The purpose of the current study was to expand upon previous research (e.g., Shillingsburg et al., 2017, 2019) by utilizing novel materials within activities and examining the use of partial textual stimuli within an interactive task to increase reports of past behavior for one child with autism. The results showed that correct reports increased following intervention across both opportunities, as well as the number of correct varying reports across activities. Additionally, because the materials used during each activity never repeated and the same question–answer combinations were never taught and reinforced, responding was most likely under multiple sources of control (i.e., the question and additional stimuli generated by Ellie’s problem solving behaviors). These results are promising in that it demonstrated a potential procedure to evoke answers about past activities.

Although children with autism may display difficulties with reports of past behavior for which there is only one correct answer (e.g., personal information, preferences of peers; e.g., Fienup et al., 2013) and reports in which the answer changes contingent upon many variables (e.g., “What did you do last weekend?”), the sources of control differ. In the first circumstance in which there is only one correct answer, eventually the verbal discriminative stimulus will evoke the answer. However, in the second circumstance, problem solving is required as an ongoing skill and recall can be conceptualized, at least from a behavior-analytic perspective, as a problem to be solved (Miguel, 2018; Palmer, 1991; Skinner, 1957, 1984). Granted, although the question will typically exert some discriminative control over the answer, it is often the case that the question alone is insufficient to evoke the response and a sequence of precurrent behaviors must be emitted at either the overt or covert level (Palmer, 1991; Skinner, 1957). Further, when verbal discriminative stimuli fail to evoke problem solving behavior, supplemental stimuli (prompts, etc.) may be required (Palmer, 1991; Skinner, 1957). Thus, rather than teaching problem solving, the experimenters expanded upon previous research to explore the use of partial textual stimuli within an interactive activity as potential supplemental stimuli to evoke it. We chose to use textual stimuli as opposed to an echoic prompt as the former allows increased exposure to the stimulus. That is, the textual stimulus is not removed from the environment after presentation, whereas the echoic prompt, requiring an echoic response is removed after initial presentation (Farber & Dickson, 2023).

Although the current results are encouraging, several limitations should be noted. First, there appeared to be a potential carryover effect for the second activity after the intervention was introduced for the first activity. However, mastery criteria were not reached until the intervention was introduced for subsequent activities. Although increases in responding for other activities after the onset of intervention for only one may be advantageous in a clinical setting, it may pose threats to the internal validity of the study. Future research should consider replicating our procedures with other designs such as a multiple baseline across participants. Additionally, future research should examine expanding the number of possible responses across activities and the number of activities to prevent a possible ceiling effect.

Second, it remains unclear what combination of components (e.g., textual stimuli, interactive activity, reinforcement) during the intervention were responsible for behavior change. Future studies should attempt to isolate these variables. Third, it is unclear which behavioral mechanisms were responsible for the increase in responding during intervention. It is possible that when verbal discriminative stimuli were not sufficient to evoke reports of past behavior, the addition of supplemental stimuli served to evoke covert behavior in the form of an intraverbal chain or covert seeing, or some combination of the two. It is also possible that partial textual stimuli evoked the emission of an autoclitic frame which may have led to covert imagining and covert tacting. In other words, after stating, “I made…” Ellie covertly imagined the craft activity, the product of which served to evoke a tact (e.g., “kite with bows”) or a covert intraverbal chain to allow correct responding (degli Espinosa et al., 2021; Miguel, 2018). Future research should attempt to isolate the behavioral mechanisms during this type of intervention.

Next, only one participant was included in the current study. Therefore, it is unclear if similar results would be obtained with participants with similar or different characteristics, for reports of past behavior involving longer delays, and for different types of activities. Further, the stacking blocks and partial textual stimuli were never removed. Therefore, the extent to which the response would maintain in the absence of the blocks is unknown. This limitation speaks to the practicality of this intervention. However, the fact that textual stimuli were not provided for the last response in the conversation sequence, suggests that the intervention might be effective in evoking reports of past behavior. Future research should examine the complete removal of textual stimuli and the feasibility of implementing the intervention in other educational settings. Additionally, formalized pretesting for tacts of the materials and activities used was not completed and may have presented a limitation to the current study. However, this may be unlikely given the participant’s scores on the VB-MAPP.

Finally, the current study did not teach a problem-solving strategy but rather explored the use of supplemental stimuli to possibly evoke problem solving. Thus, future research should examine potential assessment measures to determine whether difficulties with reports of past behavior are (a) a function of deficits in attending; (b) influenced by the prompts or stimuli utilized; or (c) related to lack of problem solving skills. Despite these limitations, the results provide preliminary evidence for the use of partial textual stimuli within an interactive task to increase reports of past behavior for a child diagnosed with autism, and it is our hope that these results will encourage future research in this area.