Which Phonological Skills, as Primary Skills, Can Be More Effectively Trained and Lead to Decoding Skills in Low-Skilled Kindergartners?.

Which phonological skills, as primary skills, can more effectively be developed and lead to decoding skills in low-skilled kindergartners: (a) segmentation/blending, or (b) rhyming/first sound identification? Sixty-one low-skilled kindergartners, who scored less than 4 items correctly in 3 out of 5 measures, were randomly assigned to one of the strategy groups, receiving instruction in small groups of 3 to 4, 20 to 30 minutes for each time, twice a week, over 10 weeks. Results showed that both groups were effective in improving target skills, as well as reading and writing readiness skills. No significant difference between the groups was found. Neither group demonstrated successful transfer of trained skills to the untrained skills, or sufficient generalization to reading or spelling novel words. The study supports other research using both training strategies, and reminds us of the challenge in promoting generalization ability. Phonological Training and Decoding Skills 3 Which Phonological Skills, as Primary Skills, Can be More Effectively Trained and Lead to Decoding Skills in Low-Skilled Kindergartners? There are 20-25% of children who have problems in reading (Liberman & Liberman, 1992). A majority of children with learning disabilities, from 75% to 95%, suffer from reading disabilities (Bateman, 1991). Broader evidence indicates that phonological awareness may be the main determinant related to reading disabilities (e.g. Fletcher, Shaywitz, Shankwieler, Liberman, Sturebing, Francis, Fowler, & Shaywitz, 1994; Liberman & Shankweiler, 1985; Stanovich & Siegel, 1994; Vellutino, Scanlon, & Spearing, 1995; Waterman & Lewandowski, 1993), although there may be a subgroup of children's reading disabilities that are caused by dysfunction of the visual system, brain injuries, or genetics. How may we prevent young children from developing reading disabilities or intervene with children at risk of reading disabilities as early as possible? Encouraged by finding the important role of phonological awareness in reading, many phonological and /or phonemic awareness training projects have been conducted. The results show that children's phonological awareness can be developed through explicit training in preschool or kindergarten (e.g. Ball & Blackman, 1988 & 1991; Bradley & Bryant, 1985; O'Connor, Jenkins, & Slocum, 1995; Lundberg, Frost, & Petersen, 1988), even for very low-skilled children (O'Connor et al., 1995; O'Connor, Notari-syverson, & Vadasy, 1996). Training strategies, however, are various. One of the issues that puzzled researchers is which phonological and /or phonemic skills influence decoding skills and ultimately reading ability. Some researchers used phonemic segmentation (e.g. cat c-a-t) or blending (e.g. c-a-t -cat) as the target training skills. With trained phonemic skills, students were taught to assign Phonological Training and Decoding Skills 4 letter sounds to the letters in a word and sound the word out (e.g. Ball & Blachman, 1988). This is called the letter-sound decoding procedure (3-ph) (see Rack, Hulme, Snow ling, & Wightman, 1994). Other researchers used sound categorization as the target training skills (i.e., categorizing word sounds by initial sounds or rhyming parts such as fat / fight or hat / sat) (e.g. Bradley & Bryant, 1985). This skill was thought to be directly related to reading by making analogues (e.g. leak -beak peak) (e.g. Goswami & Bryant, 1990; Goswami & Mead, 1992). This is called the procedure of using analogues (SC) (see Rack et al., 1994). The study tested this experimental question: Which phonological skills, as primary skills, can more effectively be trained and lead to decoding skills in low-skilled kindergarten students: (a) letter-sound decoding with 3-phoneme segmentation and blending as the target skills, or (b) reading by analogues with sound categorization (rhyming & first sound identification) as the target skills? Specifically, this study examined the following questions: Would 10 weeks of instruction with one of the instructional strategies be more effective than the other in raising the trained phonological skills in the low-skilled kindergarten students? Would 10 weeks of instruction with one of the instructional strategies be more effective than the other in transfer trained phonological skills to the untrained phonological skills? And would 10 weeks of instruction with one of the instructional strategies be more effective than the other strategy in promoting generalization of developed phonological skills to reading and spelling tasks? Phonological Training and Decoding Skills 5 Methods Participants One hundred and seventy-five children from 7 kindergarten classes were assessed in late February, 1997. The basic criteria for the participants were that they scored less than 40% in 3 out of the following 5 phonological measures: a) segmentation, b) blending, c) rhyming, d) syllable deletion, and e) first sound identification. Segmentation, as the primary screening measure, must be one of the 3 measures. In addition, they could not recognize more than one word in the Woodcock-Johnson Letter-Word Identification Subtest. These criteria meant that the students selected would be both nonreaders and very poor in phonological skills. Sixty-four lowskilled kindergarten students were finally selected for the training. Since 3 of the selected students moved out, 61 participants remained. A randomized blocking procedure was applied. The selected samples were blocked by the five teachers who participated, so that the different teachers' instructional effects could be controlled. Students selected from each teacher were randomly assigned to one of the two treatment conditions: 31 to the letter-sound decoding procedure with segmentation and blending as the target skill, and 30 to the analogue procedure with sound categorization (rhyming & first sound identification) as the target skill. Assessment Measures Peabody Picture Vocabulary Test-Revised The Peabody Picture Vocabulary Test-Revised (PPVT-R) (Dunn & Dunn, 1981) was used in the pretests to assess students' receptive vocabulary. The PPVT-R is a norm-referenced, individually administrated achievement test on subject's receptive vocabulary, conducted through