In multilingual sub-Saharan African countries, many children attend school and learn to read in a language that they do not speak at home. This mismatch between home and school language may contribute to poor learning outcomes, including low literacy rates. Bilingual education that includes a local language of instruction has become more prevalent in an effort to improve primary school children's learning. Indeed, high-quality bilingual programs are associated with favorable language, literacy, and learning outcomes (Berens et al., 2013;Bühmann & Trudell, 2007;Takam & Fassé, 2020). Yet, little is known about how home and school language environments support skilled reading in multilingual communities with low literacy rates. In this study, we assessed primary school children's (N = 830) French and local language (Abidji, Attié, Baoulé, Bété) phonological awareness, vocabulary, and oral language comprehension skills and French reading skills. Further, we explored differences in quality between monolingual French and bilingual local language-French schools that may contribute to differences in children's language and literacy performance. We found that bilingual local language-French homes were associated with better language outcomes than local language-only homes, reflecting advantages associated with early bilingual exposure. On the other hand, monolingual French schools were associated with better language and literacy outcomes than bilingual local language-French schools. We found that monolingual French schools were of higher quality than bilingual schools, likely contributing to the discrepancies in language and literacy results. Our results emphasize the importance of monitoring program quality to allow children to reap the benefits associated with bilingual education. Educational Impact and Implications StatementThe present study suggests that bilingual home environments (French-Ivorian language) are advantageous for children's spoken Ivorian language skills and for their French literacy skills. However, children from monolingual French schools had better spoken language and literacy skills overall than children from bilingual schools. Monolingual schools had access to better resources and greater support, indicating that high quality education is important for children's literacy outcomes.

Few standardized language assessments are adapted to different cultural and linguistic contexts to assess children’s first language (L1) abilities. We introduce the Ivorian Children’s Language Assessment Toolkit for measuring phonological awareness, vocabulary, oral comprehension, and tone awareness in the Abidji, Attié, and Baoulé languages of Côte d’Ivoire. Six hundred and three primary-school children (age 4–14) completed language assessments in their L1 and French. The toolkit provided a reliable and comprehensive assessment of children’s language abilities. We observed age- and grade-related increases in all subtest scores. Still, children scored higher in their L1 compared to French, highlighting the need for language assessments in a bilingual’s two languages to achieve an accurate measure of children’s language abilities. The ability to benchmark children’s scores relative to age- and grade-norms are discussed in the context of language of instruction education policies as well as the potential use of age- and grade-norms in identifying children with language impairment and/or children who are at risk for reading difficulties due to poor language skills.

Technology-based remote research methods are increasingly widespread, including learning assessments in child development and education research. However, little is known about whether technology-based remote assessments remain as valid and reliable as in-person assessments. We developed a low-cost phone-based language and literacy assessment for primary-school children in low-resource communities in rural Côte d’Ivoire using voice calls and SMS. We compared the reliability and validity of this phone-based assessment to an established in-person assessment.679 5th grade children completed language (phonological awareness, vocabulary, language comprehension) and literacy (letter, word, pseudoword, passage reading and comprehension) tasks in-person and by phone. Reliability (internal consistency) and predictive validity was high across in-person and phone-based tasks. In-person and phone-based assessments were moderately to strongly correlated. Phonological awareness and vocabulary skills measured in-person and by phone significantly predicted in-person and phone-based letter, word and pseudoword reading. Oral language and decoding skills measured in-person and by phone also significantly predicted in-person and phone-based passage reading and comprehension scores.Our phone-based assessment was a reliable and valid measure of language and reading and feasible for low-resource settings. Low-cost technologies offer significant potential to measure children’s learning remotely, contributing to greater inclusion of remote and low-resource populations in education research.

Risks associated with school dropout have been studied in West Africa, yet more research is needed to understand what protective factors can be associated with academic resilience (i.e., remaining in school despite facing adversity). At the beginning of our longitudinal study in rural Côte d’Ivoire, 1195 students (Mage=10.75, SDage=1.42) were enrolled in fifth grade. Two years later, 7% of students had dropped out. Characteristics related to the child (e.g., child labour), family (e.g., socioeconomic status), and school (e.g., teacher quality) were first examined distinctly. We then applied a cumulative risk (CR) framework to examine child-, family-, and school-level CR and their interactions. To understand academic resilience, we used findings from our risk analysis to identify a “high-risk enrolled” subset of children and compared them to the children who dropped out. Children who dropped out were older, involved in more child labour, had poorer literacy, owned fewer books, and attended schools with poorer learning environments. Child-level CR most strongly predicted dropout (b=-.860, OR=.424); however, children with low child-level CR were more likely to drop out when family-level CR was high (b=.227, OR=1.250). School characteristics (better infrastructure and teachers) were protective for children who were at high risk of dropout yet remained enrolled. Child-, family-, and school-level factors all contributed to dropout and these factors interact to affect dropout. Although child- and family-level factors contribute significantly to dropout, certain school factors may mitigate these risks and promote academic resilience.

Literacy and numeracy are correlated throughout development and require many of the same underlying cognitive skills. We explored the predictors of literacy and numeracy and their covariance (overlap between the two skills) in rural Côte d’Ivoire. Many Ivorian children are old for their grade due to late enrollment and grade repetition, leading to variation in age-for-grade (i.e., how far a child’s age differs from official age for a given grade). Structural equation models were used to examine how cognitive components and individual-level factors predict literacy and numeracy covariance. Phonological awareness, numerical magnitude, ordinality, working memory, and inhibitory control were cognitive predictors of covariance. Additionally, age was positively related to covariance and age-for-grade was negatively related to covariance. Post-hoc multi-group modelling revealed literacy and numeracy covariance was significant for the on-time group (students in the correct grade based on official age) but not the late group (students in a lower grade). Our results suggest that literacy and numeracy covariance depend on the maturity of underlying cognitive components.