Large-scale education data are collected via complex sampling designs that incorporate clustering and unequal probability of selection. Multilevel models are often utilized to account for clustering effects. The probability weighted approach (PWA) has been frequently used to deal with the unequal probability of selection. In this study, we examine…

There is growing pressure to make efficacy experiments more useful. This requires attending to the twin goals of generalizing experimental results to those schools that will use the results and testing the intervention's theory of action. We show how electronic records, created naturally during the daily operation of technology-based…

Randomized experiments are considered the gold standard for causal inference because they can provide unbiased estimates of treatment effects for the experimental participants. However, researchers and policymakers are often interested in using a specific experiment to inform decisions about other target populations. In education research,…

Recent methods to improve generalizations from nonrandom samples typically invoke assumptions such as the strong ignorability of sample selection, which is challenging to meet in practice. Although researchers acknowledge the difficulty in meeting this assumption, point estimates are still provided and used without considering alternative…

Recently, Stephen Gorard has outlined strong objections to the use of significance testing in social research. He has argued, first, that as the samples used in social research are almost always non-random it is not possible to use inferential statistical techniques and, second, that even if a truly random sample were achieved, the logic behind…

This research examined the issue of nonresponse bias and how it was reported in nonexperimental quantitative research published in the "Delta Pi Epsilon Journal" between 1995 and 2004. Through content analysis, 85 articles consisting of 91 separate samples were examined. In 72.5% of the cases, possible nonresponse bias was not examined in the…

This article addresses the problem of small nonprobability samples in research in the education of the deaf and hard of hearing in the face of a current and increasing emphasis on "scientifically based research" as required by recent No Child Left Behind (NCLB) federal legislation. The authors examine the gains and losses in information…

The general question examined by this study was whether the tendency of subjects to ignore the known score in giving the best guess for a sample mean was due to a descriptive heuristic such as representativeness or to a mechanistic one such as active balancing. Two experiments were conducted. In Experiment 1, subjects estimated: (1) the mean of a…

In some educational research studies--particularly longitudinal studies requiring a probability sample of schools and spanning a wide range of grades--it is desirable to so select the sample that schools at different levels (e.g., elementary and secondary) "correspond." This has often proved unachievable, using standard methods of selecting school…

This article shows how the methods of survival analysis (also known as event history analysis) lend themselves to the study of the timing of educational events. Using examples from teacher attrition and student dropout research, survival methods are introduced for building statistical models of the risk of event occurrence over time. (Author/SLD)

Advances in computer technology are making it possible for educational researchers to use simpler statistical methods to address a wide range of questions with smaller data sets and fewer, and less restrictive, assumptions. This digest introduces computationally intensive statistics, collectively called resampling techniques. Resampling is a…

This paper provides an overview of common sampling methods (both the good and the bad) likely to be used in community college self-evaluations and presents the results from several simulated trials. The report begins by reviewing various survey techniques, discussing the negative and positive aspects of each method. The increased accuracy and…