Abstract:
Language is a complex compendium of skills which involve both recognition and production. Reading Disabilities usually involve deficient spelling ability yet remediation has traditionally focused on remediating reading without regard to spelling. If remediation is structured around what is known about the links in language, it may be possible to utilise these links for more effective and efficient instruction.
This study tested the hypothesis that a programme which emphasised the correspondences in language would result in improved transfer between language skills. The study investigated the transfer effects of a computerised oral spelling procedure, on reading, written spelling and spelling discrimination. The oral spelling procedure involved teaching the sequence of letter names corresponding to a spoken word. The methodology incorporated a multiple-baseline design across words, nested within a nonconcurrent multiple-baseline across participants. Fourteen subjects aged 9-l to l3-5 years participated in the project.
A fading technique was employed to draw attention to the sequential nature of words and to aid the memory process. Each word was segmented into individual letters, then into letter clusters (onset, nucleus and rime; and onset and rime). The participant was asked to provide the missing letter name, or the missing sound then the letter names. This procedure drew attention to the correspondences within words at the level of visual letter and letter names as well as at the level of visual letters, letter names and letter sounds.
The results supported the hypothesis that oral spelling training can result in improved written spelling, reading, and spelling discrimination. The results are discussed with respect to the literature on the links in language. Specifically, they are discussed with respect to Stimulus Equivalence, a model which predicts generalisation in language based on the mathematical properties of equivalence relations. This model is critiqued and a new model, Stimulus Correspondence is proposed as a more accurate description of the links in language and a mote complete description of generalisation within language processes.