Effects of spontaneous phonemic awareness knowledge, contextualized phonemic awareness instruction, and decontextualized phonemic awareness training on phonemic awareness, reading, and spelling development in kindergarten students

Includes bibliographical references (pages [207]-217)Thirty-two second-semester English-Language-Learner (ELL) and monolingual kindergartners, with a mean age of 70.53 months ( SD = 3.70) from a Midwestern suburban school, were assigned to two treatment and two control groups. Twenty-four subjects met selection criteria; they identified 20 or more upper-case letters, wrote 15 or more letters, and spelled representing no more than initial phonemes (sounds). The remaining eight subjects were a control group of found inventive spellers (FIS group); they represented beginning, middle, and ending phonemes in their pretest spellings. All subjects were administered phoneme segmentation, phoneme deletion, invented spelling, word identification, and word attack assessments at the pretest and posttests 1 and 2. Triplets, matched on spelling pretest scores, were randomly assigned to three groups: decontextualized phoneme segmentation training (PS group); contextualized invented spelling instruction (IS group); and a control group with no inventive spellers (without FIS group). The two treatment groups received twenty, 20-minute training sessions over an 8-week period for a total of 6 hours and 40 minutes. Descriptively, the FIS group had the highest means on all measures. The IS group's phoneme segmentation mean at posttest 2 (M = 14.75, SD = 5.26) approached that of the FIS group (M = 16.00, SD = 6.09). The IS group had the highest spelling pretest-posttest-1 and pretest-posttest-2 mean gains. In each group, an ELL subject was the top scorer for most literacy measures. Repeated measures ANOVA showed the IS group was the only group that had significant phoneme segmentation gains between both the pretest and posttest 1, F(3, 28) = 7.29, p = .03, and between posttest 1 and posttest 2, F(3, 28) = 6.85, p = .04. With pretest scores used as the covariate, ANCOVA found a significant difference between groups on spelling at posttests 1 and 2, F(3, 27) = 2.98, p = .05, and F(3, 27) = 4.20, p = .02, respectively; the IS group's adjusted posttest means were the highest at both spelling posttests.Ed.D. (Doctor of Education