Measurement of Team Decision-Making Through Generalizability Theory

The reliabilities and errors of judgmental-based assessment outcomes derived through use of the System to Plan Early Childhood Services (SPECS; Bagnato & Neisworth, 1990) were investigated within the framework of the generalizability theory of measurement (Cronbach, Gleser, Nanda, & Rajaratnam, 1972). The SPECS was designed explicitly to produce collaborative/team decisions with regard to child development status and need and routinely involves more than one rater. Overall (with the exception of the unusual situation in which only one rater was employed), scores from SPECS generally had adequate levels of reliability under the interpretation scenarios investigated.