Two young children with mental retardation displayed inconclusive patterns of destructive behavior when a functional analysis was conducted using the procedures described by Iwata et al. (1982/1994). A second functional analysis incorporated modifications to the social attention condition that were based on interview data from care providers and descriptive observations. Results of the modified social attention condition indicated that a specific 'quality' of attention was needed to identify the maintaining contingency for destructive behavior within an experimental analysis. Results are discussed in terms of considerations for further assessment when traditional analog functional analysis conditions are ineffective in isolating maintaining contingencies for destructive behavior.