Agentic education · Learning science

Actionable companion · Cognitive load & attention

Learning with agency: The effects of using an auto-subtitle system in educational videos to facilitate learning for secondary school students: learning comprehension, cognitive load, and satisfaction

AI-assisted editorial implementation guide, separate from the original research. Proposed actions are not claims or recommendations attributed to the source authors. No independent peer review of this guide is asserted.

Read original paperSearch paper agentCC BY 4.0

Evidence and context

Title and abstract screening; stage labels describe reported settings, not verified numerical ages. Design labels require full-text appraisal. Reported stage(s): Secondary school. Verify the participant population and task before transferring the findings.

Original authors’ abstract, reproduced under the linked source license:

Abstract While subtitles are considered a primary learning support tool for people who cannot understand video narration in foreign languages, recent advancements in artificial intelligence (AI) technologies have played a pivotal role in automatic subtitling on online video platforms such as YouTube. This study examines the effects of three different types of subtitles in the Thai language (i.e., auto-subtitles, edited subtitles, and no subtitles) on learning comprehension, cognitive load, and satisfaction to determine whether it is feasible to use AI technology as an auto-subtitles system to facilitate online learning with educational videos. To that aim, 79 Thai secondary school students from three Mathayom 5 (Grade 11) computer science classrooms participated in this study. This study used the static group comparison, which is the Posttest-Only Control Group Design. The results of this study found that the auto-subtitles system that generates Thai language subtitles for English educational videos has greater feasibility of implementation to facilitate online learning when compared to editorial subtitles by Thai natives. Therefore, Thai subtitles generated by the auto-subtitles system in English educational videos can facilitate students’ learning comprehension, cognitive load, and satisfaction. Keywords: Auto-subtitles, Subtitles, Educational videos, Learning comprehension, Cognitive load, Satisfaction

What kind of evidence is this?

Supporting AI learning science · Comparative / pre-post study reported

Stage: Secondary school. Screening labels describe reported context and study-design language. They are not a risk-of-bias rating or proof of effectiveness for an age band.

Study methods · source passage

Students who participated in this experiment were 79 students in grade 11 (Mathayom 5) from the secondary demonstration school under the faculty of education of a public university in Bangkok. All of the students studied science and mathematics programmes as well as distributed levels of educational achievement in each classroom. The demographic characteristics of participants in terms of age, gender, and the number of students in each experiment group are shown in Table 2. This study used stratified sampling (Cohen et al., 2007) to assign a single treatment to each of the three classrooms. There were two groups of experimental treatments (i.e., 1 with auto-subtitles and 1 with edited subtitles) and a single control group with no subtitles. Teaching was conducted during regular teaching hours. All participants in this research project acknowledged the guideline of the experiment and filled out a consent form with authorisation from their guardians. According to the experiment constructed on the actual date and time of the class, the experiment was blended into the classroom’s learning activities as Classroom Action Research (CAR). All data collected were highly protected and confidential. The participants’ privacy was maintained by replacing classroom numbers with A–C and applying for student orders instead of student identification numbers.

Limitations and uncertainty · source passage

This study had limitations related to the sample population. During the Covid-19 pandemic recruiting samples from schools ready for online learning and experimentation learning activities is a concern. In particular, it must include the requirements of schools and participants equipped with devices and network connections. As a result, the experiment of the sample group was limited to only one school in Thailand’s capital city, Bangkok. The sample group contained students with high background knowledge and advanced English skills due to social context. Moreover, the sample size was small, and the experimental period was limited. Plus, it was necessary to remove some students’ records who did not participate in both trials due to the students’ leaving the class for vaccination on the day of the experiment.

Design the system before the pilot

Use this study to inform a learning mechanism or implementation question. It is not direct evidence that autonomous agents improve learning. Keep the proposed agent behavior separate from what the source actually tested.

  • Limit simultaneous instructions and unnecessary dialogue. Compare task performance and a suitable mental-effort measure; easier-feeling work is not automatically better learning.

A 90-day district pilot

  1. Days 1–15 · Appraise. Have the curriculum lead and a teacher read the methods and limitations. Record the study design, sample, comparator, outcomes and evidence gaps. Define a local question that the study can help inform.
  2. Days 16–30 · Design. Ask learners to attempt a problem unaided, compare an agent hint, explain a revision and verify the final reasoning against a trusted source. Make the permitted use explicit for each assignment. Name an accountable owner and set the baseline, success threshold and budget before implementation.
  3. Days 31–60 · Test. Start with a small voluntary cohort, preserve an accessible non-AI route and require teacher review. Use synthetic or de-identified examples; approve data handling and tool terms through district processes.
  4. Days 61–90 · Decide. Measure unaided transfer and delayed retention, calibration of confidence and the ability to identify an incorrect AI explanation. Record whether help-seeking becomes answer copying. Compare with baseline, document uncertainty and stop or revise if learning worsens, access is unequal or critical errors persist.

The board decision brief

Report the learning problem, evidence quality, participating population, total cost including review time, measured benefit, errors, access differences and remaining uncertainties. Make a dated recommendation to stop, revise or expand. A positive perception survey alone does not establish learning improvement.

Attribution and reuse

Sivakorn Malakul; Innwoo Park. The effects of using an auto-subtitle system in educational videos to facilitate learning for secondary school students: learning comprehension, cognitive load, and satisfaction. Smart Learning Environments (2023-01-10). https://doi.org/10.1186/s40561-023-00224-2

Companion prepared 2026-10-30. Original authorship and DOI are preserved. The abstract is an attributed source extract; the implementation plan is new AI-assisted editorial material, offered under the same CC BY 4.0 license. This guide has not undergone independent peer review.