Search

The first purpose of the study reported in this paper was to determine whether subjects trained in phonemic segmentation and spelling would have an advantage in segmenting, which would establish a causal relationship between segmenting and reading. The second purpose was to see whether trained subjects were better blenders than control subjects. Subjects were 22 middle-class first-grade students at a New York City independent school. The students were assigned to two treatment groups using stratified random assignment with strata based on kindergarten teachers' predictions of reading success. The Test of Auditory Analysis Skills, the Block Segmentation Test, and the Roswell-Chall Blending Test were administered for pre- and posttest measures. First graders were trained for 6.5 months to segment and spell with wooden blocks and to spell on the computer. Control subjects were trained to read letters and blend words, and to read text on the computer. Results indicated that trained subjects were significantly better than controls by posttest on measures of segmentation and blending, even though blending was not taught explicitly to experimental subjects. Results suggest that blending is associated with skilled reading and is facilitated by segmentation training. (Three tables of data are included and 30 references are attached. One appendix contains the Block Segmentation Test.) (Author/MG)