Abstract
Research on science, technology, engineering, and mathematics (STEM) education for students with neurodevelopmental disorders (NDD) focuses on vocabulary and content knowledge, often overlooking science practices and extended oral scientific skills. Using a single-case design, this study examined the efficacy of a video-enhanced intervention package to teach science practices to two students with NDD. In a secondary analysis, students’ scientific utterances were measured using event recording focusing on mean utterance length and the use of scientific words. The intervention package was effective in teaching science practice skills and extended oral scientific skills. The results inform practice implications and future research directions.
Keywords
Introduction
Promoting students’ science, technology, engineering, and mathematics (STEM) literacy has been at the forefront of educational interest across Ontario (Ontario Ministry of Education, 2022). Most STEM-related educational strategies target neurotypical students, leaving students with neurodevelopmental disorders (NDD) without differentiated support to engage with STEM curricula (Jenson et al., 2024; Therrien et al., 2011). Approximately 5% of Canadian students aged 5 to 14 years have a disability, 74% of whom have an NDD such as autism, attention deficit hyperactivity disorder (ADHD), intellectual disability (ID), and specific learning disabilities (Miller et al., 2013). Research from a Canadian context guiding equitable access to STEM learning for these students remains limited (Ariza & Hernández, 2025). The Ontario Ministry of Education emphasizes an integrated approach to STEM education, with equal representation of science, technology, engineering and mathematics (Ontario Ministry of Education, 2022). This approach equips students with real-world problem-solving skills (Youth Science Canada, 2011). For students with NDD, developing problem-solving skills has real-life implications that can be applied across home, school, and community contexts (Knight et al., 2020).
Inquiry-based learning is central to STEM education, allowing students to explore natural phenomena through experimentation (Youth Science Canada, 2011). Ontario’s STEM curriculum emphasizes hands-on experimentation and student-led learning, guided by the Smarter Science Framework (Ontario Ministry of Education, 2022; Youth Science Canada, 2011). The framework outlines key scientific practices: initiating and planning, performing and recording, analyzing and interpreting, and communicating (Youth Science Canada, 2011). Students with NDD often have challenges with cognitive demands and sequential problem-solving required for science practices and therefore benefit from additional support, such as structured guidance, to participate meaningfully in inquiry-based activities (Brigham et al., 2011; Jenson et al., 2024; Therrien et al., 2011).
Most research focuses on teaching content over science practice skills for students with NDD (Knight et al., 2020). Previous literature shows that systematic instruction is an evidence-based method for teaching science practices to students with NDD (Knight et al., 2020). Visual activities schedules (VAS) and tools such as the KWHL chart (i.e., what do you
Methods
Participants, Interventionist, and Setting
The study protocol was reviewed and approved by Western University Research Ethics Board. Three students from Southwestern Ontario aged 8 to 11 years with NDD who had Individualized Education Plans (IEP) at school were recruited. One participant dropped out before completing baseline.
Talia was a 9-year-old white female participant diagnosed with ADHD in the fourth grade who had a non-identified IEP. On the Domain Level Caregiver form of Vineland Adaptive Behavior Scale, Third Edition (VABS-3), Talia ranked in the 27th percentile in Communications, 19th for Daily Living Skills, and the 21st for Socialization, with an overall behavior composite in the 16th percentile. Talia had a prior interest in electricity and robotics.
Carlos was a 10-year-old white male participant diagnosed with autism and auditory processing disorder in the fifth grade. Carlos had an IEP under the identification of an intellectual exceptionality. On the Domain Level Caregiver form of the VABS-3, Carlos ranked in the 10th percentile in Communications, 55th for Daily Living Skills, and 39th for Socialization, with an overall adaptive behavior composite in the 25th percentile. According to his mother, Carlos found STEM challenging and he was often discouraged by complex tasks.
A Master of Arts student in School and Applied Child Psychology served as the primary interventionist and data collector. The study was conducted via Zoom.
Materials
Twenty-one integrated STEM lesson plans, based on the Ontario STEM curriculum and including the Smarter Science Framework science practices outlined in Table 1, were developed by a certified elementary school teacher completing a Master of Arts degree in Education. Materials for each STEM unit (e.g., Ozobots, electrical snap circuits, and building materials) were delivered to participants’ homes.
Target Skills, Definitions, and Corresponding Smarter Science Framework Science Practices.
Using Microsoft PowerPoint, personalized video-enhanced VAS were developed (e.g., Spriggs et al., 2015). The VAS included 10 slides each with a static picture of a target skill which could be clicked to play VBM clips. Each clip ranged from 5 to 10 s and depicted two child actors modeling the target skill with the skill written below (e.g., Step 3: State what you know). Three exemplars of each skill were created.
Dependent Variables
In single-case experimental designs, each participant serves as their own control, and baseline data collection functions as the primary method for establishing initial skill levels. In this study, we measured two key outcomes: (1) the percentage of independently completed science practice steps, and (2) oral scientific skills, including the number and length of utterances and use of scientific vocabulary.
STEM Practice Skills
The interventionist collected data on the percentage of independent, correct STEM practice steps completed in the task analysis. Responses were marked as correct if the student began the step within 5 s and incorrect if the student made an error or did not respond within 5 s.
Oral Science Skills
A research assistant recorded all phrases that reflected at least one of the Smarter Science Framework practices (see Table 1). She also recorded the total number of words per utterance and per session, and the number of utterances per session. Each occurrence of scientific vocabulary introduced as part of the lessons spoken by the student was also recorded.
Interobserver Agreement and Treatment Fidelity
To meet quality standards for single-case research, IOA and fidelity data were collected for more than 20% of sessions, as recommended by single-case design quality standards (Tate et al., 2015). These measures were collected for 35% of the sessions. A research assistant was trained to collect point-by-point inter-observer agreement (IOA) data and achieved a minimum of 90% reliability with the primary data collector before beginning. A different research assistant completed procedural fidelity using a checklist outlining the steps of intervention (i.e., setting up the environment, introducing the skill, playing the video model, waiting for a response, and delivering appropriate consequences). The scores were calculated by dividing the number of agreements by the number of agreements and disagreements and multiplying by 100. IOA scores below 90% were resolved via discussion. IOA for task analysis and oral science skills averaged 97% (range = 92%–100%) and 95% (range = 91%–100%) respectively, and treatment fidelity data averaged 99% (range = 98%–100%).
Experimental Design
Given the rarity of the population and the constraints imposed by the COVID-19 pandemic, a single-case design was selected as a methodologically appropriate and pragmatic approach to allow for experimental analysis while maintaining feasibility. A multiple-probe across-participants design was used to evaluate the efficacy of the video-enhanced intervention package in teaching target STEM practice and oral science skills. This design allows each participant to serve as their own control. The staggered introduction of the intervention across time and repeated replications of the effect at different time points minimizes threats, such as maturation, to internal validity (Ledford & Gast, 2024).
Procedures
Pre-Assessment Interview
Students who met inclusion criteria completed a 1-hr pre-assessment interview with caregivers which included a demographic questionnaire. Participants were asked to observe a short video to ensure they could attend to it for at least 20 s.
General
One-hour individual sessions were held approximately two times per week for each participant for approximately 4 months. All sessions began by reviewing the session rules. It followed with the interventionist stating, “Today we are learning about (unit concept),” followed by five short experiments. During all study phases, participants started each unit by watching a short video on the STEM concepts for the lessons.
Baseline
During baseline, the interventionist was present and interacting with participants in the same one-to-one format used during intervention, but without delivering the active components of the intervention package. The interventionist instructed the student to retrieve a specific STEM kit from their materials. The student was first given a chance to explore the materials unprompted, then the interventionist provided information for the experiment (e.g., “In this experiment, we are going to use the materials to build a structure to support a load of books”). Students were told to “try their best” to complete the experiment. If the participant did not respond within 5 s, the interventionist provided a verbal prompt (i.e., “What is next?”). No planned reinforcement was provided.
Video-Enhanced Visual Activity Schedule Intervention Package
During the intervention phase, the interventionist introduced the KWHL chart. Students retrieved the requested STEM kit and were shown the VAS with embedded VBM clips. As the student worked through each step in the VAS, the interventionist stated, “We are going to learn how to (target behaviour). Watch the video of the students showing you an example of how to do this skill, then it will be your turn.” Next, the facilitator contrived a situation to demonstrate the target skill. The interventionist provided social praise for correct responses and for incorrect answers, stating, “Nice try, next time remember to (target behaviour), just like they did in the video.” A system of least-to-most intrusive prompting was used. Fading procedures were planned to be applied systematically; however, participants self-faded the embedded VBM.
Post-Training Probes
Post-training probe sessions occurred after participants mastered the target practice skills (i.e., correctly demonstrating eight of the 10 target skills) using the video-enhanced intervention package. Participants used the video-enhanced VAS intervention package support in this phase as needed. The interventionist contrived situations to evoke the target behaviors and recorded participant responses.
Generalization and Maintenance Sessions
Generalization and maintenance were assessed at 3-week follow-up sessions with a new facilitator and novel experiments. Participants’ performances on target practice skills were recorded during intervention sessions. Generalization and maintenance sessions were conducted in the same manner as intervention sessions; participants could use the video-enhanced VAS as needed.
Social Validity
Social validity was assessed after the study was complete using four questions modified from previous research by Yakubova et al. (2020). A research assistant conducted semi-structured interviews with students and their caregivers (i.e., Did you like the activities in the study?; What did you like/What did you not like?; Was it easy to learn using the materials?; Would you like to use these strategies again?; Is there anything else you would like to tell us about your participation in the study?). Responses were analyzed using descriptive coding to summarize participants’ and caregivers’ reported likes, dislikes, and overall impressions of the intervention.
Results
Figure 1 displays the percentage of steps in the task analysis completed independently. The number of steps completed independently increased during intervention for both students (Talia: 92%–100%; Carlos: 89%–94%). Participants self-faded the VBM before fading procedures were implemented. During the second training session, Talia self-faded the VBM component of the intervention. During the third training session, Carlos self-faded the VBM clips, using it only if he required additional support. Both participants met mastery criteria in the third training session.

Percent correct of science practices and STEM knowledge.
During post-training, both participants used the static VAS to complete steps. Post-training performance remained stable (range = 92%–100%). Talia used it to support making predictions, planning an experiment, describing results, and stating what she learned. Carlos used the static VAS to support stating what he knew, asking questions, making a prediction, planning an experiment, describing results, and stating what he learned. Accuracy during maintenance and generalization remained high (range = 96%–100%).
Oral Science Skills
Figure 2 shows the number of utterances and the mean length of utterances (MLU) during baseline and intervention. The number of utterances increased from baseline to post-baseline. For Talia, they ranged from 6 to 14 during baseline and increased to 30 to 45 once the intervention was introduced. For Carlos, they ranged from 7 to 37 utterances during baseline to 34 to 50 post-baseline. The MLU was variable during baseline and intervention, ranging from 6.23 to 19.08 for Talia and 6.75 to 25.71 for Carlos.

Utterance measures.
Both participants uttered more new scientific words post-baseline than in baseline and most of the newly used words were reused. Talia used seven different scientific words in baseline. She uttered 31 more post-baseline (for a total of 38) and reused 63% of words. Carlos uttered 11 scientific words during baseline and 17 additional post-baseline (for a total of 29) and reused 55% of words.
Social Validity
The following outlines participants’ social validity responses. Talia enjoyed participating in the study stating that, “They [the experiments] were super fun. I liked the noisy things, like the snap circuits.” She indicated that she disliked the video models, as it made experiments longer and expressed a preference not to use them in the future. Her mother stated that “It was a great experience. Overall, anyone can benefit from this.” She noted an increased interest in Talia’s STEM hobbies but mentioned challenges with virtual sessions due to technological issues. Carlos enjoyed the study, especially the space unit, and found the VAS helpful for completing experiments stating, “It was easier to learn with it.” He also expressed wanting to use the VAS in the future. However, he preferred fewer sessions per week. Carlos’s mother said the reinforcement procedures were motivating for him, and “He liked using the visual schedule as he struggles with reading.”
Discussion
The current study examined the effectiveness of a video-enhanced VAS intervention package on the science practice and oral science skills for students with NDD. The intervention included a video-modeling VAS, KWHL chart, prompting hierarchy, and reinforcement procedures. Overall, we found that students in the study learned and applied science practice skills across STEM lessons while improving oral science skills. Importantly, both students and caregivers found the intervention package helpful in increasing STEM interest and completing experiments.
The intervention package was designed for differentiated support, offering different options for the different student learning needs. We investigated the intervention as a package, yet students varied in their reliance on the specific components. For example, Talia often used the written prompts to complete each step, whereas Carlos used the static photos and videos more frequently which may have been due to difficulties with reading. Spriggs et al. (2015) found that providing both static and embedded video modeling offered students varied levels of support as needed. In the current study, students did not engage with the KWHL chart, possibly because the instructor did not prompt its use. Other research supports knowledge charts combined with prompting and time delay (Knight et al., 2020). In the context of a classroom, the different components of the intervention package (i.e., verbal prompt, visual prompt of pictures and words, and video-model exemplar) can support different learners. In practice, it may be valuable for educators to do a pre-assessment before using the intervention package as a whole to individually tailor supports.
The intervention also improved oral science skills, including the number and length of scientific utterances. Students uttered more scientific vocabulary post-baseline, and most of the newly used scientific words were repeated at least once. The repeated words were often used in new contexts, where the students applied their previous knowledge to new concepts. The target skills included a communication component, such as asking questions, describing the materials and the results, stating what was known and what was learned, and making predictions. Students in the current study faced some difficulties formulating predictions, highlighting a potential area of further support. Previous research has shown that vocabulary cards and social scripts can support scientific communication, offering structured guidance for teachers with limited STEM experience and new teachers who may feel unprepared to support students with NDD (Knight et al., 2018).
This study has several limitations. Although the design met most quality indicators for sound methodological research (Cook et al., 2014), the lack of three-tier experimental effects limits the study’s generalizability. However, some authors suggest three demonstrations may be “overly stringent” (Lanovaz & Turgeon, 2020, p. 614). There was also a risk of bias due to a lack of blinding in collecting IOA and treatment fidelity and there were breaks between sessions for Talia due to sickness.
Looking ahead, future studies should evaluate the intervention package within an inclusive classroom setting where collaboration is essential to STEM learning (Youth Science Canada, 2011). Given that Talia and Carlos met mastery criteria in the first training session and self-faded the embedded video models in the second and third session, the intervention package may be feasible in a group setting such as classroom. Although some students may require intensive one-to-one support to learn the skills, future research could examine the package as a tier 2 intervention, such as with a group of 3 to 4 students and one educator or as a classroom wide intervention. Additional support could focus on emotional regulation and distress tolerance skills, as STEM often involves revisiting problems and learning from mistakes (Youth Science Canada, 2011). In the current study, Carlos responded well to prompting and reinforcement, although he expressed frustration with complex experiments. While research on social-emotional and STEM learning for neurotypical students exists (Garner et al., 2018), it is limited for students with NDD.
Overall, the findings offer promising implications for inclusive STEM education. VBM might be a practical and feasible way to support students in the classroom while reducing the demand for in vivo modeling on educators. Combining VAS and other prompting and reinforcement procedures could help students’ on-task behavior during sequential steps in learning (Knight et al., 2013). The intervention supports calls to action for research in STEM learning for students with NDD to include science practices and STEM content as an interdisciplinary whole (Knight et al., 2020).
Footnotes
Ethical Considerations
This study was reviewed and approved by Western University’s Non-Medical Research Ethics Board (NMREB) [Protocol Number 118771].
Consent to Participate
All participants’ caregivers provided written consent, and participants provided assent to take part in the study.
Funding
The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This research was supported by the Social Sciences and Humanities Research Council of Canada (SSHRC) Insight Development Grant [Grant Number 430-2018-00304].
Declaration of Conflicting Interests
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Data Availability Statement
This dataset is not available outside the institution where it is hosted. Requests to access this dataset should be directed to
