Özel Sayı: Robert J. Mislevy'nin Katkıları

Özet (abstract)

This special issue honors the work of Robert J. Mislevy, whom I had the pleasure of knowing from the time he arrived at ETS in 1984 until his untimely passing in 2025.Shortly before his arrival, Bob's name became known to me during an IRT workshop given by Martha Stocking in late 1983.One morning, Martha came bounding into the room and announced, "We've just hired a rising star to help with our IRT research.His name is Bob Mislevy." Bob would have only been in his early thirties at that point but his reputation was already fast preceding him.And rise, his star certainly did.Across more than four decades, Bob helped reshape educational measurement.He did so by inventing new methodology and novel conceptual frameworks, uniquely drawing ideas from a stunning diversity of fields.He was a validity theorist, psychometrician, learning scientist, and philosopher rolled into one, an individual who enjoyed and, indeed thrived from, working on practical assessment problems because that work was the grist for his innovation mill.The articles collected here illuminate some of the breadth and depth of his influence-from the statistical infrastructure for large-scale group-score assessments to the epistemological foundations of educational measurement to a general design methodology for creating the most basic, as well as the most complex, digital performance assessments.John Donoghue opens the issue by documenting Mislevy's pivotal role in the development of group-score assessment methodology, the problem that constituted one of Bob's first assignments upon his arrival at ETS.What now is commonplace-marginal maximum likelihood estimation, conditioning on background variables, and the use of plausible values-was in the early 1980s a radical rethinking of how population-level inferences could be drawn responsibly from sparse individual data.As Donoghue describes, the architecture introduced in the 1984 NAEP reading assessment did more than solve an urgent technical problem; it established a durable framework for producing, reporting, and using assessment results at scale, one that is still employed in some form on virtually every group-score assessment.Drew Gitomer's article situates evidence-centered design (ECD) as one of Mislevy's most foundational and influential contributions.ECD reframes assessment as an argument from evidence, linking claims about learners to the tasks we design and the data we collect.Coming from Bob's insight that complex performance assessments required a structured design methodology, Gitomer traces ECD's intellectual lineage and practical impact.Gitomer shows how ECD provided the field with a common language for both assessment development and validity, one that continues to be used by assessment programs around the world.The essay by James Pellegrino, Mark Wilson, and Kadriye Ercikan offers a personal and historical perspective, focusing on Mislevy's influence on the National Research Council report, Knowing What Students Know.Among other things, the report popularized the "Assessment Triangle" (cogni-

Yazarların özeti; kaynağından alınmıştır. Educational Measurement Issues and Practice, 2026 · DOI ↗