Conventional linear and equipercentile procedures are reviewed with an emphasis on their utility for equating achievement tests pitched at different levels of difficulty. It is argued that the equipercentile procedure is superior to the linear procedure, but that neither is very satisfactory for the vertical equating problem. (Author/JKS)

The arbitrary nature of growth trends in cognitive variables is illustrated. Two metrics, grade equivalent and item-response theory representations, both of which preserve the order of performance levels in test data, produced different pictures of cognitive growth, and differences were seen to arise from differences in the scaling models. (SLD)

Estimates of reading ability for second through fourth grade Title I pupils were obtained through grade-level and out-of-level testing. Results show that the method of testing affects the students' scores, the identification of pupils needing remedial reading instruction, and evaluation of the reading program. (JKS)

An experiment investigated the extent to which the results of out-of-level testing may be biased because the child given an out of level test may have had a significantly different curriculum than the children given in-level tests. Item analysis data suggested this was unlikely. (CTM)