Two observed-score methods that assume equally reliable congeneric tests were compared for the external anchor test equating model. Each derivation was compared to existing test equating methods based on different assumptions. Using samples of Australian secondary school students (N=about 6,000), the methods were compared empirically with existing…

The effect of test length on predictive validity is examined empirically. For four tests, the curve of validity against test length had a very gentle slope for the longer tests and all tests could be reduced by more than 60 percent without appreciable decreases in validity. (Author/BW)

Equating error was estimated using the same test with three linear equating methods in three paradigms--single-link equating; circular equating through a chain; and pseudo-circular equating, in which a test is equated to itself through equating chains. Over 31,000 Psychometric Entrance Test scores of college applicants in Israel were analyzed.…

Mokken models of monotone homogeneity and double monotonicity and the Rasch model are compared using data from 990 young adult examinees taking a Dutch verbal intelligence test--the Verbal Analogies Test. The model of monotone homogeneity was found suitable for basic testing; more sophisticated applications appear to require parametric models.…

A new model is presented that explains reaction time fluctuations in prolonged work tasks. The model extends the so-called Poisson-Erlang model and accounts for long-term trend effects in the reaction time curve. The model is consistent with Spearman's hypothesis that inhibition increases during work and decreases during rest. (TJH)

In this study, 590 third graders from 12 Dutch schools took 32 tests indicating 16 semantic Structure-of-Intellect (SI) abilities and 1 of 4 reading comprehension tests, involving either multiple-choice or open-ended items. Results indicate that item type for reading comprehension is congeneric with respect to SI abilities measured. (TJH)