Traditional statistical model evaluation typically relies on goodness-of-fit testing and quantifying model complexity by counting parameters. Both of these practices may result in overfitting and have thereby contributed to the generalizability crisis. The information-theoretic principle of minimum description length addresses both of these…

The big data imposes the key problem of generalizability of the results. In the present contribution, we discuss statistical tools which can help to select variables adequate for target level of abstraction. We show that a model considered as over-fitted in one context can be accurate in another. We illustrate this notion with an example analysis…

Intraclass correlation coefficients (ICCs) are commonly used in behavioral measurement and psychometrics when a researcher is interested in the relationship among variables of a common class. The formulas for deriving ICCs, or generalizability coefficients, vary depending on which models are specified. This article gives the equations for…

The purpose of this study was to investigate the methods of estimating the reliability of school-level scores using generalizability theory and multilevel models. Two approaches, "student within schools" and "students within schools and subject areas," were conceptualized and implemented in this study. Four methods resulting from the combination…

We propose that the continuing controversies over the use of quasi-"F"-ratios in psycholinguistic research might be circumvented, if not resolved, by the judicious application of Generalizability Theory (GT) analyses. We argue that GT is a logical extension of the basic rationale behind repeated measures Analysis of Variance (ANOVA) and the…

Increased demands for colleges and universities to engage in outcomes assessment for accountability purposes have accelerated the need to bridge the gap between higher education practice and the fields of measurement, assessment, and evaluation. The "Handbook on Measurement, Assessment, and Evaluation in Higher Education" provides higher…

Results from a performance assessment in which 68 high school students wrote essays support the use of latent variable modeling for estimating reliability, concurrent validity, and generalizability of a scoring rubric. The latent variable modeling approach overcomes the limitations of certain conventional statistical techniques in handling…

This study examined the difficulty level of essay topics used in the large-scale assessment of writing in relation to five different scoring models, and sought to determine what effects the scoring models would have on passing rates. In model one, examinee's score is the direct result of a score assigned by the reader or the sum of scores assigned…

The teacher assessment that is the subject of this paper is an essentially informal activity. The teacher assesses the student by posing questions, observing activities, and evaluating work in a planned or ad hoc way. The information obtained may be partial or fragmented, but repeating such assessments over time will allow the buildup of a solid…