Traditional nursing education in China has prioritized the development of students’ clinical operational skills, communication and coordination abilities, and critical thinking (Fu et al., 2022). However, there is a relative lack of emphasis on metacognitive awareness among nursing students (Pu et al., 2020). This oversight results in students’ deficiencies in cognitive processes such as learning planning, management, monitoring, adjustment, and evaluation, which in turn leads to a shortfall in their self-directed learning capabilities (Jin and Ji, 2021). In light of the rapidly changing landscape characterized by new technologies, digitalization, and automated educational environments, Chinese scholars (Sha et al., 2022) are working to bolster students’ metacognitive awareness through blended teaching reforms that integrate online and offline modalities in nursing practice training courses (Sha et al., 2022). These initiatives are designed to encourage the development of lifelong learning habits among nursing students.

In recent years, the advent of intelligent nursing services enabled by technologies such as cloud computing, big data, and the Internet of Things has led to heightened expectations for the intelligence, personalization, and precision of nursing education in China (Liu et al., 2023). The cultivation of lifelong learning among nursing students has become essential (Wong et al., 2021). Studies have shown that nursing students with elevated levels of metacognitive awareness are more capable of evaluating their learning situations, planning learning actions, and adjusting learning strategies, which in turn improves learning efficiency and promotes the development of positive study habits (Desender et al., 2021; Du et al., 2023). Nursing students who possess higher levels of metacognitive awareness can better understand and assimilate knowledge across various disciplines, supporting the establishment of lifelong learning habits and the ongoing enhancement of professional knowledge and skillsets to meet the evolving demands of the medical landscape (Gholami et al., 2016).

Currently, despite the integration of virtual simulation teaching, high-quality online classrooms, and other training modalities to augment nursing students’ metacognitive abilities, there remains a paucity in the depth of analysis and discourse regarding the content of metacognitive awareness within nursing education (Barranquero-Herbosa et al., 2022). As a result, there is an imperative need for an accurate assessment tool to gauge the efficacy of metacognitive awareness training among nursing students. Such a tool is crucial for establishing a theoretical framework that can inform the development of more precise educational strategies aimed at enhancing nursing students’ metacognitive awareness.

In China, three prevailing scales have been employed to measure metacognitive awareness among Chinese nursing students: the Metacognitive Ability Scale for College Students, developed by Chinese scholar Kang Ye et al. (2018), the Metacognitive Inventory for Nursing Students by Hsu (2010), and the State Metacognitive Inventory by O’Neil and Abedi (1996). These instruments assess metacognitive abilities directly, offering an indirect reflection of changes in metacognitive awareness through these abilities. However, they do not directly investigate and elucidate the multifaceted dimensions of metacognitive awareness specific to nursing students. Consequently, there is a need for research to employ internationally recognized metacognitive awareness scales and to conduct psychometric evaluations within the Chinese linguistic context to ascertain their validity. Adopting this method will facilitate precise monitoring of fluctuations in nursing students’ metacognitive awareness, providing a robust theoretical framework for comparative studies on metacognitive awareness among students from varying cultural backgrounds, thereby enriching cross-cultural inquiry in the field of nursing education.

Cross-cultural adaptation and validation of C-MAS were conducted in a methodological cross-sectional study (Huang et al., 2020).

In adherence to the standard guidelines for sample size (Myers et al., 2011), which recommend a ratio of at least 5:1 between sample size and parameters, the study was conducted with 52 items in mind. To account for potential invalid samples, the sample size was increased by 15%. Consequently, the required sample size was determined to be 299, ensuring it met the maximum threshold.

Study participants were recruited through convenient sampling from four universities in eastern, southern, western, and northern China between January and July 2022. The inclusion criteria were: (i) full-time nursing undergraduates and (ii) willingness to provide informed consent for participation. The exclusion criterion was students pursuing a degree upgrade from a junior college to a bachelor’s level. The study was approved by the Ethics Committee of Huzhou University (Approval No.: 202012-JG02), with voluntary participation and confidentiality assured. A total of 592 nursing undergraduates completed the survey, with an average age of 21.24 years (standard deviation [SD], 1.65). The sample comprised 512 females (86.5%) and 80 males (13.5%). In terms of academic standing, 142 (24.0%) were freshmen, 140 (23.6%) were sophomores, 186 (31.4%) were juniors, and 124 (20.9%) were seniors.

The translation and adaptation process of the MAS was based on Brislin’s translation model (Jones et al., 2001), following the two phases below.

Phase I comprised four steps. (i) Forward translation: the scale was independently translated into Chinese by two researchers proficient in English and Chinese. Subsequently, the researchers evaluated the congruity between the two translated versions through discussions with the research team, leading to the development of the final version. (ii) Back translation: two non-nursing researchers independently translated the scale back into English. One researcher held a doctoral degree in English translation, while the other possessed 10 years of experience in metacognition research and education in an English-speaking country. The two translated versions were compared, and their similarities and differences were analyzed. (iii) Cultural adaptation: a panel of nine experts reviewed the concept and connotation of the translated original scale, and scored the content relevance. This expert panel included two professors with expertise in psychological nursing, two professors specializing in nursing education management, three teachers with over 15 years of metacognition research, and two frontline nurses with more than 10 years of clinical experience. Leveraging their professional theoretical knowledge, educational background, and clinical expertise, the experts independently assessed each scale item for clarity of expression, conceptual equivalence, and content relevance. Each item was evaluated individually, and any ambiguous aspects were revised accordingly. (iv) Content validity: content validity was determined through expert scoring (Polit and Beck, 2014), with the above-mentioned nine experts participating in the content validity review.

Phase II comprised two steps. (i) A convenience sampling approach was used to enlist 30 nursing students for the evaluation of their understanding of the Chinese version of the Metacognitive Awareness Scale (C-MAS). The goal was to ensure that the scale could be easily understood by Mandarin-speaking nursing students. Participants were asked if they needed assistance understanding and responding to the questionnaire items. All items were fully understood, and participants completed the questionnaire without issue. (ii) Utilizing factor analysis to examine the factorial structure of the Chinese version of the metacognitive awareness scale. A total of 592 nursing students were recruited, and the samples were randomly divided into Group A and Group B based on matching criteria such as grade and gender. Data from Group A were subjected to exploratory factor analysis (EFA), while data from Group B underwent confirmatory factor analysis (CFA). Each group consisted of 296 nursing students.

The study received support from the management department of the surveyed school. The researcher provided a detailed explanation of the study’s purpose and significance to the participants, and each participant signed an informed consent form. Paper versions of the questionnaire were distributed in sealed envelopes to the participants in their classrooms. Participants were asked to answer the questionnaire on the spot and return it immediately. The completeness of each questionnaire was checked, and any missing items were addressed by asking the participants again and supplementing the information promptly. The researchers confirmed that all answers were complete. To ensure the completeness and authenticity of the responses, four researchers, uniformly trained, were responsible for distributing and collecting the survey questionnaires. The targeted sample size was 598 nursing students, and a total of 592 valid questionnaires met the criteria, resulting in a valid response rate of 98.99%.

Statistical analysis of the collected data was performed using IBM SPSS 22.0 and Mplus software version 7.4 (Muthen and Muthen, Los Angeles, CA, United States). Descriptive statistics, including frequencies and percentages, were utilized to summarize the characteristics of the respondents’ basic information. The reliability of the C-MAS was evaluated through item-total correlation coefficients and internal consistency, as measured by Cronbach’s alpha coefficients.

Construct validity was assessed through exploratory factor analysis (EFA) and confirmatory factor analysis (CFA). EFA was employed to explore the factor model of the C-MAS, utilizing principal component analysis and the maximum variance method. The sampling adequacy was determined using the Kaiser–Meyer–Olkin (KMO) measure, with a KMO value exceeding 0.8, and Bartlett’s sphericity test showing a p-value less than 0.01. Furthermore, additional criteria were applied to refine the factor structure and determine the optimal number of factors: eigenvalues greater than 1.0, factor loadings exceeding 0.50, and a percentage of total variance explained (Hou et al., 2004; Zhao et al., 2023).

The CFA was used to confirm the prior explored model obtained through EFA. Model fit was assessed using the following indices: (a) Chi-square/degrees of freedom (χ²/df) ratio, ideally falling between 1 and 3. (b) Root mean square error of approximation (RMSEA) with values less than 0.08. (c) Standardized root mean square residual (SRMR) values less than 0.06, and RMSEA with a 90% confidence interval (RMSEA 90%CI). (d) Normed fit index (NFI), incremental fit index (IFI), and goodness of fit index (GFI) with values greater than 0.90 (Zhao et al., 2023).

Convergent validity was evaluated by examining the fit of the subscales using average variance extracted (AVE) and composite reliability (CR). An AVE exceeding 0.5 is considered satisfactory, while a CR surpassing 0.6 is deemed acceptable, both indicating sound structural reliability (Xue and Zhang, 2017). In line with the methodology proposed by Marsh and Hocevar (1985), two models were identified: an 8-factor first-order model and an 8-factor second-order model, to assess their adequacy. A target coefficient value approximating 1 suggests that the second-order model could effectively substitute the first-order model, thus enhancing the overall precision of the model.

Content validity index (CVI) was calculated to assess content validity. Nine experts participated in the validity analysis, with each expert providing a score for the relevance of each item on a scale of 1 (not relevant) to 4 (definitely relevant). An item is considered valid if its content validity score is 0.80 or higher (Luo et al., 2019).

The findings revealed that the average score for each item measuring metacognitive awareness among nursing students surpassed 2.50. This implies that the overall level of metacognitive awareness was above the midpoint, as detailed in Table 2. Furthermore, each item was subjected to an item-total correlation test. The outcomes showed that the scores for each item were significantly and positively correlated with the total scale score, with all correlations exceeding 0.4, suggesting the absence of multicollinearity. Additionally, every item achieved statistical significance (p < 0.05). Consequently, all 52 items fulfilled the pre-established criteria and were retained, with no items being removed.

Nine experts, including professionals from the fields of psychiatric nursing, frontline clinical nursing, nursing education management, and research related to metacognition, evaluated the relevance of each item to metacognitive awareness. Employing a Likert 4-point rating scale, where 4 points denoted high relevance, 3 points indicated moderate relevance, 2 points suggested low relevance, and 1 point reflected irrelevance. The content validity index at the item level was calculated based on the ratings provided by the nine experts. The results indicated that the Content Validity Index at the scale level (S-CVI) was 0.912, and the Content Validity Index at the item level (I-CVI) ranged from 0.855 to 1.