Skip to main content
NCBI home page
The Analysis of Verbal Behavior logoLink to The Analysis of Verbal Behavior
. 2022 Dec 5;38(2):190–198. doi: 10.1007/s40616-022-00176-7

The Effects of Teaching a Problem-Solving Strategy on Recalling Past Events with a Child with Autism

Stephanie H Keesey-Phelan 1, Judah B Axe 1,✉, Ashley L Williams 1,2
PMCID: PMC9747991  PMID: 36618971

Abstract

Problem-solving strategies, such as visual imagining and self-questioning, may assist children with autism spectrum disorder (ASD) in recalling past events. In the current study, at the start of each session, a 7-year-old boy with ASD engaged in a novel activity with a behavior therapist who took pictures of the activity. Ninety minutes later, a different therapist asked the participant to describe the prior activity. The intervention consisted of showing the participant pictures of the activity, telling him to close his eyes and imagine the activity, modeling asking and answering seven questions (e.g., “Who was there?” “What was one thing that happened?”), prompt fading, and reinforcement. Following the intervention, recall statements increased.

Keywords: Memory, Problem solving, Recall, Self-questioning, Visual imagining

Children with autism spectrum disorder (ASD) exhibit delays in answering questions about past events such as, “How was your visit to the zoo?” (Andersen et al., 2013). Academic learning, performance on assessments, and social skills require recalling past events, yet there is limited research on this skill (e.g., Fienup et al., 2013; Shillingsburg et al., 2017, 2019). According to Skinner (1957), recalling past events is evoked by (a) the request to do so, (b) “events within the speaker generated by the question,” and (c) “a history of earlier conditioning” (p. 143). A stronger conditioning history (i.e., the pairing of reinforcing or punishing events and relevant behaviors) increases the likelihood of producing private discriminative stimuli (SDs) that evoke recall (Palmer, 1991). For example, the private SDs that evoke a response to the question, “Where did you go to college?” are likely more readily available than the private SDs necessary to respond to the question, “What did you do three Saturdays ago?”

This last question is likely a problem in that it is a stimulus condition that does not immediately evoke behavior to produce a solution (Skinner, 1953, 1957). To answer, or engage in problem solving, one behaves in ways that produce or supplement the SDs that will evoke the response (Axe et al., 2019; Donahoe & Palmer, 1994; Palmer, 1991; Skinner, 1953, 1957, 1968). One problem-solving behavior might be covertly imagining your calendar (Kisamore et al., 2011) and producing a covert image that will evoke the response (Kritch & Bostow, 1993). Additionally, engaging in self-questioning, such as, “What was the date? What did I do last Saturday?” (Park & Gaylord-Ross, 1989), produces answers that function as covert SDs that evoke recall.

The purpose of the current study was to extend prior research on recall (Fienup et al., 2013; Shillingsburg et al., 2017, 2019) by using Skinner’s (1953, 1957, 1968) and Palmer’s (1991) analysis of problem solving and the analysis that forgetting is a function of a loss of stimulus control. Specifically, we evaluated the effects of visual imagining (Aguirre & Rehfeldt, 2015; Kisamore et al., 2011) and self-questioning (Park & Gaylord-Ross, 1989) on the frequency of recall statements with a child with ASD.

Method

Participant

Frank was a White, 7-year-old boy diagnosed with ASD. Frank’s verbal repertoire matched Level 2 on the Verbal Behavior Milestones Assessment and Placement Program (VB-MAPP; Sundberg, 2014) with emerging skills in Level 3. Frank engaged in the full range of verbal operants, including manding for information and multiply controlled intraverbals.

Setting and Materials

The study took place at an ABA clinic with two targeted locations. One was a conference room (4.9 x 4.3 m) for the “activity,” with a white board, credenza, and table with 10 chairs (Frank spent no other time in this room). There was a poster of an elephant on the door, thus named the “elephant room.” The second location was a kitchen (3 x 4.3 m) for the recall request; the kitchen had a table and three chairs. Materials for the elephant room activity included toys, pencils, markers, paper, balloons, finger puppets, and toy camping items. Additional materials included a digital camera with a 6.9 cm screen used to take pictures of the activity to use for the visual imagining portion of training, a timer on a smartphone, a video camera, and a token board.

Dependent Variable and Experimental Design

The dependent variable was the frequency of recall statements, defined as responses that corresponded with events that occurred during the activity and included at least a subject and a verb. We required at least a subject and a verb as we hypothesized a lower likelihood of social reinforcement for responses containing only subjects or verbs (e.g., “Stephanie, played, fell down, lost, happy, me”) compared to responses containing subjects and verbs (e.g., “Stephanie was there. We played bowling. All my pins fell down. I was happy”). In accordance with the trained self-questions in each tier of the concurrent multiple baseline design across sets of self-questions (see Table 1), we categorized recall statements as (a) people and materials (e.g., “Meg was there”), (b) feelings (e.g., “I felt excited”), and (c) events (e.g., “I won the game”). Mastery within each training level was answering all self-questions one time per level.

Table 1.

Self-questions and picture protocols

Tier Self-Question Picture
1 People/Materials Who was there? RA1 took a picture of RA1 and Frank together.
What was there? RA1 took a picture of the materials used in the activity.
2 Feelings How did I feel? At the end of the activity, RA1 asked Frank how he was feeling. For example, if Frank said that he was feeling “happy,” RA1 said, “show me a happy face,” and took a picture of Frank.
How did [RA1] feel? At the end of the activity, RA1 stated how he or she felt. For example, RA1 would say, “I am feeling excited, so I am going to take a picture of myself looking excited!” RA1 varied responses regarding how they felt.
3 Events What is one thing that happened? RA1 took a picture of the first “memorable” event.
What else happened? RA1 took a picture of the second “memorable” event.
What is one more thing that happened? RA1 took a picture of the third “memorable” event.
Open in a new tab

Interobserver Agreement and Procedural Integrity

Interobserver agreement (IOA) and procedural integrity (PI) data were collected for nearly 100% of sessions. Mean IOA was 100% for baseline probes, 96.5% (range, 57–100%) for probes during training, and 100% for generalization probes. Across baseline probes, training in Tier 1, training in Tier 2, training in Tier 3, and generalization probes, mean PI with the activity was 94% (range, 90-100%), 92% (range, 70–100%), 92% (range, 73–100%), 96% (range, 73–100%), and 88% (range, 81%–91%), respectively. Mean PI during probes and training was 96% (range, 87–100%), 79% (range, 50–100%), 76% (range, 36–100%), 91% (range, 71–100%), and 80% (range, 60%–100%), respectively.

Procedures

Activity

All sessions began with Research Assistant (RA)1 saying to Frank, “Let’s go to the elephant room.” RA1 introduced themselves and presented a novel activity, such as making pretend pizza, drawing superheroes, eye spy, balloon tennis, finger puppets, and camping. Frank could tact each activity and associated feelings. The activities were also age-appropriate, and they aligned with Frank’s school curriculum. Although we did not formally assess preference, Frank often made positive comments during the activities (e.g., “This is cool!”). RA1 explained the activity, guided Frank through it, and made sure there were at least three “memorable” events designed to have reinforcing or punishing value, such as someone winning, calling out cheating, and a scary ghost story. RA1 took seven pictures with the digital camera that corresponded to the seven self-questions we taught in the visual imagining and self-questioning training (Table 1).

Probe

Approximately 90 min after the activity, RA2 brought Frank and RA1 into the kitchen. RA2 said, “Frank, I want to ask you a question. Tell me about what you did in the elephant room.” Regardless of the accuracy of Frank’s responses, RA2 responded, “Okay thanks.” When Frank stopped responding for 10 s, RA2 ended the session. In Sessions 8, 9, and 10, Frank said, “I don’t know” immediately after the recall request. As we hypothesized that this was due to the use of extinction procedures in the probes, in Session 11 through the end of the study, we added a reinforcement system that was familiar to Frank. To implement this, RA2 asked the probe question and said, “If you can remember some things now, you can earn your tokens and go back to group right away!” Contingent on correct responses, RA2 delivered tokens, exchangeable for the reward of returning to the final, fun activity in the social skills group; Frank always earned the tokens.

Visual Imagining and Self-Questioning Training

Training occurred in each tier (see Table 1) based on incorrect or no answers in the probe. For example, if Frank responded to the probe question by saying, “There were markers, popsicle sticks, and paper,” thus answering Self-Question (SQ) 2, RA2 provided training on SQ 1 only. After initial training, if there were missed responses in the probe, training occurred in order of the tiers (i.e., first Tier 1 SQs, then Tier 2 SQs, then Tier 3 SQs).

Following a probe with an incorrect or no response, RA2 said, “Frank, I want to ask you a question. I want you to tell me about what you did in the elephant room. But before you answer, I’m going to ask [RA1] first to help you remember.” RA2 turned to RA1 and said, “[RA1], tell me about what you did in the elephant room.” The RAs provided and faded the following prompts using most-to-least prompting: Level 1: Model, Visual Prompt, Vocal Prompt; Level 2: Visual Prompt, Vocal Prompt; Level 3: Vocal Prompt; and Level 4: No prompts and no Error Correction 3 (see below). Contingent on answering all target questions correctly once per level, we faded to the next level.

Model

RA2 displayed the photograph corresponding to the question being taught (e.g., the picture corresponding to “Who was there?”) on the screen of the digital camera and said, “Look at the picture of who was there in the elephant room” (italics indicate emphasis). After 5 s, RA2 changed the screen to show the picture corresponding to “What was there” and said, “Look at the picture of what was there in the elephant room.” Next, RA2 said, “[RA1], I want you to close your eyes and imagine who was there and what was there in the elephant room.” RA2 pushed the power button on the camera to turn it off and said, “This is what it looks like when you close your eyes.” With RA1’s eyes still closed, RA2 said, “[RA1], I want you to tell me who was there and what was there in the elephant room.” RA1 modeled asking and answering those questions (e.g., “Who was there? Frank and Becky were there”). RA2 said, “That’s right. Thanks for telling me about who was there in the elephant room.”

Visual Prompt

RA2 turned to Frank and said, “Okay Frank, it’s your turn. I want you to tell me about what you did in the elephant room.” RA2 displayed the picture and said, “Look at the picture of who was there in the elephant room,” and left the picture on the display for 5 s. RA2 then displayed the second picture and said, “Look at the picture of what was there in the elephant room” (5 s). RA2 then said, “I want you to close your eyes and imagine who was there and what was there in the elephant room. I want you to tell me about what you did in the elephant room.” Contingent on correct responses here and in the remaining prompt levels, RA2 delivered enthusiastic praise and a token.

Vocal Prompt

RA2 said, “The first question is X (e.g., who was there)? You ask and answer it.” RA2 used least-to-most prompting to prompt Frank to ask the question, but RA2 did not prompt the answer. For example, if Frank did not immediately repeat, “Who was there?” RA2 prompted “Wh…,” then “Who…,” then “Who wa….,” and finally, “Who was there?” RA2 introduced the more intrusive prompt after 3 s if Frank did not respond or if he responded incorrectly. After stating a question, Frank had 5 s to respond.

Error Correction

Contingent on errors at any level of prompting, RA2 used a progressive error correction (EC) sequence and stopped when Frank responded correctly. For EC 1, RA2 said, “Close your eyes and try to imagine [question] (e.g., who was there) in the elephant room.” For EC 2, RA2 showed Frank the picture and said, “Look at the picture (3–5 s). Now close your eyes and imagine [question] – now you ask and answer.” For EC 3, RA2 asked the question to RA1 who modeled a response and then turned to Frank and said, “Now that [RA1] has helped us remember, Frank, you tell me [question] – now you ask and answer.” An incorrect response at this point resulted in progressing to the next question.

Generalization Probes

After answering all self-questions in VI/SQ training Level 4, we administered three generalization probes that were identical to those described above except the first one was in the waiting room, the second one was with a novel therapist, and the third one was with Frank’s mother in the waiting room.

Results and Discussion

In Tier 1 (people/materials) and Tier 2 (feelings), baseline responding was consistently 0 (Fig. 1). In Tier 3 (events), baseline responding was often 0 but also variable with 1, 2, and 3 recall statements. Upon VI/SQ Level 1 training in all three tiers, responding was 0. Tier 1 reached its maximum (2) in Level 4; Tier 2 reached its maximum (2) in Level 3 but was lower and more variable in Level 4; and Tier 3 reached its maximum (3) in Levels 3 and 4. Frank emitted 7, 7, and 3 responses in the generalization probes.

Fig. 1.

Fig. 1

Open in a new tab

The frequency of Frank’s recall statements during probes across baseline, the four levels of training, and the three types of statements (people/materials, feelings, events). Note: BL, baseline, VI/SQ, visual imagining/self-questioning training, Gen, generalization. Closed circles represent probe statements; the closed square was RA2 requesting recall in the waiting room; the closed diamond was a novel therapist requesting recall in the kitchen; and the open diamond was Frank’s mother requesting recall in the waiting room

Teaching a child with ASD the problem-solving strategies of visually imagining and self-questioning resulted in increased frequencies of recall statements. These results extend the literatures of teaching children with ASD to recall past events (Shillingsburg et al., 2017, 2019) and the problem-solving strategies of visual imagining (Aguirre & Rehfeldt, 2015; Kisamore et al., 2011) and self-questioning (Mann & Karsten, 2020; Park & Gaylord-Ross, 1989).

Teaching recall may occur (a) at the time of the event, (b) between the event and the recall request, and (c) at the time of the recall request. Fienup et al. (2013) recorded responses during the event (a). Shillingsburg et al. (2017, 2019) faded the latency between the event and the recall request (b). We intervened at the time of the recall request (c), which appears most aligned with the behavior-analytic conceptualization of recall as problem solving during the problem (Palmer, 1991; Skinner, 1953). Regarding Skinner’s (1957) definition of recall (outlined above), we arranged conditioning during the activity (i.e., reinforcement of play behaviors in the form of a preferred activity and attention), a recall request, and visual imagining and self-questioning to generate covert SDs to evoke the recall responses.

Motivation seemed to impact Frank’s responding. Given that we hypothesized limited responding in Sessions 8, 9, and 10 due to extinction in the probes, we added a reinforcement system during probes. Future researchers should arrange positive reinforcement from the start of a study to control for motivation, which may not affect internal validity (LeBlanc et al., 2003).

There are several limitations to address in future studies. First, low ranges of IOA and PI data reflect the complexity of the procedures. Future researchers should streamline the procedures and ensure data collectors are trained to criterion. Second, the variable data in Tier 3 baseline may have been due to weakly established responses or changes in the reinforcing qualities of the activities (Palmer, 1991). Third, using only one location for activities (elephant room) and recall requests (kitchen) limited the promotion of generalization across settings. Finally, the absence of baseline generalization data limits conclusions regarding generalization across settings.

Future researchers may also evaluate other problem-solving strategies to increase recall, such as self-rules (Root et al., 2017; Sautter et al., 2011); for example, “First I think about who I played with, then I think about what we played with, and then I try to remember something that happened.” A more generative problem-solving procedure may be teaching students to take their own pictures or notes using smartphones, as note-taking has been shown to improve the recall of students (Austin et al., 2002) and jurors (Thorley, 2016). Future researchers may increase recall by reinforcing verbal behavior, such as tacting the materials, during the initial activity. Finally, future researchers may gain information about covert behavior by recording behavioral correlates, such as lip movements, looking up, and closing eyes (Aguirre & Rehfeldt, 2015).

Acknowledgments

We thank the following people for their assistance with data collection on this project: Samantha Benoit, Andrea Buroff, Laura Cray, Hannah Greenwood, Kendra Guinness, Aimee Morrison, Lauren Werth, and Tali Rudy Zaltzman.

Funding

We also thank the Verbal Behavior Special Interest Group for providing funding to support this project.

Data Availability

Data sharing not applicable to this article as no datasets were generated or analyzed during the current study

Declarations

Conflicts of Interest

We have no conflicts of interest relevant to this article to disclose.

Footnotes

Publisher’s Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

References

  1. Aguirre AA, Rehfeldt RA. An evaluation of instruction in visual imagining on the written spelling performance of adolescents with learning disabilities. The Analysis of Verbal Behavior. 2015;31(1):118–125. doi: 10.1007/s40616-015-0028-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Andersen PN, Hovik KT, Skogli EW, Egeland J, Øie M. Symptoms of ADHD in children with high-functioning autism are related to impaired verbal working memory and verbal delayed recall. PloS One. 2013;8(5):e64842. doi: 10.1371/journal.pone.0064842. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Austin JL, Lee MG, Thibeault MD, Carr JE, Bailey JS. Effects of guided notes on university students’ responding and recall of information. Journal of Behavioral Education. 2002;11(4):243–254. doi: 10.1023/A:1021110922552. [DOI] [Google Scholar]
  4. Axe JB, Phelan SH, Irwin CL. Empirical evaluations of Skinner’s analysis of problem solving. The Analysis of Verbal Behavior. 2019;35(1):39–56. doi: 10.1007/s40616-018-0103-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Donahoe JW, Palmer DC. Learning and complex behavior. Allyn & Bacon; 1994. [Google Scholar]
  6. Fienup DM, Shelvin KH, Doepke K. Increasing recall of information of children diagnosed with Asperger’s Syndrome: Utilization of visual strategies. Research in Autism Spectrum Disorders. 2013;7(12):1647–1652. doi: 10.1016/j.rasd.2013.09.015. [DOI] [Google Scholar]
  7. Kisamore AN, Carr JE, LeBlanc LA. Training preschool children to use visual imagining as a problem-solving strategy for complex categorization tasks. Journal of Applied Behavior Analysis. 2011;44(2):255–278. doi: 10.1901/jaba.2011.44-255. [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Kritch KM, Bostow DE. Verbal responses to past events: Intraverbal relations, or tacts to private events? The Analysis Verbal Behavior. 1993;11:1–7. doi: 10.1007/BF03392882. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. LeBlanc, L. A., Miguel, C. F., Cummings, A. R., Goldsmith, T. R., & Carr, J. E. (2003). The effects of three stimulus-equivalence testing conditions on emergent US geography relations of children diagnosed with autism. Behavioral Interventions, 18(4), 279–289. 10.1002/bin.144
  10. Mann CC, Karsten AM. Efficacy and social validity of procedures for improving conversational skills of college students with autism. Journal of Applied Behavior Analysis. 2020;53:402–421. doi: 10.1002/jaba.600. [DOI] [PubMed] [Google Scholar]
  11. Palmer DC. A behavioral interpretation of memory. In: Hayes LG, Chase PN, editors. Dialogues on verbal behavior. Context Press; 1991. pp. 261–279. [Google Scholar]
  12. Park H, Gaylord-Ross R. A problem-solving approach to social skills training in employment settings with mentally retarded youth. Journal of Applied Behavior Analysis. 1989;22(4):373–380. doi: 10.1901/jaba.1989.22-373. [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Root WB, Rehfeldt RA, Castro M. Some refinements of a methodology for examining the influence of overt and covert self-rules on task completion. Journal of Developmental and Physical Disabilities. 2017;29(3):475–488. doi: 10.1007/s10882-017-9538-0. [DOI] [Google Scholar]
  14. Sautter RA, LeBlanc LA, Jay AA, Goldsmith TR, Carr JE. The role of problem solving in complex intraverbal repertoires. Journal of Applied Behavior Analysis. 2011;44(2):227–244. doi: 10.1901/jaba.2011.44-227. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Shillingsburg MA, Cariveau T, Talmadge B, Frampton S. A preliminary analysis of procedures to teach children with autism to report past behavior. The Analysis of Verbal Behavior. 2017;33(2):275–282. doi: 10.1007/s40616-017-0085-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Shillingsburg A, Marya V, Bartlett B, Thompson T, Walters D. Teaching children with autism spectrum disorder to report past behavior with the use of a speech-generating device. The Analysis of Verbal Behavior. 2019;35(2):258–269. doi: 10.1007/s40616-019-00112-2. [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Skinner BF. Science and human behavior. Macmillan; 1953. [Google Scholar]
  18. Skinner BF. Verbal behavior. Prentice Hall; 1957. [Google Scholar]
  19. Skinner BF. The technology of teaching. Appleton-Century-Crofts; 1968. [Google Scholar]
  20. Sundberg, M. L. (2014). Verbal Behavior Milestones Assessment and Placement Program: The VB-MAPP. AVB Press. [DOI] [PMC free article] [PubMed]
  21. Thorley, C. (2016). Note taking and note reviewing enhance jurors’ recall of trial information: Juror memory. Applied Cognitive Psychology, 30(5), 655–663. 10.1002/acp.3240

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Data Availability Statement

Data sharing not applicable to this article as no datasets were generated or analyzed during the current study

Articles from The Analysis of Verbal Behavior are provided here courtesy of Association for Behavior Analysis International

RESOURCES