📄 other

Resource Adequacy and Uneven Equity Gradients in Creative Thinking: A Multilevel Analysis of PISA 2022 Indonesia

This multilevel analysis of PISA 2022 Indonesian data reveals that while educational resource shortages significantly hinder student creative thinking and equity gradients vary across schools, the mediating role of creative self-efficacy observed in high-performing systems is not yet active in this context.

Original authors: Miftah Aria Kusuma, Hari Purnomo, Winda Nur Cahyo, Widodo Brontowiyono

Published 2026-07-23

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Miftah Aria Kusuma, Hari Purnomo, Winda Nur Cahyo, Widodo Brontowiyono

Original paper licensed under CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

Imagine the human mind as a garden. For a long time, scientists have been fascinated by how to make this garden bloom with "creative thinking"—the ability to come up with ideas that are both new and useful. Think of it like solving a puzzle in a way no one else has thought of before, or inventing a game with rules you just made up. For years, researchers believed that if you could just convince a student, "You are a creative genius!" (a concept called creative self-efficacy), their garden would suddenly burst into flowers. It was a popular theory: boost the student's confidence, and the creativity follows.

But there's a catch. A garden needs more than just a confident gardener; it needs soil, water, and sunlight. If the soil is rocky and the water is cut off, no amount of believing you can grow a rose will make one appear. This is the corner of science this paper explores: the messy, real-world conditions of schools, especially in places where resources are tight. The big question isn't just "How do we make kids feel creative?" but "What does a school actually need to have before a kid can even start being creative?" It's a question that matters because if we get the conditions wrong, we might be wasting time telling kids to "believe in themselves" when the real problem is that the classroom doesn't have enough books, tools, or teachers to let them try.

The Great School Garden Experiment

This study is like a massive detective story set in Indonesia, looking at the gardens of 404 different schools and the minds of over 12,000 students. The researchers used a super-powerful magnifying glass called "hierarchical linear modeling" (a fancy way of saying they looked at how students are nested inside schools, rather than just looking at students as isolated islands). They wanted to see what actually makes a student's creative thinking score go up or down.

First, they checked the soil. They found something surprising: 24.2% of the difference in creativity between students wasn't about who the student was, but about which school they attended. It's as if one student walked into a garden with rich, dark soil and another walked into a patch of concrete, and the difference in their harvest was huge. This proved that the school itself is a massive player in the game.

The Missing Water and the Broken Hose

The researchers tested three main theories about what helps a school garden grow:

Creative Programs: Does having a drama club, a music class, or a creative writing circle help?
Creative Environment: Does the school principal say, "We love weird ideas here!"?
Resource Shortage: Does the school lack basic stuff like books, computers, or qualified teachers?

Here is where the story gets interesting. The "Creative Programs" and "Creative Environment" were like promising seeds. The data showed they might help (the numbers were positive), but the researchers couldn't say for sure because there weren't enough schools in the sample to be 100% certain. It's like seeing a sprout but not having enough rain to know if it will become a tree.

However, one factor was crystal clear: Resource Shortage. This was the only thing that reached a level of certainty the researchers could trust. When schools reported a shortage of educational materials, facilities, or staff, the students' creative thinking scores dropped significantly. The study found that for every step up in resource shortage, the creative thinking score went down by nearly 1 point. It's a hard fact: if you don't have the basic tools (the water and the soil), the garden simply won't grow, no matter how many "creative clubs" you try to start.

The Confidence Myth

Now, let's talk about the "confidence" theory. The researchers asked: "Does a student's belief in their own creativity (Creative Self-Efficacy) act as the bridge that turns a good school environment into a creative student?" In high-performing countries, this bridge usually works. But in this study, the bridge was missing.

The data showed that creative self-efficacy did not connect the school environment to the student's creativity. The researchers suggest this isn't because the theory is wrong, but because the "bridge" needs a solid foundation to stand on. In schools where students haven't had enough chances to succeed creatively (because of the resource shortages mentioned earlier), their confidence might not be strong enough to actually change their performance yet. It's like trying to build a house on a swamp; the blueprint (the theory) is fine, but the ground isn't ready to hold it.

The Uneven Playing Field

Finally, the study discovered something unexpected about fairness. Usually, we think that a student's family background (how much money their parents have) affects their creativity in the same way everywhere. But this study found that the "gap" between rich and poor students changes depending on which school they go to. In some schools, the gap is wide; in others, it's smaller. This means that the "creativity gap" isn't fixed; it depends on the specific school's conditions.

The Bottom Line

So, what's the takeaway for our garden? The study suggests that before we try to teach kids to "think outside the box" or boost their confidence, we need to make sure the box itself isn't broken. In Indonesia, the biggest lever for change isn't a new program or a pep talk; it's fixing the resource shortage. Until schools have the basic materials and staff they need, the more advanced psychological tricks might not work. The garden needs water before it can bloom.

Technical Summary: Resource Adequacy and Uneven Equity Gradients in Creative Thinking

Problem Statement
While creative thinking has been established as a priority outcome in international assessments (PISA 2022), the school-level conditions predicting performance in low-performing systems remain poorly understood. Indonesia's average score (19.5) falls significantly below the OECD mean (~33), masking substantial variation between schools. Preliminary analysis indicates that 24.2% of the variance in creative thinking lies between schools rather than between students within the same school. Existing literature presents conflicting findings: some studies suggest a negative correlation between extracurricular participation and creative scores, while others in high-performing East Asian systems identify creative self-efficacy (CSE) as a key mediator between school environment and performance. However, it is unclear if these mechanisms hold in low-performing contexts where resource constraints and historical marginalization of creative instruction may alter the developmental trajectory of efficacy beliefs.

Methodology
This study employs a two-level Hierarchical Linear Modeling (HLM) approach on the PISA 2022 Indonesian dataset, comprising 12,340 students nested within 404 schools.

Data Handling: To address the complex survey design, the analysis pools all ten plausible values for creative thinking using Rubin's rules. Standard errors and degrees of freedom are adjusted following Barnard (1999) and weighted using official PISA sampling weights (WeMix).
Variables:
- Outcome: Creative thinking scores (PV1–PV10).
- Level-1 Predictors: Creative self-efficacy (CREATEFF), Socioeconomic status (ESCS), and Gender.
- Level-2 Predictors: School creative program availability (CREACTIV), School creative environment (aggregated student perceptions, CREENVSC), School openness to creativity culture (OPENCUL), and Educational resource shortage (EDUSHORT).
Analytic Strategy: Four sequential models were fitted: a null model (M0), a model with Level-1 covariates (M1), a model adding school creative predictors (M2), and a full model (M3).
Mediation Testing: A 2-1-1 multilevel mediation framework (Preacher et al., 2010) was used to test if school-level predictors influence creative thinking indirectly through student-level CSE. Indirect effects were evaluated via Monte Carlo simulation (10,000 replications) to account for non-normal distributions of product coefficients.
Additional Analyses: Random slope extensions were tested via likelihood ratio tests, and cross-level interactions (ESCS × School Predictors) were examined to detect Matthew effects. A random forest analysis served as a descriptive, model-free check on variable importance.

Key Results

Variance Decomposition: The null model yielded an Intraclass Correlation Coefficient (ICC) of .242, confirming that school membership accounts for nearly a quarter of the variance in creative thinking, justifying multilevel analysis.
Direct Effects:
- Educational Resource Shortage (EDUSHORT): This was the only school-level predictor to reach conventional statistical significance ( $\beta = -0.991, p = .043$ ). A one-standard-deviation increase in resource shortage predicts a nearly one-point drop in creative thinking scores, independent of student SES.
- Creative Programs and Environment: School creative program availability (CREACTIV) and school creative environment (CREENVSC) showed positive point estimates ( $\beta \approx 0.85$ ) comparable to the magnitude of the resource shortage effect. However, their confidence intervals were wide (including zero), rendering them statistically inconclusive due to the limited number of Level-2 units (404 schools).
- Student-Level Predictors: SES was the strongest predictor ( $\beta = 0.955, p < .001$ ). Female gender was associated with slightly lower scores ( $\beta = -0.917$ ). Notably, creative self-efficacy (CREATEFF) showed no significant association with creative thinking scores ( $\beta = -0.138, p = .576$ ).
Mediation: The hypothesized mediation pathway (School Environment $\to$ CSE $\to$ Creative Thinking) was statistically null. Monte Carlo confidence intervals for indirect effects fully included zero, indicating CSE does not mediate school-level effects in this sample.
Equity Gradients: A likelihood ratio test revealed that the relationship between SES and creative thinking varies significantly across schools ( $p < .001$ ). The socioeconomic gradient is not uniform; some schools appear to mitigate the SES gap more effectively than others.
Cross-Level Interactions: Tests for Matthew effects (whether creative programs disproportionately benefit higher SES students) were non-significant, though the study notes limited statistical power to detect small interaction effects with the current sample size.

Significance and Claims
The paper posits that resource adequacy is a precondition for creative thinking development in low-performing systems. The findings suggest that in contexts where basic educational resources are scarce, sophisticated psychological interventions (such as boosting self-efficacy) or specific creative programs may not yield measurable returns. The study challenges the generalizability of the "self-efficacy pathway" documented in high-performing East Asian systems, proposing a boundary condition: efficacy beliefs may only become a functional driver of performance once a system crosses a minimum threshold of resource adequacy and accumulated creative experience.

Furthermore, the study highlights that equity in creative thinking is not a fixed system-wide parameter but is school-dependent. The variation in the SES gradient across schools suggests that identifying the specific conditions in schools that successfully narrow this gap is a critical policy direction. The authors conclude that for systems like Indonesia, addressing structural resource shortages is likely a more immediate and effective lever for improvement than focusing exclusively on programmatic or psychological interventions, which may be ineffective until basic adequacy is established.

Limitations Acknowledged
The authors explicitly note that the study is cross-sectional, precluding causal inference. The inconclusive results for creative programs and environment are attributed to statistical power limitations (404 schools) rather than a lack of effect. Additionally, the null finding regarding creative self-efficacy may be influenced by measurement invariance issues across cultural contexts, a factor requiring future investigation. Finally, the study uses cluster-based standard errors rather than the full Balanced Repeated Replication (BRR) method, which could slightly affect p-values near the significance threshold.

Technical Summary: Resource Adequacy and Uneven Equity Gradients in Creative Thinking

More like this