Assessing the construct validity of the therapy outcome measure for pre-school children with delayed speech and language

Abstract This paper explores the construct validity of a measure of clinical outcomes, the Therapy Outcomes Measures (TOM). The work took place within a randomized controlled trial of pre-school children with speech and language difficulties in community clinics. Assessments of 159 children pre-randomization covered aspects of the children's expressive and receptive language, phonology, attention, play and socialization. The analysis investigated the relationship between the TOM and the various assessments. The sample, which included a range of primary speech and language difficulties, was stratified according to children's baseline scores on receptive and expressive language and their phonology. This made possible an assessment of the TOM against the pattern of the children's difficulties rather than against a single criterion. The pattern of correlation found between the TOM impairment ratings and the assessments reflected the child's predominant difficulty suggesting that the TOM ratings do have constr...