The low-stakes character of international large-scale educational assessments implies that a participating student might at times provide unrelated answers as if s/he was not even reading the items and choosing a response option randomly throughout. Depending on the severity of this invalid response behavior, interpretations of the assessment…

Large-scale standardized tests are regularly used to measure student achievement overall and for student subgroups. These uses assume tests provide comparable measures of outcomes across student subgroups, but prior research suggests score comparisons across gender groups may be complicated by the type of test items used. This paper presents…

The Standards for Educational and Psychological Testing indicate that validation of assessments should include analyses of participants' response processes. However, such analyses typically are conducted only to supplement quantitative field studies with qualitative data, and seldom are such data connected to quantitative data on student or item…

Using generalizability theory as a framework, the problem of assessing the content validity of standardized achievement tests is considered. Four designs to assess test-item fit to a curriculum are described, and procedures for determining the optimal number of raters and schools in a content-validation decision-making study are considered. (TJH)

Item statistics between illustrated and written items drawn from the same content areas were compared using F ratios. The results indicated: that illustrated items performed slightly better than matched written items; and that the best performing category of illustrated items was tables. (Author/BB)

The expected reliability of a multiple choice test is maximized by the use of three alternative items. (Author)

The validity of inferences based on achievement test scores is dependent on the amount of effort that examinees put forth while taking the test. With low-stakes tests, for which this problem is particularly prevalent, there is a consequent need for psychometric models that can take into account differing levels of examinee effort. This article…

A conception of invalidity as bias is related to content validity for standardized achievement tests. A method of estimating content bias for each of three content domains (a priori, curricular, and instructional) based on the specification of a content taxonomy is also proposed. (Author/CM)

A structure for describing different approaches to testing is generated by identifying five dimensions along which tests differ: test uses, item generation, item revision, assessment of precision, and validation. These dimensions are used to profile tests of reading comprehension. Only norm-referenced achievement tests had an inference system…

Investigates the convergence of three measures on a representation of cognitive structure and examines the correspondence between these representations and the representation of subject matter structure. (Author/DEP)

The purpose of this study was to gather empirical evidence to compare the reliabilities and concurrent validities of multiple choice and true-false tests that were written to measure understandings and relationships in the same content areas. (Author)

The history of the relation between achievement tests and curriculum programs is reviewed, and it is concluded that content specialists are best qualified as sources of curricular goals to specify content, kinds of attainment, and standards. The specification of instructional intents is also considered from the perspective of modern cognition…

The Stanford Achievement Test Forms E and F were judged to be one of the best achievement batteries for assessing basic skills taught in grades one through nine. The test publisher provides several booklets in addition to the administration manual. These include the Norms Booklet, Handbook of Instructional Strategies, and Guide to Classroom…

The validity of a comparison of mean test scores for two groups and of a longitudinal comparison of means within each group is assessed. Using LISREL, factor analyses are used to test the hypotheses of similar factor patterns, equal units of measurement, and equal measurement accuracy between groups and across time. (Author/DWH)

In general, faults did make the items easier, with the largest effects occurring for cue and grammar contaminations. (Authors)