Addressing the global burden of hearing loss, which affects over 1.5 billion people, requires public access to reliable health information. The absence of quality, evidence-based, and accessible information has been recognized as a significant public health impediment (Tordrup et al., 2022; World Health Organization, 2021). Wikipedia, as one of the most accessed sources of health information, presents a strategic opportunity for disseminating evidence-based knowledge on hearing health (Morata et al., 2024; Santangelo et al., 2022; Smith, 2020). The coverage of scientific content in Wikipedia varies across disciplines and language domains in terms of the availability and quality of collaboratively documented content on audiology (Lewoniewski et al., 2023; Montilla et al., 2023; Yasseri et al., 2022).

Integrating Wikipedia editing into higher education health curricula has been proposed and implemented as a mechanism to address information gaps (Azar, 2023; Khalid et al., 2023). This active learning approach positions students as contributors to knowledge production rather than passive recipients (Maggio et al., 2020; Tham et al., 2021). Wikipedia editing contributes to the development of scientific communication and digital literacy skills (Kennedy, 2021; Stine, 2022). Studies have examined various methods and educational models (see Wikipedia:Academic studies of health information on Wikipedia [https://w.wiki/6r8U] and Wikipedia:Academic studies of Wikipedia in education [https://w.wiki/9Muc]) demonstrating Wikipedia's role as a validated, crowdsourced platform for health communication and literacy competence development (Davis et al., 2023; Morata et al., 2024; Smith, 2023). Most studies have not replicated educational methods to validate and assess student perceptions of Wikipedia editing assignments (Bridges & Dowell, 2020; Ceballos et al., 2021). Geographic, cultural, and educational level biases are possible. Available evidence is mostly derived from English-language contexts or restricted generalization due to limited sample sizes (Ruprechter et al., 2023). Uncertainty remains regarding whether students in diverse educational, geographical, and linguistic contexts perceive Wikipedia as a learning model in digital writing environments in a similar manner (Park & Bridges, 2022; Tham et al., 2021). Addressing this problem is essential for informing the use of Wikipedia as a tool to disseminate literacy and health knowledge among students as knowledge producers (Jemielniak, 2019).

Evaluating student perceptions of Wikipedia activities using a previously validated instrument allows for the comparison across diverse contexts. This approach also enables testing the scalability and adaptability of the assessment method and assists in identifying facilitators and barriers to Wikipedia-based tasks in higher education.

We examined if the selected instrument can detect differences in students' perceptions, differing from those found in existing literature for comparisons of traditional training methods. In addition, we evaluate the contribution of Wikipedia editing assignments in promoting competencies such as critical information analysis, health communication, and digital knowledge production. This study evaluates the experiences of undergraduate audiology students from a Brazilian public university in a Wikipedia editing assignment, using an adapted course perceptions questionnaire.

This study uses a cross-sectional and descriptive design. The study received Ethics Committee approval (process numbers 6.114.935 and 6.823.110). All participants signed an informed consent form, following Resolution 196/1996 of the Brazilian National Research Ethics Commission. The reporting of this study adheres to the STROBE (Strengthening the Reporting of Observational Studies in Epidemiology) guidelines for cross-sectional studies.

We used an adapted version of the instrument designed to evaluate Wikipedia assignments (Ceballos et al., 2021). The original instrument (in English) was previously applied to a sample of graduate students (N = 10). Selected questions from the original questionnaire were adapted for use in the context of Wikipedia's educational activities held at Brazilian universities. The original version was translated into Brazilian Portuguese by the research team.

The instrument (Table 1) contains 10 selected items with responses to be given on a six-point Likert scale: Strongly Agree, Agree, Neutral, Disagree, Strongly Disagree, and No Response. Some items from the original questionnaire were omitted in the Portuguese adaptation to align with the specific Wikipedia assignments and coursework expectations for deliverables.

The questionnaire examined students' views on Wikipedia in university classrooms. It assesses: (i) training and material usability [Q1]; (ii) critical thinking through analysis, evidence use, and bias assessment [Q2, Q3, Q6]; (iii) health communication by writing for lay readers [Q4, Q5]; (iv) problem-solving and creativity in editing [Q7]; (v) becoming a knowledge producer and understanding collective knowledge building [Q8, Q9]; and (vi) participation and comparison to traditional methods [Q10]. For the purpose of this study, the 'traditional assignments' used for comparison consisted of standard theoretical lectures followed by written examinations and oral presentations. Translation centered on linguistic and semantic equivalence for instructional use and was validated by the original authors (Portuguese adaptation in Appendix 1).

Data collection was conducted using Google Forms, and responses were tabulated in Google Sheets. Descriptive statistics summarized the frequency and residuals of each questionnaire response category. For inferential analysis, responses were classified into three groups: positive (strongly agree and agree), neutral, and negative (strongly disagree and disagree). Response trends were analyzed using Chi-square Goodness of Fit tests by comparing observed distributions to an equal probability model. This statistical approach was selected to identify items where student perceptions significantly coalesced into a distinct consensus pattern, differentiating them from random or fragmented opinion distributions. Standardized residuals were examined post-hoc (absolute values > 1.96) to determine the direction and strength of these consensus (Disagreement, Neutrality, or Agreement) contributing to the significant chi-square results. Reported measures include the χ² statistic, significance (p < .05) with Holm-Bonferroni correction (p-adjusted) to control the Family-Wise Error Rate (multiple testing), Cohen's w effect size (W-benchmarks: .10 small, .30 medium, .50 large), 95% confidence intervals (CI), and statistical power (Type II error, 1 − β, threshold set at .80). In cases where the Chi-square test yielded inconclusive results regarding consensus, a post-hoc one-sample Wilcoxon signed-rank test was applied to confirm whether the response distribution significantly deviated from the neutral midpoint, reporting the Hodges-Lehmann estimator with 95% CI and effect size (r). Psychometric reliability of the adapted instrument was explored by calculating Cronbach's alpha (α) for the competency domains (Q2-Q9). This analysis described the intervention's consistency rather than validating the scale's structure. Furthermore, to investigate the relationship between material usability (Q1) and the perceived value of the assignment (Q10), a Spearman rank correlation (rho) was computed. All analyses were conducted using R (version 4.4.3) on the R Studio IDE (Posit Software, version 2025.09.2 + 418), using the Quarto system for reproducible reporting.

The response patterns for all questionnaire items are displayed in Table 2, illustrating the distribution of student selections across the six-point Likert response categories.

The psychometric evaluation of the competency domains (Q2–Q9) demonstrated high internal consistency, with a Cronbach’s α of .92 (95% CI [.89, .95]). Table 3 provides an overview of the inferential analyses, including concordance trends, chi-square statistics, effect-size estimates, and post-hoc standardized residuals used to identify which response categories contributed most to the observed deviations from the expected distribution.

The proportion of students who found the training, handouts, and assignments easy to understand and helpful (Q1) was 50.8% (30/59; 95% CI [37.5, 64.1]). Post-hoc analysis indicated that questions Q2 through Q9, which had large effect sizes (w > .50), achieved robust statistical power (1-β > .99). The remaining questions with medium effect size (w ≈ .30) had lower statistical power (1-β < .80): Q1 (.760) and Q10 (.359). Adjustment for multiple comparisons using the Holm-Bonferroni method confirmed the robustness of the findings. This validation ensures that the observed statistical patterns represent genuine response trends rather than artifacts of multiple testing.

The sample reported strong agreement on the need for critically evaluating information sources (Q2: 79.7%, 47/59; 95% CI [67.2, 89.0]; w = .991, 95% CI [.725, 1.238]) and assessing articles for missing information or bias (Q6: 91.4%, 53/58; 95% CI [81.0, 97.1]; w = 1.232, 95% CI [.966, 1.482]). Statistical analysis confirmed these results with high significance and robust power (1-β > .99).

The perceived need for using evidence to justify edits was also considered strong (Q3: 88.1%, 52/59; 95% CI [77.1, 95.1]; w = 1.164, 95% CI [.901, 1.413]). The necessity to communicate effectively with lay audiences (Q4) garnered 89.8% agreement (53/59; 95% CI [79.2, 96.2]; w = 1.199, 95% CI [.936, 1.448]), whereas the significance of acquiring writing skills for lay readers (Q5) attained the highest agreement rate in the study at 93.2% (55/59; 95% CI [83.5, 98.1]; w = 1.271, 95% CI [1.008, 1.520]). All items in this domain maintained high statistical significance (p < .001).

Students identified the assignment as an authentic problem-solving exercise requiring both disciplinary knowledge and creativity (Q7: 83.1%, 49/59; 95% CI [71.0, 91.6]; w = 1.058, 95% CI [.794, 1.306]). They also reported agreement regarding the shift from knowledge consumer to producer (Q8: 74.6%, 44/59; 95% CI [61.6, 85.0]; w = .877, 95% CI [.610, 1.124]) and recognition of collective knowledge creation (Q9: 71.2%, 42/59; 95% CI [57.9, 82.2]; w = .847, 95% CI [.580, 1.093]). This shift demonstrated a statistically robust consensus (p < .001, 1-β > .99), with strong effect sizes indicating that students perceived the task as having effectively facilitated their transition to active roles in scientific knowledge construction.

Competency domains (Q2-Q9) and material usability (Q1) demonstrated significant consensus, evidenced by 'Agreement' standardized residuals exceeding the threshold (SR > 1.96). This pattern did not occur for the assignment's comparative value (Q10), consistent with the non-significant Chi-square result (p-adjusted = 0.184), no response category exhibited |SR| < 1.96, confirming an undefined tendency.

In order to validate this finding, a supplementary Wilcoxon signed-rank test confirmed that the Q10 median response did not significantly deviate from the neutral midpoint (median = 3.00, 95% CI [2.5, 3.5], p = .901), with a negligible effect size (r = .016). Furthermore, a strong positive correlation was found (Spearman’s r = .67, 95% CI [.50, .79], p < .001) between the perceived usability of materials (Q1) and the overall value attributed to the assignment (Q10).

The globalization of innovative educational tools, such as Wikipedia-based assignments, is likely to necessitate significant adaptation to local linguistic and cultural contexts. This study's findings highlight a disconnect between the perception of the pedagogical value of such assignments and their practical implementation in a non-Anglophone setting.

Statistical analysis identified two distinct perception patterns. Competency domains (Q2–Q9) exhibited strong positive residuals, confirming a robust consensus on pedagogical benefits. In contrast, material usability (Q1) and comparative value with traditional assignments (Q10) presented fragmented distributions. Q1 retained statistical significance despite lower agreement (50.8% versus 49.1% for neutral and disagreement combined), whereas Q10 failed to reject the null hypothesis (random distribution), indicating a strictly polarized opinion (39% for both agreement and disagreement) rather than a defined trend. This suggests that implementation challenges were structural, not pedagogical. The high internal consistency found in the competency domains (Cronbach’s α = .92, 95% CI [.89, .95]) confirms that students perceived these skills, critical thinking, communication, and knowledge production, as an interconnected pedagogical construct, validating the adaptation of the instrument for this context.

Participants showed substantial consensus on the Wikipedia task's value for developing critical audiology competencies. This consensus is most pronounced in critical thinking, health communication, and problem-solving. Divergence from previous literature did not involve core competencies. Instead, it centered on instructional material usability and the task's comparative relevance.

Students reported that the Wikipedia assignment had significantly enhanced critical thinking and information analysis (Q2: 79.7%; Q6: 91.4%). Their positive perceptions (w > .99) further suggest that the methodology promotes scientific rigor and careful source selection. This is underscored by the specific consensus on the need to use evidence to build a case for edits (Q3: w = 1.164, 95% CI [.901, 1.413]), a skill essential for evidence-based audiology practice. These findings align with the reference study (Ceballos et al., 2021), when most graduate students also recognized the task's value for developing critical thinking. The consistency across educational levels and geographic contexts indicates that Wikipedia editing tasks broadly and reliably foster analytical competencies.

The proportion of students who found the training, handouts, and assignments easy to understand and helpful (Q1) was 50.8% (30/59; 95% CI [37.5, 64.1]), lower than the 70% reported in the reference study (Ceballos et al., 2021). Statistical significance was observed (p-adjusted = .025), with moderate statistical power (1-β = .760) and effect size (w = .386, 95% CI [.084, .624]). This lower agreement rate suggests challenges in creating and driving training approaches, such as those based on the Wiki Education Foundation training modules, for the educational context. Possible explanations include translation and cultural adaptation issues beyond semantic equivalence, differences in digital literacy, limited prior experience with platforms such as Wikipedia, and variations in academic background between undergraduate and graduate students (Bawack & Kala Kamdjoug, 2020; Selwyn & Gorard, 2016). This finding has important methodological implications, indicating that scaling Wikipedia-based tasks in non-Anglophone contexts requires substantial contextual adaptation addressing linguistic, cultural, and educational level differences.

Questions comparing the assignment to traditional methods (Q10) revealed the largest discrepancies with the reference study. Only 39% of Brazilian students agreed that the project added more value than a traditional assignment (23/59; 95% CI [26.5, 52.6]), compared to 70% in the reference study. In contrast to the other domains, the Chi-square test indicated no significant consensus (p-adjusted = .184), a finding validated by a subsequent Wilcoxon signed-rank test (p = .901). This statistical neutrality highlights a sharp contrast: while students recognize the specific skill-building value of the task, the perceived difficulty of the materials (Q1) likely offsets the perceived comparative advantage over traditional methods, resulting in a strictly polarized opinion. The discrepancy suggests that undergraduates in the studied non-Anglophone context may not perceive the intrinsic or extrinsic value of contributing to a global knowledge platform as superior to their perceived value of a traditional assignment.

Interpreting these divergences requires consideration of several interrelated factors. Educational level is significant; graduate students generally possess greater academic autonomy and a more developed understanding of public knowledge dissemination. This principle is applicable even in contexts with varying levels of technical and professional proficiency, as seen in the distinction in autonomy and professional practice between undergraduate and graduate education (Al Gharsi et al., 2024). In contrast, undergraduates may prefer traditional structures such as exams and reports, perceiving innovative tasks as more laborious without clear benefits. This outcome is strongly driven by instructional friction, evidenced by the strong correlation (Spearman’s r = .67, 95% CI [.50, .79]) between material usability and perceived value. This statistical link supports the hypothesis that difficulties with the instructional materials increased the cognitive load required of the students, which in turn obscured the assignment's pedagogical benefits. This finding diverges from the established literature, which typically reports that students positively perceive innovative and non-traditional methodologies (Melo et al., 2022).

We must also consider the potential variability in perception among undergraduates. The rapid, short-term context of the activities may influence this variability, potentially making it difficult for some students to fully understand its pedagogical value. Furthermore, students possess diverse prior writing experiences; those accustomed primarily to formal academic assignments perceive the public-facing, encyclopedic style of Wikipedia editing differently (Vetter et al., 2019). While cultural and societal factors, such as the perceived reach of the Portuguese versus English Wikipedia, might play a minor role in motivation, the variance in student background and activity framing is likely a driving factor of motivation (Azar, 2023).

These results have implications for the scalability and adaptation of Wikipedia editing tasks. The strong consensus on the development of critical thinking and health communication reinforces Wikipedia assignments in educational settings as an effective approach for promoting these essential competencies across diverse contexts. However, the findings also indicate that contextual adaptation is essential. The lower usability of materials demonstrates that translation alone is inadequate; successful implementation in non-Anglophone contexts requires cultural adaptation, additional instructional support, and specific modules on digital literacy.

Comparative perceptions are intrinsically context-dependent. Although the coursework assignment promotes valuable skills, students may not view it as superior to traditional methods, particularly when usability barriers exist. Explicit connections between the assignment and professional learning objectives, such as knowledge translation skills valued in public health, are therefore essential.

This study provides evidence that confirms the applicability of the proposed assessment instrument. Brazilian undergraduate audiology students showed strong consensus on the value of Wikipedia editing for developing essential competencies, including critical thinking, health communication, and knowledge production, thus validating the first dimension of the hypothesis. In contrast, results also reveal significantly different perceptions regarding the assignment's relevance compared to traditional methods, with students perceiving no additional value or greater engagement. These findings suggest that, while Wikipedia editing is considered effective for promoting competencies in diverse contexts, successful implementation requires careful adaptation of instructional materials and explicit alignment with professional objectives. With appropriate resourcing and sensitivity to cultural and educational specificities, this methodology holds substantial potential as a scalable framework in the evaluation of health education initiatives.