The research combined a systematic review, quantitative research and the modified Delphi technique. Three postgraduate students and three interdisciplinary experts conducted the systematic review, which included 19 eligible studies and identified four core themes. The authors then developed and statistically tested a regression model based on Corley’s moral distress theoretical model. This model was statistically significant (F = 34.173, P < 0.001, R² = 0.229), although the abstract does not provide enough methodological detail to interpret the model’s predictive or causal significance.

Cross-sectional quantitative data were collected in eastern China. Eighteen specialists in nursing management, clinical nursing and nursing ethics subsequently completed two rounds of modified Delphi consultation. The expert authority coefficient was 0.87. Following these consultations, the researchers finalized the three-level ethical training framework, centered on nursing ethics education.

The abstract describes the program as a feasible, evidence-based intervention intended to improve nurses’ moral competence for coping with moral distress. However, the reported work was a program-construction study rather than a clinical trial of training effectiveness: no participant outcomes after delivery of the program are presented, and no clinical trial number was applicable. The evidence is also limited to the reported review, cross-sectional data from eastern China and consultations with 18 specialists. Accordingly, the results support the construction of the training framework but do not establish that using it reduces moral distress or improves patient care. The article was available as an accepted version subject to further edits, and the findings summarized here are based on the supplied abstract rather than a full assessment of the article.

Subject of Research: Development of a nurse moral-competence training program for coping with moral distress

Article Title: Construction of a training program for clinical nurses coping with moral distress based on Corley’s theoretical model

Article References: Meng, J., Wang, Y., & Liu, Y. (2026). Construction of a training program for clinical nurses coping with moral distress based on Corley’s theoretical model. BMC Nursing. https://doi.org/10.1186/s12912-026-05473-3

Image Credits: AI Generated

DOI: 10.1186/s12912-026-05473-3

Keywords: nursing ethics, moral distress, moral competence, training program, mixed-methods study, systematic review, Delphi method, clinical nurses, healthcare education, Construction, training, program

Researchers began with policy and institutional documents, published research and interviews with 20 public health professionals in Hebei Province. These sources informed 49 candidate indicators concerning practice-based learning. Two groups of experts then assessed each candidate’s importance, scientific soundness and feasibility through Delphi rounds involving 20 experts in the first round and 18 in the second. The researchers summarised ratings descriptively and used tie-corrected Kendall’s W to assess concordance in the experts’ relative rankings.

The second-round mean ratings ranged from 3.89 to 4.67 across the three assessment dimensions, while every coefficient of variation was no greater than 0.25. Expert agreement on ranking was low: Kendall’s W was 0.0529 for importance, 0.0569 for scientific soundness and 0.0474 for feasibility, and all associated P values exceeded 0.05. The results therefore indicate that experts gave generally high ratings to the retained indicators without consistently ranking them in the same order.

After the Delphi process, the framework contained three primary dimensions, nine secondary domains, 19 tertiary subdomains and 45 fourth-level indicators. A fuzzy analytic hierarchy process assigned Process the largest primary weight, at 0.481, followed by Input at 0.266 and Output at 0.253. The global weights were based on pairwise judgments made by the first author, with calculations jointly checked by the first and corresponding authors; products of local weights were then normalised to obtain global indicator weights. These values represent the weighting model’s priorities rather than measured effects on learning outcomes.

At the indicator level, specification of practice-mentor guidance time received the highest normalised global weight, 0.080508. Dual-qualified faculty ranked next, followed by four indicators with equal weights concerning practical training and assessment. The framework consequently provides a structured set of candidate conditions for reviewing how Chinese MPH programmes organise practical learning, including supervision and assessment within the wider degree. Because the study did not test implementation across programmes or students, its authors conclude that empirical evaluation is needed to determine whether the arrangements are delivered as intended and associated with students’ public health performance.

Subject of Research: Expert-informed evaluation of practice-based learning conditions in Chinese Master of Public Health programmes

Article Title: Evaluating the conditions for practice-based learning: an expert-informed Input–Process–Output framework for Master of Public Health programmes in China

Article References: Tang, L., Shi, H., Li, B., Lan, J., & Tian, J. (2026). Evaluating the conditions for practice-based learning: an expert-informed Input–Process–Output framework for Master of Public Health programmes in China. BMC Medical Education. https://doi.org/10.1186/s12909-026-10429-x

Image Credits: AI Generated

DOI: 10.1186/s12909-026-10429-x

Keywords: public health education, Master of Public Health, China, practice-based learning, Delphi method, student supervision, programme evaluation, fuzzy analytic hierarchy process, medical education, Evaluating, conditions, practice-based

The research team, working across Italy, Switzerland and Germany, confronted a problem that has long troubled invasion biologists: control programmes are typically designed to maximise ecological and economic outcomes, while the welfare of the target animals is treated as an afterthought or ignored altogether. Recent work in conservation decision science has argued that animal welfare should be evaluated as an explicit objective alongside biodiversity benefits, so that trade-offs can be made transparently rather than by default. To do that for slider turtles, however, managers first need to know how much pain, distress or deprivation each technique actually causes, and for a reptile species, that evidence base is remarkably thin.

The study adapted the Five Domains framework, a widely used model in animal welfare science. Domain 1 covers nutrition, including food and water deprivation; Domain 2 covers environmental challenges such as heat, cold or lack of shelter; Domain 3 covers health, injury and disease; Domain 4 covers behavioural and social restrictions; and Domain 5 represents the overall mental consequence, the suffering that emerges from the first four domains. Building on an earlier Australian framework for assessing the humaneness of vertebrate pest control, the researchers defined eight management actions: capture with fyke or minnow traps, capture with basking traps, transportation in covered boxes, containment in semi-natural enclosures, containment in intensively managed facilities, shooting with a rifle, lethal injection, and freezing of eggs removed from wild nests.

Crucially, each action was assessed across explicit time intervals, a significant modification of the original framework. Trapping scenarios ranged from 12 hours to seven days depending on trap type, transportation from 3 to 24 hours, captivity from one year to more than a decade, and lethal methods from the instant of application to beyond 30 minutes until loss of consciousness. The underlying logic is that welfare impacts are not static: a turtle held briefly in a funnel trap may experience little more than behavioural unease, while one left for two days in warm water faces dehydration, overheating, aggression from other trapped individuals and escalating distress.

Because empirical data on turtle welfare barely exist, the team collated whatever evidence they could, including temperature logs and camera footage from field programmes in Hungary and Switzerland, necropsy records, husbandry guidelines and studies of embryonic development, and then turned to structured expert elicitation. Following a modified Delphi protocol, 24 experts scored the expected welfare impact of every action and timeframe on a five-step scale from No impact to Extreme, rating their own confidence after each answer. They then discussed anonymised results in multiple online rounds in English, German and Italian before revising their scores, a cycle designed to expose linguistic misunderstandings and prevent groupthink while preserving the diversity of judgement.

The results offer reassurance for well-run operations and warnings for sloppy ones. Experts broadly agreed that trapping causes no or mild suffering within the first 24 hours, and that transport of up to six hours at appropriate temperatures of 20 to 24 degrees Celsius is similarly benign. Intensively managed captivity, following husbandry standards developed in Italy, was judged to produce predominantly no or mild impacts even over many years. On the lethal side, all experts agreed that a properly placed rifle shot causing immediate loss of consciousness inflicts no suffering, and lethal injection with prior sedation was rated as causing none or mild suffering only. Egg freezing attracted relatively low impact estimates, on the reasoning that cold blocks nerve conduction, although the panel flagged that late-stage embryos, in which neural pathways may already be forming, remain a genuine uncertainty.

Time emerged as the single most important amplifier of harm. Statistical models confirmed that the probability of impacts reaching Moderate or Severe rose steadily, and often nearly linearly, with the duration of every action. Opinions diverged sharply beyond standard practice: once fyke traps were left unchecked beyond 24 hours, or basking traps beyond a few days, or transport stretched to 12 or 24 hours, some experts predicted Severe or even Extreme suffering. The greatest spread of opinion concerned semi-natural containment over decades, where overcrowding, disease transmission, poor water quality and the absence of veterinary care could accumulate into serious welfare decline, yet five of thirteen experts judged decade-long containment as Severe or Extreme while others remained far less concerned.

Expert confidence averaged a relatively high 80 percent overall, but it dropped in revealing patterns. Confidence was higher for non-lethal methods than for lethal ones, highest for transport and lowest for egg freezing, and it fell significantly as estimated impact levels increased and, after controlling for impact, over the longest timeframes. The weakest ground of all was Domain 4, behavioural and social welfare, where experts confessed that the inner states of a non-mammalian species are notoriously difficult to read. Veterinarians tended to rate impacts slightly higher than reptile specialists and caretakers, though the difference was not statistically significant, and their judgements were generally less polarised.

From the discussion rounds emerged a set of concrete best practices that managers can apply immediately. Traps should be checked at least every 24 hours for fyke traps and every three days for basking traps, and placed to offer shade and prevent overheating, with particular care around post-hibernation and breeding periods when densities and activity spike. Transport boxes should be compartmentalised or lined with cotton bags to maintain moisture, avoid mixing turtles of different sizes, and never exceed a few hours. Captive facilities must limit densities, provide varied microhabitats, basking sites, nesting opportunities and balanced diets, since inadequate nutrition is linked to bladder stones, shell deformities and ear infections in captive sliders. Shooting should only be performed by trained marksmen, and injection protocols must account for the ectothermic physiology of reptiles, whose slow metabolism in cold or stressed animals can delay drug action, making verification of anaesthesia, for example by testing the corneal reflex, essential before administering lethal agents.

The authors are careful to note that their findings reflect expert opinion rather than experimental proof, and that they assessed individual welfare only, leaving aside bycatch, multi-step interventions and the numbers of animals affected. Even so, the study demonstrates that formal elicitation can convert scattered, partly intuitive knowledge into a transparent and reproducible baseline for policy, showing decision-makers exactly where consensus is strong and where ignorance reigns. It also identifies where new research would pay off fastest: field studies quantifying first-shot success and time to insensibility in shooting, longitudinal comparisons of captivity facilities, and developmental work on when reptile embryos become capable of sensing pain. As invasive species management intensifies worldwide under pressure from biodiversity loss, the lesson from the humble slider turtle is that conservation and compassion need not be adversaries, provided the welfare costs of every capture, journey and end-point are counted honestly and minimised by design.

Subject of Research: Animal welfare assessment of lethal and non-lethal management methods for invasive slider turtles

Article Title: Animal welfare implications of management strategies for invasive freshwater turtles

Article References: Moser, L. A., Smith, B. R., Matthys, S., Kenup, C., Spongano, I., Cavatoni, S., Dal Zotto, M., Ferrari, N., Ferri, V., Hoby, S., Lloyd Jr., R. B., Moretti, A., Oneto, F., Origgi, F., Ottonello, D., Santicchia, F., Zanirato, N., Zuffi, M. A. L., & Canessa, S. (2026). Animal welfare implications of management strategies for invasive freshwater turtles. NeoBiota, 109, 345-366. https://doi.org/10.3897/neobiota.109.204683

Image Credits: AI Generated

DOI: 10.3897/neobiota.109.204683

Keywords: invasive species, Trachemys scripta, animal welfare, Five Domains framework, expert elicitation, herpetology, conservation management, euthanasia, reptile welfare, trapping, captivity, Delphi method

The instrument in question, formally titled the Assessment of Professional Skills of Occupational Therapy Students, was originally developed in Spanish-speaking contexts to evaluate the clinical competencies of students in the field. Occupational therapy, a profession focused on helping people regain the ability to perform everyday activities, has been expanding rapidly in China, and the World Federation of Occupational Therapists has set international expectations for what entry-level practitioners should be able to do. Existing Chinese-language options, however, have been limited. Tools such as clinical competency evaluation instruments, clinical internship evaluation tools, the revised Student Practice Evaluation Form, and competency-based fieldwork evaluations for occupational therapists exist in various forms, but none of them included a validated Chinese version of the CPTO. That gap meant Chinese educators had no way to benchmark their students against an internationally recognized measure of professional skill.

Translating a psychological or educational test is far more involved than running text through a dictionary. Words must carry not only the same literal meaning but the same connotations, and the underlying constructs measured by each item must survive the journey across cultures. To manage this, the team followed a structured four-step translation procedure, a standard approach in cross-cultural instrument adaptation that typically involves forward translation by independent translators, reconciliation of the versions, back-translation by linguists blind to the original, and review by a committee of experts to resolve discrepancies. This systematic process is designed to establish semantic, idiomatic, experiential, and conceptual equivalence between the source and target versions, ensuring that a Chinese-speaking student responding to an item is effectively answering the same question as a Spanish-speaking counterpart.

With a draft Chinese version in hand, the researchers turned to a modified Delphi process, a consensus-building method in which a panel of experts reviews items over repeated rounds while the organizers collect ratings and feedback. Nine experts participated in two rounds of consultation, five of whom, or 55.56 percent of the panel, were women. Their professional experience was substantial, averaging 21 years of practice in their respective fields. The panel judged each item for importance and for clarity, and the statistical signals from the second round were encouraging. The mean Spearman’s rho value, a measure of ranking agreement among experts, reached 0.88 with a standard deviation of 0.21, and the improvement between rounds was statistically significant, with a Z statistic of -2.33 and a p value below 0.05. The semi-interquartile range, which captures how tightly expert ratings clustered, averaged just 0.05 in the second round, indicating that consensus had converged to a narrow band. Every item passed the importance threshold set for retention.

Content validity was then quantified using the content validity index, a metric widely used in instrument development. At the item level, the I-CVI values for all items fell between 0.8 and 1.0, meaning that at least four out of every five experts considered each item relevant and essential. At the scale level, the average S-CVI reached 0.98 in the second Delphi round, an exceptionally high figure that comfortably exceeds the commonly cited benchmarks of 0.80 for the item level and 0.90 for the scale level. In plain terms, the expert panel regarded the translated items as a faithful and complete representation of the professional skills the instrument is meant to measure.

The next stage involved putting the instrument in front of actual students. All 103 students invited from two universities agreed to participate, a 100 percent response rate that strengthens the representativeness of the pilot data. The researchers analyzed the responses using SPSS version 26.0 and R version 4.3.1, applying both classical and modern psychometric techniques. A central question in validation is whether the structure of the instrument, meaning the pattern of relationships among its items and sections, matches the theoretical model of professional competence it is supposed to reflect. Confirmatory factor analysis, a technique that tests how well observed data fit a predefined factor structure, demonstrated a good fit to the data, with standardized factor loadings ranging from 0.55 to 0.98. These loadings indicate how strongly each item relates to its intended underlying skill domain, and values in this range suggest that the items hang together coherently.

Reliability, the second pillar of psychometric quality, was assessed from several angles. Inter-rater agreement, which reflects whether different evaluators assign consistent scores to the same performance, produced an intraclass correlation coefficient of 0.787 for the total scale, with section-level values ranging from 0.817 to 0.961. Test-retest reliability, which examines whether scores remain stable when the instrument is administered on separate occasions under similar conditions, was even stronger, with a total-scale ICC of 0.896 and section-level values between 0.759 and 0.955. Internal consistency, measured by Cronbach’s alpha, reached 0.950 for the total scale, a figure indicating that the items collectively measure a unified construct with very high precision. Correlations between the total score and the scores of the individual sections, labeled the s-section, c-section, and p-section, were 0.43, 0.77, and 0.84 respectively, all statistically significant at p below 0.001. This pattern suggests that each section contributes meaningfully to the overall score while also capturing distinct information of its own.

The distribution of student performance offered a further sanity check on the instrument’s usefulness. When the 103 participants were classified into the CPTO’s four performance bands, roughly a quarter fell into each category: 24.27 percent scored in the low range of 0 to 150, 25.24 percent in the medium range of 151 to 154, 25.25 percent in the high range of 155 to 183, and 25.24 percent in the excellent category of 184 or above. For an assessment tool intended to differentiate among learners, this near-uniform spread is a favorable property, because a test that piles most examinees into a single band cannot distinguish strong performers from weak ones. The even dispersion across the Chinese student sample suggests the translated instrument can meaningfully stratify professional skill levels in this population.

The authors conclude that the Chinese version of the CPTO demonstrates good reliability and validity for assessing the professional skills and clinical competencies of occupational therapy students. The study received ethical approval from the Fujian Medical University Biomedical Research Ethics Review Committee under number 2024 Fuyi Ethics Review No. 172, was conducted in accordance with the Declaration of Helsinki, and was funded by undergraduate educational teaching research projects from Fujian Medical University and Fujian Province, with the funders playing no role in the study’s design or conduct. The work, accepted on 2 October 2026 and published on 19 October 2026 under a Creative Commons Attribution 4.0 license, arrives at a moment when rehabilitation professions across Asia are professionalizing quickly and education programs are under pressure to demonstrate measurable outcomes. A validated Chinese-language measure of professional competence gives Chinese occupational therapy programs a tool that aligns with international standards, supports fair and consistent evaluation during clinical placements, and opens the door to cross-national comparisons of student readiness. For a profession whose entire purpose is helping people reclaim independence in daily life, ensuring that its newest practitioners are competently assessed is not an academic nicety but a matter of patient safety and workforce quality.

Subject of Research: Translation and psychometric validation of the Chinese version of an occupational therapy student professional skills assessment instrument

Article Title: Translation, reliability and validity of the assessment of professional skills of occupational therapy students (Chinese version)

Article References: Wang, Z., Han, P., Liu, X., Guo, W., Zhong, H., Mo, Y., Liu, Z., & Lin, C. (2026). Translation, reliability and validity of the assessment of professional skills of occupational therapy students (Chinese version). BMC Medical Education. https://doi.org/10.1186/s12909-026-10552-9

Image Credits: AI Generated

DOI: 10.1186/s12909-026-10552-9

Keywords: occupational therapy, psychometrics, translation, reliability, validity, Delphi method, content validity index, confirmatory factor analysis, medical education, China, clinical competency, assessment instrument

The research, led by Qian Wu and colleagues at The Second People’s Hospital of Hefei, affiliated with Bengbu Medical University and Anhui Medical University, set out to answer a deceptively simple question: what should emergency nurses learn, and when should they learn it, as they progress from novice to expert? The team’s answer is a structured framework spanning four professional levels, designated N1 through N4, that maps specific competencies onto each stage of a nurse’s development. Rather than treating training as a one-time hurdle to be cleared at hire, the framework envisions education as a continuous arc that follows the entire career trajectory.

Kolb’s Experiential Learning Cycle, developed by educational theorist David Kolb in the 1970s, proposes that effective learning proceeds through four recurring phases: concrete experience, reflective observation, abstract conceptualization, and active experimentation. In practical terms, a learner acts, reflects on what happened, distills the experience into general principles, and then tests those principles in new situations. For emergency nursing, this model has an obvious appeal. Clinical work in the emergency department is inherently experiential, but without structured reflection and conceptual grounding, experience alone can harden into habit rather than expertise. The study’s authors argue that embedding Kolb’s cycle into a tiered curriculum gives training both a psychological engine and a career-long roadmap.

To build the framework, the researchers used a mixed-methods design that began with a systematic review of the literature on emergency nursing competencies and training. They supplemented this with semi-structured interviews, drawing on the lived perspectives of frontline nurses to capture competencies that published literature might overlook. From these two evidence streams, they assembled a preliminary draft of the training framework, which then faced its most demanding test: scrutiny by a panel of seasoned experts.

That test took the form of a two-round Delphi consultation, a well-established technique for building consensus in fields where hard experimental data are scarce. Eighteen experts, each with more than ten years of emergency nursing experience, were recruited from four provinces across China. In each round, the experts rated the importance of every proposed framework item on a five-point Likert scale. The consensus criteria were deliberately strict: an item had to achieve a mean importance score of at least 4.0, a coefficient of variation no greater than 0.25, and an agreement rate of at least 75 percent among experts scoring the item 4 or higher. Items that fell short were revised or removed, and the revised framework was sent back to the panel for a second round of evaluation.

The engagement and authority of the expert panel proved robust. Response rates were 100 percent in both rounds, a striking figure in Delphi research where attrition is common, and the authority coefficients, a measure of expert credibility based on self-assessed familiarity and judgment, were 0.913 and 0.917 across the two rounds, well above the thresholds typically considered acceptable. The degree of coordination among expert opinions, assessed with Kendall’s W, indicated that the panel converged rather than merely coexisted with disagreement. After two rounds of refinement, every remaining item achieved consensus.

The final framework is a three-level architecture comprising 5 first-level items, 15 second-level items, and 48 third-level items. In plain terms, the top level captures broad domains of emergency nursing competence, the middle level breaks those domains into distinct capability areas, and the finest level specifies concrete, trainable skills and knowledge points. This granularity matters for curriculum designers: a training director can look at the third-level items and translate them directly into learning objectives, simulation scenarios, and assessment checklists, rather than guessing how an abstract competency like clinical judgment should be taught in practice.

What distinguishes this framework from earlier competency lists is its integration of two axes at once. One axis is the competency itself, from foundational clinical skills to advanced decision-making and leadership. The other is the nurse’s professional stage, from N1, the novice just entering emergency practice, to N4, the expert who mentors others and handles the department’s most complex cases. By crossing these axes, the framework ensures that the same domain, say resuscitation management, is taught at increasing depth and autonomy as the nurse advances. The experiential learning cycle supplies the pedagogical method at each tier: novices accumulate concrete experiences under supervision and learn to reflect on them, while senior nurses conceptualize patterns across cases and experiment with system-level improvements.

The study’s authors are candid about its limits. The Delphi method produces consensus among experts, not proof of effectiveness. Whether the framework actually improves patient outcomes, accelerates competency development, or reduces turnover when deployed in real emergency departments remains an open question that will require empirical validation in future research. The framework is also rooted in the Chinese nursing context, with its specific N1-to-N4 professional grading system, so international adoption would require adaptation to local career structures and regulatory requirements. Still, the authors position the work as a theoretical foundation for standardized training design and competency certification, the scaffolding on which future intervention studies and policy can be built.

The significance of the study extends beyond emergency nursing. Health systems worldwide struggle with the transition from classroom to bedside, and with retaining nurses long enough for them to become experts rather than burning out in their first years. A framework that treats learning as a structured, career-long cycle of experience, reflection, conceptualization, and experimentation offers a template that other specialties, from intensive care to oncology, could adapt. For a profession under strain, the message is quietly radical: expertise is not an accident of time served, but an outcome that can be engineered, tier by tier, with the right theory behind the training.

Subject of Research: A Delphi-based tiered core competency training framework for emergency nurses grounded in Kolb's Experiential Learning Cycle

Article Title: A tiered training program for emergency nurses’ core competencies based on Kolb’s Experiential Learning Cycle: A Delphi study

Article References: Wu, Q., Chen, C., Cai, W., Cheng, F., & Jiang, Y. (2026). A tiered training program for emergency nurses’ core competencies based on Kolb’s Experiential Learning Cycle: A Delphi study. BMC Nursing, 25(1), Article 846. https://doi.org/10.1186/s12912-026-05152-3

Image Credits: AI Generated

DOI: 10.1186/s12912-026-05152-3

Keywords: emergency nursing, Kolb's Experiential Learning Cycle, Delphi method, core competencies, tiered training framework, nursing education, competency certification, professional development, expert consultation, China, curriculum design, BMC Nursing

The study’s central discovery is a striking divergence between importance and influence. When 18 Vietnamese experts ranked six stakeholder groups by perceived importance, Central Government came first and Tourism Enterprises second. But when the researchers traced the directional flow of influence through the governance system, a different picture emerged. Tourism Enterprises, despite their enormous systemic prominence, turned out to be net receivers of influence, absorbing more pressure from other actors than they exert. Meanwhile, Knowledge and Research Institutions, which received the lowest importance weight of all six groups, occupy a net-influencing position, quietly shaping the system far beyond their apparent weight. Local Government also sits in the net-influencing camp, while Local Communities and Tourists are classified as effect-oriented, meaning they are largely acted upon rather than acting.

Technically, the study is built on an integrated three-method framework that has become a workhorse of complex decision analysis. First, a two-round Delphi process was used to validate the stakeholder architecture: six groups and twelve sub-factors, refined through iterative expert consensus until judgments stabilized. The Delphi method, developed originally at the RAND Corporation for forecasting, works by anonymously aggregating expert judgments across successive rounds, allowing participants to revise their views without group pressure. Second, the Analytic Hierarchy Process, or AHP, estimated the relative importance of each stakeholder group and sub-factor by structuring the problem into a hierarchy and deriving priority weights from pairwise comparisons. Third, the Decision-Making Trial and Evaluation Laboratory, or DEMATEL, assessed the directional interdependencies among the six stakeholder groups, converting expert judgments about who influences whom into a mathematical map of cause-and-effect relationships across the entire system.

DEMATEL is the method that gives the study its relational teeth. Unlike AHP, which produces a static ranking, DEMATEL builds a directed influence matrix from expert assessments of pairwise relationships, then decomposes it into two measures for each actor: prominence, the total degree of systemic involvement, and net influence, the balance between influence exerted and influence received. Actors with positive net influence are classified as causes, driving the system, while those with negative net influence are effects, being driven. This distinction is what allows the study to separate systemic prominence from directional power, a separation that most stakeholder analyses conflate. The result is a system-level portrait of governance that reveals not just who matters, but who moves whom.

The study’s conceptual contribution is equally significant. Rather than treating digital transformation as an autonomous force that automatically drives sustainability, the authors conceptualize it as a governance-embedded coordination mechanism, a tool whose effects depend on the institutional architecture within which it operates. This reframing matters because much of the digital sustainability literature assumes that technology itself, whether booking platforms, smart destination management systems, or data dashboards, will inevitably push tourism toward greener outcomes. The Vietnamese case suggests otherwise: digital tools coordinate stakeholders, but the direction of coordination is set by the underlying power structure, not by the technology.

The divergence between importance and influence has direct policy implications. Central Government’s dual position, first in AHP importance and strongest in net influence, confirms its role as the system’s primary driver, consistent with Vietnam’s state-led governance tradition. But the fact that Tourism Enterprises, the second most important group, are net receivers of influence suggests that the private sector’s enormous practical role in delivering sustainable tourism is not matched by corresponding directional power. Knowledge and Research Institutions, meanwhile, occupy a paradoxical position: lowest in perceived importance, yet net influencers of the system. This suggests that expert knowledge quietly shapes governance outcomes even when institutions themselves are undervalued in formal policy discourse.

Local Government’s net-influencing position is another notable finding. In Vietnam’s governance system, provincial and municipal authorities implement tourism policy on the ground, mediating between national directives and local realities. Their classification as net influencers confirms that this mediating role carries real directional power, not merely administrative weight. By contrast, Local Communities and Tourists, the groups most directly affected by tourism’s sustainability outcomes, are classified as effect-oriented, meaning they receive more influence than they exert. This asymmetry is precisely the kind of structural imbalance that participatory governance frameworks are designed to correct, and the study’s findings give that correction a concrete target.

The study extends stakeholder salience research, a tradition rooted in Mitchell, Agle, and Wood’s classic 1997 framework, which identified power, legitimacy, and urgency as the attributes that make stakeholders salient. By incorporating a relational, system-level perspective, the Vietnamese study adds a fourth dimension: directional influence within a network of interdependencies. A stakeholder can be salient in the classic sense, powerful and legitimate, yet still be structurally positioned as an effect rather than a cause. This network view is particularly relevant under digital transformation, when booking platforms, social media, and destination management systems create new channels of influence that do not map neatly onto traditional institutional hierarchies.

For Vietnam’s national sustainable tourism governance system, the study provides policy-relevant insights for strengthening institutional coordination, knowledge support, and stakeholder participation. The findings suggest that policymakers should not assume that formal importance translates into directional power, and that governance reforms should target the specific asymmetries the study identifies: elevating the directional role of Tourism Enterprises, sustaining the quiet influence of Knowledge and Research Institutions, and creating channels through which Local Communities and Tourists can shift from effect-oriented to cause-oriented positions. Under digital transformation, when the coordination mechanisms themselves are being rebuilt, these structural corrections are more feasible than ever.

The study’s methodological rigor, validated through a two-round Delphi consensus process and confirmed by institutional research guidelines at Thai Nguyen University of Economics and Business Administration, gives its findings credibility. The research involved voluntary expert participation without clinical intervention or identifiable personal data, and informed consent was obtained from all participants. Published open access under a Creative Commons Attribution 4.0 license, the study is citable and carries a permanent DOI, ensuring that its findings on the hidden wiring of sustainable tourism governance will be accessible to researchers, policymakers, and practitioners worldwide. As digital transformation continues to reshape tourism governance globally, the Vietnamese case offers a template for mapping stakeholder influence that other nations can adapt, revealing not just who matters in sustainability, but who actually moves the system.

Subject of Research: Stakeholder influence mapping in sustainable tourism governance under digital transformation

Article Title: Stakeholder influence in sustainable tourism under digital transformation through an integrated Delphi, AHP and DEMATEL framework

Article References: Nguyen, N. H., & Shyu, W.-H. (2026). Stakeholder influence in sustainable tourism under digital transformation through an integrated Delphi, AHP and DEMATEL framework. Discover Sustainability. https://doi.org/10.1007/s43621-026-04922-y

Image Credits: AI Generated

DOI: 10.1007/s43621-026-04922-y

Keywords: sustainable tourism, digital transformation, stakeholder governance, AHP, DEMATEL, Delphi method, Vietnam, tourism policy, stakeholder salience, governance, Stakeholder, influence

The new tool, called the AI Literacy Scale, is a 33-item questionnaire organized around four distinct dimensions: Technical Understanding, Ethical Awareness, Practical Ability and Social Impact. Rather than treating AI literacy as a single vague skill, the scale breaks it into measurable components that map onto what students must actually know and do. Technical Understanding probes grasp of how AI systems work, Ethical Awareness covers bias, privacy and responsible use, Practical Ability addresses applying AI tools in learning and clinical contexts, and Social Impact examines the broader consequences of AI for healthcare systems and society.

What makes the study notable to measurement specialists is the theoretical scaffolding behind the item pool. The researchers drew on UNESCO’s AI Competency Framework for Students, the Technological Pedagogical Content Knowledge model and the Kirkpatrick evaluation model, supplementing these frameworks with a systematic literature review and qualitative interviews with students. Items were then refined through a Delphi process, in which expert panels iteratively rate and revise candidate questions, followed by content validation and pilot testing. This layered approach is designed to ensure that every question on the final scale has a defensible theoretical and empirical reason for being there.

To test the instrument, the team collected responses from 473 health professions students and split the sample in two. The first subsample of 241 students was used for exploratory factor analysis, a statistical technique that searches for hidden structure in the data and reveals how many independent dimensions the questionnaire is actually measuring. The second subsample of 232 students was reserved for confirmatory factor analysis, which tests whether the four-factor structure identified in the first half holds up in fresh data, alongside a Rasch rating scale analysis that examines each item individually.

The results were strong by conventional standards. The four-factor solution explained 63.676 percent of the total variance, meaning nearly two-thirds of the differences between students could be captured by the four dimensions. Internal consistency, measured by Cronbach’s alpha, reached 0.979, an exceptionally high value indicating that the items hang together coherently. Confirmatory factor analysis showed acceptable model fit, with a chi-square to degrees of freedom ratio of 2.175, incremental fit index of 0.917, Tucker-Lewis index of 0.908, comparative fit index of 0.916 and root mean square error of approximation of 0.071, all within or near accepted thresholds for a well-fitting measurement model.

The Rasch analysis added a different and arguably more demanding layer of evidence. Unlike classical factor analysis, Rasch modeling converts raw questionnaire responses onto a common interval scale, or logit scale, allowing researchers to compare the ability of each person with the difficulty of each item directly. In this study, person reliability was 0.96 and item reliability was 0.95, both excellent figures that indicate the scale produces highly reproducible measurements of students and stable difficulty estimates for items.

But the Rasch analysis also exposed a limitation that classical statistics would have missed. The person-item map, a visual plot that places students and questions along the same difficulty axis, showed that the average student scored 1.66 logits above the average item difficulty of 0.00 logits. In plain terms, the students were, on average, too skilled for the questions being asked. The scale has relatively few items hard enough to distinguish among the most capable students, a problem known as limited upper-end targeting. The authors suggest several possible explanations: the items may simply not be difficult enough, the sample may have had unusually strong AI backgrounds, students may have accumulated substantial prior AI experience outside the curriculum, or health professions programs may not yet emphasize advanced AI competencies in their teaching.

This finding carries practical consequences for how the scale should be used. The researchers conclude that the AILS is currently better suited to group-level assessment of foundational and intermediate AI literacy than to identifying students with advanced competence. For curriculum designers and education researchers, that means the questionnaire can reliably tell whether a cohort is building baseline AI knowledge and skills, but it cannot yet serve as a fine-grained diagnostic for expertise at the top of the distribution. The team recommends developing more challenging items and validating the scale in independent, multicentre samples before broader deployment.

The study also makes a methodological argument that extends well beyond AI literacy. Combining exploratory and confirmatory factor analysis with Rasch measurement gave the researchers two complementary lenses: factor analysis confirmed the internal structure of the construct, while Rasch analysis revealed how individual items function and whether the difficulty of the test matches the ability of the population. The authors argue that this combined approach illustrates the potential value of Rasch analysis for evaluating item functioning and person-item targeting in health professions education more generally, where many existing questionnaires have never been examined at the item level.

The timing of the work reflects a growing urgency in medical education. As AI reshapes healthcare, new competency requirements are falling on health professionals who must understand, critique and safely deploy these systems, yet assessment instruments tailored to this population have remained scarce and item-level measurement evidence scarcer still. A validated scale gives educators a common yardstick for evaluating whether curricula are working, for benchmarking students across institutions and for tracking progress as AI content is woven into training. The researchers caution that applying the AILS in other professional education domains would require discipline-specific adaptation and revalidation, but within health professions education the instrument now offers one of the most thoroughly tested measures available. The study was conducted in accordance with the Declaration of Helsinki, approved by the Academic Ethics Committee of Zhuhai Campus of Zunyi Medical University, and funded by the university’s Future Distinguished Faculty Development Program.

Subject of Research: Development and validation of an AI literacy questionnaire for health professions students using factor analysis and Rasch measurement

Article Title: Development and validation of the AI literacy questionnaire for health professions students using Rasch analysis

Article References: Li, Y., Ye, Q., Qin, W., & Liu, N. (2026). Development and validation of the AI literacy questionnaire for health professions students using Rasch analysis. BMC Medical Education. https://doi.org/10.1186/s12909-026-10613-z

Image Credits: AI Generated

DOI: 10.1186/s12909-026-10613-z

Keywords: AI literacy, health professions education, Rasch analysis, questionnaire validation, psychometrics, medical education, scale development, confirmatory factor analysis, UNESCO AI Competency Framework, assessment instruments, artificial intelligence, Delphi method

The research team, led by Jorge Sanabria-Z of the School of Architecture, Art and Design at Tecnológico de Monterrey, together with Pamela Olivo, Amadeo Arguelles of the Instituto Politécnico Nacional, and Inna Artemova of the Universidad de Guadalajara, set out to address a conspicuous gap. Science museums are increasingly positioned as venues for non-formal sustainability education, yet systematic frameworks for integrating AI-supported virtual laboratories into these environments remain scarce. Virtual laboratories themselves are not new; simulations and remote-access experiments have been used in formal education for years. What distinguishes the new framework is its focus on the museum context, where audiences are heterogeneous, visits are voluntary and often brief, and the goals blend entertainment, education, and civic engagement around socioecological challenges such as climate change, biodiversity loss, and resource management.

Methodologically, the study took a Delphi-oriented approach, a technique traditionally used to build consensus among experts through iterative rounds of reflection and feedback. In this case, the process was internal and designer-led, combining two rounds of multidisciplinary reflection with synthesis assisted by a large language model. The authors used GPT-4, developed by OpenAI, as a methodological instrument to synthesize expert inputs into preliminary dimensional frameworks and to propose supporting AI tools and references across the two iterative rounds. Crucially, the team documented that all LLM-generated outputs were critically reviewed, discussed, and refined by the expert panel, with references independently verified against academic databases. This transparent handling of AI assistance reflects a growing awareness in the research community that generative models can accelerate synthesis but must remain under rigorous human oversight, particularly when the output is meant to guide institutional design decisions.

From this consensus process, the researchers identified thirteen dimensions that collectively characterize AI-integrated virtual laboratories in museum settings. These are contextual understanding, operational sustainability, research support, educational design, audience personalization, technological infrastructure, physical-digital interaction, operability, ethics, participation, accessibility, interculturality, and international collaboration. Each dimension is paired with possible AI capabilities and intended outcomes, creating a matrix that institutions can use as a checklist during planning. The breadth of the list is telling: it signals that building a virtual laboratory is not merely a software engineering problem but a socio-technical challenge that touches pedagogy, ethics, cultural inclusion, and long-term institutional viability all at once.

Several of the dimensions deserve closer attention because they highlight tensions unique to the museum environment. Contextual understanding refers to the system’s capacity to ground scientific content in the local ecological and social realities of visitors, something particularly relevant for a museum in Guadalajara, a metropolitan region facing pressing water and environmental pressures. Audience personalization leverages AI’s ability to adapt explanations, difficulty levels, and interactive scenarios to individual visitors, from schoolchildren encountering a concept for the first time to adults with prior scientific training. Physical-digital interaction addresses the hybrid nature of the museum experience, where virtual experiments must complement rather than replace tangible exhibits, and where the flow between gallery floors and digital interfaces determines whether the technology feels integrated or intrusive.

Operability and technological infrastructure form the practical backbone of the framework. Virtual laboratories powered by generative AI demand reliable connectivity, computational resources, and maintenance plans that many cultural institutions, especially in low- and middle-income countries, must budget for carefully. Operational sustainability extends this concern into the future, asking how a laboratory will be updated, staffed, and financed beyond its launch. The inclusion of these dimensions in a conceptual framework is a quiet but important corrective to technology-driven enthusiasm, reminding designers that a dazzling AI demonstration that fails after six months serves neither the museum nor its audience. The framework’s emphasis on operability suggests that ease of use for museum educators, not just visitors, was treated as a design priority.

Ethics, accessibility, and interculturality round out the human-centered dimensions, and their presence reflects contemporary debates about the deployment of AI in public-facing institutions. Ethical considerations in this context include data privacy when systems personalize content, the transparency of AI-generated explanations, and the risk of presenting algorithmic outputs as authoritative scientific truth. Accessibility demands that virtual laboratories serve visitors with disabilities, a legal and moral requirement that is often easier to state than to implement in immersive digital environments. Interculturality pushes further, asking how a laboratory can respect and incorporate Indigenous and local knowledge systems alongside Western scientific frameworks, a question of particular resonance in Mexico, where the intersection of diverse knowledge traditions and formal science education is an active area of scholarship and policy.

Participation and international collaboration connect the framework to the broader movement of citizen science, in which members of the public contribute observations, data, and even research questions to genuine scientific projects. The study explicitly frames AI-supported virtual laboratories as platforms for citizen science participation oriented toward socioecological challenges, meaning that visitors would not merely consume simulations but potentially engage with real environmental data and contribute to ongoing research efforts. International collaboration, meanwhile, acknowledges that socioecological phenomena such as migration, drought, and pollution cross borders, and that museums can act as nodes in transnational networks of shared virtual infrastructure and open educational resources. The framework was supported by the Secretaría de Ciencia, Humanidades, Tecnología e Innovación (SECIHTI), Mexico, under its 2025 Call for Humanistic Research, grant reference IH-2025-G-125, awarded to the LúmiaLab project.

The authors are notably candid about the limits of their contribution. The framework is contextually grounded in the LúmiaLab project and the Museum of Environmental Sciences in Guadalajara, but it has not yet been independently or empirically validated. It is offered as a structured theoretical-practical reference for subsequent co-design, piloting, and external validation, rather than as evidence of educational effectiveness or generalizability. This honesty matters in a research landscape where AI-in-education claims often outrun the data. The team also disclosed a second form of AI assistance: a large language model, Claude, developed by Anthropic, was used during manuscript preparation to assist with structural organization, drafting of specific sections, and language editing, based on the authors’ conceptual directions and prior drafts. The authors reviewed, verified, and revised all such content, and no AI tool is listed as an author, with the researchers taking full responsibility for the scientific content and the accuracy of references.

For museum professionals, educators, and technologists watching the rapid evolution of generative AI, the study offers a sober starting point rather than a spectacle. Its thirteen dimensions function less as a blueprint than as a set of questions that any institution hoping to build an AI-supported virtual laboratory should be able to answer: Who is the audience and what do they need? How will the system respect ethical boundaries and cultural diversity? Can the institution sustain the technology over time? And how will anyone know whether the laboratory actually improves learning or engagement? By framing these questions before implementation, the Mexican research team has contributed a reference point for a field that is being built in real time, one in which the museum, that most public of scientific spaces, may become a proving ground for how artificial intelligence and human curiosity can productively meet.

Subject of Research: Conceptual framework for AI-supported virtual laboratories in science museums for sustainability and citizen science education

Article Title: Guidelines for AI-supported virtual laboratories in science museums

Article References: Sanabria-Z, J., Olivo, P., Arguelles, A., & Artemova, I. (2026). Guidelines for AI-supported virtual laboratories in science museums. Discover Sustainability. https://doi.org/10.1007/s43621-026-04907-x

Image Credits: AI Generated

DOI: 10.1007/s43621-026-04907-x

Keywords: virtual laboratories, science museums, artificial intelligence, generative AI, citizen science, sustainability education, non-formal education, socioecological challenges, LúmiaLab, Guadalajara, Delphi method, educational design

The central problem the researchers set out to solve is one that has long troubled evaluation and assessment research in health communication. When scientists or clinicians create visual materials—illustrations, artworks, or diagrams intended to convey health information—there is often no standardized, construct-based way to judge whether those images actually communicate biological meaning. Judgments tend to be subjective, varying from one reviewer to the next, and the criteria behind them are rarely made explicit. The team grounded their solution in two established theoretical frameworks: the Biopsychosocial Model, which frames health as an interplay of biological, psychological, and social factors, and Visual Literacy Theory, which describes how viewers construct meaning from images. From these foundations, they built a 20-item scale designed to measure how biologically interpretable a visual stimulus is.

Developing the scale was not a solitary act of authorship but a structured consensus exercise. The researchers employed a two-stage modified Delphi process, a method widely used in instrument development to distill expert opinion into stable, agreed-upon items. In the first stage, a joint panel considered both the images and the candidate items together, allowing reviewers to see how proposed criteria behaved when applied to actual stimuli. In the second stage, an independent panel evaluated the items in isolation, without the images, to test whether each item stood on its own as a clear and meaningful criterion. This dual-panel design was intended to strip out items that only made sense in context and to retain those that could function as reliable, reusable measurement tools.

The final instrument produces scores across four dimensions. Three of these are biomedical subscales, capturing different facets of biological interpretability in health-related imagery, while the fourth, Non-Medical Interpretability, assesses meaning that lies outside the biomedical domain. With the scale in hand, the team turned to the question at the heart of the study: could multimodal AI systems—models that process both images and text—apply this framework in a reproducible way, functioning as standardized raters that do not fatigue, disagree idiosyncratically, or drift in their criteria?

To answer it, the researchers assembled a small but carefully controlled stimulus set: six artworks, four of them health-themed and two serving as controls. Three multimodal AI models and four researchers, who served as an internal human reference, evaluated the images using a standardized protocol. The protocol’s design reveals how seriously the team took the methodological pitfalls of AI assessment. Each model was accessed through independent researcher-operated accounts, memory-isolation procedures were used to prevent the systems from carrying information from one evaluation into the next, and assessments were repeated across two separate days. This allowed the team to measure not only whether different accounts produced similar scores, but whether the same system produced stable scores over time.

The statistical approach was deliberately matched to the limitations of the data. Because the image type variable spanned only six stimuli—four health-themed and two controls—the researchers summarized health-versus-control differences at the image level using Cohen’s d, a standardized measure of effect size, and treated human–AI differences descriptively at the rater level rather than claiming statistical significance. Reproducibility was assessed with intraclass correlation coefficients, or ICCs, which quantify how strongly scores agree across raters or occasions, supplemented by Spearman correlations. This restraint matters: with so few images, the study cannot deliver definitive population-level conclusions, and the authors are explicit that their findings are preliminary methodological evidence rather than established generalizations.

Within that constraint, the pattern of results is instructive. The AI-generated scores were descriptively higher for health-themed images than for control images on the three biomedical subscales, with image-level Cohen’s d values ranging from 0.91 to 1.38—effect sizes that, in a larger sample, would typically be considered large. On the Non-Medical Interpretability subscale, the difference was smaller, at d = 0.49. With only four health-themed and two control images, none of these differences reached statistical significance, with p values ranging from 0.19 to 0.60. In other words, the machines appeared to distinguish health-themed from control material in the expected direction, particularly on biomedical criteria, but the sample was too small to confirm that the distinction was real rather than chance.

Reproducibility told a more nuanced story. Across researcher-operated accounts, agreement was excellent for the three biomedical subscales, with ICCs between 0.90 and 0.97—figures that suggest the AI systems, when run under controlled conditions, applied the biomedical criteria with remarkable consistency. Agreement on the Non-Medical Interpretability subscale was moderate, at an ICC of 0.57, indicating that the machines were considerably less consistent in judging meaning outside the biomedical domain. Most strikingly, agreement across assessment days was poor, with ICCs ranging from just 0.04 to 0.46. Whatever produces this day-to-day instability—whether model updates, sampling variability in the models’ generative processes, or subtle differences in prompt handling—it represents a genuine obstacle to using AI raters as stable measurement instruments without further safeguards.

The comparison with human raters added another layer of complexity. The four researchers’ internal reference scores ran higher than the AI-generated scores on the biomedical subscales for health-themed images, with mean differences of 1.41 to 1.96 points, suggesting the humans saw more biological interpretability in the health-themed artworks than the machines did. On Non-Medical Interpretability, the pattern reversed: the human scores were lower than the AI scores by 2.02 to 2.19 points, meaning the machines detected more non-medical meaning than the human panel. This divergence is intriguing, because it hints that AI systems and trained human raters may weight different aspects of an image differently, and that calibration between human and machine judgment is not automatic. It also flags the Non-Medical Interpretability subscale as the weakest component of the instrument; the authors report that it showed substantial internal-consistency problems, meaning its items did not hang together as a coherent measure, and it requires further refinement.

The authors are careful about what their study does and does not establish. The findings, they conclude, provide preliminary methodological evidence that multimodal AI can reproducibly apply a structured interpretability framework to controlled visual stimuli, supporting the feasibility of using AI as standardized non-human raters within controlled methodological settings. But the caveats are substantial: day-to-day stability must be improved, the Non-Medical Interpretability subscale needs rework, the stimulus set was tiny and artificial, and the human reference was an internal panel rather than a diverse group of independent external raters. Broader validation using authentic health communication materials—real patient education graphics, public health campaigns, clinical illustrations—and independent external human raters remains necessary before AI raters could be trusted in applied settings. Even so, the study opens a provocative line of inquiry for a field that increasingly relies on visual media: if the criteria for biological interpretability can be made explicit enough for a machine to apply them consistently, then the machines may eventually become not just tools that make images, but disciplined judges of what those images actually communicate about health.

Subject of Research: Using multimodal artificial intelligence as standardized non-human raters to assess biological interpretability of health-related visual stimuli

Article Title: Operationalizing biological interpretability in health-related visual stimuli using multimodal artificial intelligence as standardized non-human raters

Article References: Ayman, D., Ozkan, H., Ayman, D., & Tanriverdi, O. (2026). Operationalizing biological interpretability in health-related visual stimuli using multimodal artificial intelligence as standardized non-human raters. Scientific Reports. https://doi.org/10.1038/s41598-026-75309-2

Image Credits: AI Generated

DOI: 10.1038/s41598-026-75309-2

Keywords: artificial intelligence, multimodal AI, health communication, visual literacy, biopsychosocial model, psychometrics, Delphi method, intraclass correlation, medical humanities, image evaluation, Scientific Reports, non-human raters

The instrument, named the CKD-Malnutrition Knowledge, Attitude and Practice scale, or CKD-MKAP, was described in a study published in the journal Nursing Open. Its development followed a multiphase quantitative design grounded in the well-established KAP framework of behavioural change theory, which holds that accurate perceived knowledge fosters positive attitudes, which in turn translate into effective dietary practices. The team began by synthesising a comprehensive literature review of clinical nutrition guidelines for kidney disease with qualitative findings from previously published semi-structured interviews with thirteen Chinese patients who had experienced malnutrition. Those interviews explored what patients actually knew about malnutrition, how they perceived their own nutritional status, and what barriers and motivators shaped their eating habits. Mapping those lived experiences onto the three KAP domains produced an initial pool of thirty-eight candidate items covering everything from awareness of symptoms and risk factors to beliefs about diet therapy and concrete self-management behaviours such as regular weighing and protein intake.

Refining that raw item pool required the structured judgement of experts. The researchers convened a Delphi panel of fifteen specialists, including ten nephrology experts, three clinical nutrition specialists and two psychometricians, drawn from eight institutions across five regions in two countries. More than eighty percent of the panellists had over a decade of clinical or research experience. Across two email-based consultation rounds conducted between May and July 2023, response rates reached 89.5 percent and 93.75 percent respectively, and the calculated authority coefficient of 0.77 signalled a high degree of expert reliability. Items were evaluated using boundary value methods incorporating full-score frequency, arithmetic means and coefficients of variation, with predefined statistical thresholds determining survival. Eleven items fell during the Delphi phase, including questions about regional and religious food cultures and a query on whether oral supplements could replace ordinary food, leaving a preliminary twenty-seven-item scale.

Before formal validation, a pilot study with twenty kidney disease patients tested the draft instrument for clarity and acceptability. Every participant completed the questionnaire, reported that the items were clear and easy to understand, and finished in eight to fourteen minutes, with no revisions requested. The team then moved to the main psychometric evaluation, distributing 330 questionnaires and analysing 303 valid responses collected between July and November 2023. The sample was predominantly male at 63.7 percent, with a mean age of 48.26 years, and spanned the full range of kidney disease severity, from early-stage patients to the 32 percent already at stage five. Crucially, the sample was randomly split into two independent subsamples of 152 and 151 participants, one for exploratory factor analysis and one for confirmatory factor analysis, a cross-validation strategy designed to ensure that the scale’s structure would hold up beyond the data that shaped it.

The statistical machinery behind the validation was considerable. Content validity was assessed with item-level and scale-level content validity indices, and the results were striking: item-level values ranged from 0.93 to 1.00 across all three domains, comfortably above the accepted threshold of 0.83, while the scale-level average reached 0.99. For the dichotomous Perceived Knowledge domain, classical test theory came into play, with difficulty indices expected between 0.31 and 0.60 and discrimination indices above 0.20. Seven of the eight knowledge items performed well, but one item asking simply whether patients knew what malnutrition was showed suboptimal difficulty and discrimination and was deleted. Exploratory factor analysis, using principal-axis factoring with oblique rotation, extracted a single dominant factor for both the Attitude domain, explaining 69.87 percent of variance with a Kaiser-Meyer-Olkin value of 0.931, and the Practice domain, explaining 61.84 percent of variance, a clean structure that aligned precisely with the theoretical KAP framework.

Confirmatory factor analysis then put the emerging structure through a harsher test. Because the data violated assumptions of multivariate normality, the researchers applied the robust maximum likelihood estimator to the Attitude and Practice models and the weighted least squares mean and variance adjusted estimator to the categorical knowledge items. The initial knowledge model fit poorly, and one item asking about the relationship between malnutrition and chronic renal disease, with a factor loading of just 0.304, was removed, dramatically improving fit indices to a comparative fit index of 0.978 and a standardised root mean residual of 0.061. The attitude and practice models required theoretically justified post hoc correlations among residuals of conceptually related items, such as those concerning professional nutritional interventions, before achieving acceptable fit. The final attitude model reached a comparative fit index of 0.982 and the practice model 0.976, both well above the 0.90 benchmark, with root mean square error of approximation values of 0.077 and 0.071 respectively.

The finished instrument is deliberately lean: twenty-five items across three dimensions, scored separately rather than summed. The Perceived Knowledge domain contains six yes-or-no items yielding scores from zero to six, covering nutritional status awareness, symptoms, risk factors, consequences, daily protein requirements and prevention. The Attitude domain uses nine items on a five-point Likert scale from strong disagreement to strong agreement, spanning beliefs about the importance of nutrition, the value of optimism and self-control, and the role of diet therapy in treatment. The Practice domain asks patients how often they perform ten behaviours, from prioritising nutrition intake and measuring weight regularly to consuming high-quality protein and adhering to dietary recommendations, on a frequency scale from never to always. Reliability figures were strong, with Cronbach’s alpha values of 0.740 for knowledge, 0.970 for attitude and 0.958 for practice, and composite reliability estimates of 0.743, 0.968 and 0.957. Discriminant validity was confirmed through the Fornell-Larcker criterion, with the square roots of average variance extracted for each domain exceeding all inter-construct correlations, which ranged from 0.443 to 0.585.

The clinical logic of the scale is straightforward but powerful. Because many malnourished Chinese kidney patients receive inadequate nutritional education, a brief questionnaire that separates what patients know from what they believe and what they actually do allows nurses and physicians to pinpoint the weakest link. A patient with solid knowledge but poor practices, for example, may benefit far more from behavioural counselling than from another informational pamphlet. The researchers also compared their instrument against existing alternatives, noting that a recently developed seventy-seven-item nutritional KAP questionnaire for non-dialysis patients reported domain alpha values of 0.861, 0.816 and 0.612, whereas the twenty-five-item CKD-MKAP achieves comparable or superior reliability with far less respondent burden, an important consideration for busy wards and outpatient clinics.

The authors are candid about the limits of their work. The validation sample came from a single city in northwest China, raising questions about generalisability across the country’s enormous regional and cultural diversity. Test-retest reliability was not assessed, so temporal stability remains unknown, and the scale’s responsiveness to change has yet to be demonstrated, meaning it cannot yet be formally adopted as an outcome measure for evaluating educational interventions without further longitudinal psychometric testing. The post hoc residual correlations in the attitude and practice models, while theoretically justified, were not independently cross-validated and may hint at local dependence or hidden subdimensions that larger studies should investigate. No formal a priori power analysis was conducted, though the observed strong communalities and factor loadings compensate for the modest sample sizes.

Even with those caveats, the arrival of a culturally adapted, psychometrically validated malnutrition KAP scale fills a genuine gap. Previously available Chinese tools examined isolated KAP dimensions or lacked rigorous validation, while international scales carry assumptions shaped by foreign socioeconomic and dietary contexts that limit their applicability to Chinese patients, whose food culture, linguistic nuances and healthcare experiences differ substantially. By embedding items rooted in the actual reported experiences of Chinese patients and then subjecting them to Delphi consensus, item analysis, exploratory and confirmatory factor analysis, the team has produced an instrument that is short enough for routine clinical use yet statistically robust. If subsequent multicentre validation confirms its stability across regions, the CKD-MKAP could become a standard part of nursing assessments for kidney patients, turning the quiet epidemic of protein-energy wasting into a measurable, targetable and ultimately preventable problem.

Subject of Research: Development and psychometric validation of a malnutrition knowledge, attitude and practice scale for chronic kidney disease patients in China

Article Title: Development and Validation of CKD‐MKAP: A Malnutrition Perceived Knowledge, Attitude, and Practice (KAP) Scale for Patients With Chronic Kidney Diseases (CKD) in China

Article References: Zhu, S., Yusoff, H., Cheng, K. Y., Yusoff, D. M., Feng, X., Zheng, W.-K., & Chen, H. (2026). Development and Validation of CKD ‐ MKAP : A Malnutrition Perceived Knowledge, Attitude, and Practice ( KAP ) Scale for Patients With Chronic Kidney Diseases ( CKD ) in China. Nursing Open, 13(10), Article e70925. https://doi.org/10.1002/nop2.70925

Image Credits: AI Generated

DOI: 10.1002/nop2.70925

Keywords: chronic kidney disease, malnutrition, KAP scale, psychometric validation, nursing, protein-energy wasting, Delphi method, factor analysis, China, nutritional assessment, scale development, patient education