Generative AI for Three-Dimensional Geometry Learning: A Systematic Review of Visualization, Spatial Reasoning, and Conceptual Understanding

# Data and Supplementary Materials## Generative AI for Three-Dimensional Geometry: A Systematic Review of Visualization, Spatial Reasoning, and Conceptual Understanding **Authors:** Tomi Purwanto, Majid, Hasan S. Panigoro**Affiliation:** Universitas Negeri Gorontalo**Manuscript submitted to:** International Journal of Mathematical Education in Science and Technology (Taylor & Francis) This repository contains the data and materials underlying the systematic review manuscript listed above, deposited in accordance with the Taylor & Francis Publicly Available Data policy. --- ## Contents ### 01_bibliographic_corpus/- `CSV_Artikel_3.csv` — Full bibliographic export (Zotero CSV format) of the initial literature search, 982 records, including titles, abstracts, DOIs, authors, publication years, and source journals/proceedings. ### 02_screening_workbook/- `37_relevant_records_screening.csv` — The 37 records judged relevant to the review topic after keyword pre-filtering (982 → 85 keyword-flagged candidates) and title/abstract screening. Each row records: record index, title, author(s), year, publication venue, DOI, whether the study explicitly addresses 3D geometry content, whether it reports a measurable student outcome, and the reasoning for inclusion/classification. Of these 37, full-text access was obtainable for 6 records (5 included as evidence after full-text review; 1 excluded as off-topic after full-text reading — see manuscript Section 3/PRISMA flow); the remaining 31 required institutional subscription access unavailable to the authors at the time of this revision. ### 03_supporting_data_figures/Underlying numeric values for the descriptive statistics, effect sizes, and figures reported in the manuscript, as required by the Taylor & Francis Publicly Available Data policy:- `Gambar2_data_effect_size.csv` — Means (1–5 scale), Cohen's d, and 95% confidence intervals for the five competency items reported in Manuscript Table 2 / Figure 2. Values reproduced with attribution from Liang et al. (2026), *Applied Sciences*, 16(14), 7106, Tables 4–5 (MDPI, CC BY 4.0).- `Gambar3_data_pilot_study.csv` — Relative task-completion speed (control vs. experimental group, N = 20) underlying Manuscript Figure 3. Derived from Petrov & Atanasova (2026), *Engineering Proceedings*, 150, 125, Section 4.4 (MDPI, CC BY 4.0). Outcomes reported only narratively in the source article (engagement, conceptual understanding, anatomical accuracy — no exact published figures) are flagged as such in the file rather than estimated or invented. **No data points were extracted from images for quantitative analysis in this review.** All values in this folder are transcribed directly from numbers published in the cited source articles' text/tables, not digitized from chart images. ### 04_fulltext_pdfs_open_access/Full-text PDF copies of the five open-access articles cited as Tier 2 (strong empirical) or contextual/verification evidence in the review:- `Liang_et_al_2026_AppliedSciences.pdf` — Liang, W., Li, Y., Zhou, Y., Zhang, J., & Wang, J. (2026). The construction of sustainable digital resources and the application of AI technology in the engineering drawing course. *Applied Sciences*, 16(14), 7106. https://doi.org/10.3390/app16147106 — MDPI, **CC BY 4.0**, redistribution confirmed permitted.- `Petrov_Atanasova_2026_EngineeringProceedings.pdf` — Petrov, P., & Atanasova, T. (2026). Integrating AI-generated 3D models into education—A methodological and practical approach. *Engineering Proceedings*, 150, 125. https://doi.org/10.3390/engproc2026150125 — MDPI, **CC BY 4.0**, redistribution confirmed permitted.- `Pek_2026_EpitesEpiteszettudomany.pdf` — Pék, J. (2026). Az ábrázoló geometria jövője az építészmérnök-oktatásban a mesterséges intelligencia világában [The future of descriptive geometry in architectural education in the age of artificial intelligence]. *Építés – Építészettudomány*, 54(3–4), 411–426. https://doi.org/10.1556/096.2026.00168 — **license not independently verified by the authors; confirm open-access redistribution terms with the publisher (Akadémiai Kiadó) before the final Zenodo deposit.**- `Vilaca_et_al_2026_CompApplicEngineering.pdf` — Vilaça, P.M.F., González-Moreno, J.C., & Faria, P.M. (2026). Capabilities of Generative Artificial Intelligence and Three-Dimensional... *Computer Applications in Engineering Education*. https://doi.org/10.1002/cae.70243 — **license not independently verified by the authors; confirm open-access redistribution terms with the publisher (Wiley) before the final Zenodo deposit.**- `Kicklighter_et_al_2024_SIGGRAPH.pdf` — Kicklighter, C., Seo, J.H., Andreassen, M., & Bujnoch, E. (2024). Empowering creativity through generative AI in digital art education in higher education. *SIGGRAPH Educators Forum '24*. https://doi.org/10.1145/3641235.3664438 — **license not independently verified by the authors; confirm open-access redistribution terms with the publisher (ACM) before the final Zenodo deposit.** > ⚠️ **Action needed before final deposit:** Please verify the redistribution license for the three PDFs marked above. If any does not permit redistribution, remove that PDF from this folder and keep only its DOI link in the manuscript's reference list and Data Availability Statement (the manuscript text already allows for this: "included where publisher licensing permits redistribution, or linked by DOI where it does not"). ### 05_figures/- `Gambar1_PRISMA_Flow.png` — PRISMA 2020 flow diagram of the screening process (982 → 951 title/abstract screened → 85 keyword-flagged → 37 relevant → 6 full-text obtained → 5 included as evidence / 1 excluded after full-text reading).- `Gambar2_EffectSize.png` — Bar chart of effect sizes (Cohen's d) for three competency dimensions, data in `03_supporting_data_figures/Gambar2_data_effect_size.csv`.- `Gambar3_PilotStudy.png` — Bar chart of relative task-completion speed, control vs. experimental pilot group, data in `03_supporting_data_figures/Gambar3_data_pilot_study.csv`.- `prisma.mmd` — Mermaid source file used to generate the PRISMA flow diagram. ### 06_manuscript/- `Manuskrip_GenAI_3D_Geometry_REVISI.docx` — Revised manuscript submitted to *International Journal of Mathematical Education in Science and Technology*, incorporating the unstructured 200-word abstract and the Data Availability Statement that this repository supports. --- ## Limitations on data availability This article type (a systematic review) does not involve primary human-subjects data collected by the authors; no datasets requiring ethical, privacy, or security restriction apply. The sole limitation concerns literature access rather than data sharing: of the 37 records judged relevant during screening, full-text access could be obtained for only 6 (5 included as evidence, 1 excluded after full-text reading as off-topic). The remaining 31 records require institutional subscription access unavailable to the authors at the time of this revision. This does not affect the availability of any data the authors generated or are entitled to redistribute, all of which is included in this repository. ## License Unless noted otherwise above, the authors' own materials in this repository (`01_bibliographic_corpus/`, `02_screening_workbook/`, `03_supporting_data_figures/`, `05_figures/`, `06_manuscript/`) are released under **CC BY 4.0** (see `LICENSE`). Third-party PDFs in `04_fulltext_pdfs_open_access/` retain their original publishers' licenses as noted per file above. ## Citation Please cite the associated manuscript when reusing this dataset: Purwanto, T., Majid, & Panigoro, H. S. (2026). Generative AI for Three-Dimensional Geometry: A Systematic Review of Visualization, Spatial Reasoning, and Conceptual Understanding. *International Journal of Mathematical Education in Science and Technology* (manuscript under revision).