Social Sciences · Education
THE EFFECT OF A SPECIAL CURRICULUM FOR DISADVANTAGED CHILDREN ON TEST‐RETEST RELIABILITIES OF THREE STANDARDIZED INSTRUMENTS<sup>1</sup>
Leo S. Goldstein, Alan R. Coller, John R. Dill, Howard S. Tilis
See on the mapAbstract
Test‐retest (stability) reliability coefficients are, by their nature, affected by what has happened to the tested subjects during the period between testings. Similarly, changes in subjects' mean performance will reflect what happened to them in the interim period. In this regard, the relative values of stability coefficients for experimental and control groups are indicative of the relative degree to which an experimental program is able to change the rankings of the students on criteria measures at pre‐ and posttreatment times. Changes in students' rankings however do not necessarily affect the level of the group's average performance. This study examines the stability reliabilities of three standardized testing instruments used in the evaluation of a special (enriched) curriculum for young disadvantaged children and the changes in their average performance over a 2‐year period. Magnitudes in test‐retest reliability differences between treatment groups seem to parallel differences in mean scores. Using the stability reliability coefficients as indices, the program discussed in this report seemed to be most effective in producing change in the skills measured by the Peabody Picture Vocabulary Test. However, significant differences in mean performance between the treatment groups indicates the effectiveness of the program not only in the skills measured by the PPVT but also in those areas measured by the Stanford Binet and the Columbia Mental Maturity Test.