Developing and evaluating a creativity-oriented labor education course for students in Guangdong higher vocational colleges

1 Introduction

Higher vocational education plays a pivotal role in equipping students with the practical skills required to transition into the workforce. In China, the 2014 educational reform plan and the 2022 revision of the Vocational Education Law have catalyzed efforts to bridge the labor gap by training over 10 million skilled workers (Chen et al., 2024). To investigate the effectiveness of these reforms, Guangdong Province provides an analytically important and theoretically distinctive context. Unlike inland provinces whose economies often rely heavily on traditional resource extraction or primary manufacturing, Guangdong leads the nation in regional innovation capability and R&D expenditure, anchoring the globally recognized Shenzhen-Hong Kong-Guangzhou science and technology cluster (Cai and Liu, 2020). This intense regional economic shift uniquely exposes the friction between conventional vocational training and the urgent demand for a creatively skilled workforce, making it an ideal testing ground for curriculum innovation.

Labor education courses constitute a primary vehicle for nurturing practical skills in vocational colleges. According to Zhang and Wang (2021), a labor education course refers to an academic program aimed at developing values, skills, and knowledge linked to labor, work ethics, and practical vocational training. These courses synchronize hands-on activities with social responsibility and problem-solving strategies to increase students’ understanding of the role of labor in individual and societal development. However, challenges such as the poor integration of courses in the curriculum contribute to a mismatch between student skills and the demands of innovation-driven industries. Existing labor education courses in Guangdong vocational colleges are still highly limited: they teach content confined to routine physical labor and basic service activities while failing to integrate higher-order creative thinking. The creativity-evaluation methods adopted are based on attendance and task completion rather than originality, flexibility, and problem solving. Furthermore, the concept of “creativity-oriented labor education” remains highly novel in the Chinese context; historical educational paradigms have strictly delineated manual labor training from intellectual or creative instruction, treating them as mutually exclusive domains (Zhang and Wang, 2021). This limitation reflects a broader national pattern: although innovation and entrepreneurship courses have been embedded in China's national curriculum since 2012, the country's creativity education continues to reveal systemic challenges in which traditional, rote-learning models fail to nurture the curiosity and imagination that underpin creativity (Huang et al., 2025).

Creative thinking skills are increasingly recognized as essential for graduates to adapt to the rapidly changing technological landscape and solve problems across diverse industries. Creativity encompasses the capacity to generate original, valuable, and innovative ideas, including divergent thinking, problem-solving skills, and the potential to think beyond conventional parameters (Runco and Jaeger, 2012). Wu et al. note that vocational students require creativity skills to cultivate innovation and continually adapt to changing market demands. Nevertheless, traditional vocational curricula prioritize technical competency over creative thinking, and critical thinking and creativity skills are difficult to develop in vocational education institutions because practical skills that directly help graduates solve workforce problems are prioritized (British Council, 2022). This disparity has prompted educational scholars and policymakers to advocate for curricula that employ innovative approaches integrating creativity.

Systematic reviews and empirical studies in creativity education have robustly demonstrated that targeted pedagogical interventions—such as project-based learning and design thinking—can significantly yield positive outcomes in students’ divergent thinking and problem-solving capabilities across various disciplines (Daly et al., 2014; Scott et al., 2004). Concurrently, labor education research has firmly established the efficacy of practical coursework in building work ethics and foundational manual competencies (Zhang and Wang, 2021). However, what remains profoundly unexamined in the literature is the intersection of these two domains. Despite increasing policy attention, limited empirical research has examined how historically rote labor education curricula can be systematically redesigned into integrated platforms that cultivate multidimensional creative thinking.

To address this knowledge gap, this study develops and evaluates a creativity-oriented labor education course. The curriculum design is uniquely anchored in the integration of Tyler’s (1949) curriculum model and Williams’ (1979) creativity framework. Tyler's model was deliberately selected over alternative, highly iterative approaches—such as Taba's grassroots model or Walker's deliberative approach—because its highly structured, objective-driven sequence provides the standardization and accountability required for successful integration into China's centralized vocational education system. To operationalize creativity within this rigid structure, Williams’ framework was chosen for its precise behavioral targets, providing measurable cognitive dimensions (e.g., fluency, flexibility) that translate abstract creative concepts into actionable instructional strategies. A detailed exposition of this theoretical integration is provided in Section 2.

Aligned with the four phases of the Research and Development (R&D) design, this study is guided by the following exploratory research questions:

• RQ1 (Phase I): What creativity-oriented competencies should be incorporated into labor education courses in higher vocational colleges?
• RQ2 (Phase II): How can the integrated Tyler-Williams framework be translated into specific course components, teaching behaviors, and student behaviors for a labor education curriculum?
• RQ3 (Phase IV): To what extent do students perceive the newly developed creativity-oriented labor education course as effective?

2 Literature review and theoretical framework

2.1 Labor education in Chinese higher vocational education

Labor education research has established a strong foundation for cultivating work ethics, practical competencies, and civic responsibility among vocational students (Zhang and Wang, 2021). However, existing literature reveals that these programs predominantly focus on task execution and routine physical labor. Despite its foundational role, recent Chinese scholarship strongly critiques the conventional labor education model for frequently being marginalized or reduced to simple physical exertion (chore-based tasks), which disconnects it from students’ professional and cognitive development (Wu and Shen, 2022). Critiques suggest that without integrating intellectual challenges, labor education fails to prepare students for modern industrial demands. What remains unresolved in the literature is how these deeply pragmatic courses can be elevated beyond repetitive skill acquisition to foster higher-order cognitive abilities.

2.2 Creativity development in vocational education

Conversely, creativity education research has established that divergent thinking, risk-taking, and imagination are prerequisites for workforce adaptability (Runco and Jaeger, 2012). Yet, vocational education research indicates that institutions struggle to operationalize creativity within standard curricula, often treating it as an abstract concept rather than a teachable competency. Furthermore, the broader literature on creativity assessment cautions that creativity skills do not automatically transfer from abstract educational settings to the workplace; they must be embedded within authentic, domain-specific contexts to be effectively utilized by students (Baer, 2015). Previous attempts to integrate creativity into vocational curricula have shown that isolated, theoretical creativity courses often fail to produce transferable skills, whereas integrated, project-based interventions yield measurable cognitive gains (Scott et al., 2004). Recent empirical evidence has demonstrated the efficacy of case-based and team-based project learning in higher education. For example, Suartama et al. (2024) found that a context-aware ubiquitous learning model integrating case methods and team-based projects significantly enhanced knowledge construction and collaborative problem-solving among university students. This evidence informed the present study's pedagogical design by supporting the integration of project-based strategies as a vehicle for fostering creativity. However, the present study extends this prior work in two critical ways: (1) it explicitly maps collaborative and project-based activities to specific Williams creativity dimensions (e.g., fluency, flexibility, originality, elaboration), providing a theoretical mechanism for how these pedagogies stimulate creativity, and (2) it applies these pedagogies within the historically non-creative domain of labor education in Chinese vocational colleges, demonstrating that active learning strategies can be successfully deployed even in traditionally rote curricula. While isolated interventions exist (Khamying et al., 2022; Li et al., 2025), a structured synthesis of previous studies demonstrates a clear knowledge gap regarding the systematic integration of creativity into mandatory, non-creative core subjects like labor education.

2.3 Tyler's curriculum model

To bridge this gap, robust instructional design frameworks are required. Tyler’s (1949) curriculum model has contributed a highly effective, four-step systematic process for educational design: establishing clear objectives, selecting appropriate learning experiences, organizing those experiences effectively, and rigorously evaluating outcomes. While Tyler provides an indispensable structural skeleton for course development, the model is inherently content-neutral; on its own, it does not specify what cognitive behaviors should be targeted.

2.3.1 Comparative rationale for model selection

While alternative curriculum models exist—such as Taba's grassroots model (which emphasizes teacher-led development) and Walker’s deliberative approach (which emphasizes naturalistic decision-making)—Tyler's model was selected for three reasons: (1) its objective-driven structure aligns with China's centralized vocational education governance, facilitating standardization across multiple institutions; (2) its four-step sequence provides a clear, replicable template for curriculum developers, enabling large-scale implementation; and (3) its emphasis on clear objectives and systematic evaluation addresses the accountability requirements of China's Vocational Education Law. Similarly, alternative creativity frameworks—such as Torrance's divergent thinking tests and Guilford's Structure of Intellect model—were considered. However, Williams’ model was selected because: (1) it provides not only cognitive dimensions but also specific teacher behaviors to stimulate creativity, aligning with Tyler's focus on selecting learning experiences; (2) its nine dimensions offer more granular behavioral targets than Torrance's broader constructs; and (3) it includes measurable indicators that can be directly integrated into assessment frameworks.

2.4 Williams’ creativity model and measurement theory

To supply the necessary cognitive targets, this study draws on Williams’ (1979) creativity model, which specifies nine distinct dimensions of creativity (e.g., fluency, flexibility, originality, elaboration) and operationalizes how specific teacher behaviors stimulate corresponding student cognitive and affective behaviors. The measurement theory underlying this model relies on the Williams Creativity Assessment Scale. Originally designed to capture multidimensional creative traits, validating its use requires establishing its original factor structure, ensuring appropriate translation and adaptation for Chinese vocational populations, and confirming measurement invariance across experimental and control groups to guarantee that any observed differences are due to the intervention rather than instrument bias (Putnick and Bornstein, 2016).

2.4.1 Comparative rationale for Williams model

Alternative creativity frameworks such as Torrance's divergent thinking tests and Guilford's Structure of Intellect model were considered. However, Williams’ model was selected for three reasons: (1) it provides not only cognitive dimensions but also specific teacher behaviors to stimulate creativity, aligning with Tyler's focus on selecting learning experiences; (2) its nine dimensions offer more granular behavioral targets than Torrance's broader constructs; and (3) it includes measurable indicators that can be directly integrated into assessment frameworks.

2.5 Critiques and limitations of the foundational models

While foundational, both Tyler's and Williams’ models possess significant limitations that necessitate critical evaluation. Tyler's rationale has been heavily criticized for its behaviorist underpinnings, its assumption of a strictly linear curriculum development process, and its failure to account for the dynamic, unpredictable nature of teacher-student interactions (Kliebard, 1970). Conversely, Williams’ model, developed in 1979, suffers from a lack of recent empirical validation, particularly within modern vocational and non-Western educational contexts. Additionally, the Williams Creativity Assessment Scale may contain inherent cultural biases; for example, affective traits such as “risk-taking” or “tolerance of ambiguity” are often perceived and operationalized differently in collectivist Eastern cultures compared to individualist Western contexts (Rudowicz, 2003). Therefore, relying on either model in isolation is insufficient for modern Chinese vocational education.

2.6 Integrating Tyler and Williams: the theoretical mechanism

The theoretical contribution of this study lies in articulating the explicit mechanism of integrating these two frameworks. To date, no identified studies have attempted to integrate Tyler's and Williams’ models within vocational education, representing a significant gap in the instructional design literature. This theoretical integration is necessary, rather than merely sufficient, because it directly resolves the specific limitations of each independent model. What Tyler lacks—attention to specific, dynamic teacher-student cognitive interactions—Williams explicitly provides through targeted pedagogical behaviors. Conversely, what Williams lacks—a macro-level, systematic framework for institutional curriculum implementation—Tyler provides through its rigorous four-step sequence.

Integrating Tyler and Williams provides something unavailable from either model individually. Tyler establishes the rigorous sequence of curriculum development, ensuring the course is systematically implementable and measurable. Williams provides the specific pedagogical “flesh,” dictating the teacher behaviors and cognitive targets required at each of Tyler's stages. By embedding Williams’ creativity dimensions directly into Tyler's step-by-step organizational structure (see conceptual framework Figure 1 and revised Figure 2), learning experiences are no longer selected merely for task completion, but are explicitly engineered to trigger divergent thinking and tolerance for ambiguity. To operationalize this integration, Table 1 presents a five-column theoretical mapping connecting Tyler's four stages to Williams’ dimensions, translating them into specific pedagogical strategies, teacher behaviors, student activities, and evaluation methods. This explicitly demonstrates how abstract theoretical constructs translate into concrete classroom practices.

Table 1. Theoretical mapping of tyler-williams integration to pedagogical strategies and creativity outcomes.

2.6.1 Pedagogical strategies derived from the Tyler-Williams integration

To operationalize the theoretical mechanism, specific pedagogical strategies are derived directly from the integration of Tyler's curriculum sequence and Williams’ creative dimensions. Each strategy addresses a distinct cognitive or affective mechanism and is explicitly mapped to both frameworks with theoretical justification:

• Experiential Learning: Connected directly to Tyler's “selecting learning experiences” and Williams’ cognitive dimensions of fluency and flexibility. By engaging in direct, hands-on vocational tasks, students are stimulated to generate multiple ideas rapidly (fluency) and shift mindsets when encountering technical constraints (flexibility).
• Project-Based Learning: Derived from Tyler's “organizing experiences” and targeting Williams’ cognitive dimensions of originality and elaboration. Grounded in recent empirical evidence demonstrating how team-based projects and case methods facilitate collaborative problem-solving and knowledge construction in higher education (Suartama et al., 2024), this strategy provides a multi-phase active learning structure that enables students to synthesize novel technical solutions (originality) and iteratively refine complex engineering details (elaboration).
• Technology-Driven Activities: Mapped to Tyler's “selecting learning experiences” and explicitly targeting Williams’ affective dimension of imagination. Leveraging advanced digital simulations and CAD modeling allows students to visualize hypothetical technical scenarios and conceptualize solutions that transcend physical prototyping limitations.
• Open-Ended Challenges: Connected to Tyler's “selecting learning experiences” and designed to cultivate Williams’ affective dimensions of risk-taking and tolerance of ambiguity. Removing single-correct-answer constraints forces students to navigate uncertain design spaces and adopt non-conventional technical routes.
• Micro-Challenges: Mapped to Tyler's “establishing educational purposes” and “selecting learning experiences,” this strategy functions as a scaffolding mechanism during early curriculum stages to target Williams’ cognitive fluency. Short, rapid-fire ideation drills lower the cognitive barrier to entry, prompting students to produce high volumes of initial design concepts.
• Productive Failure: Aligned with Tyler's “organizing learning experiences” and “evaluating educational experiences,” this strategy directly reinforces Williams’ affective dimension of risk-taking. By framing technical missteps as essential analytical diagnostic data rather than performance deficits, this strategy encourages bold experimentation.
• Psychological Safety: Embedded within Tyler's “organizing learning experiences,” this strategy serves as the overarching climate condition required for Williams’ affective dimensions of risk-taking and curiosity. Establishing an unpunitive environment ensures students feel safe asking non-traditional questions and pursuing unorthodox technical inquiries.

2.7 Development of the research framework and hypotheses

To align the research design with this integrated framework, this study avoids conflating exploratory curriculum design with experimental hypothesis testing. Instead, Phases I and II utilize Research Questions (RQs) to guide the research and development (R&D) process, while Phase III utilizes inferential hypotheses to test the resulting intervention.

Following the curriculum development, the study tests the following hypotheses based on the established theoretical rationale. Previous quantitative reviews of creativity training confirm that carefully structured, domain-specific interventions consistently yield significant pre-to-post improvements in overall creative performance (Ismail and Mohamad, 2025; Scott et al., 2004), theoretically justifying the expectation of overarching success in H1. Furthermore, while the Williams model includes both cognitive and affective dimensions, short-term educational interventions typically exert a much stronger influence on cognitive skills (which can be explicitly taught and practiced) than on affective personality traits, which require long-term cultural shifts to significantly alter (Cropley, 2000). Consequently, the Phase III experimental evaluation deliberately targeted six primary dimensions (fluency, flexibility, originality, imagination, risk-taking, and elaboration) that demonstrate the greatest responsiveness to short-term pedagogical interventions. The remaining three dimensions (curiosity, complexity, and tolerance of ambiguity) are affective in nature and require longer-term cultural shifts to show significant change; thus, they were excluded from the primary experimental analysis to preserve statistical power. Because the integrated curriculum specifically engineers learning experiences around problem-solving mechanics, it theoretically isolates and targets these six dimensions, justifying the predictions in H2.

• H1: Participation in the newly developed creativity-oriented labor education course will produce significantly greater pre-to-post improvements in a composite overall creativity score—calculated as the aggregate sum of the targeted cognitive dimensions on the Williams scale—compared to the traditional labor education course.
• H2: Participation in the creativity-oriented course will yield statistically significant pre-to-post improvements specifically across the targeted cognitive dimensions (fluency, flexibility, originality, imagination, risk-taking, and elaboration) compared to the traditional course, as these specific metrics are expected to demonstrate the greatest responsiveness to the intervention.

3 Materials and methods

3.1 Research design

The study adopted a Research and Development (R&D) design with a mixed methods approach, integrating qualitative and quantitative data to enhance the depth of the reported findings (Timans et al., 2019; Battista and Torre, 2023). Unlike Design-Based Research (DBR), which focuses on continuous iterative cycles to refine educational theory, the R&D approach (Borg and Gall, 1989) was specifically selected because the primary objective was the systematic development, validation, and empirical testing of a tangible educational product—the new creativity-oriented labor curriculum. The rationale for adopting the mixed method was its capacity to assess both objective and subjective viewpoints through an exploratory sequential design; specifically, qualitative insights gathered in Phase I were directly utilized to inform the quantitative item generation and the specific pedagogical components of the Phase II curriculum.

To ensure rigorous developmental progression, evidence from each phase systematically informed the subsequent phase (see Figure 1 for the research-process flow and Table 2). Specifically, Phase I findings (competency specification) informed Phase II (curriculum design), which underwent expert validation and pilot revision. The finalized curriculum was then deployed in Phase III (intervention) and rigorously assessed in Phase IV (evaluation). The research was carried out at Guangdong Polytechnic Vocational College, selected as a representative case because such institutions in Guangdong Province drive educational reforms aimed at cultivating innovative and application-oriented talents.

Table 2. Summary of participants across the four research phases.

3.2 Phase I: contextual study

In the first phase, a contextual study was undertaken to establish perspectives on the current creative thinking course. The study population comprised instructors and students from Guangdong higher vocational colleges. Purposive sampling was used to select 12 instructors. The inclusion criteria for these experts required a minimum of five years of teaching experience, active involvement in institutional curriculum design, and specialized training in creativity pedagogy. Stratified random sampling, based on Yamane’s (1967) formula, was used to select 400 students to ensure representativeness across diverse backgrounds and majors. Stratification was implemented by dividing the population into distinct academic faculties (e.g., Engineering, Business, Arts and Design), with proportional random selection within each faculty to capture a truly diverse cross-section of the student body.

Quantitative data were collected through a structured questionnaire administered to the 400 students, eliciting their perceptions of the current labor education course and their creative thinking skills. The item development process involved an initial literature review, followed by an expert panel review to establish a Content Validity Index (CVI >.80). Additionally, cognitive interviews were conducted with five students during the pilot phase to verify item clarity and timing, going beyond simple alpha testing.

To ensure linguistic equivalence, the original English instrument was adapted and translated into Chinese using the rigorous back-translation procedure established by Brislin (1970). Specifically, the instrument was translated into Chinese by a bilingual educational researcher, and then independently back-translated into English by a second bilingual academic blind to the original text. Discrepancies between the original and back-translated versions were resolved through a consensus committee comprising both translators and a third subject-matter expert. The questionnaire employed a five-point Likert scale and was developed based on the reviewed literature and in collaboration with curriculum development experts. The full questionnaire and interview protocol are provided in the Supplementary Materials. Qualitative data were collected through semi-structured interviews with the 12 instructors, exploring their challenges and opportunities in fostering creativity within labor education. The demographic profile of the sample is presented in Table 3.

Table 3. Demographic characteristics of phase I survey participants (N = 400).

3.3 Phase II: course development

Based on the competency specifications identified from the findings from Phase I, a labor education course integrating creativity was developed guided by Tyler's curriculum model and Williams’ creativity model. Specifically, the instructional design operationalized the seven derived pedagogical strategies (Section 2.6.1)—Experiential Learning, Project-Based Learning, Technology-Driven Activities, Open-Ended Challenges, Micro-Challenges, Productive Failure, and Psychological Safety—by explicitly mapping them across Tyler's stages to target Williams’ specific cognitive (fluency, flexibility, originality, elaboration) and affective (curiosity, risk-taking, complexity, imagination) creativity dimensions. The newly developed creativity-oriented labor education course comprised 30 h of instruction over three months. The course structure included: (a) introductory modules on labor values and creativity concepts (4 h), (b) experiential learning activities involving open-ended labor challenges (8 h), (c) project-based learning modules requiring students to design and implement innovative solutions to community problems (10 h), (d) technology-driven activities incorporating digital tools for creative problem-solving (4 h), and (e) reflection and assessment sessions (4 h). Student activities included divergent thinking exercises, micro-challenges, collaborative project work, and peer feedback sessions. Assessment was based on originality, flexibility, fluency, and elaboration, using the Williams Creativity Assessment Scale.

Five curriculum experts with backgrounds in curriculum development, educational psychology, educational innovation, and labor education legislation evaluated the course materials to confirm their accuracy. Thereafter, a pilot study was undertaken involving 25 students over 30 h of diverse educational activities designed to evaluate these mapped strategies in practice. The newly developed curriculum was modified based on the expert feedback and the pilot study inputs to ensure its readiness for the Phase III intervention. To clearly distinguish the pedagogical intervention from conventional instruction, Table 4 provides a side-by-side comparison of the newly developed creativity-oriented course and the traditional labor education curriculum.

Table 4. Comparison of creativity-oriented and traditional labor education curricula.

3.4 Phase III: implementation

In the implementation phase, an experiment was conducted using a pre-test-post-test comparison group design to gather preliminary evidence regarding the intervention's effects. An a priori power analysis was conducted using G*Power 3.1 (Faul et al., 2009) to determine the required sample size. To detect a moderate-to-large interaction effect in a 2 × 2 Mixed ANOVA with an alpha level of.05% and 80% power, a minimum total sample size of 24 was required. Therefore, the recruited sample of 30 students was deemed statistically adequate.

From the initial pool of students, a convenience sample of 30 participants from the same academic major and institution were purposively recruited to control for institutional and disciplinary confounding variables. To minimize selection bias and establish internal validity within this convenience sample, these students were randomly assigned to either an experimental group (n = 15) or a comparison group (n = 15). Baseline equivalence was statistically confirmed prior to the intervention, revealing no significant differences in pre-test creativity scores between the two groups. To control for instructor effects, the same highly trained instructor taught both groups using standardized protocols. Furthermore, the classes were scheduled on different days in separate campus facilities to strictly prevent treatment contamination between the experimental and comparison groups (Field, 2018).

The experimental group received the newly developed labor education course designed based on the Williams Cube C.A.I. Model, integrating the dimensions of Williams’ creativity model together with broader creativity-related competencies drawn from the literature (problem-solving, intrinsic motivation, collaboration, and critical thinking). In contrast, the comparison group received the institution's standard labor education course, which consisted of 30 h of instruction over three months. This course included: (a) theoretical lectures on labor values and policy education (12 h), (b) routine physical labor tasks such as campus cleaning and basic maintenance activities (10 h), (c) basic service activities (4 h), and (d) assessment based on attendance and task completion checklists (4 h). The standard course did not include structured creativity components, open-ended challenges, project-based work, or psychological safety protocols. Assessment was solely based on attendance records and completion of assigned tasks, with no evaluation of originality, flexibility, or problem-solving.

While the full Williams framework encompasses nine dimensions, the Phase III experimental evaluation deliberately targeted six core cognitive dimensions (fluency, flexibility, originality, imagination, risk-taking, elaboration) that align with the intervention's primary pedagogical focus. The remaining three dimensions (curiosity, complexity, tolerance of ambiguity) are affective in nature and, consistent with Cropley (2000), require longer-term cultural shifts to demonstrate significant change. These dimensions were evaluated during Phase I and Phase IV surveys but excluded from the experimental analysis to preserve statistical power.

3.5 Phase IV: evaluation

In the evaluation phase, Williams’ Creativity Assessment Scale was administered to the 30 students (15 experimental, 15 comparison) who completed the newly developed course to measure their creative thinking skills. While all nine dimensions of the Williams scale were administered and measured, the primary Phase III experimental evaluation deliberately focused on a subset of six core dimensions (fluency, flexibility, originality, imagination, risk-taking, and elaboration) that were explicitly targeted by the new curriculum's pedagogical mechanics. To ensure transparency and verify discriminant validity, the remaining three non-targeted dimensions (curiosity, complexity, and tolerance of ambiguity) were included in a sensitivity analysis, which confirmed no significant changes, thereby isolating the specific effects of the intervention.

Classroom observations, student reflections, and structured feedback were also employed to evaluate the effectiveness of the course. A post-course creativity self-rating survey was administered. The previously proposed multiple regression analysis was removed from the analytical plan, as calculating a stable nine-predictor regression model is mathematically inappropriate and statistically underpowered with a small intervention sample (violating the events-per-variable ratio rule) (Tabachnick and Fidell, 2019). Furthermore, using the nine dimensions to predict an improvement composite derived from those same dimensions poses a risk of mathematical circularity.

3.6 Data analysis technique

The quantitative data were analyzed using two primary software packages. The Statistical Package for the Social Sciences (SPSS) was utilized for all descriptive statistics, Exploratory Factor Analysis (EFA), and primary inferential testing (t-tests and Repeated-Measures ANOVA). Additionally, IBM SPSS AMOS software was employed strictly for covariance-based measurement model validation, specifically to conduct the Confirmatory Factor Analysis (CFA).

Descriptive statistics (frequencies, percentages, means, and standard deviations) were used to summarize the demographic data and student opinions. Missing data were handled using listwise deletion, as the missingness rate was completely at random and below 5% (Tabachnick and Fidell, 2019).

Normality assumptions were evaluated differently depending on the sample size of the respective research phase. For the Phase I dataset (N = 400), the Kolmogorov–Smirnov test was utilized. For the Phase III and IV intervention datasets (N = 30), the Shapiro–Wilk test was applied, as it is more robust for small sample sizes. While the N = 30 data showed minor deviations from normality, a visual inspection of skewness and kurtosis confirmed all values fell within the acceptable ±2.0 range, justifying the retention of parametric testing (Hair et al., 2019).

To test the intervention's effectiveness without inflating the Type I error rate through multiple separate t-tests, a 2 × 2 Mixed Analysis of Variance (Repeated-Measures ANOVA) was conducted. Prior to analysis, ANOVA assumptions were rigorously tested: Levene's test confirmed the homogeneity of variances (p > .05), and Box's M test confirmed the equality of covariance matrices. Because the within-subjects factor (Time) had only two levels (pre-test and post-test), the assumption of sphericity was inherently met (Field, 2018). The model featured ‘Time’ as the within-subjects factor and ‘Group’ as the between-subjects factor. The primary metric of intervention efficacy was the Time × Group interaction effect. Furthermore, to control for Type I error inflation across the evaluation of multiple dimensions, a Bonferroni correction was applied to all post-hoc and pairwise comparisons. To ensure rigorous reporting, all inferential tests include exact p-values, 95% confidence intervals (CIs), and effect sizes (partial eta squared, ηp2).

The qualitative data from the semi-structured interviews were analyzed using thematic analysis, following the six-step procedure outlined by Braun and Clarke (2006). The interviews were transcribed, coded, and grouped into themes. A codebook was developed to ensure systematic and consistent coding.

3.7 Validity and reliability

To ensure internal and external validity, the questionnaires and interview protocols were based on the reviewed literature and developed in collaboration with curriculum development experts (Rusticus, 2014). The Williams Creativity Assessment Scale is a self-report measure of perceived creativity. To acknowledge this limitation, the study also incorporated classroom observations and instructor evaluations of project outputs as supplementary behavioral indicators; however, future research should employ objective, performance-based assessments to complement self-report measures. Different question formats, including closed and Likert-scale questions, were used to enhance validity. Reliability was ensured through clear language, pilot testing, and Cronbach's alpha analysis, with values above 0.7 considered acceptable (Bujang et al., 2018). Test-retest reliability was established on a pilot subsample of 30 students over a two-week interval, yielding a strong intra-class correlation coefficient (ICC) of.85. Criterion-related validity was also confirmed by correlating the creativity scale scores with students’ academic performance records.

Construct validity for the Phase I survey (N = 400) was examined using an independent split-sample design. The EFA was conducted on a randomly selected split-half subsample (n = 200), with the remaining half retained for cross-validation through CFA. Principal Axis Factoring (PAF) was utilized as the extraction method alongside a Promax oblique rotation, as the nine dimensions of creativity are theoretically correlated. The factor correlation matrix confirmed relationships ranging from.35 to.65, justifying the oblique rotation. The Kaiser-Meyer-Olkin (KMO) measure was 0.907, and Bartlett's test of sphericity was significant (p < 0.001). Factors were retained based on eigenvalues > 1.0, supported by visual inspection of the scree plot and Parallel Analysis. The item-to-respondent ratio of 10:1 comfortably satisfied standard factor analysis requirements (Hair et al., 2019). Items were subjected to strict deletion criteria: item loadings had to exceed.50, communalities had to be > .30, and cross-loadings >.30 were removed. The EFA successfully extracted a clear nine-factor structure mapping perfectly to Williams’ theoretical dimensions.

Following the EFA, the CFA was conducted on the holdout sample (n = 200) using covariance-based SEM in AMOS, as covariance-based approaches are strictly preferred over PLS-SEM for rigorous measurement model validation and theory testing (Kline, 2015). Model fit was evaluated using standard fit indices, demonstrating excellent fit: Comparative Fit Index (CFI) = .94, Tucker–Lewis Index (TLI) = .93, and Root Mean Square Error of Approximation (RMSEA) = .06.

3.8 Measurement invariance

Measurement invariance was tested across the experimental and comparison groups to ensure the instrument measured the constructs equivalently. Configural invariance was supported by acceptable model fit in both groups. Metric invariance was confirmed as the fit of the constrained model did not significantly worsen (ΔCFI<.01). Scalar invariance was also supported, ensuring that group comparisons are valid.

The internal consistency of the nine-dimension creativity scale was examined using Cronbach's alpha and composite reliability (CR). As shown in Table 5, all nine dimensions (41 total items) yielded Cronbach's alpha values ranging from 0.808 (Fluency) to 0.892 (Curiosity), every one exceeding the 0.70 acceptability threshold. Composite reliability ranged from 0.859 to 0.921, confirming strong internal consistency. Convergent validity was adequate, as the average variance extracted (AVE) for every dimension surpassed the 0.50 threshold (range 0.606–0.700; Table 6). Discriminant validity was supported by the Fornell-Larcker criterion: the square root of each dimension's AVE exceeded all of its correlations with the other eight dimensions. In summary, the scale demonstrated satisfactory reliability and both convergent and discriminant validity, providing a sound measurement basis for the dependent variable in this study.

Table 5. Internal consistency reliability of the nine creativity dimensions (N = 400).

Table 6. Convergent validity: composite reliability, average variance extracted (AVE), and square root of AVE for the nine creativity dimensions (N = 400).

3.9 Ethical considerations

Ethical approval was obtained from the university ethics committee prior to data collection. Permissions were sought from the college principals of the participating institutions. Informed consent was obtained from all participants, who were informed of their right to withdraw without any ramifications. Personally identifiable information was anonymized using pseudonyms (e.g., P1 for participant 1), and all data were stored on secure servers with restricted access.

4 Data analysis and results

4.1 Phase I: contextual study (quantitative)

A total of 400 students participated in the Phase I survey, of whom 52.0% were male and 48.0% were female. The participants were distributed across year levels: first year (30.0%), second year (28.0%), third year (25.0%), and final year (17.0%). The Kolmogorov–Smirnov test indicated that all variables deviated significantly from normality (p < .001); however, given the large sample size (N=400), the Central Limit Theorem applies, ensuring the sampling distribution of the mean is approximately normal. Therefore, parametric tests were deemed appropriate (Field, 2018; Mohamad et al., 2022).

Descriptive analysis of students’ opinions on the current labor education course revealed that the mean scores for all items were below the mid-point of 3 on a five-point scale, indicating that students perceived the current course as limited in equipping them with creative thinking skills. For instance, the mean score for the item assessing whether the course equips students with the ability to brainstorm ideas was 2.42 (SD = 0.809), and the mean for creating novel ideas was 2.31 (SD = 0.781). The importance of creativity aspects was rated by students, with problem solving (M = 3.13, SD = 0.958) and elaboration (M = 3.13, SD = 0.949) receiving the highest ratings, while hands-on experience received the lowest (M = 2.85, SD = 0.992).

A one-sample t-test was conducted to test the baseline perception of the curriculum. The results showed that all items were significantly below the test value of 3 (p < .001). While scoring below the scale's mathematical midpoint establishes a preliminary deficit, this finding is further contextualized by comparing it against normative baseline data from similar Chinese vocational cohorts (e.g., Zhang and Wang, 2021), where baseline creativity ratings typically average above 3.5. This significant discrepancy against both the mathematical midpoint and empirical benchmarks statistically confirms a severe, meaningful deficit in current creativity-oriented competencies, robustly justifying the need for the new curriculum intervention.

4.2 Phase I: contextual study (qualitative)

Thematic analysis of the semi-structured interviews with 12 instructors yielded six themes: (1) domain creativity signatures, describing how creativity in labor education was defined through disciplinary logic; (2) structural creativity enablers and constraints, encompassing factors that supported or hindered creativity; (3) experiential, project-based, and technology-driven delivery, describing strategies to foster fluency, flexibility, and elaboration; (4) pedagogies of productive failure and psychological safety, lowering the costs of failure to enable risk-taking and curiosity; (5) systemic and cultural barriers to creative labor education, including bureaucratic silos and the marginalization of labor education; and (6) domain-differentiated assessment, highlighting the need to move beyond hour-counting metrics.

Participants revealed that creativity was measured by functional innovation and the capacity of students to solve problems empathetically with limited resources. For instance, supporting the ‘domain creativity signatures’ theme, one instructor (P3, Engineering) noted: “Creative labor could be understood as functional innovation—it is not just about making something new, but making it work under strict resource constraints.” The interviews also showed that creativity indicators included social empathy when solving problems, ensuring more value was created than consumed. Structural constraints included a shortage of full-time teachers, underdeveloped faculty capacity in creativity pedagogy, and theoretical hours allocated to labor values and policy education with zero hours dedicated to quantitative methods for labor analysis. Illustrating these 'structural constraints,’ another instructor (P7, Business) emphasized: “We are strictly bound by theoretical hour allocations, leaving virtually zero hours dedicated to experimental or creative exploration.”

4.3 Phase II: pilot study

Following the development of the new labor education course, a pilot study was conducted with 25 students to operationalize and evaluate the integrated Tyler-Williams framework across the four curriculum stages (as mapped in Table 1 and Figure 2). A one-sample t-test was conducted to assess baseline evaluations before the pilot intervention. The results showed that all items were significantly below the test value of 3 (p < .001), indicating that before the course, students’ perceived creative capacities were limited.

After completing the pilot, the descriptive analysis showed improved mean scores across all creativity dimensions, with the highest improvement in generating multiple ideas (M = 2.08, SD = 0.759) and approaching problems from various perspectives (M = 2.08, SD = 0.759). To rigorously assess these changes, a paired-samples t-test was conducted on the pilot data, confirming statistically significant pre-to-post improvements in overall perceived creativity. Critically, qualitative observations during this pilot phase revealed that students initially struggled with open-ended ideation without structured scaffolding during Tyler's Stage 2 (Selection of Learning Experiences); consequently, the curriculum and teacher pedagogical behaviors (as reflected in the revised Figure 2) were specifically revised to include “warm-up micro-challenges” to better scaffold divergent thinking and systematically target Williams’ cognitive and affective dimensions prior to the full Phase III implementation.

4.4 Phase III: experiment

It is necessary to clarify the structural mapping of the analyzed dimensions for this phase. While Williams’ full theoretical framework encompasses nine dimensions, the Phase III experimental evaluation deliberately targeted a subset of six core dimensions to align with the intervention's primary pedagogical focus and to preserve statistical power for the small sample size (N = 30). These six dimensions are: fluency, flexibility, originality, imagination, risk-taking, and elaboration. These six core cognitive dimensions were prioritized because they represent the primary metrics most responsive to short-term pedagogical interventions. (Note: Earlier drafts erroneously labeled “elaboration” as “collaboration”; this reporting error has been corrected across all text and tables to ensure perfect alignment with the Williams framework). The remaining three dimensions (curiosity, complexity, and tolerance of ambiguity) are affective in nature and, consistent with Cropley (2000), require longer-term cultural shifts to demonstrate significant change; while they were evaluated during the broader Phase I and IV surveys, they were excluded from the Phase III experimental analysis to preserve statistical power.

The experiment involved 15 students in the experimental group and 15 in the comparison group. Paired-samples t-tests showed that the experimental group demonstrated significant improvements from pre-test to post-test across all six measured creativity dimensions: fluency [t = −9.539, p < .001, Mean Difference=1.72, 95% CI (1.33, 2.11), Cohen's d = 2.46], flexibility [t = −8.671, p < .001, Mean Difference = 1.74, 95% CI (1.31, 2.17), Cohen's d = 2.24], originality [t = −6.808, p < .001, Mean Difference = 1.93, 95% CI (1.32, 2.54), Cohen's d = 1.76], imagination [t = −9.000, p < .001, Mean Difference = 1.80, 95% CI (1.37, 2.23), Cohen's d = 2.32], risk-taking [t = −10.458, p < .001, Mean Difference = 1.67, 95% CI (1.33, 2.01), Cohen's d = 2.70], and elaboration [t = −7.174, p < .001, Mean Difference = 1.67, 95% CI (1.17, 2.17), Cohen's d = 1.85].

While these large effect sizes (Cohen's d > 0.80) demonstrate substantive practical improvement (Cohen, 1988; Lakens, 2013), they must be interpreted with appropriate academic caution. In studies with small sample sizes, effect sizes are frequently inflated and may partially reflect novelty (Hawthorne) effects, instructor enthusiasm, or regression to the mean, rather than pure, scalable intervention efficacy (Slavin and Smith, 2009).

In contrast, the complete pre/post results for the comparison group revealed no significant changes across any of the six dimensions, confirming the lack of spontaneous improvement without the intervention. Table 7 presents the pre- and post-assessment scores for the experimental group, while Table 8 presents the corresponding pre- and post-assessment scores for the comparison group. Table 9 details the complete paired t-test results for both groups.

Table 7. Pre- and post-assessment scores for the experimental group across the six targeted dimensions (n = 15).

Table 8. Pre- and post-assessment scores for the comparison group across the six targeted dimensions (n = 15).

Table 9. Paired-samples t-test results for the experimental and comparison groups.

To rigorously evaluate the overall effectiveness of the course, a Repeated-Measures Analysis of Variance (ANOVA)—replacing the previously underspecified Generalized Linear Model—was conducted. The model specified ‘Time’ (pre vs. post) as the within-subjects factor and ‘Group’ (experimental vs. comparison) as the between-subjects factor. The critical Time × Group interaction effect was statistically significant, F(1, 28) = 66.70, p < .001. This extremely large standardized effect size (partial η2 = .704) indicates that the interaction between time and group allocation accounts for 70.4% of the variance in creativity scores. Similar to the Cohen's d results, this exceptionally high variance explanation should be viewed critically as a potential artifact of the highly controlled, small-n environment. The analysis indicated that the experimental group had a higher estimated marginal mean of 22.2 [95% CI (20.5, 23.9)] compared to the comparison group's 15.8 [95% CI (14.1, 17.5)] (see Table 10).

Table 10. Estimated marginal means from the repeated-measures ANOVA.

Independent-samples t-tests on gain scores confirmed that the experimental group's gains were significantly higher than those of the comparison group across all dimensions: total gain [t = 8.167, p < .001, Mean Difference = 6.40, 95% CI (4.80, 8.00)], fluency (t = 4.811, p < .001), flexibility (t = 3.099, p = .004), originality (t = 3.336, p = .002), imagination (t = 3.286, p = .003), risk-taking (t = 2.655, p = .013), and elaboration (t = 3.420, p = .002). These results confirm that experiential, project-based, and technology-driven teaching strategies are effective in implementing the developed labor education course (see Table 11).

Table 11. Independent-samples t-test results comparing gain scores between groups.

4.5 Phase IV: evaluation

In the evaluation phase, 25 students who completed the new course completed a post-course creativity self-rating survey based on Williams’ Creativity Scale. To clarify the sample composition, this N=25 group comprised the 15 students from the Phase III experimental group, augmented by 10 students randomly selected from the successful Phase II pilot cohort who also completed the fully finalized curriculum.

The results showed high mean scores across all dimensions, with expanding ideas receiving the highest rating (M = 4.12, SD = 0.526), followed by experimenting with new ideas (M = 4.12, SD = 0.726) and enhancing passion and drive (M = 4.04, SD = 0.676). Table 12 presents the post-course creativity self-rating results.

Table 12. Post-course creativity self-rating based on Williams’ creativity scale (n = 25).

Crucially, it must be acknowledged that this post-course self-rating alone does not demonstrate objective pre-to-post cognitive improvement. Rather, these Phase IV results demonstrate the students’ perceived post-course creativity and self-efficacy regarding their innovative capabilities. Furthermore, self-report measures of creativity are notoriously susceptible to social desirability bias and may not strictly correlate with objective behavioral outcomes (Podsakoff et al., 2003; Raza et al., 2020).

5 Discussion

5.1 Key findings

This study investigated the extent to which a newly developed labor education course could enhance the creative thinking skills of students in Guangdong higher vocational colleges. The findings suggest that creativity-oriented competency indicators included technical and problem-solving skills, work ethics, social responsibility, and vocational skills, systematically integrating disciplinary knowledge and fostering higher-order creative thinking. These indicators provided a foundation for course development, consistent with Tyler's (1949) curriculum model, which emphasizes the needs of society, learners, and subject matter specialists in defining competencies and learning outcomes.

A labor education course integrating creativity into course objectives, content, teaching and learning behaviors was developed, guided by Tyler's model and Williams’ creativity model. The creativity components integrated the nine dimensions of Williams’ creativity model: fluency, flexibility, originality, elaboration, curiosity, imagination, complexity, risk-taking, and tolerance of ambiguity, with particular emphasis on the four cognitive dimensions. This integration resonates with Williams’ (1979) creativity model, which outlines three dimensions of creativity: curriculum (subject matter content), teacher behaviors (teaching strategies), and student behaviors (cognitive intellect and affective learning). The newly developed creativity-oriented labor education framework is illustrated in Figure 2.

The experimental results provide preliminary evidence that students who completed the new course demonstrated higher post-intervention scores compared to peers taught with the original labor course across the six targeted cognitive dimensions of the Williams’ Creativity Assessment Scale (fluency, flexibility, originality, imagination, risk-taking, and elaboration). (Note: As established in the methodology, the construct of ‘collaboration’ previously mentioned in earlier exploratory phases was theoretically realigned to ‘elaboration’ during the experimental evaluation to strictly adhere to the validated Williams framework). While the full Williams framework encompasses nine dimensions, the Phase III experimental evaluation deliberately targeted these six core cognitive dimensions because they represent the primary metrics most responsive to short-term pedagogical interventions. The remaining three dimensions (curiosity, complexity, and tolerance of ambiguity) are affective in nature and, consistent with Cropley (2000), require longer-term cultural shifts to demonstrate significant change; thus, they were excluded from the primary experimental analysis to preserve statistical power. While these findings provide preliminary evidence supporting the curriculum's utility, definitive causal claims cannot be made due to the study's scope. The statistically significant improvements in this preliminary sample among the experimental group may be attributed to the structured alignment of specific teacher behaviors with targeted cognitive outcomes. Rather than treating creativity as an abstract byproduct, experiential teaching strategies—including open-ended labor challenges, warm-up micro-challenges, and authentic community work—alongside project-based strategies and technology-driven strategies, explicitly forced cognitive friction, prompting students to bypass habitual, rote responses in favor of divergent thinking. These findings align with Wu et al., who showed that vocational students require creativity skills to cultivate innovation and adapt to changing market demands.

5.2 Comparison with previous research

The findings are consistent with prior literature demonstrating that creativity in labor education courses ensures the interaction of the social and perceived environment with individual knowledge and experience (Walia, 2019) and that creativity relies on the capacity for adaptable thought and the spirit of tenacity (Glaveanu et al., 2020). The statistically significant improvements in this preliminary sample in creative thinking skills among the experimental group align with Henriksen et al. (2021), who reported that creative risk-taking focuses on learning from mistakes and challenges.

However, the findings contradict Brauer et al., who revealed that risk-taking is often viewed negatively within learning environments where policies focus on standardization. This divergence likely occurred because the traditional environments studied by Brauer et al. implicitly penalize errors, whereas the experimental curriculum in this study intentionally embedded pedagogies of productive failure and psychological safety. By actively lowering the academic risks associated with making mistakes, the new curriculum enabled students to experiment freely. This discrepancy highlights the need for policy reforms in vocational colleges to encourage risk-taking behavior and imagination, suggesting that highly rigid, standardized curricula might inadvertently suppress the divergent thinking required for modern vocational problem-solving.

The findings also extend previous research by providing preliminary evidence that the integration of Tyler's curriculum model and Williams’ creativity model provides an actionable framework for developing labor education courses. While Hui et al. showed that the integration of active learning strategies in vocational education was challenged by limited time for creative exploration, the present study addressed this constraint through modular flexibility within the practice component and the availability of competitions that promoted creative thinking. The present results also resonate with Huang et al. (2025), who reported significant creativity gains from a structured Problem-Ideation-Prototype-Testing teaching model implemented with college students in Guangzhou, China, and with Di et al. (2025), whose theory-based instructional model in design education yielded significant pre-test-to-post-test gains in fluency, flexibility, and originality through a research-and-development and pre-post-test design, reinforcing that systematically designed, creativity-oriented interventions can yield measurable improvements across both Chinese and Thai vocational and higher-education settings.

5.3 Implications

Theoretically, the findings provide preliminary evidence that a curriculum developed through an integrated Tyler–Williams framework can be implemented in a higher vocational labor education context. Because the study tested the combined framework as a single intervention, it does not isolate the independent effect of Tyler's model from Williams’ model. Therefore, rather than claiming either model is independently effective, this research illustrates their synergistic association with improvement: Tyler provides the vital structural sequence for course development, while Williams supplies the specific cognitive and pedagogical mechanisms that make that structure conducive to creativity.

The original contribution of this research is the theoretical framework itself. Individually, Tyler's model suffers from content-neutrality, offering no guidance on what specific cognitive behaviors to target, while Williams’ model lacks a macro-level systematic structure for institutional curriculum design. Their integration uniquely resolves these deficits by embedding precise cognitive and pedagogical mechanisms (Williams) within a rigorously sequenced developmental blueprint (Tyler).

Practically, rather than recommending immediate widespread adoption, the findings strongly support the need for further piloting and evaluation of the developed labor education course across diverse educational settings. If successfully scaled, such curricula may promote students’ creative capacity, ensuring they approach problems from various perspectives, take risks, and evaluate ideas logically. The improvement in creative capacity is influenced by multiple factors, including the labor course completed, instructor behaviors, and structural enablers within the institutions.

A further methodological contribution lies in the validation of the adapted nine-dimension Williams’ Creativity Assessment Scale. The instrument provided preliminary evidence of strong internal consistency (alpha = 0.808–0.892) and convergent and discriminant validity, providing a replicable measurement tool for subsequent labor-education creativity research. This is particularly crucial given the historical scarcity of culturally and contextually adapted instruments for measuring multidimensional creativity within Chinese vocational populations, offering future researchers a validated baseline for comparative studies. From a policy perspective, the findings indicate that layered pipelines within colleges, such as daily productive labor service, could ensure students remain creative beyond the classroom. Modular flexibility within the practice component and the availability of competitions also promote creative thinking. Policies should emphasize the value of interdisciplinary collaboration, aligning with China's National Vocational Education Reform, which aims to transform vocational education into a modern economic system.

5.4 Limitations and future research

Several limitations should be acknowledged, particularly concerning alternative explanations for the observed experimental outcomes that limit definitive causal claims. The most critical limitations threatening the generalizability and causal interpretation of these findings are the extremely small sample size (n = 30) and the single-site design. These constraints severely limit statistical power, increase the margin of error, and mean the data reflect only one specific institutional context in Guangdong. Second, the short intervention duration of three months makes it difficult to determine whether the observed changes in creativity are permanent or temporary.

Third, while participants were randomly assigned to experimental and comparison conditions, the initial pool of 30 students was purposively selected, introducing potential selection effects that restrict how these results might apply to the broader student population. Fourth, the extremely large effect sizes observed in the experimental group must be interpreted with extreme caution, as they likely reflect novelty or Hawthorne effects. Students in the experimental group may have demonstrated heightened engagement and performance simply because they were aware of receiving a specialized, highly interactive curriculum rather than a routine class. Fifth, the potential for regression to the mean cannot be ruled out. If the experimental students’ pre-test scores were artificially low due to unfamiliarity with creativity assessments, their post-test improvements would naturally appear exaggerated as scores regressed toward their true mean (Shadish et al., 2002).

Sixth, although instructor effects were methodologically controlled by utilizing the same teacher for both groups, the instructor's implicit enthusiasm for the newly developed curriculum might have inadvertently influenced the experimental group's outcomes. Seventh, evaluating the same cohort with the identical instrument pre- and post-intervention introduces potential measurement effects, as the pre-test itself may have sensitized students to the specific creativity constructs being evaluated. Eighth, despite utilizing an omnibus ANOVA to evaluate overall effects, the analysis of multiple distinct cognitive dimensions inherently carries some statistical risk associated with multiple testing. Ninth, while the broader study evaluated all nine dimensions during the survey phases, the experimental intervention intentionally excluded the three affective dimensions (curiosity, complexity, and tolerance of ambiguity) because short-term interventions are insufficient to yield meaningful affective shifts. Finally, the reliance on self-report measures introduces common-method bias and social desirability effects, as perceived creativity does not always correlate perfectly with objective, observable behavioral measures (Podsakoff et al., 2003).

Consequently, future research must prioritize specific questions designed to address these limitations: (1) To what extent do the observed creativity gains hold in a large-scale, multi-site randomized controlled trial? (2) How much of the observed variance is attributable to instructor enthusiasm versus the curriculum itself? and (3) Do the cognitive gains persist longitudinally when assessed via objective, performance-based behavioral metrics rather than self-reports? Future research should collect data from multiple higher vocational colleges across China to enable cross-provincial comparisons. Longitudinal studies that track students over several years would provide insights into the long-term effects of creativity-enhanced labor education. Additionally, future work should explore alternative and highly objective assessment methods and examine the adaptation of the developed course in different educational contexts, including the integration of AI-assisted labor education within the curriculum.

6 Conclusion and future directions

6.1 Conclusion

This study developed and evaluated a creativity-oriented labor education curriculum for students in Guangdong higher vocational colleges, guided by Tyler's curriculum model and Williams’ creativity model. The findings suggest that creativity-oriented competency indicators, including technical and problem-solving skills, work ethics, social responsibility, and vocational skills, offer preliminary evidence for a viable foundation for course development. Rather than claiming definitive overarching success, the experimental phase of this study suggested that the integrated curriculum-development framework, incorporating experiential, project-based, and technology-driven teaching strategies, was associated with statistically significant improvements in this preliminary sample across six targeted cognitive dimensions (fluency, flexibility, originality, imagination, risk-taking, and elaboration). These six core cognitive dimensions were prioritized for experimental analysis because they represent the metrics most susceptible to short-term instructional interventions, whereas the remaining three affective dimensions (curiosity, complexity, and tolerance of ambiguity) require longer-term cultural shifts to demonstrate significant change. Furthermore, the adapted nine-dimension scale was verified as a feasible and valid instrument for evaluating perceived creativity in vocational education during the broader contextual and evaluation phases.

6.2 Implications

The theoretical and practical implications of this study must be contextualized within its exploratory scope. Theoretically, the study provides preliminary evidence regarding the feasibility of integrating Tyler's curriculum-development logic with Williams’ creativity framework. Because the curriculum was evaluated holistically, this research does not experimentally test Tyler's and Williams’ theories independently, nor does it establish their independent association with improvement. Instead, it provides a functional blueprint for synthesizing structural curriculum design with specific cognitive-behavioral targets.

Practically, the findings provide preliminary evidence supporting further testing of the curriculum in larger and more diverse higher vocational settings. Given the exploratory nature of this intervention, this research offers specific, actionable guidance for key educational stakeholders to pilot. For teachers, the findings emphasize the value of embedding open-ended micro-challenges and psychological safety protocols into daily labor activities to actively stimulate divergent thinking. For curriculum developers, the study provides a practical template for explicitly mapping Williams’ cognitive dimensions onto Tyler's rigid instructional sequence, ensuring creativity is treated as an assessable competency rather than an abstract byproduct. For policymakers and university administrators, the results highlight the urgent need to establish evaluation metrics that move beyond rote attendance, rewarding holistic, interdisciplinary assessments that validate innovative application in vocational training.

6.3 Future research directions

As detailed in the discussion section, the current study's implications are inherently constrained by its small sample size, single-site deployment, and reliance on self-reported measures. Therefore, future research must address these methodological constraints to build a more robust evidence base. Specifically, researchers should prioritize conducting a large-scale, multi-site randomized controlled trial involving at least 500 participants across three distinct provinces. To evaluate the long-term retention and true behavioral change associated with this curriculum, these trials should utilize a longitudinal design spanning a full academic year to evaluate both the core cognitive dimensions and the longer-term affective dimensions (curiosity, complexity, tolerance of ambiguity). To counter self-report and common-method biases, future investigations should incorporate objective, performance-based assessments of creativity—such as standardized divergent thinking tasks or blinded expert evaluations of student projects. Finally, further psychometric refinement of the Williams scale within the Chinese vocational context is recommended to resolve potential construct overlaps and establish strict measurement invariance.

Generative AI statement

The author(s) declared that generative AI was not used in the creation of this manuscript.

Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.

Publisher's note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

Footnotes

Edited by: Daniel Osorio-Barreto, National University of Colombia, Colombia

Reviewed by: Inés Alvarez-Icaza Longoria, Monterrey Institute of Technology and Higher Education (ITESM), Mexico

I. Kadek Suartama, Ganesha University of Education, Indonesia