Religious diversity education: raising children's awareness of religious diversity through augmented reality

Taiwan is among the most religiously diverse regions globally, shaped by historical colonization and contemporary immigration, with a population practicing a wide range of religions. Multicultural education is seen as a way to mitigate conflicts arising from cultural misunderstandings by fostering respect, tolerance, and intercultural competence. Religious education in Taiwan has developed relatively late and faces challenges such as limited textbook coverage and rigid pedagogy that can dampen student interest and learning. Prior work suggests religious education should cultivate respect for diverse beliefs and that innovative, engaging methods are needed. Interactive technologies have improved learning in various domains but have seen limited application in religious education. This study asks whether integrating AR into religious diversity education for elementary students can improve learning outcomes compared to traditional paper-based materials, and examines students’ acceptance of the technology and their learning experiences.

Surveys indicate Taiwan hosts extensive religious plurality, with 22 major religions and high diversity. Multicultural and religious education can reduce prejudice and promote respect and tolerance (Kim, 2012; Moore, 2009; Ubani, 2013; White, 2004). In Taiwan, religious education has only recently been integrated into higher education and remains underdeveloped in elementary curricula, often limited in scope and employing rigid teaching methods that impede engagement (Hsu & Kao, 2011; Chien & Shiu, 2016). Alternative pedagogies, such as topic- and inclusion-based curricula (Lin, 2013), children’s literature (Green & Oldendorf, 2005), and theatre-in-education (Koukounaras-Liagis, 2011), have shown promise in shifting attitudes toward diversity. Interactive technologies have demonstrated effectiveness across disciplines in increasing interest, motivation, participation, and learning outcomes (e.g., Chen & Hsu, 2020; Sahin & Yilmaz, 2020; Teo et al., 2022; Li, Chen & Kang, 2022), yet applications in religious education are scarce. Mobile and AR-based learning in informal settings can support engagement and knowledge gains (Huizenga et al., 2009; Sandberg et al., 2011; Tosti et al., 2014; Wu et al., 2013). AR can provide immediacy and reduce cognitive overload in exhibitions (Aitamurto et al., 2018; Chang et al., 2013). This study addresses the gap by evaluating AR-assisted religious diversity education.

Design: Quasi-experimental comparative study using mixed methods during the 2020 Taipei Lantern Festival Religious Education Exhibition (Feb 8–16, 2020). Participants were randomly assigned to an experimental group using an AR system and a control group using a paper-based brochure. Assessments included a knowledge test (pre- and post-), a Generic Learning Outcomes (GLOs) questionnaire, open-ended questions for both groups, and a Technology Acceptance Model (TAM) questionnaire for the experimental group. Participants: N=50 elementary students (grades 3–4), proficient in traditional Chinese; 28 females, 22 males. Experimental group: n=24 (9 males, 15 females; ages 8: n=4, 9: n=7, 10: n=13). Control group: n=26 (13 males, 13 females; ages 8: n=5, 9: n=15, 10: n=6). Prior Lantern Festival attendance: 39 students (experimental: 20; control: 19). Informed consent obtained from participants and guardians. Materials: Exhibition themed ‘Praying Together to Safeguard Taipei’ highlighting Taipei’s multicultural religious heritage, integrating AR and interactive installations. AR system developed with Unity, Vuforia, Adobe Illustrator; device ASUS Zen Pad 8.0 (Z380M). Seven 3D lanterns and nine QR codes served as image targets; scanning revealed religion-specific introductions (e.g., Islam, associated practices). The paper brochure paralleled AR content: (1) Taiwan’s multicultural characteristics, (2) nine religious symbols, (3) introductions to nine religions, (4) 100 years of Taiwanese religious architecture. Visual design was unified across conditions. Procedure: Step 1: Consent and pre-test knowledge assessment (~20 min). Step 2: Exhibition learning session (40 min). Experimental group used AR tablets to explore and meet four learning objectives; control group used the brochure with identical objectives. Step 3: Post-test measures: knowledge test, GLOs questionnaire, and open-ended questions for both groups; experimental group additionally completed the TAM (~30 min). Instruments:

• Knowledge test: 20 items (10 T/F, 10 multiple-choice), 5 points per item, total 100. Constructed by the research team with cultural scholars and elementary teachers; validated by three experts. Content aligned with the four learning objectives. Same items used pre- and post-.
• GLOs questionnaire: 19 items on five constructs—Knowledge and Understanding (KU, 4 items), Skills (S, 3), Attitudes and Values (AV, 5), Enjoyment, Inspiration and Creativity (EIC, 4), Action, Behaviour and Progression (ABP, 3)—seven-point Likert scale (1–7). Items adapted from prior research and reviewed by two museum curators, one HCI researcher, and one lantern craftsperson.
• TAM questionnaire (experimental group): 19 items covering Perceived Usefulness (PU), Perceived Ease of Use (PEOU), Attitude Toward Using (ATU), and Behavioral Intention to Use (BIOU), seven-point Likert scale; items adapted from Davis (1989) and Salloum et al. (2019), reviewed by the same expert panel.
• Open-ended questionnaire: Five prompts on perceived learning about multi-religious society, differences from school learning, preferred materials and reasons, whether materials enhanced understanding and why, and additional feedback. Analysis: Quantitative analyses conducted in IBM SPSS v20. Reliability assessed via Cronbach’s alpha. Independent-samples t-tests compared groups on pre/post knowledge tests and GLOs. One-sample t-tests (test value=4) assessed TAM constructs. Qualitative responses were thematically analyzed by two researchers with subsequent discussion to synthesize insights.

Reliability: Cronbach’s alpha indicated good internal consistency: GLOs α=0.882; TAM α=0.931. Objective learning (knowledge test): Pretest means did not differ significantly between groups (Experimental M=72.71, SD=7.94; Control M=72.50, SD=15.38; t=0.061, df=38.065, P=0.952). Posttest scores were significantly higher for the experimental group (Experimental M=92.92, SD=6.24; Control M=85.58, SD=7.91; t=3.620, df=48, P=0.001), indicating superior knowledge gains with AR-assisted learning. Subjective learning (GLOs): Experimental group scored significantly higher than control on KU (M=6.00 vs 4.97; t=4.758, df=47.559, P<0.001), AV (M=5.97 vs 4.87; t=4.189, df=48, P<0.001), EIC (M=6.19 vs 4.89; t=6.185, df=47.649, P<0.001), and ABP (M=5.67 vs 4.86; t=3.505, df=48, P=0.001). No significant difference on Skills (M=5.38 vs 5.25; t=0.632, df=48, P=0.530). Technology acceptance (experimental group, one-sample t-test vs 4): High acceptance across all constructs: PU M=5.313 (SD=1.083; t=8.578, df=49, P<0.001; 95% CI: 1.006 to 1.621 above 4), PEOU M=5.176 (SD=0.983; t=8.461, P<0.001; CI: 0.897 to 1.455), ATU M=5.280 (SD=1.209; t=7.486, P<0.001; CI: 0.936 to 1.624), BIOU M=5.096 (SD=1.092; t=7.098, P<0.001; CI: 0.786 to 1.406). Qualitative insights: Both groups reported increased understanding of different religions. AR users emphasized higher interest, fun, freshness, and real-time access to information via scanning, which facilitated learning and motivated continued exploration; some suggested adding animations and sound. Control participants valued the clarity and convenience of the brochure. Informal, exhibition-based learning was perceived as more engaging and self-directed than classroom learning, with visible 3D exhibits enhancing engagement.

The findings directly address the research question by demonstrating that AR-assisted religious education significantly improves both objective knowledge acquisition and subjective learning outcomes compared to traditional brochure-based learning for elementary students. AR’s interactive, immediate feedback and virtual-physical integration likely enhanced engagement, motivation, and comprehension, aligning with literature that digital tools can raise interest and reduce cognitive load in informal learning contexts. The significant gains in Knowledge and Understanding, Attitudes and Values, Enjoyment/Inspiration/Creativity, and Action/Behaviour/Progression suggest AR not only transmits information effectively but also positively shapes affective and behavioral dimensions crucial for religious diversity education, such as respect and motivation for further exploration. The non-significant difference in Skills may reflect the broader, longer-term nature of skill development and measurement challenges within the GLOs framework. High TAM scores indicate strong acceptance and feasibility of AR among children, supported by familiarity with similar technologies. Overall, integrating AR into religious diversity education appears to transform students from passive to active learners, supports immediate comprehension, and may mitigate cognitive overload during complex cultural content learning.

This study shows that integrating augmented reality into religious diversity education for elementary students significantly enhances both objective knowledge gains and subjective learning outcomes relative to a paper-based approach, with high student acceptance of the technology. AR’s interactivity and immediacy increase motivation, engagement, and understanding and may help reduce cognitive load during learning about complex, diverse religious traditions. These results fill a gap in the literature by evidencing the effectiveness of interactive technology in religious education and suggest practical value for educators seeking to modernize curricula. Future research should expand participant demographics and contexts, incorporate longitudinal designs to assess skill development and retention, and examine additional factors (e.g., self-efficacy, prior tech exposure) that may mediate or moderate AR’s impact.

The study was conducted in a single city (Taipei) with a limited sample of grades 3–4 students from a specific ethnic context, constraining generalizability. The sample size (N=50) was restricted by COVID-19 constraints. Potential external influencing factors (e.g., self-efficacy, prior familiarity with AR) were not fully accounted for. The short intervention duration may be insufficient to detect changes in broader skill constructs. Future work should include larger and more diverse samples across regions and ages, consider longitudinal follow-ups, and model additional mediators and moderators of learning outcomes.