Abstract
Children with a well-established and elaborate mathematics vocabulary perform better in assessments and engage more easily in mathematics discussions and have more cognitive resources available for higher-order mathematics tasks (Lin, Peng and Zeng 2021). The terms mathematics-specific language / mathematics language and mathematics-specific vocabulary / mathematics vocabulary are often used interchangeably (e.g. Lin et al. 2021; Turan and De Smedt 2022). We therefore define mathematics vocabulary as terms, phrases, abbreviations, and symbols essential for reasoning and problem-solving and suggest that early exposure to mathematics vocabulary predicts later academic success. Mathematics vocabulary thus contributes to mathematics learning (Purpura, Schmitt, Napoli, Dobbs-Oates, King, Hornburg, Westerberg, Borriello, Bryant, Anaya, Kung, Litkowski, Lin and Rolan 2021) and is also a reliable predictor for future performance on mathematics assessments (Bezuidenhout 2018). Previous research highlights the effectiveness of using dialogue reading books in home settings to develop mathematics vocabulary and skills (Purpura et al. 2021). However, little research has been conducted on implementing this method in classrooms. This study aimed to translate an existing series of three English dialogue reading books for mathematics vocabulary into Afrikaans and assess their impact on Grade RR learners’ mathematics vocabulary and mathematics performance over a period of four weeks. Unlike prior studies, this intervention is classroom-based, offering insights into its effectiveness in larger group settings.
This study considers the effectiveness of Afrikaans-translated dialogue reading books for mathematics vocabulary and skills through a theoretical lens of two integrated frameworks. Firstly, to maintain functional equivalence in a translated mathematics vocabulary text, comparability is considered the key factor. This study used decentralisation (Sechrest, Fay and Zaidi 1972) and a double-focus approach (Erkut, Alarcón, Coll, Tropp and García 1999), involving an interdisciplinary team of experts, including linguists, teachers, parents, and researchers, to ensure both linguistic and functional accuracy. Secondly, dialogic reading, a structured approach where the reader actively engages children in the story, enhances mathematics vocabulary and concepts. Techniques like CROWD and PEER (Lonigan and Whitehurst 1998) encourage children to become co-storytellers rather than passive listeners. Research has shown that dialogue reading improves mathematics vocabulary development, number concepts, counting, and cardinality (Gibson, Gunderson and Levine 2020). To assess the effectiveness of this approach, Bezuidenhout (2021) suggested five key components for storybooks: age-appropriate topics, relatable characters, specific vocabulary, clear constructs, and suitable illustrations. In this study, we used these principles to show that the Afrikaans translation of The little elephants’ big adventures is a suitable dialogue reading series for the development of young children’s mathematics vocabulary.
The series consists of three books. In Too many pillows, three friends go camping, and Benjamin packs too many pillows, flashlights, and peanuts, making his backpack very heavy. Lucy helps him unpack so that only the necessary items remain, lightening his load. Each illustration facilitates discussions about comparing quantities. In Just enough eggs, Lucy and Benjamin bake a cake for Bear’s birthday. They discuss the recipe and solve mathematics problems in the kitchen. Each storyline provides examples for families and educators on how to spark maths conversations in everyday situations. Each page of the stories contains three questions, each with a different purpose. In Picnic with some peanuts, the characters divide the apples, lemonade, and peanuts fairly among themselves. Ants carry Bear away while the elephants read and the concerned elephants go looking for Bear just to find him and some missing peanuts with the ants, Bear’s new friends. Illustrations clearly depict mathematics concepts like division by making all quantities of items visible.
In the first phase of the study, we discuss how the Bezuidenhout (2021) model was used to evaluate the effectiveness of the three Afrikaans books and how the translation process was handled. By describing the process, teachers and researchers can use a dual-focus approach (Erkut et al. 1999) and an interdisciplinary team of translators to translate teaching materials in the future.
The second phase was a conceptual replica of Purpura et al. (2021), but in a classroom setting with large groups of children. Four Grade RR classes (mean age at pretest was 6,34 years) at an Afrikaans public school in Gauteng participated in the study. Thirty-five children participated in the intervention while 26 children were in a business-as-usual group. After a training session with all participating teachers, all children’s i) mathematics, ii) mathematics vocabulary, iii) working memory, and iv) similarity recognition skills were tested during a pretest. After a four-week dialogue reading intervention, all participating children’s mathematics vocabulary and mathematics skills were assessed. No delayed post-test was conducted due to time constraints at the school. During the intervention, each teacher in the intervention group read the books to their children over a period of four weeks. The first book was read at least three times during the first week, the second book at least three times in the second week, and the third book at least three times in the third week. All three books were read at least once again during the final intervention week. The suggested questions at the bottom of each page were used for discussions. Some of the teachers expanded on the questions and discussed more than what was covered by the specific question.
The normality of distribution for continuous variables were determined and the means of all variables were compared. To determine the effect of the intervention, independent t-test and Mann-Whitney U test as a non-parametric alternative to t-test, were conducted to compare the two groups (intervention vs comparison group) at two time points (pretest vs post-test). Differences within each group (intervention and comparison group) between the pretest and post-test were also examined, as well as associations between variables. All analyses were performed using computer software called Jamovi (Version 2.5, 2024).
There were no differences between the intervention and comparison groups in terms of background variables (age, gender, siblings, and recognition agreement) or measurement instruments, including mathematics skill and mathematics vocabulary at both the pretest and post-test (all p values > 0,05). The results indicated that the intervention did not lead to a statistically significant improvement in the mathematics skills of the intervention group compared to the comparison group. Although the average for mathematics vocabulary was higher in the intervention group (mean difference = 1,03) compared to the comparison group (mean difference in the intervention group = 0,88), the difference was not significant, with a small effect size. Mathematics vocabulary in the intervention group (Mdn = 21,50) significantly differed between the pretest and post-test (Mdn = 23), W = 75, p = 0,010, r = −0,57, indicating a large effect. However, the mathematics vocabulary in the comparison group (Mdn = 22) at the pretest did not significantly differ from the post-test (Mdn = 22), W = 78, p = 0,114, r= −0,38, indicating a medium effect. The significant improvement in mathematics skills, as well as mathematics vocabulary, after the teacher workshops and intervention programs, calls for further investigation into the impact of teacher training programmes on teachers’ intuitive inclusion of mathematics vocabulary during their mathematics instruction.
Spearman correlation showed a significant positive relationship between mathematics skills (W = 96, p = 0,061) and mathematics vocabulary (W = 0,89, p < ,001) at the pretest, r = 0,34, p < ,01. Positive association between mathematics skills and mathematics vocabulary during both the pretest (r = 0,34, p = ,008) and post-test (r = 0,27, p = ,035) were also determined. Further, there were significant predictive relationships between mathematics skills at the pretest and mathematics vocabulary at the post-test (r = 0,30, p = ,018) and between mathematics vocabulary at the pretest and mathematics skills at the post-test (r = 0,42, p < ,001). These correlations highlight the importance of explicit teaching of mathematics vocabulary (Riccomini, Smith, Hughes and Fries 2015).
Although the books did not have a positive effect on the mathematics vocabulary or skills of the class groups, we consider this an important contribution to science and education, allowing future studies to be more focused in design. Future studies could also focus on evaluating the impact of the same Afrikaans series in home environments, in small groups, and one-on-one teaching situations. The implementation of dialogue reading books in other languages will also be explored in small groups and home environments.
We conclude that although the intervention did not lead to significant differences between the intervention and comparison groups for both mathematics skills and mathematics vocabulary, significant changes over time were observed for both measures, regardless of the group. Furthermore, several correlations between study variables were identified, which may justify further investigation in future research.
Keywords: dialogue reading; mathematics competencies; mathematics vocabulary; preschool