Individuals with autism and other developmental disabilities often have difficulty learning auditory-visual conditional discriminations that are important for early communication and generalization may be restricted. Programs for teaching these individuals often involve the fading (gradual change) of stimuli in small steps across trials. Failure to establish desired discriminations occurs when the fading does not direct attention to the relevant critical aspects of the stimuli.
The research described here illustrates new attention-shaping procedures for teaching complex auditory-visual discriminations and assessing generalization. The major purpose was to begin evaluation of the dual-modality transfer procedures with abstract stimulus sets that would rule out pre-experimental learning as an explanation of cross-modal (i.e., visual-to-auditory) transfer of stimulus control. In one procedure, auditory samples (pairs of same and different tones) were added to visual stimuli (pairs of same and different forms that participants matched already) presented as samples and comparisons. Across-trial contrast fading then gradually ‘vanished’ the visual samples, in order to establish conditional control of comparison selections by the pairs of same and different tones. Generalization then was assessed using new frequencies of the tones. Another procedure introduced a novel dynamic fading method. The visual cues that already controlled responding were erased actively within trials (like apparent movement) and cumulatively across trials. After fading, conditional control was demonstrated by the auditory samples only. Generalization was tested using tones with frequencies different from the training stimuli.