The research team, led by Tara Soleimani and colleagues at Kashan University of Medical Sciences and partner institutions across five Iranian universities of medical sciences, applied a family of techniques known collectively as multi-criteria decision analysis, or MCDA. Rather than asking faculty which teaching method they simply prefer, the approach forces a structured comparison of methods against explicit evaluation criteria, each assigned a numerical weight. The result is not an opinion poll but a transparent, reproducible scoring system in which every judgment is documented, every weight is justified, and the final ranking can be stress-tested to see whether it survives changes in assumptions. It is the same analytical machinery used in engineering, healthcare resource allocation, and environmental policy, here turned on the classroom.

The study’s first task was to decide what actually matters when judging a teaching method. The researchers conducted a scoping review of studies published between 2011 and 2025, following the PRISMA-ScR reporting guidelines to systematically identify evaluation criteria that previous educational research had deemed important. They then convened an expert panel of medical education specialists to refine this long list into a workable framework: seven domains containing 25 items in total. The resulting researcher-made instrument underwent formal face and content validation, with the team reporting content validity ratios and item-level content validity indices to demonstrate that experts agreed the items were relevant, clear, and essential. Only criteria that survived this statistical scrutiny were carried forward into the weighting phase.

With the criteria framework established, the study moved to its analytical core. Five experts completed pairwise comparisons, judging the relative importance of the criteria two at a time, and these judgments were analyzed using the fuzzy analytic hierarchy process, or FAHP. The fuzzy variant is a meaningful refinement of the classical method developed by Thomas Saaty in the 1970s. Instead of forcing experts to commit to a single crisp number when comparing two criteria, FAHP allows responses to be expressed as triangular fuzzy numbers, which capture a range of uncertainty around each judgment. Human evaluators are notoriously imprecise when asked whether one criterion is, say, exactly three times more important than another; fuzzy mathematics absorbs that hesitation rather than discarding it. The team also computed consistency ratios for both the middle values and the geometric mean bounds of the fuzzy judgments, ensuring that the pairwise comparisons were internally coherent rather than contradictory.

The weighting results are revealing in themselves. Quality of the educational experience received the highest weight at 0.2987, followed by perceived effectiveness at 0.2292 and interaction at 0.1795. Together, these three domains account for roughly seven-tenths of the total decision weight, a clear signal that the faculty experts prioritized the depth and impact of learning over logistical convenience. Two criteria were explicitly treated as cost factors, where lower values are better: time limitation and the need for technology. This is a crucial design choice, because a teaching method that scores brilliantly on engagement but consumes scarce faculty hours or demands infrastructure a department lacks must be penalized accordingly. The framework does not pretend that resources are unlimited; it builds scarcity directly into the arithmetic.

The second analytical stage, simple additive weighting or SAW, translated these weights into a final ranking. Fifteen faculty members who had at least three years of teaching experience and hands-on familiarity with all three methods rated how well each method performed on each criterion. The study deliberately required prior use of lecture, flipped classroom, and problem-based learning, so the ratings reflect lived experience rather than secondhand impressions. Each method’s scores were multiplied by the domain weights and summed, with the cost criteria subtracted in the appropriate direction. Problem-based learning emerged decisively in first place with a weighted score of 0.9701, the flipped classroom followed closely at 0.9148, and the traditional lecture trailed at 0.6068. The gap between the two active-learning formats was narrow; the gap between active learning and lecturing was not.

Dissecting the contributions behind the scores shows why the ranking came out as it did. Problem-based learning produced the largest weighted contributions in four of the seven domains: quality, interaction, perceived effectiveness, and performance. This aligns with what educational theory would predict, since PBL is explicitly engineered to place learners in the driver’s seat, forcing them to reason through clinical uncertainty, articulate their reasoning to peers, and defend their decisions. The flipped classroom claimed the largest contribution in flexibility, a reflection of its hybrid design that decouples content delivery from class time and lets students control the pace of their initial learning. The lecture, meanwhile, dominated in only one domain: availability. It remains the cheapest and most logistically forgiving method ever devised, capable of reaching unlimited numbers of students with minimal technology and preparation. The lecture’s problem is not that it does nothing well, but that what it does well matters least to the experts who weighted the criteria.

A ranking is only as credible as its robustness, and the researchers subjected theirs to an unusually thorough battery of sensitivity analyses. They varied the weights of all seven domains and found that the ordering of the three methods remained unchanged in every scenario examined. They examined rater-specific rankings individually and found high agreement among the fifteen faculty evaluators, quantified by Kendall’s coefficient of concordance of 0.804 with a p-value below 0.001, indicating that the consensus was statistically strong rather than an artifact of averaging. Finally, they ran 10,000 bootstrap resamples of the data, and problem-based learning retained its first-place position in 99.99 percent of them. For a field in which teaching recommendations often rest on single-institution surveys or narrative opinion, this level of analytical resilience is notable.

The authors are careful about what their findings do and do not establish. This is a faculty-perspective study, a cross-sectional snapshot of expert judgment rather than a measurement of what students actually learn. The conclusions explicitly call for prospective studies to evaluate learner outcomes and clinical competence, acknowledging that the ultimate test of any teaching method is whether it produces safer, more capable anesthesiologists. The framework is also context-specific, built from criteria identified in a scoping review and weighted by experts within a particular set of institutions, so departments elsewhere would need to consider their own educational priorities, available resources, and implementation conditions before importing the ranking wholesale. What the study offers is not a mandate but a decision-support tool: a validated, transparent method that any department could replicate with its own faculty and its own constraints.

Even with those caveats, the implications are hard to ignore. Anesthesiology education is a domain where the stakes of poor preparation are measured in operating room outcomes, and where simulation, case discussion, and team training have long been recognized as essential. A formal analysis showing that experienced faculty, when forced to weigh quality, effectiveness, and interaction against time and technology costs, place active learning far above the lecture adds quantitative weight to the ongoing shift away from passive instruction. The near-tie between problem-based learning and the flipped classroom is perhaps the most practically useful finding, because it suggests departments lacking the resources for full PBL implementation can capture much of the benefit through the more flexible flipped format. As medical schools worldwide confront crowded curricula, constrained budgets, and rising expectations for competency-based training, this study demonstrates that the choice of teaching method need not be a matter of tradition or taste. It can be measured, weighted, stress-tested, and decided, one transparent criterion at a time.

Subject of Research: Multi-criteria decision analysis of faculty priorities for lecture, flipped classroom, and problem-based learning in anesthesiology education

Article Title: Faculty priorities for lecture, flipped classroom, and problem-based learning in anesthesiology education: a fuzzy AHP–SAW multi-criteria decision analysis

Article References: Soleimani, T., Akbari, H., Hajijafari, M., Raeyat Mohtashami, A., & Mahdian, M. (2026). Faculty priorities for lecture, flipped classroom, and problem-based learning in anesthesiology education: a fuzzy AHP–SAW multi-criteria decision analysis. BMC Medical Education. https://doi.org/10.1186/s12909-026-10555-6

Image Credits: AI Generated

DOI: 10.1186/s12909-026-10555-6

Keywords: anesthesiology education, problem-based learning, flipped classroom, lectures, multi-criteria decision analysis, fuzzy analytic hierarchy process, simple additive weighting, medical education, active learning, faculty priorities, decision support, teaching methods

Water regulation is one of the least visible yet most consequential services that ecosystems provide. Forested slopes intercept rainfall, slow surface runoff, and encourage infiltration into soils and aquifers, which in turn sustains baseflow in rivers during dry seasons and buffers downstream communities against floods and landslides. When forests are cleared for agriculture, plantations, or settlements, the landscape loses much of this sponge-like behavior. Rainfall that once soaked gradually into the ground instead races over compacted surfaces, carrying topsoil with it, swelling rivers quickly, and leaving less water stored for the months when it is needed most. Sulawesi, with its rugged central highlands, four peninsulas, and dense network of small watersheds, is particularly sensitive to these shifts because so many communities depend on short, steep catchments for irrigation and drinking water.

To capture this capacity spatially, the study employed a weighted spatial index built on a Simple Additive Weighting approach, a multi-criteria decision technique that combines several input layers into a single score. In this case, the layers included natural landscape characteristics, natural vegetation, and land cover data for each of the three reference years. Each land cover type was assigned a weight reflecting its contribution to water regulation, so that intact natural forest scored highest, while agricultural land and built-up areas scored progressively lower. The weighted layers were then summed to produce a continuous index of potential water regulation capacity, which was classified into categories ranging from very high to very low. This approach allowed the author to compare like with like across a 32-year span and to pinpoint where on the island the losses were concentrated.

The results reveal a clear geographic pattern of decline. Central Sulawesi and South Sulawesi contributed most strongly to the reduction in high-capacity areas, consistent with the intensive forest conversion and agricultural expansion that have characterized those provinces in recent decades. South Sulawesi, home to the densely populated southwestern lowlands and the Makassar urban corridor, has long experienced pressure from rice cultivation, cocoa and other perennial crops, and urban growth. Central Sulawesi has seen extensive land clearing linked to agriculture and, in some areas, plantation development. By contrast, Southeast Sulawesi and North Sulawesi showed partial recovery of high-capacity classes in the later period, suggesting that regrowth of secondary vegetation, protection efforts, or shifts in land management may have locally restored some regulatory function even as the island-wide trend remained negative.

Underlying these capacity changes is a fundamental shift in what covers the land. The study found that vegetated land cover types that strongly support water regulation declined from 75 percent of the island in 1990 to 54 percent in 2022. That is a loss of roughly a fifth of the island’s vegetated surface in a single generation, largely driven by conversion to agricultural and built-up areas. The magnitude of this transition matters because the relationship between vegetation and water regulation is not linear at the margins. Small losses of forest in headwater zones can disproportionately affect downstream flow regimes, and once soils are degraded, the hydrological functions they supported can take decades to recover even if vegetation returns.

The spatial heterogeneity that the study documents is itself scientifically important. Sulawesi is not losing its water regulation capacity uniformly; it is losing it in patches, corridors, and hotspots that correspond to provincial boundaries, watershed networks, and the geography of agricultural frontiers. This heterogeneity is precisely what makes the mapping approach valuable for policy. Rather than treating the island as a single unit with a single trend, the weighted index produces provincial- and watershed-level hotspots where the decline is most acute, and therefore where interventions such as forest protection, restoration, and ecosystem service-based zoning could be prioritized with the greatest expected benefit. In practical terms, a watershed whose index has slipped from high to moderate is a candidate for targeted restoration of riparian buffers and headwater forest, while a province showing widespread contraction of high-capacity classes may warrant stronger land use controls.

The study’s methodology also illustrates a broader trend in ecosystem service science: the use of relatively simple, transparent, and repeatable indices rather than data-hungry process models. Complex hydrological simulations can capture the physics of rainfall-runoff partitioning in detail, but they demand extensive calibration data that are often unavailable in tropical regions. A Simple Additive Weighting index, by contrast, requires only consistent land cover and landscape datasets, which are increasingly available from national mapping agencies and satellite programs. The trade-off is that the index measures potential capacity rather than actual water fluxes; it indicates where the landscape is structurally capable of supporting rainfall retention, baseflow, and watershed resilience, not how many cubic meters of water are being stored in any given year. For regional planning purposes, however, this potential-based framing is often exactly what is needed, because it identifies where the hydrological infrastructure of the landscape is being dismantled before the consequences appear in stream gauges.

The findings sit within a wider body of evidence on deforestation in Indonesia and its consequences. Sulawesi has been identified in previous research as a priority island for conservation because of its exceptional endemism and the rapid loss of its forests, with documented impacts on primate habitat and unique ecosystems. Studies of deforestation patterns in Sulawesi between 1990 and 2018 have shown persistent pressure on forest cover, and work on mangrove decline in South Sulawesi has highlighted how coastal ecosystem services have been degraded alongside terrestrial ones. The new analysis extends this picture into the hydrological domain, showing that the same land conversion processes that threaten biodiversity are also dismantling the island’s capacity to regulate water, with direct implications for flood risk, dry-season water supply, and agricultural stability for millions of people.

What makes the study particularly timely is the policy context. Indonesia has experimented with forest moratoria, peatland protection, and payment-for-ecosystem-service schemes, and researchers have debated how cost-effective these instruments are in reducing emissions and protecting services. Mapping exercises like this one provide the spatial evidence base that such policies require. If restoration funds are limited, they should flow to the watersheds where the loss of regulation capacity is steepest and where downstream exposure is greatest. The study’s outputs, which identify provincial- and watershed-level hotspots, are designed to support exactly that kind of prioritization, and the author notes that they could inform ecosystem service-based zoning as part of provincial spatial planning.

There are also lessons here for how tropical islands elsewhere are monitored. Sulawesi’s story, a 15-percentage-point decline in high-capacity land over three decades, driven overwhelmingly by the replacement of vegetation with agriculture and settlements, is a pattern that repeats across Southeast Asia, Madagascar, the Caribbean, and many other island systems where steep terrain concentrates both water services and land use pressure. The demonstration that a weighted spatial index built from freely available land cover data can track these changes consistently over more than three decades offers a template for other data-limited regions. As climate change intensifies both droughts and extreme rainfall in the tropics, the landscapes that quietly regulate water are becoming more valuable, not less. Sulawesi’s shrinking high-capacity core is a quantitative reminder that this natural infrastructure is finite, spatially specific, and, once lost, slow and expensive to rebuild.

Subject of Research: Land cover change and water regulation ecosystem services on Sulawesi Island, Indonesia

Article Title: Spatial heterogeneity of landscapes and land cover in Sulawesi Island in relation to water regulation ecosystem services

Article References: Nursaputra, M. (2026). Spatial heterogeneity of landscapes and land cover in Sulawesi Island in relation to water regulation ecosystem services. Environmental Monitoring and Assessment, 198(10), Article 1086. https://doi.org/10.1007/s10661-026-15892-z

Image Credits: AI Generated

DOI: 10.1007/s10661-026-15892-z

Keywords: ecosystem services, water regulation, land cover change, spatial heterogeneity, Sulawesi Island, deforestation, Simple Additive Weighting, watershed resilience, forest restoration, Indonesia, landscape ecology, remote sensing