Trend scoring constructed response items (i.e. rescoring Time A responses at Time B) gives rise to two-way data that follow a product multinomial distribution rather than the multinomial distribution that is usually assumed. Recent work has shown that the difference in sampling model can have profound negative effects on statistics usually used to…

In generalizability theory studies in large-scale testing contexts, sometimes a facet is very sparsely crossed with the object of measurement. For example, when assessments are scored by human raters, it may not be practical to have every rater score all students. Sometimes the scoring is systematically designed such that the raters are…

The Angoff method requires experts to view every item on the test and make a probability judgment. This can be time consuming when there are large numbers of items on the test. In this study, a G-theory framework was used to determine if a subset of items can be used to make generalizable cut-score recommendations. Angoff ratings (i.e.,…

Developing indicators that reflect important aspects of school and classroom environments has become central in a nationwide effort to develop comprehensive programs that measure teacher quality and effectiveness. Formulating teacher evaluation policy necessitates accurate and reliable methods for measuring these environmental variables. This…

We examined how rater language background affects the scoring of short-answer, open-ended test items in the assessment of English language learners (ELLs). Four native English and four native Spanish-speaking certified bilingual teachers scored 107 responses of fourth- and fifth-grade Spanish-speaking ELLs to mathematics items administered in…

Two relatively simple methods for estimating the condition standard error of measurement (SEM) for nonlinearly derived score scales are proposed. Applications indicate that these two procedures produce fairly consistent estimates that tend to peak near the high end of the scale and reach a minimum in the middle of the raw score scale. (SLD)

Considers the situation in which content or administrative considerations limit the way in which a test can be partitioned to estimate the internal consistency reliability of the total test score. Demonstrates that a single-valued estimate of the total score reliability is possible only if an assumption is made about the comparative size of the…

Examined the stability of the standard error of measurement and the relationship between the reliability coefficient and the variance of both true scores and error scores for 170 school districts in a state. As expected, reliability coefficients varied as a function of group variability, but the variation in split-half coefficients from school to…

This overview of the role of error and tolerance for error in measurement asserts that the generic precision associated with a measurement procedure is defined as the root mean square error, or standard error, in some relevant population. This view of precision is explored in several applications of measurement. (SLD)

This study provides empirical evidence about the sampling variability and generalizability (reliability) of a statewide performance assessment for grade six. Results for 600 students at individual and school levels indicate that task-sampling variability was the major source of measurement error. Rater-sampling variability was negligible. (SLD)