Baseline data were collected for the play scenarios that were selected for the intervention phase (Play Scenarios A), as well as the scenarios selected to assess generalization (Play Scenarios B) for each participant. During each 15-min baseline session, the child played with two sets of toys (i.e., either Play Scenarios A or B). Baseline probes were taken for a minimum of three sessions for each participant, for both Play Scenarios A and B, and continued until there was little variability in the data with no evident trend of increasing performance (i.e., maximum variability of three multi-symbol messages between the highest and lowest data point).