The Impact of Bilingual Scaffolding on the Acquisition and Retention of Specialized Medical Terminology

Abstract

Introduction

This study investigates the impact of bilingual instruction compared to L2-only instruction on the acquisition and retention of specialized medical vocabulary.

Methods

50 medical and health science students were divided into two groups: A control group receiving English-only instruction and an intervention group receiving bilingual instruction. Using a longitudinal quasi-experimental design, both groups were assessed on vocabulary breadth and depth at three intervals: immediately, one week, and one month following instruction.

Results

The findings revealed that while initial gains at T1 were similar between groups, the intervention group receiving bilingual instruction consistently achieved significantly higher scores and superior retention rates at the one-week and one-month follow-ups. The advantage of bilingual instruction was most salient in long-term depth of knowledge, where learners maintained more durable mental representations of complex terms.

Discussion

These findings suggest that first language scaffolding reduces cognitive load, which may make abstract medical terminology more accessible. Ultimately, the study advocates for translanguaging as a beneficial pedagogical strategy to enhance academic success in English-medium instruction medical contexts.

Introduction

In academic contexts, the mastery of specialized vocabulary is essential for effective comprehension and knowledge across disciplines.¹ These specialized lexicons consist of low-frequency and domain-specific words that are fundamental to academic communication.^2,3 Research has established a significant relationship between vocabulary acquisition and academic success, as the proficiency in both general and academic vocabulary is critical for success in any subject area.⁴ Furthermore, developing this lexical mastery improves the knowledge acquisition process and increases students’ level of engagement with their subject matter.^5,6 Theoretical frameworks suggest that learners’ first language (L1) can be a helpful resource as it allows them to relate new second language (L2) vocabulary to concepts and ideas they already know in their L1.⁷ This adoption of certain aspects of learners’ L1 had the potential of leading to beneficial acquisition strategies in academic contexts.⁸ In domain-specific fields, using L1 can support comprehension and help maintain immediate and long-term vocabulary retention by clarifying challenging terminology.^9,10

Bilingual scaffolding is a pedagogical approach that strategically incorporates learners’ L1 as a supportive resource during L2 instruction to facilitate comprehension, reduce cognitive demands, and build conceptual bridges between existing knowledge and new learning.¹¹ This approach draws on the notion that learners’ L1 can be a helpful resource as it allows them to relate new L2 vocabulary to concepts and ideas they already know in their L1.¹² Vocabulary acquisition in English for Specific Purposes (ESP) contexts refers to learning domain-specific lexical items required for professional communication in specialized fields.⁷ The mastery of the English language vocabulary is vital for both ESL and EFL learners,¹³ especially for learners of ESP.¹⁴

The present study explores how learners interact with specialized vocabulary in their L1 compared to their L2. Specifically, this study aims to determine the effectiveness of bilingual instruction compared to L2-only instruction in the acquisition and retention of specialized medical vocabulary. The study seeks to identify effective pedagogical strategies for teaching specialized vocabulary in bilingual contexts by analyzing the outcomes of vocabulary retention rates and depth of understanding.

Theoretical Underpinning

Vygotsky’s¹⁵ Sociocultural Theory (SCT) offers a theoretical basis for understanding how bilingual scaffolding enhances the acquisition of specialized vocabulary. SCT suggests that development is a mediated process driven by social interactions. Language functions as a key mediating tool by facilitating the growth of higher-order cognitive skills through social and material engagement. This view emphasizes that language learning is fundamentally social, with cognitive growth occurring through interactions within the Zone of Proximal Development (ZPD). The ZPD is the space between what learners can achieve alone and what they can achieve with guidance. In the current investigation, guidance is operationalized through bilingual scaffolding provided to the intervention group. Scaffolding aligns with ZPD by facilitating students’ progression from their current proficiency level to a more advanced stage.

The cognitive mechanisms underlying bilingual scaffolding can be explained through Cognitive Load Theory (CLT) and the limitations of working memory. CLT posits that learning is constrained by the limited capacity of working memory and promotes instructional strategies that minimize cognitive load, which allows learners to devote more resources to the construction and integration of new knowledge structures.¹⁶ In ESP contexts, processing unfamiliar medical terminology solely in L2 may impose additional cognitive demands on learners, particularly when the concepts are novel. Consequently, it may reduce the available cognitive resources necessary for the acquisition and consolidation of new vocabulary. Bilingual scaffolding, which includes providing definitions and explanations in the learners’ L1, may help alleviate such processing difficulties by facilitating access to pre-existing conceptual knowledge and allowing connections between new L2 words and familiar L1 schemas.^16,17 The study by Yuan and Tang¹⁸ on bilingual subtitles supports this by demonstrating that the inclusion of L1 subtitles reduces cognitive load, with the L1-bilingual condition outperforming the other conditions in recall of word form and meaning and eliciting the lowest overall cognitive load.

The Relationship Between Reading and Vocabulary

Reading is an invaluable means of expanding knowledge and an effective source of written input for vocabulary acquisition.¹⁹ Historically, vocabulary knowledge has been viewed as the most crucial determinant of reading comprehension.^20,21 Within the field of second language acquisition (SLA), there is a consensus that increased exposure to diverse texts leads to gains of more terms in the L2.^22–24 Whether L2 learners acquire vocabulary through incidental learning while reading^25,26 or through intentional learning,^27,28 researchers agree that vocabulary enhancement is an inevitable result of the reading process.

Modern perspectives, such as Perfetti’s²⁹ lexical quality hypothesis, suggest that vocabulary acquisition is a product of developing high-quality mental representations of words. This framework posits that reading comprehension is hindered or facilitated by the quality of a learner’s lexical knowledge. This notion supports the reciprocal link between word-level processing and text-level understanding.³⁰ This approach encourages learners to infer meaning from context rather than relying on memorization. Similarly, extensive reading allows learners to engage with diverse words in natural contexts, which facilitates deeper comprehension and retention.⁷ Thus, undoubtedly, context serves as an important tool in learning new words, as it provides the necessary clues to predict and solidify word meanings. The relationship between vocabulary and reading is reciprocal.^30,31 Given this reciprocal link, reading and vocabulary should not be treated as isolated domains, as vocabulary should not be viewed as merely a supplement to reading.

In English as a second language (ESL) classrooms, L2-only instruction has been the preferred approach.^32–34 By encouraging real-time use of the language, this practice has been regarded as a facilitator of more engaging learning experiences and of enhanced students’ linguistic proficiency. Nevertheless, recent research suggests that the first language (L1) may serve as an effective resource for learning the L2 in different academic settings.^8,35–39

Specialized Vocabulary in the Medical Field

English has emerged as the lingua franca of the international health care community, supported by the globalization of medical publishing and education.⁴⁰ It acts as a unifying medium that transcends geographical and linguistic boundaries.⁴¹ However, for L2 learners, the acquisition of specialized medical terminology presents unique challenges. Many medical terms derive from Latin and Greek roots, which creates morphological complexity that can overwhelm learners unfamiliar with these classical languages. Additionally, the unfamiliar pronunciation patterns and density of specialized vocabulary in medical texts can result in cognitive overload, particularly for L2 learners who must simultaneously process new content knowledge and linguistic information. The challenges may include linguistic inequality and limited accessibility.⁴⁰

Medical students encounter a high density of specialized vocabulary daily and are expected to use it in high-stakes clinical settings.⁴² Mastering technical vocabulary is essential for academic communication in such fields, as this vocabulary represents domain-specific lexis required to master content.^2,3 However, medical terminology is considered one of the most challenging types of ESP because of its diverse linguistic roots.⁴³ Heming and Nandagopal further suggest that for L2 learners, the lack of linguistic competence acts as a barrier to academic achievement. Furthermore, the study by Gablasova⁴⁴ suggests that L1 users have a lower depth of knowledge of specialized terms than L2 users. This means that they frequently struggle to provide detailed information about word meanings.^10,44

To overcome these challenges, L2 learners employ various vocabulary learning strategies (VLS) to acquire and retain unfamiliar words.⁴⁵ These strategies include memorization, repetition, and note-taking.^13,46 In medical education, specific techniques such as analyzing word roots and affixes⁴⁷ as well as utilizing bilingual dictionaries⁴⁸ have proven to be effective. Moreover, advanced learners tend to use more elaborate strategies, such as translating target words into their L1 or connecting new terms to personal experiences to ensure long-term retention.⁴⁹ Recent studies emphasize that incorporating the L1 can serve as an effective resource for learning L2 vocabulary in academic settings.^8,37–39

The Study

The existing literature on vocabulary acquisition in bilingual contexts reveals several areas of consensus alongside notable gaps. There is broad agreement that vocabulary knowledge is fundamental to academic success,¹ and that both incidental and intentional learning contribute to lexical development.⁶ However, perspectives diverge on the role of L1 in L2 vocabulary learning. While L2-only instruction has been more common, some studies examined whether L1 use functions as a facilitator or barrier in L2 vocabulary learning³⁷ and suggested that the debate remains unresolved. Moreover, Gablasova⁴⁴ compared learning outcomes through L1 and L2 for specialized vocabulary from textbook reading and provided evidence that bilingual approaches may have advantages for technical vocabulary acquisition.

Despite the established importance of specialized vocabulary in academic achievement, a gap in the body of literature exists in relation to the retention of such vocabulary in bilingual settings. Although previous studies investigated the acquisition of vocabulary and its relationship to academic success, few studies explored how learners retain this vocabulary over time within bilingual instructional contexts. In high-stakes medical fields, there is a need for such investigation as linguistic barriers are a primary cause of student attrition,⁵⁰ yet pedagogical strategies involving L1 support remain under-explored.

The present study addresses these research gaps by exploring how medical and health science L2 English learners interact with and retain specialized medical vocabulary when provided with bilingual instruction compared to L2-only instruction. The study employs a longitudinal design to measure learning outcomes at three intervals: immediately after instruction, one week later, and one month later. Specifically, this study analyzes both the breadth and depth of lexical knowledge to determine if incorporating L1 leads to higher retention rates of complex technical terms. By focusing on medical and health science students in an English as a medium of instruction (EMI) context, the study aims to provide empirical evidence for pedagogical strategies that can help students overcome linguistic barriers in relation to specialized medical terminology.

Three primary research questions guided this investigation:

1. To what extent does bilingual instruction, compared to English-only instruction, influence students’ breadth of specialized medical vocabulary across three time points: immediately after instruction, one week later, and one month later?

2. To what extent does bilingual instruction, compared to English-only instruction, affect students’ depth of specialized medical vocabulary across immediate, one-week, and one-month post-instruction intervals?

3. How does the rate of retention in both the breadth and depth of specialized medical vocabulary differ between bilingual and English-only instruction over the course of the three assessment points?

Methodology

Research Design

This study employed a longitudinal quasi-experimental design to investigate the impact of instructional language on the acquisition and retention of specialized medical vocabulary. The study is structured as a comparative analysis between two groups: a control group receiving English instruction (L2-only group) and an intervention group receiving Arabic and English instruction (bilingual group). The independent variable in this design is the language of instruction (monolingual vs bilingual), and the dependent variables are the breadth and depth of lexical knowledge. This approach allows the study to capture whether a word is recognized, as well as the quality and accuracy of students’ understanding. To isolate the effect of the L1 as a facilitator, both groups were balanced for English proficiency and utilized the same medical text. Additionally, acquisition is measured at three points: an immediate posttest (T1) for initial gain, a one-week delayed posttest (T2) for short-term retention, and a one-month delayed posttest (T3) for long-term retention.

The use of authentic medical texts with lexical familiarization (definitions) mimics the reading-to-learn environment typical of health education, which ensures the ecological validity of the findings. Regarding the measurement of depth, we followed the framework established and used in prior studies, and operationalized depth as the ability to recall specific meaning components of a word, which allows for a nuanced analysis compared to simple recognition tasks. The language used to assess participants’ vocabulary knowledge was carefully considered. All participants, regardless of group, completed the posttest and both delayed posttests in English only. This decision was made to align the testing procedures with the goals of the study, which focused on measuring participants’ understanding of specialized medical vocabulary in English (the language used in their academic program).

Participants

The participants were 50 Saudi male medical and health science students (ages 18–20) enrolled in a preparatory year program at a college using EMI. Having completed their first semester of intensive English, including courses in academic writing, reading, and oral skills, all participants had established a foundational level of proficiency. The study took place in a natural language classroom setting. A relatively homogeneous sample was selected to reduce variability associated with age, L1 background, and English proficiency. All participants were Saudi undergraduate students whose L1 is Arabic. To ensure group comparability, students were divided into the English-only group (n = 25) and the bilingual group (n = 25). Proficiency was controlled using vocabulary size test (VST),⁵¹ which assessed knowledge of the 5000 most frequent English words. Results indicated no significant difference in language proficiency between groups, with an average vocabulary size of 3550 words. The L2-only group received the learning material in English only, while the bilingual group received it through a bilingual approach (English and Arabic).

This study was approved by the institutional review board at King Abdullah International Medical Research Center at King Saud bin Abdulaziz University for Health Sciences (Protocol Number: NRR25/086/1). All ethical precautions were carefully considered. Students’ participation was voluntary, and they were notified of their freedom to withdraw at any time during the data collection process. The participants’ names and information were replaced with pseudonyms and pseudo-identifying information.

Materials

A 1000-word medical text titled “Surgery” was selected from The Language of Medicine in English.⁵² This text was chosen for its density of specialized vocabulary that was neither overly familiar nor obscure. The text was used without modification, and all specialized terms appeared in bold, as they did in the original source. The English definitions of these terms were taken directly from the same textbook in which the terms appeared. For the bilingual group, these definitions were translated into Arabic using medical sources to ensure accuracy and consistency. A lexical analysis was conducted using AntWordProfiler software against BNC/COCA headword lists. The analysis revealed that 94.29% of the words were within the top 5000 frequent words. This suggests that the text was moderately challenging yet accessible for the participants’ measured proficiency levels. A pilot test confirmed the readability of the material.

Target Words

30 specialized medical terms were identified in the text, 10 of which were randomly selected as the target words (TWs). The remaining 20 words served as distractors. There were no major differences between TWs and distractors in terms of difficulty or relevance to the context. Definitions of TWs were sourced from the original textbook and reliable medical sources to ensure accuracy. TWs were distributed randomly throughout the text to prevent participants from identifying them based solely on order of presentation. Moreover, the number of TWs’ appearances in the text was not controlled. This choice was made to preserve the natural flow of the reading material and maintain coherence with the context.

Procedure

The study was conducted over one month, including the pretest, intervention, and posttests. Prior to conducting the study, a pretest was administered to measure prior knowledge using a 3-point familiarity scale adapted from Paribakht and Wesche.⁵³ The results of the pretest confirmed that participants had zero knowledge of the 10 target words.

The intervention included a reading stage as well as an instructional stage. The intervention took place in a classroom setting over approximately 35 minutes. In the reading phase, both groups read the English text aloud. A teacher corrected mispronunciations but delayed explicit vocabulary instruction until after the reading phase to ensure students engaged with the terms in the context first. Following the reading phase, the instructional phase started, where students were provided with a vocabulary list of the 30 specialized terms. The L2-only group received definitions in English only, while the bilingual group received definitions in English and Arabic.

After the reading session, participants completed an immediate computer-based posttest after the initial acquisition that included 30 multiple-choice questions related to the content of the text. 10 of these questions focused on the meanings of the TWs, while the remaining 20 served as distractors. For example, questions followed the following format: “What is X (target word)?” Students were then required to select two correct meaning components from a list of five options. The purpose of this format (choosing two answers) was designed to go beyond surface-level recognition by requiring participants to identify multiple aspects of a term’s meaning. Such a task provides deeper insight into how well students had grasped the specialized vocabulary.^54,55 A second follow-up test was administered four weeks later, again using the same 30 questions without alteration. This third measurement point provided an additional check on longer-term retention of both breadth and depth of vocabulary knowledge, while still keeping testing conditions identical across all sessions.

Data Analysis

The analysis included descriptive and inferential statistical analyses to evaluate the impact of bilingual instruction on vocabulary breadth and depth. To answer the research questions on the effectiveness of bilingual and L2-only instruction on the breadth and depth of specialized medical vocabulary across three assessment intervals (T1, T2, and T3), the study employed descriptive statistics, linear mixed-effect models (LLM), and post Hoc comparisons.

We first calculated the means (M) and standard deviations (SD) for both vocabulary breadth and depth scores. This provided an initial overview of performance trends for each group at every time point. Additionally, mean retention rates were calculated to assess long-term stability across groups. Retention rates were operationalized as the percentage of vocabulary knowledge acquired at T1 that was maintained at the T2 and T3 intervals.

To account for the repeated measures, linear mixed-effects models were fitted for the first two research questions (breadth and depth). Each model included a random intercept for participants to control for within-subject correlation and variance across the assessment points. The models examined the following fixed effects: Instructional group (comparing the performance of the bilingual group against the L2-only group, time (evaluating changes across assessment intervals as a reference point to determine significant changes at T1 and T3), and interaction effects (assessing the interaction between group and time to determine if the instructional method influenced the trajectory of vocabulary acquisition and retention over time). Separate LLMs were fitted for vocabulary breadth scores, depth scores, and retention rates. Regarding retention, two linear mixed-effects models were used. It was measured as the percentage of initially acquired vocabulary (T1 scores) retained at one-week (T2) and one-month (T3) follow-ups, varied by instructional group (L2-only vs bilingual) for both breadth and depth of vocabulary knowledge.

Following the mixed-effects models, post hoc comparisons with Bonferroni corrections were conducted to provide more granular insight into group differences. The p-values reported for the post hoc comparisons throughout the manuscript are the adjusted p-values. Because the Bonferroni correction is already factored into these reported values, they are evaluated directly against the standard significance threshold of α = 0.05. These pairwise comparisons were utilized to identify statistically significant differences between the bilingual and L2-only groups at each time point (T1, T2, and T3) for both breadth and depth. Additionally, they isolate specific intervals where bilingual instruction provided a statistically robust advantage in knowledge retention.

Results

Descriptive Statistics

Table 1 displays the means and standard deviations for students’ vocabulary breadth and depth scores across three assessment points: T1, the one-week delayed T2, and the one-month delayed T3.

Table 1.

Means and Standard Deviations for Breadth and Depth Scores

Instruction Group	Time Point	Breadth Score		Depth Score
		M	SD	M	SD
L2-only	T1	8.88	1.39	14.92	4.51
	T2	8.08	1.96	12.44	5.54
	T3	8.00	3.14	11.20	5.87
Bilingual	T1	9.32	0.99	16.96	3.49
	T2	9.44	1.00	17.60	2.99
	T3	9.24	1.16	16.84	3.75

The findings indicate that the bilingual group consistently achieved higher vocabulary breadth scores than the L2-only group at all three time points (Figure 1). At T1, the bilingual group’s mean score (M = 9.32, SD = 0.99) exceeded that of the L2-only group (M = 8.88, SD = 1.39). This advantage persisted at T2, with the bilingual group exhibiting a slight increase (M = 9.44, SD = 1.00), while the L2-only group’s performance declined (M = 8.08, SD = 1.96). By T3, the bilingual group maintained stable performance (M = 9.24, SD = 1.16), whereas the L2-only group showed further deterioration (M = 8.00, SD = 3.14).

Figure 1.

Mean breadth scores.

A parallel pattern emerged for vocabulary depth (Figure 2). At T1, the bilingual group demonstrated superior depth knowledge (M = 16.96, SD = 3.49) compared to the L2-only group (M = 14.92, SD = 4.51). By T2, the bilingual group’s mean depth score increased (M = 17.60, SD = 2.99), while the L2-only group’s mean decreased (M = 12.44, SD = 5.54). At T3, the bilingual group exhibited only a marginal decline (M = 16.84, SD = 3.75), whereas the L2-only group’s scores continued to drop (M = 11.20, SD = 5.87).

Figure 2.

Mean depth scores.

Vocabulary Breadth

The results in Table 2 revealed a significant main effect of group (b = 1.36, p < 0.01), indicating that students in the bilingual group, on average, outperformed those in the L2-only group in vocabulary breadth. There was also a significant main effect of time at T1 (b = 0.80, p < 0.05), which suggests performance improved immediately after instruction compared to the reference time point (T2). The effect of Time at T3, however, was not significant (p = 0.83), indicating no meaningful change in scores relative to T2 at the one-month follow-up. The interaction between group and time was not statistically significant at either T1 (p = 0.08) or T3 (p = 0.82), though the interaction at T1 approached significance, hinting at a possible moderating effect of bilingual instruction immediately after the intervention.

Table 2.

Linear Mixed-Effects Model Results for Breadth Scores

	Coefficient	P-value
Intercept	8.08	<0.001
Group (bilingual)	1.36	0.007
Time (T1)	0.80	0.032
Time (T3)	−0.08	0.828
Group (bilingual) x Time (T1)	−0.92	0.080
Group (bilingual) x Time (T3)	−0.12	0.818

To further investigate these effects, post hoc comparisons with Bonferroni corrections were conducted to compare group differences at each time point. At T2, the bilingual group scored significantly higher in vocabulary breadth than the L2-only group (p < 0.01). A similar pattern was observed at T3, with the bilingual group again outperforming the L2-only group (p = 0.015). However, at T1, the group difference did not reach significance (p = 0.383). These findings suggest that while both groups performed similarly immediately after instruction, the bilingual group demonstrated significantly higher vocabulary breadth at the one-week and one-month follow-up assessments. This highlights the effectiveness of incorporating L1 definitions in enhancing learners’ acquisition of medical terminology over time.

Vocabulary Depth

Table 3 shows a significant main effect of group (b = 5.16, p < 0.001), with the bilingual group demonstrating significantly greater depth of vocabulary knowledge overall compared to the L2-only group. There was also a significant main effect of time at T1 (b = 2.48, p < 0.001), indicating an immediate improvement in vocabulary depth following instruction compared to the reference point (T2). The effect at T3 approached significance in the negative direction (b = −1.24, p = 0.08), suggesting a possible decline in depth over time, although this was not statistically significant. A significant interaction between Group and Time was found at T1 (b = −3.12, p < 0.01), indicating that the immediate gains in vocabulary depth were smaller for the bilingual group compared to the L2-only group. No significant interaction was observed at T3 (p = 0.629), suggesting that group differences at the one-month mark were not moderated by time.

Table 3.

Linear Mixed-Effects Model Results for Depth Scores

	Coefficient	P-value
Intercept	12.44	<0.001
Group (bilingual)	5.16	<0.001
Time (T1)	2.48	<0.001
Time (T3)	−1.24	0.080
Group (bilingual) x Time (T1)	−3.12	0.002
Group (bilingual) x Time (T3)	0.48	0.629

Post hoc comparisons with Bonferroni corrections provided further insight into the group differences at each time point. At T2, the bilingual group scored significantly higher in vocabulary depth than the L2-only group (Mean difference = 5.16, p < 0.001). This difference became even more pronounced at T3, where the bilingual group maintained superior performance compared to the L2-only group (Mean difference = 5.64, p < 0.001). However, at T1, the group difference did not reach statistical significance (p = 0.112). Overall, these findings indicate that while both groups showed gains in vocabulary depth immediately after instruction, the bilingual group consistently outperformed the L2-only group at the one-week and one-month follow-up assessments. This underscores the sustained effectiveness of bilingual instruction alongside L2 ones.

Vocabulary Retention Rates

In relation to breadth retention, the model revealed a marginally significant main effect of Group (b = 10.03, p = 0.08), which indicates a trend toward higher retention in the bilingual group compared to the L2-only group, although this did not reach conventional levels of statistical significance. The main effect of Time (T3 vs T2) was not significant (p = 0.68), nor was the Group × Time interaction (p = 0.92), suggesting that the rate of decline in breadth retention over time was similar for both groups. Descriptive statistics supported this trend (Table 4), as the bilingual group showed slightly higher retention rates at both T2 (Mean = 101.96, SD = 12.05) and T3 (Mean = 99.56, SD = 11.36) compared to the L2-only group (Mean = 91.93, SD = 21.18 at T2; Mean = 90.14, SD = 30.19 at T3) (See Figures 3 and 4). These results suggest that while the bilingual instruction may have had a positive effect on breadth retention, the effect was not statistically robust.

Table 4.

Means and Standard Deviations for Breadth and Depth Retention Rates

Group	Time	Breadth Retention Rate		Depth Retention Rate
		M	SD	M	SD
L2-only	T2	91.93	21.18	83.68	25.32
	T3	90.14	30.19	75.39	30.23
Bilingual	T2	101.96	12.05	106.24	22.09
	T3	99.56	11.36	100.23	17.51

Figure 3.

Mean breadth retention rates.

Figure 4.

Mean depth retention rates.

For depth retention, Table 5 demonstrates that the linear mixed-effects model revealed a significant main effect of Group (b = 22.56, p < 0.01), indicating that students in the bilingual group retained significantly more in-depth knowledge of medical vocabulary than those in the L2-only group. The main effect of Time approached significance (b = −8.29, p = 0.146), suggesting a possible decline in retention over time, although this was not statistically conclusive. The Group × Time interaction was not significant (p = 0.776), which indicates that the observed group differences in depth retention were consistent across both time points.

Table 5.

Linear Mixed-Effects Model Results for Breadth and Depth Retention Rate

	Breadth Model		Depth Model
	Coefficient	P-value	Coefficient	P-value
Intercept	91.93	<0.001	83.68	<0.001
Group (Bilingual)	10.03	0.077	22.56	0.001
Time (T3)	−1.79	0.677	−8.29	0.146
Group (Bilingual) x Time (T3)	−0.61	0.920	2.27	0.776

Post hoc comparisons with Bonferroni corrections confirmed that the bilingual group outperformed the L2-only group in depth retention at both time points. At T2, the bilingual group retained significantly more in-depth knowledge than the L2-only group (Mean difference = 22.6, p < 0.01), and this difference remained significant at T3 (Mean difference = 24.8, p < 0.001). These findings indicate that bilingual instruction significantly enhanced students’ ability to retain an in-depth understanding of specialized medical vocabulary over time. While improvements in breadth retention were observed in the bilingual group, the advantage was more clearly pronounced and statistically robust in measures of depth retention.

Discussion

The study demonstrates the effectiveness of bilingual scaffolding by providing essential cognitive support and enhancing semantic understanding. Based on CLT,¹⁶ learning is limited by working memory capacity, and Arabic definitions likely reduced the extraneous load that overwhelms L2 learners when dealing with complex medical terms. This allowed students to focus more on building new knowledge. The scaffolding also provided a bridge to familiar L1 schemas, facilitating access to prior knowledge, supported by Yuan and Tang,¹⁸ who found that L1-bilingual conditions decrease cognitive processing demands. This pedagogical approach deepens semantic understanding by helping students develop high-quality mental representations.²⁹ Using L1 definitions may have provided semantic clarity and bridged word- and text-level understanding. It also acknowledges learners’ linguistic backgrounds, creating a more comfortable environment and reducing anxiety in EMI contexts. The bilingual group’s consistent scores suggest a more uniform learning experience that lessened misunderstandings and increased confidence.

RQ1: Vocabulary Breadth

The findings related to vocabulary breadth suggest that bilingual instruction may offer measurable benefits for learning specialized medical vocabulary in an EMI setting. Across all three time points (T1, T2, and T3), the bilingual group consistently outperformed the English-only L2-only group. This consistent performance advantage, supported by a significant main effect for Group, indicates that the inclusion of learners’ L1 may facilitate broader word recognition and understanding when processing new medical terms. Although the L2-only group showed a slight decline between the immediate and one-week tests, the bilingual group exhibited a modest gain during this period and sustained a gradual increase into the third time point. These findings align with previous research suggesting that bilingual support improves learners’ ability to understand and retain new vocabulary,^38,39 particularly when dealing with content-rich domains such as medicine.

The bilingual group’s advantage may stem from the better access to word meanings via Arabic definitions and explanations. Nation⁷ states that effective vocabulary learning involves repeated and meaningful exposure to words in context. The bilingual support likely eased the cognitive effort of understanding L2 words while still engaging with English texts. Perfetti’s lexical quality hypothesis²⁹ suggests that strong mental word representations lead to successful vocabulary learning and reading comprehension. The L1 scaffolding in this study likely helped build these representations by providing Arabic definitions, giving semantic clarity for complex medical terms. This aligns with Perfetti’s idea that word processing and text comprehension are interconnected; with a clearer understanding of individual words, students can better grasp the overall medical text.

It is important to interpret these findings with measured optimism. While the bilingual group consistently outperformed the L2-only group, the interaction between Group and Time was not statistically significant. This may suggest that the rate of improvement over time did not differ meaningfully between groups, even though the bilingual group maintained a clear overall advantage. Post hoc Bonferroni comparisons clarified this pattern, revealing that while no significant group difference was found at T1, the bilingual group outperformed the L2-only group at both T2 and T3. This suggests that the advantage emerged gradually and became more pronounced over time, reflecting a widening performance gap as students progressed beyond immediate recall. One possible interpretation is that bilingual support helped learners consolidate and retain more vocabulary knowledge over time, particularly as they re-encountered or processed the material more deeply after initial exposure. Additionally, external factors such as individual learner differences, reading motivation, and prior exposure to English may have influenced the trajectory of growth. Nonetheless, these results highlight the potential value of incorporating bilingual elements into vocabulary instruction for specialized academic content. As this study focused on EMI learners in a medical context, where terminology is often abstract and discipline-specific, the use of Arabic may have played a key role in making new vocabulary more accessible and learnable over time.^9,43

RQ2: Vocabulary Depth

The results related to vocabulary depth further support the effectiveness of bilingual instruction in enhancing learners’ specialized vocabulary knowledge. Although both groups demonstrated gains immediately after instruction, the bilingual group consistently outperformed the L2-only group across all time points, with statistically significant differences observed at both the one-week and one-month delayed assessments. While the difference at T1 did not reach statistical significance, the bilingual group’s higher mean scores suggest that initial advantages may have existed but required time to consolidate. The main effect of the group was significant, which indicates a clear overall advantage for the bilingual approach in fostering deeper word knowledge. This trend reinforces prior research that highlights how bilingual strategies, particularly those incorporating L1 meaning explanations, can strengthen learners’ ability to grasp and retain more nuanced dimensions of vocabulary.¹⁰

One notable aspect of the results is the stability of the bilingual group’s depth performance across time, contrasted with the L2-only group’s steady decline. This may suggest that bilingual support not only enhances depth of understanding during the learning phase but also helps preserve this knowledge over time. Given that depth involves knowledge of a word’s form, meaning components, and usage as discussed by Laufer and Hulstijn,⁵⁶ learners in the bilingual group may have benefited from greater semantic clarity and richer mental representations of the terms they encountered. In contexts like medicine, where vocabulary often includes polysemous or morphologically complex items, such clarity can be essential. Bravo and Cervetti⁹ point out that technical words often require both content familiarity and linguistic support, and the bilingual format used in this study may have provided the necessary scaffolding to bridge both.

The significant interaction at T1 suggests that the L2-only group may have experienced a sharper immediate gain, possibly due to greater focus on the L2 input during instruction. However, this gain was not sustained over time, and the bilingual group’s performance remained more consistent across delayed assessments. This pattern was further confirmed through post hoc Bonferroni comparisons, which showed no significant group difference at T1, but significant differences favoring the bilingual group at both T2 and T3. These results imply that while English-only instruction may yield short-term improvements, bilingual instruction appears to foster more enduring depth of understanding. As Liu³⁷ and Al-Ahdal and Aljabr⁸ have noted, the L1 can serve as a cultural and cognitive tool that enhances learners’ engagement with specialized content, making abstract or unfamiliar terms more comprehensible. For learners navigating complex academic material in a second language, this kind of support may be especially critical in achieving deeper levels of lexical mastery.

RQ3: Vocabulary Retention

The findings related to vocabulary retention offer important insight into how instructional approaches influence the sustainability of vocabulary learning. Overall, bilingual instruction appeared to provide learners with a retention advantage, particularly in relation to the depth of specialized vocabulary knowledge. While both groups experienced some degree of decline over time, the bilingual group consistently demonstrated higher retention rates than the L2-only group across the one-week and one-month follow-up assessments. This trend was particularly robust in the depth scores, where the bilingual group significantly outperformed the L2-only group at both time points.

These patterns align with cognitive load theory,¹⁶ which states learning is limited by working memory capacity, and effective instruction reduces extraneous load to help learners process and integrate information. The L2-only group may have faced higher extraneous load processing complex medical terms entirely in L2, explaining their quick initial gains but steep decline, indicating less stable learning. Conversely, the bilingual group had Arabic definitions, likely reducing cognitive burden and allowing more resources for lexical learning, resulting in better retention. Research supports this: Liao⁵⁷ found bilingual subtitles reduce cognitive load and improve retention and understanding. Li et al⁵⁸ showed multimodal bilingual input enhances vocabulary by easing processing demands, especially in language-heavy areas. These findings suggest bilingual scaffolding stabilizes learning, especially for deep understanding requiring semantic clarity. Remarkably, the bilingual group’s retention for both breadth and depth exceeded 100% at T2 and stayed near or above 100% at T3, indicating they may have continued consolidating or expanding their vocabulary after instruction. This underscores the lasting effect of bilingual support on lexical acquisition.

In contrast, the difference in breadth retention between the two groups was more modest. Although the bilingual group maintained slightly higher retention rates across both follow-up periods, these differences did not reach statistical significance. This may indicate that while bilingual instruction offers some benefit in helping learners recall broader lexical meanings, its impact is more pronounced when learners are required to retain multiple layers of a word’s form, meaning, and use, as is the case with vocabulary depth. This distinction highlights the cognitive complexity of depth knowledge and the ways in which instructional scaffolding through L1 can reinforce that complexity, as Bhooth et al⁵⁹ demonstrated the role of L1 in enabling learners to operate within their ZPD by facilitating deeper lexical processing.

The absence of a significant interaction between Group and Time indicates the bilingual group’s advantage was stable over time, not widening relative to the L2-only group. They consistently retained more at T2 and T3, indicating the development of durable mental representations, especially for depth knowledge. The lower variability in scores implies a more uniform learning experience, which supports the idea that bilingual explanations reduce misunderstandings and boost confidence. These findings align with research showing that access to linguistic and conceptual scaffolding helps in long-term vocabulary retention.^8,10,37,44 Overall, bilingual instruction has a sustained effect, enhancing long-term retention, particularly in more complex word knowledge.

Limitations

While the main findings support bilingual instruction’s effectiveness, several factors affect interpretation. All participants were male students with similar backgrounds, limiting generalizability to other groups. Different backgrounds might influence vocabulary acquisition approaches. The small sample size, typical for medical English classes, may not represent broader populations. Although baseline English proficiency was controlled via VST, individual differences could influence outcomes. The proficiency assessment focused on common words and showed no group differences, but limited variation may restrict applicability to learners with different levels. The vocabulary depth assessment was consistent but might not reflect students’ full lexical knowledge. Future research could include larger and more diverse samples and examine how proficiency levels or various bilingual supports impact comprehension and retention.

Implications

These findings contribute to the discourse on bilingual education by providing longitudinal evidence on the effectiveness of L1 for L2 learning, in line with prior studies.³⁷ While L1 support may offer initial gains comparable to L2-only instruction, it appears to be more critical for long-term mastery of specialized knowledge. Furthermore, by demonstrating that the most pronounced bilingual scaffolding effects occurred in vocabulary depth, this study provides an empirical example of how Vygotsky’s ZPD¹⁵ can be operationalized within specialized academic contexts. L1 can serve as a helpful mediating tool, aiding the development of robust mental representations.²⁹ These outcomes suggest that instead of solely using L2 in EMI, incorporating translanguaging approaches could improve cognitive efficiency and academic equity in professional education.⁶⁰

The findings of this study contribute to the theoretical understanding of vocabulary acquisition in EMI contexts by elucidating the role of bilingual instruction in enhancing vocabulary breadth, depth, and retention. The results support the involvement load hypothesis,⁵⁶ which suggests that tasks requiring higher cognitive involvement, such as those incorporating L1 scaffolding, can lead to more effective vocabulary learning. The study also resonates with the findings of Xue et al,⁶¹ who demonstrated that bilingual experience, particularly through L1 associations, enhances novel word learning by leveraging existing linguistic frameworks. These theoretical insights underscore the cognitive advantages of integrating the L1 in EMI settings, particularly in complex domains such as medical education.

The practical implications of this research are especially relevant for instructional design in content-rich EMI programs. Gablasova¹⁰ found that even advanced L2 learners struggle with specialized vocabulary, which can limit their academic engagement. Consistent with this, the present study supports the integration of bilingual materials as a pedagogical strategy. Dillon et al⁶⁰ showed university students in EMI performed better on comprehension and translation with dual-language materials, highlighting the value of translanguaging in reducing cognitive load and improving outcomes, which aligns with vocabulary gains in this study’s bilingual group. These findings advocate incorporating bilingual scaffolding in EMI, especially in medical and technical fields where content complexity can hinder learning. L1 support should be seen as a strategy to improve access and equity, not a compromise to immersion.

Conclusion

This study was set to examine the effects of bilingual versus English-only instruction on the breadth, depth, and retention of specialized medical vocabulary in an EMI context. The results clearly indicate that bilingual support positively influenced students’ vocabulary learning, with the most pronounced and sustained effects observed in depth and long-term retention. While both groups showed some improvement, the bilingual group consistently outperformed the L2-only group across key measures, highlighting the pedagogical value of integrating L1 resources into EMI environments. These findings contribute meaningfully to theories of vocabulary acquisition and offer practical direction for designing more inclusive and effective content instruction. By affirming the cognitive and academic advantages of bilingual scaffolding, the study reinforces the role of translanguaging as a legitimate and beneficial practice within second-language education.

Funding Statement

This research did not receive any funding.

AI Disclosure

The authors disclose the use of AI, specifically Grammarly, to check for grammar and enhance readability.

Data Sharing Statement

Data is available upon reasonable request, subject to the research center’s approval.

Ethical Statement

All participants provided their informed consent to participate in the study.

Disclosure

The authors declare no competing interests in this work.