? This is not driven by theory as much as by my previous question on the relationship between all these measures. In other words, are children with better sound recognition more likely to decode overall (not just within languages).

I would also refrain on using “causal” language here as these are all cross-sectional data, and you did not test reverse causality.

 

Discussion

The summary of findings is well written, but the discussion and implications makes interpretations too far beyond the study design and data. This is an important contribution, but you need a strong limitations section. Specifically, you only have cross-sectional data, so any claims of transfer need to be restrained. Here are some examples to edit:   

 

• On line 392, remove that sentence or edit it to say your data suggests the relationship, rather than shows the direction goes from sound to letter.
• Similarly, lines 399 and 400, since both languages are tested together, your data cannot support this without a strong caveat.
• Your important findings are limited to 4-year-olds, and cannot speak directly to the Grade 5 and 6 policy change. Even though I generally agree with your point, these data are limited to pre-school.

Author Response

Evidence for language policy in government pre-primary schools in Nigeria: Cross-Language Transfer and Interdependence.

Response to Reviewer 1 Comments

This paper presents the findings from a study focused on Hausa and English sound recognition and decoding for 4-year-olds in Nigeria. The study presents correlational analyses between the four measures, as well as a path-analysis suggesting oral language skills predict decoding within language, and decoding in one language predicts decoding in the other language.

Overall, I think this paper would make an important contribution to the literature on multilingual development in Nigeria with some revisions.

Thank you very much for taking the time to review this manuscript. Please find below our responses and corresponding revisions/corrections that we have highlighted in red in the re-submitted file. We thank you for making the article stronger with your suggestions.

“My most important comment is the following:

The overall study design is cross sectional, and the four measures are highly correlated. These two facts necessitate the authors to be very clear about the limitations of findings when it comes to cross-linguistic transfer. Transfer implies skills developed in one language at one time transfer to another language at another time. Very few studies have the luxury of these longitudinal data, which is the case here. This does not mean the study is not important, or does not “suggest” transfer, but you need a strong limitations section.

Thank you for pointing this out and we totally agree that this makes the paper stronger. We have added in the limitations section thus:

 

5. Limitations and Future Direction

The current findings should be interpreted keeping in mind the following limitations. First, the present findings are likely to be generalisable to pre-primary grade children in Nigeria with similar demographic characteristics: pre-primary grade children learning to read in Hausa and English and attending government provided schooling in northern Nigeria. It is plausible that a variety of social and contextual factors may have shaped the extent and direction of cross linguistic influences on the measured outcome variables. These influences for example may include differences in language exposure, literacy practices at home and school, and the sociolinguistic status of each language. In our present study such factors were not systematically examined. Future research could adopt a more comprehensive approach to investigating the role of social and contextual variables in shaping reading development in both the first language (L1) and the second language (L2). Second, this study in the context of L1 to L2 reading development and its bidirectionality is a cross-sectional design which offers only a static view of the relationship between the two languages. This may make it difficult to determine the direction or timing of cross-language transfer. A correlation between L1 phoneme awareness and L2 decoding observed at a single time point cannot reveal whether L1 skills supported the emergence of L2 decoding, whether the reverse occurred, or whether both are driven by shared underlying abilities or environmental influences. Cross-sectional data also obscure potential threshold effects, whereby transfer may only occur after a certain level of proficiency in one language is reached, and cannot capture shifts in transfer direction that may occur over time. Interpretations of cross language transfer through cross sectional designed projects remain tentative and susceptible to confounding factors. 

Below are some more specific comments.

Introduction

The introduction and literature review are concise and well organized. Only a few small suggestions.  

The first paragraph on linguistic diversity in Nigeria is very important. It would be helpful to note a bit more about the languages. Specifically, since your study is located in Hausa-speaking communities, it would be helpful to say more about the language, how it differs from the other large groups you mention (Igbo and Yoruba), and how the three relate (or not) to English (e.g. script, sounds, etc.)

 

Again thank you we have accordingly added in the paragraph below.

 

Hausa, Igbo and Yoruba differ significantly in their linguistic structure, tonal qualities and cultural contexts (Igboanusi, 2006). Hausa, a tonal language, spoken by over 63 million people, belongs to the Chadic group of the Afro-Asiatic language family, significantly influenced by Arabic. Yoruba, a tonal language with three distinct tones (high, mid and low) has over 47 million speakers. Igbo also part of the Niger-Congo language family is tonal with numerous dialects including Umuahia and Owerri. Spoken by over 46 million Igbo grammar and syntax are distinct, with unique verb structures that do not typically accommodate consonant clusters found in languages including English. Currently, all three languages primarily use Latin-based alphabets for formal education and general use. However, Hausa and Yoruba historically were written using the Ajami, a form of Arabic script, whilst Igbo utilising Nsibidi ideograms (Heine and Nurse, 2000). When learning to read using phonemes, the main sound differences between Hausa and English come from how each language maps letters to sounds and the kinds of sounds each uses. For example, in Hausa pitch/tone distinguishes words, so tone marking or awareness is part of decoding. Also, Hausa orthography is more phonemic making decoding relatively straightforward in comparison to English with less regular spelling where letters or groups of letters can represent different sounds.

 

Heine, B., and Nurse, D. (2000). African Languages: An Introduction. Cambridge: Cambridge University Press.

 

Igboanusi, H. (2006). A Comparative Study of the Pronunciation Features of Igbo English and Yoruba English Speakers of Nigeria, English Studies, 87(4), 490-497. https://doi.org/10.1080/00138380600768221

 

Malah, Z., and Rashid, S. M. (2015). Contrastive Analysis of the Segmental Phonemes of English and Hausa Languages, International Journal of Languages, Literature and Linguistics, 1(2), 106-112 https://doi.org/10.7763/IJLLL.2015.VI.21

 

 

The second paragraph starts with a human rights framework, which is important, but perhaps you could give a nod to the pedagogical reasons to support L1 instruction (which you detail later)

 

Again thank you, we have added the following:

 

Supporting L1 instruction in the early years had been found to be beneficial. Pedagogical reasons include the use of a familiar language allowing for active student participation by reducing anxiety and increasing confidence. Using L1 enhances comprehension and can acts as a cognitive scaffold. Building rapport and a positive classroom culture between the teacher and students through a more inclusive environment which is student centred can lead to better test performance and school retention rates (de Galbert, 2023; Seid, 2016; World Bank, 2010).

 

de Galbert, P. G. (2023). Language transfer theory and its policy implications: exploring interdependence between Luganda, Runyankole-Rukiga, and English in Uganda, Journal of Multilingual and Multicultural Development, 44(1), 1-19, https://doi.org/10.1080/01434632.2020.1788038

Seid, Y. (2016). Does Learning in Mother Tongue Matter? Evidence From a Natural Experiment in Ethiopia. Economics of Education Review 55: 21–38, doi:10.1016/j.econedurev.2016.08.006.

World Bank. (2010). Studies on Experimental Bilingual Education in Senegal. Washington, DC: World Bank.

 

Section 1.1 is important to note the context. However, most of the projects/initiatives you mention focus on elementary schools. Could you find examples or discuss the ECE context, since this is the focus of your study

Again thank you we have added some examples regarding the ECE context.

USAID and UNICEF are both active in supporting early childhood development (ECD) across Africa, with a shared emphasis on preparing children for formal schooling through language-rich, play-based learning. USAID’s initiatives often target pre-primary and lower primary grades, integrating early literacy and local-language instruction with teacher training and system strengthening. Examples include the Ethiopia Pre-primary in Emergencies project (2024–2029), funded at $35 million by USAID and the LEGO Foundation, which delivers mother-tongue, play-based education for crisis-affected children aged 4–6; Rwanda’s Twiyubake Community ECD Centres (2015–2020), which provided Kinyarwanda storytelling, songs, and emergent literacy activities to 3–6-year-olds; the Consortium for Pre-Primary Data and Measurement in Africa (2019–present), which improves monitoring of language and literacy outcomes in five African countries; and Uganda’s planned Integrated ECD in Basic Education program (2024–2029), expected to invest up to $45 million in school readiness and home-based language development. UNICEF’s work complements these efforts through national policy support, ECD center operation, and integrated services that combine education with health, nutrition, and protection. In Ethiopia, Rwanda, and Uganda, UNICEF backs community-based ECD centres and pre-primary programs that promote early language skills in local languages, strengthen parental engagement, and smooth the transition into Grade 1. Together, these agencies contribute to expanding access to quality early learning, building the foundational literacy and language skills essential for later academic success (Apio, et al., 2024; Spier et al., 2021; Raikes et al., 2020).

 

Apio J., Williams, E. M., Adrupio, S., Acquah S., and Lawson, L. (2024). Mapping early childhood development research outputs in sub-Saharan Africa: Uganda country report. REAL Centre: University of Cambridge and ESSA.

Spier, E., Dias, P., Ranjit, V., Rothbard, V., and Toungui, A. (2021). Evaluation of UNICEF Early Childhood Development (ECD) Programming in the Eastern and Southern Africa Region. Arlington, VA: American Institutes for Research.

Raikes, A., Koziol, N., and Burton, A. (2020). Measuring quality of pre-primary education in sub-Saharan Africa: Evaluation of the Measuring Early Learning Environments scale. Early Childhood Research Quarterly, 53, 571-585. https://doi.org/10.1016/j.ecresq.2020.06.001

 

Section 1.2 is very important theoretically. You do a good job of identifying some studies in Africa. Another recent cross-linguistic study in Uganda between two Ugandan languages and English could help as another example, but also to build on the issue of transfer in small problem spaces (letter sounds) vs large problem spaces (comprehension) discussed by Snow and Kim (2007): https://www.tandfonline.com/doi/full/10.1080/01434632.2020.1788038

Section 1.2 provides a great amount of detail on a few studies. It would be more important to give a bit less detail, and highlight the parts theoretically important to cross-linguistic transfer. For example, studies focused on younger children show transfer in small problem spaces. We have very limited evidence of transfer in larger problem spaces (Kim and Piper is an exception). If you make arguments about Grade 5 and 6, you need to point out that none of the studies with African languages have evidence here.

 

We have taken out the arguments around grade 5 and 6 in the paper and have added the following:

 

Within literacy there are both large and small problem spaces (Snow and Kim, 2007). Small problem spaces including letters, phonemes and decoding are easier to teach and test. The large problem spaces within literacy allowing for high levels of lexical knowledge include vocabulary and meaning-making strategies (de Galbert, 2023). In his study of 3,561 second grade students in 150 schools in Uganda de Galbert (2023) considers cross-linguistic transfer between two Bantu languages – Luganda and Runyankole-Rukiga – and English not only focusing on small problem spaces, that is the transfer of decoding skills but also large problem spaces through the transfer of reading comprehension. Regarding test scores in reading comprehension, oral language, letter knowledge and pseudoword the research shows there is a statistically significant correlation between all four indicators in L1 and L2. The research also finds there is a cross-linguistic transfer from L1 to L2 in reading comprehension (de Galbert, 2023).

 

de Galbert, P. G. (2023). Language transfer theory and its policy implications: exploring interdependence between Luganda, Runyankole-Rukiga, and English in Uganda, Journal of Multilingual and Multicultural Development, 44(1), 1-19, https://doi.org/10.1080/01434632.2020.1788038

Snow, C. E., and Kim, Y.-S. (2007). Large Problem Spaces: The Challenge of Vocabulary for English Language Learners. In R. K. Wagner, A. E. Muse, and K. R. Tannenbaum (Eds.), Vocabulary acquisition: Implications for reading comprehension (pp. 123-139). New York, New York: The Guilford Press.

Materials and Methods

This is very clear. You could mention the extent to which the instrument you used (Phonics Screening Check) has been validated in the UK, and how it was adapted in Hausa by academics at the Federal University of Gusa, and whether there was a validity study done.

Thank you again we have added about the validity of the measures thus:

Duff et al., (2014) when assessing the validity and sensitivity of the Phonics Screening Check find it to be strongly correlated with other literacy skills and sensitive to identifying at risk readers. The phonics screening check shows convergent and discriminant validity. According to the Standards and Testing Agency (2012) the internal reliability of the PSC is a=0.96. A pilot undertaken with 50 early years students in Zamfara State, Nigeria show that for the Hausa test the internal reliability of the Hausa sounds to be a=0.90 and Hausa words, a=0.78.

Duff, F. J., Mengoni, S. E., Bailey, A. M., and Snowling, M. J. (2014). Validity and sensitivity of the phonics screening check: implications for practice, Journal of Research in Reading, 38(2), pp. 109-123. https://doi.org/10.1111/1467-9817.12029

 

Standards and Testing Agency (2012). Phonics screening check: 2012 technical report, London: Standards and Testing Agency. https://assets.publishing.service.gov.uk/media/5a74caeded915d502d6cb0ca/Phonics_screening_check-2012_technical_report.pdf

Results

The results of the first three research questions are very clear. However, the biggest takeaway is that all 4 measures are very strongly correlated. This makes sense, but it begs the question “are all these measures pointing to the same overall construct for these 4-year-olds?” In other words, can you correlate these to other measures (e.g. SES measures, age, etc.).

For RQ4, you developed an SEM model that allows English sounds to predict English letters, and similar in Hausa. Did you test cross paths as well? This is not driven by theory as much as by my previous question on the relationship between all these measures. In other words, are children with better sound recognition more likely to decode overall (not just within languages).

With regards to other measures, we have put this as a limitation in the newly added section, i.e., we did not measure the student’s socio-economic status through this cross sectional study.

As for RQ4 as shown in Table 2 all of the scores are highly correlated showing strong associations. Generally cross path analysis is used with longitudinal data and therefore we decided not to use this type of path model (see Kenny, 1979 ‘Correlation and Causality’ and Kline, 2016 ‘Principles and Practice of Structural Equation Modelling’)

I would also refrain on using “causal” language here as these are all cross-sectional data, and you did not test reverse causality.

We have deleted ‘causal’ from research question 4 and taken out ‘causal’ from the results and discussion sections. Thank you.

Discussion

The summary of findings is well written, but the discussion and implications makes interpretations too far beyond the study design and data. This is an important contribution, but you need a strong limitations section. Specifically, you only have cross-sectional data, so any claims of transfer need to be restrained. Here are some examples to edit:   

• On line 392, remove that sentence or edit it to say your data suggests the relationship, rather than shows the direction goes from sound to letter. [Done we have added ‘These data suggest the relationship between’ ]
• Similarly, lines 399 and 400, since both languages are tested together, your data cannot support this without a strong caveat. [Done we have added this These findings, although from a cross sectional study and with its limitations, support Durgunoğlu (2002)
• Your important findings are limited to 4-year-olds, and cannot speak directly to the Grade 5 and 6 policy change. Even though I generally agree with your point, these data are limited to pre-school.

 

We have deleted the part about grade 5 and 6 as well as added in ‘at the pre-primary stage’.

 

Submission Date

23 April 2025

Date of this review

02 May 2025 17:32:02

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

Dear authors,

You have presented a large scale study that can potentially contribute to the transformation of the early childhood language education in Nigeria. However, I am not convinced on the importance of this study with regards to its RQs and findings.

In your discussion section, you note in lines 382-385 that there is limited research of such nature in an African setting and fewer in early childhood. I don't think this was well argued in the literature review section of the manuscript. In contrast, you have cited a number of studies that support the argument of skills transfer. Perhaps, this has to do with me being a qualitative sociolinguistic researcher, but when I read your RQs, I was convinced of the positive answer to each, and I was not quite intrigued by them. 

I did find your point in lines 405-408 to be interesting, and I wonder if your manuscript would benefit from referencing Otheguy and Garcia's translanguaging theory as linguistic repertoire being a unitary system. The studies you have cited for now, most of them are outdated, and recent literature on translanguaging and such might bring new insights.

My other recommendations have to do with overall writing and formatting: 

After reading your abstract, I was unsure about the topic and purpose of the manuscript.

The language of the manuscript itself is fine, but the structuring could be improved: e.g. introduction sounds repetitive and can be more concise. Lit Review has some information that belongs to the background section under methodology. Your discussion section can certainly be expanded upon and I highly encourage to think how can you revisit your lit review to inform the claims you make in the discussion section. Implications that come up toward the end of the manuscript come as a surprise considering all the efforts on MT education by the government. Perhaps, all of these are a matter of clarity. 

Author Response

Please see the attachment. 

Author Response File: Author Response.pdf

Reviewer 3 Report

Comments and Suggestions for Authors

This paper makes a valuable contribution by providing substantial empirical evidence from the African early childhood education context. It employs a strong sample size (N = 851) and applies robust statistical methods. Notably, it demonstrates a significant bi-directional relationship between Hausa and English decoding skills, offering important support for cross-linguistic transfer theories in emergent bilinguals.

The paper adopts a narrowly focused theoretical framing that reduces the scope, depth, and interpretive flexibility of its findings, while limiting its engagement with the broader scholarly debate on cross-linguistic influence. The research exhaustively reiterates the Interdependence Hypothesis without critically assessing its limitations or situating it within a comparative theoretical context. Furthermore, the author fails to acknowledge or integrate alternative frameworks such as the Linguistic Threshold Hypothesis and the Transfer Facilitation Model, which could have provided a more balanced and comprehensive analysis.

 

The sample does not adequately represent the broader population of Northern Nigeria, as the study focuses exclusively on government pre-primary schools, disregarding the inclusion of private institutions, other public-school, and urban–rural diversity. This sampling limitation raises concerns about the generalizability of the findings to the wider population. The researcher fails to account for key confounding variables, including teacher training quality, family language practices, and participants’ socio-linguistic and socio-economic status. Neglecting these factors limits the validity of the findings and weakens the study’s explanatory power.

 

Certain claims are made without adequate consideration of the longitudinal model. Since the study primarily employs a cross-sectional design, the role of longitudinal evidence is minimal, thereby limiting the validity and robustness of the inferences drawn. The limited dataset presented by the researcher does not accurately reflect Nigeria’s national language policy. Consequently, the findings cannot be generalized to the entire population, as the study draws on a geographically restricted and linguistically narrow sample. Beyond the core linguistic focus, the researcher overlooks important extra-linguistic factors such as orthography, curriculum development, teacher capacity, and language standardization. While the statistical results appear robust and practically relevant, the study does not address their educational significance, effect size within real classroom contexts, or practical applicability for policymakers. This omission raises concerns about the study’s overall utility and impact in informing language education policy.

Author Response

Please see the attachment

Author Response File: Author Response.pdf

Round 2

Reviewer 3 Report

Comments and Suggestions for Authors

I sincerely appreciate the time and effort you put into carefully incorporating the suggested feedback. Your thoughtful revisions not only improved the clarity and flow of the work but also demonstrated your strong commitment to producing high-quality results. It is evident that you engaged critically with the feedback, which has added significant value and refinement to the overall piece

Author Response

Thank you so much for all of your comments. They really added to our paper. Very much appreciated.