An investigation of the fluency paradigm: the effects of accuracy training before rate-building and incremental increases in response rates on skill retention, endurance, stability, application and adduction

Fluency has recently been operationalized in terms of the acquisition of performance rates that predict a number of learning outcomes, depicted in the acronym RESAA, which represents skill retention, endurance, stability, application and adduction (Johnson and Layng, 1996). The RESAA model has not yet been adequately researched under controlled, experimental conditions. A preliminary study (Study 1) compared two rate-building procedures, under experimental conditions, with five Year 2 children with a mean age of six years eight months and seven pre-primary children with a mean age of four years seven months. The effects of practice and reinforcement were controlled. Long-term follow-up RESAA measures were conducted three months after the completion of the intervention. The major study in this research project (Study 2) is an empirical investigation of the effects on RESAA measures of increasing the performance rates of a component skill in reading to specific, incremental rate aims with twelve Year 2 children aged between six years eight months and eight years one month who were categorized into three levels of reading ability. Speeded practice was compared to slow-paced constrained-rate practice. The effects of practice and reinforcement were controlled. The utility of learning channel analysis for defining measures of application and adduction, and for measuring adduction on two composite tasks involving topographically dissimilar sensory and response dimensions was examined. Long-term follow-up RESAA measures were conducted three months after the completion of the intervention. The results of Study 1 indicated a procedure in which accuracy and rate were trained simultaneously was more efficient in increasing component skill rates and produced higher rates on the RESAA measures than training accuracy to 100% in a stage before rate-building commenced for the Year 2 children and two pre-primary children. Training accuracy to 100% before rate-building was marginally more efficient for five of the pre-primary children. Adduction was greater for a one learning-channel cross than for a two learning-channel cross. The results of Study 2 demonstrated that systematic increases in component skill rates were produced by both the rate-building and constrained-rate procedures, although higher rates were produced by the rate-building procedures for eleven of the twelve children. Higher training rates of the component skills produced concurrently higher rates on repeated RESAA measures during the intervention and on RESAA follow-up measures. Adduction was greater for a two learning-channel cross than for a one learning-channel cross. The level of reading ability of the children did not influence training rates of the component skill but did affect performances on the RESAA measures. Comparisons indicated that different training rates predicted different RESAA outcomes for all of the children.