Data available from the Fall and Spring administration of the Stanford Achievement Test: Intermediate I: Form X and the Otis-Lennon Mental Ability Test for all Title I pupils in the State of New Hampshire plus a random sample for the entire state, made possible the item-by-item comparison of Fall and Spring performance on the same groups of children. Analysis of data was limited to one grade and five tests out of the complete battery. The purpose of the exploratory data analysis was to examine: relative gains by item, proportion of items attempted which were answered correctly, guessing, the prediction of a total score from a time-limited score on the basis of item analysis data available for the State of New Hampshire, and other incidental findings that might provide leads in constructing improved tests, especially for the measurement of Fall and Spring changes in educational achievement and particularly for the disadvantaged pupils as represented by the Title I selection within New Hampshire. Ways are now being sought to guarantee that children will work for the allowed time at the level effective for them, while at the same time predicting their score on an untimed test of "n" items. (Author/BJG)