Chinese university EFL learners’ English for General Academic Purposes: relationships between target needs and self-efficacy

Chinese higher education institutions are increasingly aiming to participate in the global academic community, driving demand for English for Academic Purposes (EAP), and particularly English for General Academic Purposes (EGAP), to complement General English. EGAP targets academic English skills such as academic listening, writing, reference skills, and others. Using Hutchinson and Waters’ target needs analysis (TNA: necessities, lacks, wants) as a framework, and building on evidence that learning goals can shape self-efficacy and that self-efficacy predicts performance, this study examines university EFL learners’ target needs and academic self-efficacy in EGAP and explores their interrelationships across majors and institution types in China. Research questions: (1) What are the characteristics of learners’ target needs and self-efficacy in EGAP? (2) What are the relationships between learners’ target needs and self-efficacy in EGAP?

Language learners’ needs have been central in ESP/EAP curriculum design, with distinctions between perceived vs felt needs and target-situation vs learning-process needs. Hutchinson and Waters (1987) proposed a TNA comprising necessities (required skills/knowledge for the target situation), lacks (gaps between current and target proficiency), and wants (self-perceived necessities). TNA integrates TSA and PSA perspectives and has been widely applied. Academic self-efficacy—beliefs about one’s capability to perform academic tasks—predicts achievement and in language learning has been studied across skills (reading, speaking, writing, listening). Bandura’s sources of self-efficacy include mastery and goal-setting; learners’ needs and goals can shape efficacy. Empirical work suggests positive links between needs and self-efficacy, but relationships between target needs (necessities, lacks, wants) and EGAP self-efficacy require further evidence. In China, college English historically emphasized General English, with EGAP marginalized. With internationalization and increased English-medium/bilingual offerings, EGAP is being introduced to supplement GE, especially in research-oriented universities. This context motivates examining learners’ EGAP needs and self-efficacy across institution types and majors.

Design: Cross-sectional survey with validated scales and structural equation modeling (SEM). Participants: Convenience sample of 1340 undergraduates from four universities in Shandong Province, China: 888 (66.3%) from a key national research-oriented university and 452 (33.7%) from provincial teaching-oriented universities. Majors: arts (n=247, 18.4%), science (n=320, 23.9%), engineering (n=580, 43.3%), medicine (n=193, 14.4%). Grades: freshmen 61.8% (n=828), sophomores 38.2% (n=512). Gender: 50.7% male (n=679), 49.3% female (n=661). Instruments: (1) Target Needs Analysis (TNA) scale based on Hutchinson & Waters (1987) and Cai (2012): three subscales—necessities (6 items; e.g., need to read professional literature in English), lacks (14 items; e.g., difficulty joining academic discussions), wants (18 items; e.g., desire to write a research proposal in English via EGAP). 5-point Likert (1–5). Higher scores indicate higher needs. (2) Learner Academic Self-efficacy (LAS) scale (initially 34 items) based on BALEAP Can Do Framework (2013), Smith & Thondhlana (2015), and Cai (2012): five dimensions—listening (3 items), speaking (6), reading (3), writing (6), study skills (4). 5-point Likert; higher scores indicate greater self-efficacy. Data analysis: EFA (principal components, varimax) and internal reliability (Cronbach’s α) using SPSS 26; CFA and SEM using Amos 26. Descriptive statistics and correlations; group comparisons via independent-samples t-tests and ANOVA (gender, grade, major, institution type). Model fit interpreted with CFI/TLI (>0.90 acceptable) and RMSEA (<0.10 acceptable). Effect sizes interpreted following Gignac & Szodorai (2016). Ethical approval was obtained; participation was voluntary.

Validity and reliability: TNA scale—KMO=0.959; Bartlett’s test χ²(703, N=1340)=42059.384, p<0.001. EFA on 38 items extracted three factors consistent with necessities, lacks, wants; one item (“study abroad”) dropped for loading <0.40. Final 37 items had loadings 0.41–0.79; variance explained 61.36%; α: necessities 0.89, lacks 0.91, wants 0.97. CFA fit: χ²=358.184, df=41, p<0.001; CFI=0.97, TLI=0.96, RMSEA=0.076. LAS scale—KMO=0.947; Bartlett’s χ²(231, N=1340)=21470.951, p<0.001. After EFA, 22 items loaded 0.50–0.89 on five factors (listening, speaking, reading, writing, study skills); variance explained 72.98%; α: listening 0.89, speaking 0.93, reading 0.82, writing 0.91, study skills 0.76. CFA fit: χ²=2097.911, df=199, p<0.001; CFI=0.91, TLI=0.90, RMSEA=0.084. Descriptive statistics: Target needs means (SD): necessities 3.59 (0.80), lacks 3.80 (0.75), wants 3.95 (0.75) – all above the midpoint. Self-efficacy means (SD): listening 2.40 (0.85), speaking 2.87 (0.83), reading 2.95 (0.83), writing 2.59 (0.79), study skills 3.42 (0.68). Correlations: Within needs—necessities correlated positively with wants (r=0.60, p<0.01, large) and negatively with lacks (r=-0.07, p<0.01, very small); lacks showed a small positive correlation with wants (r=0.11, p<0.01). Within self-efficacy—dimensions were positively intercorrelated with large effect sizes. Between needs and self-efficacy—necessities correlated positively with all self-efficacy factors (medium), except writing was smaller (r=0.16, p<0.01). Lacks correlated negatively with all self-efficacy factors: listening r=-0.34, speaking r=-0.38, reading r=-0.34, writing r=-0.39 (all p<0.01; large), and study skills r=-0.08 (p<0.01; very small). Wants correlated positively with self-efficacy, strongest for study skills (r=0.30, p<0.01; large). Group differences: Gender—females scored higher than males on necessities (t(1338)=-4.21, p<0.001) and wants (t(1338)=-4.81, p<0.001). Institution type—key university students scored higher than ordinary university students on necessities (t(1338)=8.77, p<0.001), lacks (t(1338)=3.77, p<0.001), wants (f(1338)=14.62, p<0.001), and self-efficacy in study skills (t(1338)=-4.15, p<0.001) with small to medium effects; ordinary university students had higher writing self-efficacy (f(1338)=-4.04, p<0.001). Grade—sophomores reported higher lacks than freshmen (d≈0.29). Major—differences in lacks (F(3,1336)=11.30, p<0.001), wants (F(3,1336)=6.55, p<0.001), listening self-efficacy (F(3,1336)=4.78, p<0.01), and writing self-efficacy (F(3,1336)=6.51, p<0.001); medical students scored higher on lacks and wants but lower on listening and writing self-efficacy (small effects). SEM relationships: Model fit acceptable: χ²=3064.855, df=467, p<0.01; CFI=0.92, TLI=0.91, RMSEA=0.064; R² for self-efficacy factors 0.15–0.25. Necessities positively predicted all self-efficacy dimensions. Lacks negatively predicted all self-efficacy dimensions. Wants negatively predicted listening (β=-0.09, p<0.05), positively predicted reading (β=0.15, p<0.001), speaking (β=0.08, p<0.05), and study skills (β=0.29, p<0.001). The path from wants to writing self-efficacy was nonsignificant and removed.

The study shows that Chinese university EFL learners perceive strong EGAP target needs (necessities and wants) but report relatively low self-efficacy in core academic language skills, with comparatively higher self-efficacy in study skills. This aligns with limited prior exposure to EGAP before university and the challenge of academic language tasks. Differences across institutions suggest research-oriented universities cultivate clearer EGAP expectations and environments, raising perceived needs and some efficacy (study skills), while students in teaching-oriented universities felt more confident in writing. Medical students manifested the highest needs (necessities/wants) yet the lowest listening and writing self-efficacy, likely reflecting the complexity of medical English. The SEM results address the second research question: higher perceived necessities are associated with greater academic self-efficacy, supporting the motivational role of specific, challenging goals. Conversely, greater perceived lacks undermine self-efficacy, consistent with the demotivating effect of wider gaps between current and target proficiency. Learners’ wants generally bolster self-efficacy—especially for reading and study skills—reflecting the positive influence of self-set goals and motivation, with the exception of a weak negative link to listening self-efficacy. Together, the findings underscore the value of aligning EGAP course design with learners’ articulated needs while scaffolding to reduce perceived gaps to strengthen efficacy.

This study contributes empirical evidence on Chinese university EFL learners’ EGAP target needs and academic self-efficacy and clarifies how necessities, lacks, and wants relate to efficacy across skills. Learners reported high target needs yet low self-efficacy, with institution type and major shaping both. SEM confirmed that necessities positively and lacks negatively predict self-efficacy, while wants show domain-specific positive associations. Practical implications include creating EGAP-rich environments, enhancing perceived relevance, supporting student goal-setting, and transforming perceived lacks into structured, attainable subgoals to build efficacy. Future research should employ longitudinal or experimental designs to establish causality between needs and self-efficacy and broaden sampling beyond a single province to enhance generalizability, potentially using stratified or multi-site national samples.

The cross-sectional, exploratory design precludes causal inference regarding the relationships between target needs and self-efficacy. The convenience sample from one economically developed coastal province (Shandong) limits generalizability across regions with differing resources and educational contexts. Future studies should use stratified sampling and longitudinal/experimental designs to test causal pathways and validate findings nationally.