• Login
• Admin

Cao, Hong Phat (2025) Predicting academic achievement in English medium instruction in Vietnam: a multidimensional approach using standardised testing and vocabulary assessment. PhD thesis, University of Nottingham.

PDF (Thesis - as examined) - Repository staff only until 25 July 2027. Subsequently available to Anyone - Requires a PDF viewer such as GSview, Xpdf or Adobe Acrobat Reader Available under Licence Creative Commons Attribution. Download (3MB)

Abstract

English-medium instruction (EMI) programmes are increasingly central to higher education (HE) worldwide, fostering internationalisation, academic collaboration, and global workforce readiness (Lasagabaster, 2022; Macaro et al., 2018). Vietnam, striving to align with this trend, faces challenges, such as the absence of consistent and standardised guidelines for using English language proficiency (ELP) assessments for admission purposes (e.g., Dang & Moskovsky, 2023; Nguyen et al., 2017; Tran & Nguyen, 2018). This thesis addresses these challenges by focusing on the role of ELP assessments–specifically the Pearson Test of English Academic (PTEA)–as tools for evaluating students’ academic readiness in EMI settings. Although widely adopted in recent years (Pearson-PTE, n.d.), the PTEA’s construct validity remains underexplored (Durrant et al., 2015; Riazi, 2013). This raises questions about its effectiveness, as meta-analyses consistently highlight the limited predictive power of standardised ELP tests (e.g., Abunawas, 2014; Gagen & Faez, 2024; Ihlenfeld & Rios, 2023). Through three interconnected studies–a systematic review of the PTEA’s construct validity, an empirical investigation into its predictive validity, and an analysis of lexical complexity in academic writing–this thesis highlights the significance of understanding the constructs measured by the PTEA alongside predictors of academic achievement within EMI programmes.

Chapter 2 serves as the first systematic review of the PTEA’s construct validity, conceptualising it as a unitary concept (Messick, 1989). Construct validity is essential for global-scale ELP tests like the PTEA to ensure accurate measurement of theoretical concepts. While numerous studies address different aspects of the PTEA’s construct validity, the results remain mixed. Using a systematic review approach, study 1 presented in this chapter evaluated 40 PTEA construct validity studies, focusing on research aims, construct characteristics, and methodologies. The findings revealed a predominant focus on a combination of language-oriented and trait-oriented constructs (e.g., gender factors and cognitive demand), with quantitative analyses being primarily employed. Discrepancies in the conceptualisation of constructs were noted, leading to variations in the theoretical frameworks adopted, such as Weir’s (2005) socio-cognitive model and argument-based validation approaches (Chapelle et al., 2008; Kane, 1992). The systematic review also identifies gaps in linking PTEA outcomes to academic achievement and underscores the need for robust, ethically diverse, and context-sensitive data collection to address test-taker variability and enhance construct validity.

Chapter 3 builds on the gaps identified in Chapter 2 by exploring the predictive validity of the PTEA in the Vietnamese EFL context. The second study examines how language proficiency (PTEA scores), vocabulary knowledge (vocabulary size test-VST, Nation & Beglar, 2007), and individual factors (autonomous learning activities, language learning attitudes and course perceptions) predict academic achievement, measured through grade point averages (GPAs). Regression analyses performed on the PTEA and VST scores of 105 undergraduate students at a teacher-training institution in southern Vietnam revealed that language proficiency and vocabulary measures together moderately predicted academic achievement, with no significant difference between using either measure independently. Additionally, partial least squares structural equation modelling (PLS-SEM) demonstrated that individual factors mediated by linguistic measures played a negligible role in predicting GPAs, highlighting the complex interplay between language-related and non-language-related constructs, as well as the multifaceted nature of academic achievement (e.g., Alamer, 2022; Bo et al., 2023). However, individual factors were found to be stronger predictors of language proficiency and vocabulary knowledge. The study calls for greater attention to language-related factors in predicting academic achievement and recommends including additional variables to better understand the multifaceted nature of academic achievement in EMI settings.

Chapter 4 investigates the role of lexical complexity in writing proficiency and academic achievement, focusing on lexical diversity and sophistication as predictors. While lexical complexity is well-recognised in the wring component of standardised ELP tests (e.g., Banerjee et al., 2007; Riazi & Knox, 2013; Nguyen et al., 2022), its role in the PTEA is underexplored (e.g., Treffers-Daller et al., 2018; Tsagari & Demetriou, 2022). Research on the predictive power of lexical complexity for academic achievement seems even scarcer (e.g., Laufer, 1994; Morris, 2003, Morris & Cobb, 2004). Therefore, the third study analyses Vietnamese EMI students’ written essays of the PTEA, examining how the lexical dimensions explain variance in writing scores and GPAs. An analysis of 102 writing samples showed that lexical diversity and sophistication moderately predicted writing proficiency but had a limited impact on academic achievement. Notably, the combination of diverse and sophisticated vocabulary was crucial for writing proficiency and, to a lesser extent, academic achievement. Advanced vocabulary usage, measured using the Common European Framework of Reference for Languages (CEFR) classifications, correlated with higher writing scores. While lexical measures are crucial for writing proficiency, their impact on broader academic outcomes appears mediated by other contributing factors due to the complex nature of academic achievement. This study provides evidence-based recommendations, such as incorporating advanced lexical items into teaching materials, prioritising academic vocabulary development, and employing a combination of lexical sophistication and diversity measures in assessment frameworks to enhance language instruction and assessment practices in EMI contexts.

Chapter 5 synthesises the findings of the three studies, discussing their theoretical, methodological, and pedagogical contributions to language assessment and EMI research. The systematic review establishes construct validity as a cornerstone for evaluating the PTEA, providing a robust theoretical foundation for interpreting its scores. The two empirical studies extend this understanding by demonstrating the predictive relationships between linguistic and non-linguistic factors and academic achievement. Theoretically, the thesis advances the understanding of construct validity for the PTEA, offering insights into its design and application in assessing academic readiness. The thesis confirms that while lexical complexity and individual factors significantly predict linguistic elements (e.g., writing and PTEA scores), they do not extend as effectively to academic achievement. Moreover, the modest predictive power of lexical complexity and individual factors for academic achievement also challenges the existing literature (e.g., Curle et al., 2024; Laufer, 1994; Morris, 2003; Morris & Cobb, 2004; Soruç et al., 2024; Thompson et al., 2022; Yuksel et al., 2023). Methodologically, the thesis pioneers the application of PLS-SEM in exploring academic achievement predictors in EMI settings, offering a nuanced approach to analysing the interconnections between multiple variables. Pedagogically, the findings inform the design of language support programmes, advocating for a holistic approach that integrates linguistic and affective dimensions to enhance students’ academic readiness. Despite its contributions, the thesis critically acknowledges the limitations while presenting the potential directions for prospective studies. Taken together, the studies presented in this thesis provide a deeper understanding of the extent ELP tests, particularly the PTEA, measure the intended constructs that students are supposed to demonstrate in HE settings as well as the predictors of academic achievement in Vietnamese EMI programmes.

Item Type:	Thesis (University of Nottingham only) (PhD)
Supervisors:	Szabo, Csaba Zoltan Lee, Barry Reynolds Beh, Anthony Kuan Jun
Keywords:	construct validity; validation; PTE academic ;systematic review; language proficiency; academic achievement; predictive validity; standardised testing; EMI, PLS-SEM; vocabulary assessment; lexical complexity; lexical diversity; lexical sophistication; writing proficiency
Subjects:	P Language and literature > PE English
Faculties/Schools:	University of Nottingham, Malaysia > Faculty of Arts and Social Sciences > School of Education
Item ID:	81259
Depositing User:	Cao, Hong
Date Deposited:	26 Jul 2025 04:40
Last Modified:	26 Jul 2025 04:40
URI:	https://eprints.nottingham.ac.uk/id/eprint/81259

Actions (Archive Staff Only)

Edit View