A test of climate, sun, and culture relationships from an 1810-year Chinese cave record.

No abstract

No abstract

No abstract

Two different types of anchors are investigated in this study: a mini-version anchor and an anchor that has a less spread of difficulty than the tests to be equated. The latter is referred to as a midi anchor. The impact of these two different types of anchors on observed score equating are evaluated and compared with respect to systematic error (bias), random equating error (SEE), and total equating error (RMSE) using SAT operational data. The results suggest that the overall bias, SEE, and RMSE when the midi anchor is used are either smaller than or very similar to those when the mini anchor test is used. The findings suggest that a midi anchor test would be preferred to a mini anchor test if equating accuracy at the ends of the score scale is not a primary concern.

This study uses historical data to explore the consistency of SAT ® I: Reasoning Test score conversions and to examine trends in scaled score means. During the period from April 1995 to December 2003, both Verbal (V) and Math (M) means display substantial seasonality, and a slight increasing trend for both is observed. SAT Math means increase more than SAT Verbal means. Several statistical indices indicate that, during the period under study, raw-to-scale conversions are very stable, although conversions for extreme raw score points are less stable than are other conversions.

We make a distinction between two types of test changes: inevitable deviations from specifications versus planned modifications of specifications. We describe how score equity assessment (SEA) can be used as a tool to assess a critical aspect of construct continuity, the equivalence of scores, whenever planned changes are introduced to testing programs. We also report on how SEA can be used as a quality control check to evaluate whether tests developed to a static set of specifications remain within acceptable tolerance levels with respect to equatability.

No abstract