The design of ScienceAR is grounded in four complementary perspectives. Constructivism posits that learners build understanding through active manipulation of information; AR operationalizes this by enabling students to rotate, zoom, and inspect biological structures (Akçayir and Akçayir, 2017). Cognitive Load Theory suggests that learning improves when extraneous load is minimized; AR reduces split attention by co-locating text and diagrams, thus supporting germane processing (Gómez-Ríos et al., 2025). Cognitive Theory of Multimedia Learning emphasizes dual-channel processing and contiguity; AR integrates words, graphics, and spatial cues, strengthening the integration of verbal and visual codes (Garzón et al., 2019). Finally, Self-Determination Theory indicates that fostering competence and autonomy enhances motivation; AR promotes both through interactivity and learner control (Amores-Valencia et al., 2023). Drawing on these frameworks, we anticipated that AR-supported instruction would (i) improve post-test performance relative to traditional methods, (ii) enhance engagement and motivation, and (iii) be perceived by teachers and students as clearer and more enjoyable.

We conducted a classroom survey to assess students' enjoyment and satisfaction with existing teaching methods. The survey questionnaire specifically assessed the aspects: participation in class discussions, student satisfaction with the teaching method, degree of student confidence, degree of student learning in biology is fun and exciting, and students contact students for clarification before answering. The survey results informed our decision to introduce AR technology as a new teaching approach. So, we decided to go with AR Technology. The survey results are shown in Figure 4.

Continuing the first survey, we conducted a second one in the classroom to extract students' perceptions of the difficulty levels of various course topics. This survey aimed to go beyond traditional exams by immersing ourselves in the different experiences and hurdles students encounter in their academic journey. The questionnaire was carefully created with diverse questions designed to get real answers to the most challenging topics. The survey consists of the most difficult issues of the biology book, such as botany, zoology, microbiology, biochemistry, and human anatomy. This survey aimed to explore the complexity of students‘ experiences and perspectives on these different but related topics.

The survey revealed clear trends regarding which biology topics students found most challenging. Human Anatomy was reported as the most difficult (55%), followed by Genetics (15%), Evolution (15%), Cell Biology (8%), and other topics (7%). These results highlight the areas where students struggle the most and guide the selection of content for AR enhancement. By involving students directly in identifying complex topics, the survey ensured that the intervention was both relevant and responsive to their learning needs. The results are shown in Figure 5.

Based on the survey results, we developed the ScienceAR Android application using Unity 3D and the Vuforia framework. The app integrated 3D models aligned with students' identified learning difficulties, enabling seamless navigation and real-time interaction with biological structures. The user interface was designed to be simple, accessible, and engaging for secondary school students. This immersive application allowed learners to visualize and manipulate complex biological content directly within their classroom lessons. A real-time classroom implementation is shown in Figure 6.

In the 9th-grade Biology course, students were divided into two groups and instructed to participate individually. Group B received lectures on selected topics through traditional methods, while the other group experienced the duplicate content delivered via Augmented Reality (AR) technology. We went to Lahore, the capital city of the province of Punjab, to experiment at City District Girls High School, Factory Area, Shahdra, Lahore, with the permission of both parents and the school administration. The school comprised seven sections for the students of the 9th class, designated as A, B, C, D, E, F, and G. A total of 421 students were enrolled across these sections, with an average of 60 students per section. To experiment, 60 students were randomly selected from different sections, forming two groups: a controlled group and an experimental group, each consisting of 30 students. Each student was assigned a unique identifier number as a primary key to facilitate tracking and record-keeping. After the first two surveys, the authors collaborated with classroom teachers to create a comprehensive experiment to assess the impact of incorporating augmented reality (AR) into their classrooms. The pre- and post-tests were 20-item paper-based assessments including multiple-choice, short-answer, and diagram-labeling questions covering the biology topics taught in the intervention. Example items included “Label the main chambers of the human heart” and “Which organelle produces cellular energy?” Surveys used 5-point Likert scales to measure engagement, confidence, enjoyment, and motivation. The complete pre- and post-surveys, topic difficulty questionnaire, and full 29-item biology test are provided as Supplementary material (S1–S4). The test was specifically constructed to measure several aspects of using AR technology, e.g., students' grasp of complex concepts and problem-solving skills. The Author collaborated with the classroom in charge to ensure a seamless alignment with the curriculum content, designed to meet the learning outcomes and objectives, and paired with the AR-enhanced approach to learning. The comprehensive nature of the test allows for a detailed analysis of AR's effectiveness in increasing student learning and academic achievement in specific areas. During the test, the classroom environment is shown in Figure 7.

The test results showed that the students in the experimental group performed better. The students in the control group obtained an average of 76.091% and those in the experimental group 81.034%. This difference represents a 5% increase in scores for the experimental group compared to the control group. The improvements observed in the test groups and the results show the benefit of the reality technology (AR) for the students. Class test findings support the effectiveness of incorporating augmented reality (AR) into classroom environments, indicating its ability to impact learning outcomes in the areas studied positively. Test results are graphically shown in Figures 8, 9.

Figure 8 shows that the experimental group achieved higher average post-test scores than the control group (81.0% vs. 76.1%). This represents a meaningful 5% improvement, confirming the effectiveness of AR-based lessons in enhancing academic performance. Figure 9 illustrates individual student scores for both groups. The distribution shows that while some variability exists, the experimental group consistently achieved higher marks, further supporting the positive effect of the AR intervention.

To assess the effectiveness of the ScienceAR application in enhancing student learning outcomes, an independent-samples t-test was conducted comparing the post-intervention scores of the experimental group (n = 30), who used the AR app, and the control group (n = 30), who received traditional instruction. Descriptive statistics (means, SDs, minimum and maximum values) for both groups are shown in Table 1. The analysis revealed that the experimental group scored significantly higher (M = 23.50, SD = 2.74) than the control group (M = 22.07, SD = 1.89), t₍₅₈₎ = 2.36, p = 0.022. The effect size was moderate (Cohen's d = 0.61), with a 95% confidence interval for the mean difference ranging from 0.22 to 2.65. These results provide clear evidence that using the ScienceAR application led to a meaningful improvement in academic performance, supporting the effectiveness of augmented reality tools in enhancing engagement and conceptual understanding in secondary biology. As summarized in Table 1, AR-supported lessons produced higher post-test scores than traditional instruction.

After successfully implementing augmented reality (AR) technology in the classroom, we conducted a post-test evaluation based on the enjoyment questionnaire and a satisfaction survey 1 to analyse its impact on the learning environment. This post-test evaluation, which included a structured and comprehensive questionnaire, aimed to provide a helpful definition of the quality aspects of an AR-enhanced learning environment. Data collected from sample responses will improve training methods and optimize the integration of AR technology to improve overall training. Survey results are shown in Figure 10. Results indicates that students rated AR lessons highly across all categories. Mean Likert scores exceeded 4.0 for enjoyment, comprehension, and motivation, demonstrating that AR was perceived as both engaging and effective.

We compared pre- and post-intervention survey scores for engagement, confidence, enjoyment, and motivation. These indicators capture how actively students participated in lessons and how their perceptions of learning changed after the AR intervention. Results are shown in Figure 11. Results demonstrates consistent improvements across all survey domains following AR implementation. Increases were most notable in student confidence and classroom engagement, highlighting the broader benefits of AR beyond academic test scores.

Quantitative results demonstrated that students who used ScienceAR outperformed their peers in the control group on post-test assessments. Although the difference was modest in absolute terms, it was statistically significant and reinforces earlier studies reporting that AR enhances student motivation, manages cognitive load more effectively, and strengthens conceptual comprehension (Fidan and Tuncel, 2019; Garzón et al., 2019). Survey results confirmed that learners perceived AR-based lessons as more engaging, motivating, and enjoyable than traditional instruction, which is consistent with previous findings that immersive learning environments foster more positive student attitudes (Geng and Yamada, 2023). Teacher observations provided additional insights into classroom dynamics. Instructors reported increased attentiveness, greater participation, and more frequent student questioning during AR lessons. These behavioral changes indicate that AR can transform classroom culture by encouraging active involvement rather than passive reception. From a theoretical perspective, these outcomes align closely with constructivist learning theory, which emphasizes that knowledge is most effectively acquired through active engagement and meaningful manipulation of information. AR enables learners to interact with biological systems in ways that promote deeper conceptual understanding.

The results also raise important questions about the underlying mechanisms driving improvements in learning outcomes. Three potential explanations should be considered: novelty, interactivity, and visualization. First, novelty effects may contribute to short-term motivation, as students are often enthusiastic when introduced to new technologies. However, novelty alone is unlikely to sustain learning gains beyond initial exposure. Second, interactivity plays a key role by giving learners agency over how they explore content. This autonomy can increase motivation and support the development of spatial reasoning. Third, visualization is critical because it transforms abstract and often invisible processes into concrete, manipulable models. The combination of interactivity and visualization appears to be the most powerful factor in facilitating understanding, since students reported that AR made complex topics clearer and easier to grasp. Future studies could investigate these mechanisms more systematically, for example by comparing AR with static 3D images or videos to isolate the effects of interactivity from novelty and visualization.

While this study focused on student outcomes, the role of teachers in implementing AR is equally important. Successful integration requires not only access to technology but also adequate teacher preparation. Teachers must be able to align AR activities with curriculum objectives, manage classroom pacing, and troubleshoot technical issues. In low-resource settings, challenges such as device scarcity, limited lesson time, and lack of professional development opportunities present additional barriers. Providing structured training that includes demonstrations, co-teaching models, and opportunities for teachers to design their own AR-enhanced lessons will be critical for sustainable adoption. Without teacher readiness, AR risks being perceived as a novelty rather than a serious pedagogical tool.

The successful deployment of ScienceAR in a resource-limited school demonstrates that AR can be feasible with modest infrastructure, such as low-cost smartphones and printed AR markers. This is an encouraging finding, yet broader issues of equity and accessibility remain. Many schools face limitations in terms of device availability, internet connectivity, and digital literacy. These factors contribute to the digital divide, where some students benefit from technological innovations while others are left behind. To address this, AR applications should be designed to function offline, be optimized for low-spec devices, and include features that support accessibility such as clear labels, high contrast visuals, and audio narration. Incorporating universal design principles can help ensure that AR is inclusive for students with diverse needs and abilities.

Although the results of this study are encouraging, several limitations should be acknowledged. First, the intervention involved a relatively small sample of sixty students from a single school, which limits the generalizability of the findings. Future research should involve larger and more diverse populations across multiple schools and regions. Second, the duration of the intervention was limited to seven days. While measurable improvements were observed in this short period, it is not clear whether the gains would persist over time or whether novelty effects contributed to the outcomes. Longitudinal studies are needed to assess the durability of learning benefits and their impact on retention. Third, this study was restricted to biology content, particularly human anatomy. Expanding AR integration into other science subjects would provide a more comprehensive picture of its cross-disciplinary effectiveness.

Future studies should adopt more comprehensive research designs. Longitudinal investigations that include delayed post-tests could measure retention and help clarify whether AR provides lasting benefits beyond the intervention period. Comparative designs that include video-based 3D instruction or static 3D images could help isolate the unique contributions of AR interactivity. In addition, multi-site studies involving schools from diverse regions would allow for stronger generalization of results. Research should also explore differential impacts across gender, socio-economic status, and prior academic achievement, since AR may benefit some groups more than others. Finally, investigating teacher training strategies and the scalability of AR in national curricula would provide practical insights for policymakers and educators.

Although this study focused on biology, the implications of AR extend to other STEM disciplines. In chemistry, AR could allow students to visualize molecular structures and reaction pathways in three dimensions. In physics, it could help learners explore forces, fields, and circuits through interactive simulations. In mathematics, AR can provide spatial visualizations of geometric transformations or three-dimensional graphs. These applications would build on the same theoretical foundations of constructivism, cognitive load management, and multimedia learning theory, making AR a versatile tool across the STEM curriculum.