Cross-cultural adaptation and validation of the clinical learning evaluation questionnaire with Chinese clinical interns

Abstract

Background

This study aimed to evaluate the reliability and validity of an adapted English questionnaire in Chinese medical schools, assessing its cross-cultural applicability in China.

Methods

A survey was conducted among clinical medical interns from four medical schools in China, collecting 216 valid responses. The questionnaire, based on the latest version of CLEQ, consisted of four dimensions and 18 items. It was translated into Chinese through a six-step forward and backward translation process. Data were analyzed using IBM SPSS Statistics 26 and IBM SPSS AMOS 28 Graphics.

Results

The findings demonstrate that the translated questionnaire is suitable for Chinese clinical medical interns. Except for the “Learning Motivation” dimension with a reliability coefficient of 0.760, all other dimensions scored between 0.8 and 1, indicating strong internal consistency and reliability. Correlation coefficients exceeded 0.5, confirming good test-retest reliability. Model fit indices indicated good compatibility (CMIN/DF = 1.749, RMSEA = 0.057, and IFI, TLI, CFI > 0.9). Validity testing showed that all dimensions, except “Learning Motivation,” had AVE values above 0.5 and CR values above 0.7,indicating good convergent validity and composite reliability. The discriminant validity of all dimensions was confirmed, as standardized correlation coefficients between each pair remained below the square root of their respective average variance extracted (AVE) values.

Conclusion

The four-factor CLEQ questionnaire shows good validity and reliability among Chinese clinical interns.

Clinical trial number

Not applicable.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12909-025-07088-9.

Introduction

Clinical learning is crucial in medical education, where students develop essential skills like history-taking, physical examination, clinical reasoning, decision-making, empathy, and professionalism [1]. However, the effectiveness of clinical education is often hindered by various challenges, making comprehensive and scientific evaluation essential to improving clinical learning quality.

Many tools [2–4] often fail to address organizational issues or the authenticity of clinical experiences. Skills-based tools (e.g.,OSCE) quantify performance through standardized scenarios, but their high cost and detachment from real clinical settings limit their ability to comprehensively assess the effects of long-term clinical practice [5]. In China, no widely validated evaluation tools exist. Researchers often design their own questionnaires, which lack scientific validation. Cultural differences, educational models, and healthcare systems also significantly influence the design, applicability, and effectiveness of evaluation tools.

The Clinical Learning Evaluation Questionnaire (CLEQ), developed by AlHaqwi AI [6] et al. in 2014,initially included 40 items covering five dimensions: clinical cases, authenticity of experiences, supervision, organization of patient encounters, and learning skills. In 2020, Nuha Alnaami [7] et al. revised the questionnaire into a simplified version with four dimensions and 18 items, validated in Iran. Previous validations of the CLEQ have demonstrated strong reliability (α = 0.87–0.92) and validity across diverse settings [7, 8], making it a scientifically sound choice for adaptation. The CLEQ’s four-dimensional structure (Clinical Cases, Organization of Patient Encounters, Supervision, and Learning Motivation) comprehensively addresses key components of clinical learning. This framework aligns well with the competency-based medical education reforms currently implemented in China [9]. The CLEQ’s brevity and focus on actionable aspects of clinical learning make it ideal for implementation in China’s large-scale medical education system, where efficient assessment tools are needed for quality improvement. However, CLEQ has not been applied in China. Although the medical education systems in China and Saudi Arabia differ in teaching methods, language requirements, clinical practice, and admission standards, they share a similar structure, including basic medical studies, clinical training, and residency programs. In recent years, China has reformed its medical education to focus more on clinical and practical skills, while Saudi Arabia emphasizes early clinical practice and conducts most of its courses in English. The CLEQ questionnaire from Saudi Arabia works in China but faces cultural and language challenges.

This study aimed to collect CLEQ data, evaluate its cross-cultural adaptability in Chinese medical education, and provide a scientific basis for the localized application of clinical learning evaluation tools. Specifically, it addressed two questions:

1. Does the CLEQ questionnaire demonstrate validity and reliability among clinical interns in China?

2. How to adapt the 18-item Clinical Learning Evaluation Questionnaire for Chinese medical students?

Materials and methods

CLEQ translation process

The translation followed Dorcas E. Beaton [10] et al.‘s six-stage forward and backward translation method, emphasizing conceptual accuracy and linguistic adaptability:

1. Forward Translation: Two translators with distinct backgrounds independently translated the questionnaire into Chinese.

   Translator 1: Yang Luhua (medical background).

   Translator 2: Zhang Ranran (English major, no medical background).

2. Synthesis: The translators and an observer (Wang Ruirui) consolidated the translations into a T-12 version, resolving discrepancies through consensus.

3. Backward Translation: Two independent translators (Christina Evans and Adam Fullerton) retranslated the T-12 version into English without referencing the original questionnaire.

4. Expert Review: An expert committee reviewed all versions, reports, and translations to finalize a pretest version.

5. Pretest: The pretest involved 30–40 participants from the target population to evaluate the questionnaire’s clarity and usability.

6. Documentation Submission: All reports and materials were submitted to the questionnaire developers for review and approval [11].

Participants and procedures

Sample size

The study initially recruited 270 clinical students from four medical schools. After excluding 54 respondents who completed the questionnaire in ≤ 85 s (indicating rushed responses), 216 participants (83 males, 133 females) were retained for analysis. The mean completion time was 171 s. The sample size satisfied the factor analysis requirement of a 1:10 respondent-to-item ratio [12], with a minimum requirement of 180 participants.

Target sample

Participants were clinical medical interns with at least six months of clinical rotation experience. Students were selected from four medical schools (Shandong First Medical University, Shaoxing University, Wannan Medical College, and Xinxiang Medical University), providing geographical and institutional diversity.

Inclusion criteria

Clinical interns enrolled in medical programs, having completed basic medical courses and entered clinical rotations. Proficiency in reading and understanding Chinese to ensure accurate questionnaire responses. Voluntary participation with signed informed consent.

Exclusion criteria

Preclinical students or those with insufficient clinical rotation experience (< 6 months). Language barriers preventing comprehension of the questionnaire. Declined participation or incomplete responses. Significant health conditions affecting clinical experience. Duplicate responses, ensuring unique data for each participant.

Sampling method

Cluster Sampling: Questionnaires were distributed online via the Wenjuanxing platform (https://www.wjx.cn/) to recruit clinical interns from medical schools in Anhui, Shandong, Henan, and Zhejiang provinces. Stratified Sampling by Institution: Participants were selected from four medical schools to ensure diverse institutional representation.

Data collection tools and procedures

The study utilized the latest English version of the CLEQ questionnaire, originally developed by AlHaqwi AI [6] et al. with 40 items. Based on Nuha Alnaami [7] et al.’s revisions, it was optimized to 18 items across four dimensions: Clinical Cases(4 items), Organization of Patient Encounters (5 items), Supervision (4 items), Learning Motivation (5 items). Each item was rated on a 5-point Likert scale (0 = Strongly Agree to 4 = Strongly Disagree). Scores were reverse-coded, with lower scores indicating more positive evaluations.

Test-retest procedure

The CLEQ was administered twice to a sample (n = 216) with a 14-day interval [13]. Dedicated time, effort, and financial resources ensured full participant retention in the questionnaire retest. The retest was conducted under identical conditions (e.g., instructions, questionnaire version, and the Wenjuanxing platform). Participants provided written informed consent for repeated assessments. All data were anonymized using unique identifiers to ensure privacy. Paired responses were securely matched via encrypted codes, with explicit agreements restricting data usage to research purposes.

Statistical procedures

Internal consistency was evaluated via Cronbach’s α (range:0–1),with thresholds:<0.6 (low; requiring revision), 0.6–0.7 (acceptable), 0.7–0.8(good), 0.8–0.9 (high), and 0.9-1 (excellent) [10, 14, 15]. The CLEQ was administered twice (2-week interval). Intraclass correlation coefficients (ICCs) with 95% confidence intervals (CIs) quantified measurement consistency. ICC interpretation followed established criteria:<0.5(poor),0.5–0.75 (moderate), and 0.75–0.9 (good) [14]. Pearson correlation was used to explore the relationships between variables. Bivariate correlations were analyzed using Pearson’s r, interpreted as: negligible (|r| <0.30), moderate (0.30–0.50), and strong (≥ 0.50) associations [14]. A strong empirical correlation (e.g.,r > 0.5) between theoretically related dimensions provides robust evidence for the logical consistency of the questionnaire design.

Bartlett’s Test of Sphericity was conducted to verify inter-item correlations and suitability for factor analysis in the CLEQ. Principal Component Analysis (PCA) explored the scale’s factor structure, with factor loadings > 0.6 deemed acceptable. To assess the model-data fit in confirmatory factor analysis (CFA), six key indices were evaluated: Chi-Square Minimum/Degrees of Freedom(CMIN/DF), Root Mean Square Error of Approximation (RMSEA), the Comparative Fit Index (CFI), Tucker-Lewis Index (TLI), Incremental Fit Index(IFI) and Normed Fit Index(NFI). All indices range from 0 to 1, with values approaching 1, indicating superior model fit for the CFI, TLI, NFI, and IFI.Lower RMSEA values (closer to 0) reflect better fit. A CMIN/DF value < 2 indicates excellent model fit (strict criterion), while values < 3 are considered acceptable under commonly applied thresholds [15].

Results

Sample characteristics

As shown in Table 1, the final sample comprised 216 clinical interns (38.4% male, 61.6% female) from four medical schools across Anhui, Shandong, Henan, and Zhejiang provinces. Participants had completed a minimum of six months of clinical rotations, with a mean questionnaire completion time of 171 s (SD = 23.4).

Table 1.

Description of the sample feature distribution

Variable	Option	Frequency	Percent
Age	19-25years old	215	99.50%
	Over 26 years old	1	0.50%
School	Shandong First Medical University	93	43.10%
	Shaoxing University	23	10.70%
	Wannan Medical University	55	25.50%
	Xinxiang Medical University	45	20.90%
Gender	Male	83	38.40%
	Female	133	61.60%

Internal consistency reliability

The reliability results show that except for the learning motivation dimension, which has a Cronbach’s α coefficient of 0.760 (good reliability), all other dimensions have coefficients ranging from 0.8 to 1, indicating high reliability. This demonstrates that the scales used in this study have good internal consistency and robust reliability. The reliability is shown in Table 2.

Table 2.

Reliability analysis of CLEQ

Variable	Cronbach’ɑ coefficient	ICC(95%CI)	Number of terms
Clinical cases	0.851	[0.82,0.88], p < 0.01	4
Learning motivation	0.760	[0.71,0.81], p < 0.01	5
Supervision	0.834	[0.79,0.87], p < 0.01	4
Organization	0.855	[0.82,0.88], p < 0.01	5
CLEQ	0.920	[0.90,0.93], p < 0.01	18

Test-retest reliability

The assessment was administered twice with a 14-day interval. Test-retest reliability was quantified using intraclass correlation coefficients (ICC). All dimensions demonstrated positive mean scores, with ICC values > 0.76,confirming good reliability. The retest reliability scores for each dimension are shown in Table 2.

Pearson correlation

Confidence intervals were calculated using the Bootstrap method (1,000 resamples).

A confidence interval that does not include 0 suggests a statistically significant correlation.

The results show significant correlations among all variables at the 99% significance level. With correlation coefficients (r) exceeding 0.5 for all pairs as shown in Table 3, it can be concluded that all variables in this analysis exhibit strong positive correlations [12, 14].

Table 3.

Results of the pearson correlation between dimensions

Variables	Pearson Correlation Coefficient	Bootstrap 95% Confidence Interval
Cases vs. Motivation	0.515**	[0.360,0.631]
Cases vs. Supervision	0.529**	[0.418,0.625]
Cases vs. Organization	0.562**	[0.457,0.650]
Motivation vs. Supervision	0.615**	[0.512,0.704]
Motivation vs. Organization	0.552**	[0.440,0.652]
Supervision vs. Organization	0.730**	[0.649,0.799]

Notes: ** indicates statistical significance at the p < 0.01 level

Factor structure

Based on the model fit indices in Table 4, the results show that CMIN/DF (chi-square/degree of freedom ratio) = 1.749, falling within the excellent range of [0,2] [16, 17]RMSEA (root mean square error of approximation) = 0.057,which is < 0.06 and close to the optimal range of 0.05 [18]. Additionally, IFI, TLI, CFI and NFI values are all within acceptable ranges. Overall, the analysis indicates that the Clinical Learning Evaluation model demonstrates good fit. Figure 1 is the CFA model diagram.

Table 4.

Model fitness test

Indicator	Reference standard	Measured result
CMIN/DF	[0–2] is excellent, (2,3] is acceptable	1.749
RMSEA	[0,0.05) is excellent, [0.05 or 0.08] is acceptable	0.057
IFI	> 0.9 excellent, > 0.8 acceptable	0.957
TLI	> 0.9 excellent, > 0.9 acceptable	0.945
CFI	[0.97,1] excellent, [0.95,0.97) acceptable	0.956
NFI	[0.95,1] excellent, [0.90,0.95) acceptable	0.904

Note. CMIN/DF = Chi-Square to Degrees of Freedom Ratio, RMSEA = root mean square error of approximation, CFI = comparative fit index, IFI = Incremental Fit Index, TLI = Tucker–Lewis index, GFI = goodness-of-fit index, NFI = Normed Fit Index

Fig. 1.

Diagram of the CFA model for validated factor analysis of the CLEQ

As shown in Fig. 1, most of the factor loadings are high (greater than 0.7), indicating that the variables can effectively explain the latent factors. From Fig. 1, it can be seen that the factor loadings for the learning motivation dimension are relatively low, falling within the moderate range (between 0.4 and 0.7) [19, 20]. Item 9 has the lowest factor loading. This dimension may require additional variables for a more comprehensive reflection of the factor. The questions in this dimension should also be more specific and targeted to reduce subjective differences. The reliability and validity of the learning motivation dimensions were low, so we have restructured six questions, which can be found in the supplementary document.

The Table 5 shows that the AVE value for the learning motivation dimension is relatively low. Overall, the analysis indicates that, except for the learning motivation dimension, all other dimensions have AVE values above 0.5 and CR values above 0.7,demonstrating good convergent validity and composite reliability [21].

Table 5.

Results of validity tests, convergent validity and combined reliability tests for each dimension of the CLEQ

Items	Clinical case	Motivation	Supervision	Organization
Clinical Cases	0.556
Motivation	0.192	0.433
Supervision	0.188	0.160	0.559
Organization	0.218	0.155	0.208	0.574
Square root of AVE value	0.746	0.658	0.748	0.758
AVE	0.588	0.378	0.589	0.571
CR	0.847	0.750	0.851	0.869

Based on the analysis results in Table 5, the standardized correlation coefficients between each pair of dimensions are all smaller than the square root of the AVE values for the corresponding dimensions. This indicates that all dimensions exhibit good discriminant validity [21].

Discussion

The Chinese version of CLEQ, originally developed by Saudi scholars, demonstrated robust psychometric properties in its first cross-cultural application among clinical interns in China, with confirmatory factor analysis (CFA) validating structural consistency and test-retest reliability exceeding recommended thresholds (ICC > 0.75). The questionnaire exhibited strong internal consistency across all dimensions, evidenced by Cronbach’s α coefficients of 0.76–0.92,consistent with previous international studies. The Cronbach’s alpha values in the study by Alnaami et al. range 0.72–0.87 and Cronbach’s alpha range 0.68–0.79 in the study by Ostovarfar et al. [7, 8]. However, the learning motivation dimension showed relatively lower reliability (ɑ=0.76),warranting further investigation. The CFA model diagram of the CLEQ is shown in the Fig. 1. The strong correlation between Supervision and Organization (r = 0.828) exceeded theoretical expectations, suggesting potential conceptual overlap in their item formulations (e.g., ambiguous distinctions between “supervision” and “organization”). The significant error correlation between e1 and e2 (r = 0.462) highlights the need to merge overlapping items assessing clinical case evaluations, thereby mitigating common method bias risks.

The strong performance of clinical cases and supervision dimensions (ɑ>0.85) reflects the structured nature of these clinical learning components. The moderate reliability (ɑ=0.76) of learning motivation stems from: Subjective perception bias: Item 5(“Enjoy clinical sessions”);Susceptible to transient emotional fluctuations (e.g.,procedural success-induced satisfaction spikes), failing to differentiate intrinsic (knowledge-seeking) vs. extrinsic motivations(academic pressure). Item 7 (“Role models”):Hierarchical cultural norms inflate ratings for authoritarian preceptors, unrelated to pedagogical effectiveness. Cultural Measurement Discrepancy: Chinese students underreport self-perceived communication confidence (Item 6). Cross-sectional designs inherently lack the temporal resolution to detect dynamic fluctuations in students’ learning motivation across educational phases [22–25]. Consequently, Item 6 has been removed from this dimension following comprehensive revision. Items 5, 7, 8, and 9 underwent rigorous cultural and psychometric refinement to enhance contextual validity and measurement precision in Chinese clinical training settings. Appendix A contains the 18-question CLEQ questionnaire, and Appendix B is the modified version of the CLEQ questionnaire after validation in China.

Potential Impacts of Limitations on CLEQ Implementation in China: The conflation of intrinsic/extrinsic motivations (Item 5) and cultural inflation of authoritarian mentor ratings (Item 7) could lead to distorted perceptions of clinical training quality. For instance, institutions might misinterpret elevated “role model” scores as evidence of effective mentorship, while actual pedagogical outcomes remain unimproved. This risks perpetuating didactic teaching models at the expense of learner-centered approaches emphasized in China’s medical education reforms. Systematic underreporting of communication confidence (Item 6), despite competent OSCE performance, may prompt unnecessary investments in communication skills training while neglecting genuine competency gaps. This contradicts the National Health Commission’s focus on practical skill development, potentially widening the theory-practice gap [9]. It could lead to premature termination of effective programs during expected motivational nadirs.

Our findings have several important implications for clinical education reform. First, the CLEQ’s ability to identify specific areas for improvement, such as clinical supervision quality and learning resource allocation, provides actionable insights for curriculum development [26]. Second, the instrument’s sensitivity to cultural and contextual factors highlights the need for localized adaptation of international assessment tools. Third, the differential performance across dimensions suggests that targeted interventions may be more effective than global approaches in enhancing clinical learning experiences [27].

Future research should address several limitations. First, longitudinal studies are needed to assess the CLEQ’s predictive validity for clinical competence development. Second, cross-institutional comparisons could elucidate the impact of different clinical training models on learning environment quality. Third, qualitative studies should explore the cultural and contextual factors influencing students’ perceptions of their clinical learning experiences.

Electronic supplementary material

Below is the link to the electronic supplementary material.

Abbreviations

CLE

Clinical learning environment

CLEQ

Clinical learning evaluation questionnaire

PHEEM

Postgraduate hospital educational environment measure

DREEM

Dundee ready education environment measure

OSCE

Objective structured clinical examination

CVR

Content validity ratio

CVI

Content validity index

RMSEA

The root mean square error of approximation

AVE

Average variance extracted

CR

Composite reliability

CFA

Confirmatory factor analysis

PCA

Principal component analysis

Author contributions

L Y drafted the main text of the manuscript. JS revised the paper. RW and L Y participated in the translation of the questionnaire. All authors participated in data collection and the supervision of this study.

Funding

Funding for this study was supported by the Anhui Province Youth Middle-aged Domestic Visiting Program Fund (Project No. JNFX2024105),the Provincial Teaching Team of Rehabilitation Therapy Technology (2021jxtd165), Research and Innovation Team Project of West Anhui University (WXSK202402) and Anhui Province University excellent top talent training project (JWFX2024029), Anhui Provincial quality engineering project of West Anhui University (2023sx096), Anhui Provincial quality engineering project of West Anhui University (2022jyxm1743), Anhui Provincial Education Science Research Project (JK23130).

Data availability

All data supporting the findings of this study are available within the mannuscript and its Supplementary Information.

Declarations

Ethics approval and consent to participate

The Ethics Committee approved this study at the Anhui Higher Institute of Traditional Chinese Medicine through Ethics Approval No. YW2024-024, a summary procedure review that does not require registration in the Clinical Trial Registry. The study protocol complied with the the World Medical Association’s Declaration of Helsinki. Written informed consent covering both participation and data publication was obtained from all participants prior to study commencement.

Consent for publication

Not Applicable.

Competing interests

The authors declare no competing interests.