Federal policy on alternate assessment based on modified academic achievement standards (AA-MAS) inspired this research. Specifically, an experimental study was conducted to determine whether tests composed of modified items would have the same level of reliability as tests composed of original items, and whether these modified items helped reduce…

Scoring rubrics are routinely used to evaluate the quality of writing samples produced for writing performance assessments, with anchor papers chosen to represent score points defined in the rubric. Although the careful selection of anchor papers is associated with best practices for scoring, little research has been conducted on the role of…

Item response theory (IRT) methods are generally used to create score scales for large-scale tests. Research has shown that IRT scales are stable across groups and over time. Most studies have focused on items that are dichotomously scored. Now Rasch and other IRT models are used to create scales for tests that include polytomously scored items.…

Increasingly, assessment practitioners use generalizability coefficients to estimate the reliability of scores from performance tasks. Little research, however, examines the relation between the estimation of generalizability coefficients and the number of rubric scale points and score distributions. The purpose of the present research is to…

When raters assign different scores to a performance task, a method for resolving rating differences is required to report a single score to the examinee. Recent studies indicate that decisions about examinees, such as pass/fail decisions, differ across resolution methods. Previous studies also investigated the interrater reliability of…

Studied the use of descriptive graphic rating scales by 11 raters to evaluate students' work, exploring different design features. Used a Rasch-model based rating scale analysis to determine that all the continuous scales could be considered to have at least five points, and that defined midpoints did not result in higher student separation…

We determined the recommendations for preparing and scoring constructed-response (CR) test items in 25 sources (textbooks and chapters) on educational and psychological measurement. The project was similar to Haladyna's (2004) analysis for multiple-choice items. We identified 12 recommendations for preparing CR items given by multiple sources,…

A process for judging the alignment between curriculum standards and assessments developed by the author is presented. This process produces information on the relationship of standards and assessments on four alignment criteria: Categorical Concurrence, Depth of Knowledge Consistency, Range of Knowledge Correspondence, and Balance of…

Test homogeneity and internal consistency reliability indices were developed on the basis of theoretical considerations of properties of hierarchical structures of data matrices. This reconceptualization, in terms of ordinal test theory, has potential for explication of the mutual relationship of test reliability and homogeneity. (TJH)

Sampling theory for the intraclass reliability coefficient, a Spearman-Brown extrapolation of alpha to a single measurement for each examinee, is less recognized and less cited than that of coefficient alpha. Techniques for constructing confidence intervals and testing hypotheses for the intraclass coefficient are presented. (SLD)

An approach is presented that can predict internal consistency of cognitively complex assessments on two dimensions, those of adding tasks with similar or different solution strategies and adding test takers with different solution strategies. Data from the 1992 National Assessment of Educational Progress mathematics assessment are used to…

Computer simulation indicated that reliability increased with the degree of similarity between underlying and observed distributions when the observed categorical distribution was deliberately constructed to match the shape of the underlying distribution of the trait being measured. Reliability also increased with correlation among variables and…

Two studies involving 368 elementary and high school students and 29 readers were conducted to investigate reader consistency, score reliability, and reader time requirements of three hands-on science performance tasks. Holistic scores were as reliable as analytic scores, and there was a high correlation between them after they were disattenuated…

Compared two Angoff standard-setting methods, percentage, and yes-no, in the work of four groups of judges (n=24) given behavioral descriptors or incidents to use in making ratings. Results indicate that passing scores based on percentage estimates were stable from initial to final ratings, but those based on dichotomous (yes-no) ratings had…

The scores on 2 distinct tests (e.g., essay and objective) are often combined to create a composite score, which is used to make decisions. The validity of the observed composite can sometimes be evaluated relative to an external criterion. However, in cases where no criterion is available, the observed composite has generally been evaluated in…