Dataset for: Symmetry and Asymmetry in AI-Mediated Visual Critical Analysis: Multi-Level Framework for Human-Computer Interaction in Photographic Critical Analysis Education

This dataset supports the article "Symmetry and Asymmetry in AI-Mediated Visual Critical Analysis: Multi-Level Framework for Human-Computer Interaction in Photographic Critical Analysis Education," published in the International Journal of Human-Computer Studies.

Research Hypothesis: The study investigates whether a structurally symmetric AI educational tool, based on the Hermeneutical and Interdisciplinary Socio-Semiotic Model for Photographic Critique (HSSPC Model), can effectively support photographic critical analysis education while accommodating asymmetric user needs through adaptive differentiation.

Data Description: The dataset contains aggregated quantitative results from a convergent mixed-methods study conducted with 128 undergraduate students enrolled in photography and audiovisual production courses at Tecnologico de Monterrey (Mexico City, Mexico) during August-December 2025. Data was collected through the Canvas learning management system via student analysis packages, reflective responses, and usage logs.

The dataset includes eight tables: (1) Adoption modalities of AI-assisted analysis (n=128); (2) Frequency of theoretical perspectives applied; (3) Overall tool valuations; (4) Most valued aspects; (5) Principal criticisms; (6) Emergent user profile distributions; (7) Metacritical consciousness types; and (8) Statistical test results (chi-square with effect sizes).

Notable Findings: Results demonstrate a 94% adoption rate with 68% positive evaluations. Three distinct user profiles emerged with asymmetric needs: Academic (40%), Pragmatic (35%), and Artistic (25%). A pedagogical paradox was identified where analytical depth constituted both the primary strength (46%) and a source of criticism (10%). Metacritical consciousness developed in 78% of students, indicating successful epistemological calibration in human-AI collaboration.

Data Interpretation: The tables can be read independently as frequency distributions and cross-tabulations. Percentages refer to the proportion of the total sample (n=128) unless otherwise noted. The metacritical consciousness data refers to a subset of 100 students (78% of the sample) who demonstrated critical awareness. Chi-square analysis (Table 8) tests the association between prior AI experience and adoption modality, with Cramer's V indicating a medium effect size.

Note: Raw qualitative data (student reflections and textual responses) are not included due to ethical and privacy considerations. Anonymized representative excerpts are provided in the associated manuscript.

Files: The dataset is provided in two formats: CSV (machine-readable, suitable for statistical analysis) and PDF (human-readable documentation with formatted tables).