Gene V Glass is an educational statistician, researcher, and policy critic whose career has been defined by a series of methodological innovations and sustained polemical engagements with the political economy of American education. He transferred to the University of Nebraska in 1959 and came under the early influence of Robert Stake, whose commitment to naturalistic and evaluative approaches to educational research would shape Glass's own eventual critique of narrow quantitative orthodoxies. He completed his doctoral work at the University of Wisconsin under Julian Stanley, one of the foremost statisticians in educational research, receiving training in experimental design, psychometrics, and time-series analysis that gave him an unusually strong technical foundation. Glass rose rapidly in the profession, serving as President of the American Educational Research Association in 1974–76 — the youngest person to hold that position at the time — and occupying faculty positions at the University of Colorado Boulder, Arizona State University, and the National Education Policy Center. Two encounters proved pivotal in shaping his intellectual trajectory: his early encounter with Robert Stake's conviction that large-scale data collections distort more than they illuminate; and, more dramatically, a chance 1978 conversation with Stake in a Denver airport in which Stake challenged Glass to question whether the statistical sophistication he had spent his career developing was actually serving the goal of understanding education. This challenge catalysed a transformation that would lead Glass from technical innovation toward critical policy analysis and eventually toward public advocacy against the privatisation and standardisation of American public schools.

Glass's most widely cited and methodologically consequential contribution is the invention of meta-analysis, a procedure he introduced in his 1976 presidential address to the American Educational Research Association, subsequently published in Educational Researcher. The address was titled “Primary, Secondary, and Meta-Analysis of Research,” and it proposed that the accumulated findings of empirical studies in any domain could themselves be treated as data for systematic quantitative analysis. Where previous approaches to research synthesis had relied on the informal judgment of expert reviewers — with all the subjectivity and inconsistency that entailed — meta-analysis offered a replicable, transparent procedure for aggregating effect sizes across studies, identifying moderator variables, and characterising the state of evidence on a question with far greater precision than narrative review permitted. The impact was immediate and transformative: meta-analysis rapidly became the standard method for systematic review in medicine, psychology, education, and the social sciences, and the Cochrane Collaboration and the Campbell Collaboration — which produce meta-analyses that directly inform health and social policy — are among its most important institutional legacies.

Central to Glass's meta-analytic methodology is the standardised effect size statistic that bears his name. The Glass effect size (ES) is calculated by dividing the difference between the mean of a treatment group and the mean of a control group by the standard deviation of the control group: ES = (X̄_t − X̄_c) / S_c. By expressing the magnitude of a treatment effect in standard deviation units rather than raw score differences, the Glass effect size permits meaningful comparison across studies that use different outcome measures, different populations, and different scales. A treatment producing an effect size of 1.0 raises the average treated student to the 84th percentile of the untreated distribution — a practically significant improvement. This metric enabled Glass and his collaborators to ask questions about educational interventions that had previously been unanswerable: not just whether a treatment worked but how large the effect was, whether it varied across populations, and how it compared to effects produced by other interventions. The Glass effect size, along with the related Cohen's d, became the lingua franca of quantitative educational research and has informed the design of evidence-based policy across multiple domains.

Glass first demonstrated the power of meta-analysis at scale in a landmark 1977 study, conducted with Mary Lee Smith and published in American Psychologist. The study aggregated the results of 375 controlled studies of psychotherapy and reached the then-controversial conclusion that psychotherapy was, on average, substantially more effective than no treatment, with an average effect size of approximately 0.68 standard deviations. The study was controversial because it included studies of varying methodological quality — a principled decision that Glass defended on the grounds that quality variation itself could be treated as a moderator variable — and because it challenged the influential Eysenck claim that psychotherapy was no more effective than spontaneous remission. The 1977 study not only settled a long-running empirical debate but established meta-analysis as a credible and powerful research instrument and demonstrated its applicability beyond education to clinical and social psychology.

Among Glass's most policy-relevant applications of meta-analysis was his synthesis of research on class size, conducted with Mary Lee Smith and published in 1979 and later expanded in the book School Class Size (1980). By aggregating the results of decades of studies examining the relationship between class size and student achievement, Glass and Smith were able to demonstrate that the relationship was not linear: reductions in class size produced small and inconsistent effects at the upper end of the range (classes of 25–40 students) but substantially larger effects when classes were reduced to genuinely small sizes (fewer than 15–20 students). These findings provided the empirical foundation for the Tennessee STAR experiment — a large-scale randomised controlled trial of class size reduction — and have continued to inform debates about the allocation of educational resources. Glass also used this research as an occasion to reflect critically on the ways in which quantitative findings are selectively cited, distorted, or ignored in the political process of educational policymaking.

In 1993 Glass, together with Gene Maeroff and others, founded the Education Policy Analysis Archives (EPAA), the first peer-reviewed, open-access academic journal in education. Launched well before open access became a movement, EPAA was a deliberate political and intellectual act: Glass believed that educational research ought to be freely available to teachers, parents, policymakers, and citizens rather than locked behind journal subscription paywalls accessible only to university-based researchers. The journal has continued to publish scholarship on education policy from an international perspective, and its founding anticipates by more than a decade the broader movement for open access to publicly funded research that has since transformed academic publishing. This initiative reflects a consistent thread in Glass's work: a conviction that the value of educational research lies ultimately in its capacity to inform democratic deliberation about schooling, not merely to advance academic careers.

• Statistical Methods in Education and Psychology (1970, with Julian Stanley)
• Design and Analysis of Time-Series Experiments (1975, with V.L. Willson and J.M. Gottman)
• “Primary, Secondary, and Meta-Analysis of Research” (1976, Educational Researcher)
• “The Benefits of Psychotherapy” (1977, with Mary Lee Smith, American Psychologist)
• School Class Size: Research and Policy (1980, with Mary Lee Smith)
• Fertilizers, Pills and Magnetic Strips: The Fate of Public Education in America (2008)
• 50 Myths and Lies That Threaten America's Public Schools (2014, with David C. Berliner)