By Matt Harrington, BCBA · Behaviorist Book Club · April 2026 · 12 min read

Teaching Children with Autism to Mand for Answers: Acquisition of Intraverbal Behavior Through Information-Seeking Responses

Overview & Clinical Significance

The ability to seek information by asking questions is a pivotal language skill that has profound implications for learning, social interaction, and independence. For children with autism spectrum disorder, deficits in information-seeking behavior represent a significant barrier to natural learning opportunities that typically developing peers access routinely throughout their day. When a child can spontaneously ask questions to fill gaps in their knowledge, they gain access to an ongoing, self-directed learning mechanism that extends far beyond the boundaries of structured instruction.

This course examines research by Ingvarsson and Hollobaugh (2010) that investigated a specific teaching procedure for developing information-seeking behavior in children with autism. The study focused on teaching four boys with autism to use the response "I don't know, please tell me" when presented with questions they could not answer, rather than engaging in error responses or escape behavior. This seemingly simple response represents a sophisticated verbal behavior that connects the mand (requesting information) to the subsequent acquisition of intraverbal behavior (answering the same questions correctly after receiving the information).

The clinical significance of this research lies in its demonstration of a practical, replicable procedure for teaching a skill that has far-reaching implications for learner independence. Traditional approaches to teaching intraverbal behavior typically involve direct instruction of specific question-answer pairs, which can be effective but does not teach the learner the general skill of seeking information when they encounter gaps in their knowledge. By teaching the mand for information as a functional prerequisite to intraverbal acquisition, this approach creates a generative learning mechanism.

The procedures used in the study, echoic prompting and constant prompt delay, are well-established teaching strategies in the behavior analytic repertoire. What makes this research particularly valuable is the innovative application of these familiar procedures to a target behavior that bridges two verbal operant classes: the mand (driven by the motivating operation of not knowing the answer) and the intraverbal (controlled by the verbal antecedent of the question). Understanding this relationship between verbal operants has important implications for how clinicians conceptualize and sequence language instruction for children with autism.

For practitioners working with children who have emerging language skills, this research provides a direct, evidence-based protocol that can be incorporated into skill acquisition programming. The study's demonstration that teaching the mand for information facilitated subsequent intraverbal acquisition suggests that investing in information-seeking behavior may be one of the most efficient ways to accelerate language development.

Background & Context

The theoretical foundation for this research lies in Skinner's analysis of verbal behavior, particularly the distinction between different verbal operant classes and the conditions under which they are acquired. In Skinner's framework, a mand is a verbal response that is controlled by a motivating operation and reinforced by a specific consequence related to that motivation. When a child encounters a question they cannot answer, the motivating operation of not knowing the answer creates the conditions for a mand for information, a request that the information be provided.

Intraverbal behavior, by contrast, is a verbal response controlled by a preceding verbal stimulus where the response does not have point-to-point correspondence with that stimulus. Answering questions is one of the most common and functionally important forms of intraverbal behavior. A child who can answer "What do you eat for breakfast?" with "Cereal" is demonstrating intraverbal control, where the verbal antecedent evokes a related but non-imitative verbal response.

The relationship between manding for information and intraverbal acquisition had been theorized but required empirical demonstration. The logic is straightforward: when a child mands for information by saying "I don't know, please tell me," the answer they receive functions as a learning opportunity. The child hears the correct answer in the context of the question, creating conditions for the intraverbal relationship to be established. Subsequent presentations of the same question may then evoke the correct intraverbal response without the need for the mand.

Echoic prompting, one of the two key procedures used in the study, involves providing a verbal model that the learner imitates. In this context, the therapist would model "I don't know, please tell me" after presenting a question the child could not answer, and the child would repeat this phrase. Constant prompt delay involves initially providing the prompt with zero delay, then introducing a fixed delay between the discriminative stimulus and the prompt, giving the learner an opportunity to respond independently before the prompt is delivered.

Prior to this research, many language programs for children with autism focused on building intraverbal repertoires through direct instruction of specific question-answer pairs. While effective, this approach has limitations in that each new piece of information requires separate teaching episodes. The mand-for-information approach offers a more efficient alternative by teaching a generalizable skill that the child can use across all unfamiliar questions, essentially teaching them how to learn rather than what to learn.

The study's participants were four boys with autism who had echoic repertoires but limited intraverbal skills, representing a population for whom this intervention was particularly appropriate. The inclusion criteria ensured that participants could imitate verbal models, which was necessary for the echoic prompting component of the procedure, but could not yet answer the target questions, providing a clear demonstration of the mand-to-intraverbal acquisition pathway.

Clinical Implications

The clinical implications of teaching children to mand for answers extend across multiple domains of language development and functional independence. Most immediately, this procedure addresses a common clinical challenge: children with autism who respond to unknown questions with errors, echolalia, silence, or escape behavior rather than seeking the needed information. Each of these alternative responses represents a missed learning opportunity and may, in the case of error responses, actually interfere with correct learning by strengthening incorrect response patterns.

The efficiency gains from teaching information-seeking behavior can be substantial. Rather than directly teaching each question-answer pair through discrete trial instruction, which requires the clinician to identify and program for each specific piece of information, the mand for information creates a mechanism through which the child acquires new information through naturally occurring interactions. Once the child reliably uses the information-seeking response, every question they encounter becomes a potential learning trial, dramatically expanding the opportunities for intraverbal acquisition.

Generalization is a critical consideration for this procedure. The study demonstrated that the mand for information generalized to novel questions that were not directly trained, which is the key finding that supports the procedure's utility as a generative language strategy. If the response only occurred for trained questions, its clinical value would be limited. The generalization to novel questions suggests that the children learned a response class, seeking information when they do not know an answer, rather than a specific stimulus-response relationship.

For treatment planning, this research suggests that the mand for information should be considered as a foundational target for children who have echoic repertoires but limited intraverbal skills. Rather than beginning intraverbal instruction by directly teaching specific question-answer relationships, starting with the information-seeking response creates a tool that the child can use to acquire intraverbal behavior more independently. This sequencing decision has implications for how language programs are structured and how treatment time is allocated.

The procedure also has implications for naturalistic teaching approaches. Once a child has learned to mand for information in a structured context, the response can be programmed for generalization to natural environments where questions arise incidentally. Caregivers, teachers, and peers can be oriented to respond to the child's information-seeking behavior by providing the requested information, creating a natural teaching loop that does not require clinician presence.

For children who already have some intraverbal skills but who respond with errors rather than information-seeking when they encounter unknown questions, this procedure can be adapted to teach discrimination between questions they can answer and questions they cannot. This discrimination is itself an important skill that supports accurate responding and efficient learning.

Ethical Considerations

The ethical considerations surrounding the teaching of information-seeking behavior in children with autism touch on several important dimensions of the BACB Ethics Code (2022). Code 2.14 (Selecting, Designing, and Implementing Behavior-Change Interventions) requires that interventions be based on the best available scientific evidence and tailored to the individual client's needs. The research base supporting the mand-for-information procedure provides the evidence foundation, but practitioners must evaluate whether each individual client has the prerequisite skills, particularly an echoic repertoire, to benefit from this approach.

Code 2.01 (Providing Effective Treatment) is supported by the selection of targets that have broad functional significance. Teaching a child to seek information when they do not know an answer is a skill that has immediate practical utility and that facilitates the acquisition of an extensive range of subsequent skills. From an ethical standpoint, prioritizing foundational skills that create generative learning mechanisms represents a responsible use of treatment time and resources.

The concept of social validity, while not a specific code, is embedded in the ethical framework of behavior analysis. The ability to ask questions when confused or uninformed is a highly socially valid skill that is expected and valued across all social contexts. Teaching this skill supports the child's inclusion in educational and social environments and contributes to their independence and self-determination. The social validity of this target provides additional ethical justification for its prioritization.

Code 2.18 (Providing Continuation, Modification, or Discontinuation of Services) requires that practitioners make data-based decisions about treatment modifications. When implementing the mand-for-information procedure, practitioners should collect data on both the mand response and the subsequent intraverbal acquisition to evaluate whether the procedure is producing the expected learning pathway. If data show that a child is manding for information but not acquiring the intraverbal responses, the procedure may need to be modified or supplemented with additional teaching strategies.

There are also ethical considerations related to the child's autonomy and self-determination. Teaching information-seeking behavior is fundamentally about empowering the learner to take control of their own learning process. This aligns with the broader ethical principle that treatment should promote independence and autonomy rather than creating dependence on clinician-directed instruction. A child who can independently seek information is less reliant on structured teaching environments and more capable of learning through natural interactions.

Code 1.06 (Being Knowledgeable) requires practitioners to stay current with the literature. The research on manding for information has continued to develop since the initial studies, with additional investigations examining variations in procedures, populations, and contexts. Practitioners who implement this approach should be familiar with the current state of the evidence, including any limitations or boundary conditions that have been identified in subsequent research.

Assessment & Decision-Making

Before implementing the mand-for-information procedure, a thorough assessment of the learner's current verbal repertoire is essential. The critical prerequisite is an echoic repertoire, the ability to imitate verbal models with reasonable accuracy. Without this skill, the echoic prompting component of the procedure cannot be implemented effectively. Assessment of echoic skills should include evaluation of the learner's ability to repeat phrases of the appropriate length and complexity, as "I don't know, please tell me" or a similar functional equivalent requires multi-word imitation.

The learner's existing intraverbal repertoire should also be assessed to identify the appropriate starting point for instruction. This assessment serves dual purposes: it identifies questions that the learner can already answer, which will serve as known items in the discrimination between known and unknown questions, and it identifies questions that the learner cannot answer, which will serve as the targets for the mand-for-information procedure. A mix of easy and challenging questions is useful for building discrimination between when to answer and when to seek information.

Motivating operations for information-seeking must be considered in the assessment. The procedure is most effective when the learner has a reason to want to know the answer. Questions related to preferred activities, upcoming events, or topics of interest may create stronger motivating operations for seeking information than arbitrary factual questions. The initial question sets should be selected to maximize the motivation to learn the answers.

The assessment should also evaluate the learner's current response to unknown questions. Some children may sit silently, others may echo the question, others may produce a stereotyped error response, and others may engage in escape behavior. Understanding the current response pattern informs both the teaching procedure and the expectations for how quickly the mand for information will replace existing responses.

Decision-making during implementation should be guided by ongoing data collection on multiple dependent variables. Track the frequency and latency of the mand-for-information response, the accuracy of the echoic following the prompt, the number of trials to acquisition for each intraverbal target, and any generalization to novel questions. These data allow the practitioner to evaluate whether the procedure is working as expected and to make modifications when needed.

If the mand is occurring but intraverbal acquisition is not following, several modifications should be considered. The number of presentations of the information after the mand may need to be increased. The format of the information presentation may need to be modified. Additional practice opportunities or supplementary teaching procedures may need to be incorporated. If the mand itself is not being acquired, the prompting procedure may need to be adjusted, the motivating operation may need to be strengthened, or the response form may need to be simplified.

The decision about when to fade the echoic prompt for the mand should be based on data showing consistent use of the mand following zero-delay prompts. The constant prompt delay is then introduced, typically at two to five seconds, and the learner's independent use of the mand before the prompt is tracked. Mastery criteria should be established for both the independent mand and the subsequent intraverbal acquisition.

What This Means for Your Practice

This research provides a practical, evidence-based procedure that can be directly incorporated into language programming for children with autism. If you work with children who have echoic repertoires but limited intraverbal skills, consider adding the mand for information as a foundational target. The procedure is straightforward to implement and uses teaching strategies, echoic prompting and prompt delay, that are already familiar to most ABA practitioners.

When implementing the procedure, start with a small set of questions related to topics that are motivating for the learner. Present the question, wait for the response, and if the child does not answer correctly, provide the echoic prompt for the information-seeking response. After the child echoes the prompt, immediately provide the answer and present the question again to test for intraverbal acquisition. Gradually introduce the prompt delay to promote independent use of the mand.

Program for generalization from the outset. Once the child demonstrates the mand reliably in structured instruction, create opportunities in less structured contexts where the child encounters questions they cannot answer. Prompt the mand if needed and reinforce its use by providing the requested information. Train caregivers and other team members to recognize and respond to the child's information-seeking behavior in natural contexts.

Monitor both the mand and the subsequent intraverbal acquisition through data collection. If the child is using the mand but not retaining the information provided, consider whether additional repetitions, varied contexts, or supplementary teaching procedures are needed. The mand is a means to an end, and the ultimate goal is the acquisition of intraverbal behavior.

Finally, consider how this procedure fits within the broader language program for each learner. The mand for information is not a replacement for direct intraverbal instruction but rather a complementary approach that teaches the child a generalizable strategy for acquiring new information. Integrating both approaches provides the child with multiple pathways for language development.