AI Versus Conventional Methods in Spatial Behavior Analysis of Educational Settings: A Comparative Review of Methodological Limitations
Year: 2026
Type: article
Study design: Systematic Review
Abstract: Spatial behaviour analysis in educational spaces is an important area of research for understanding user-space interaction and evaluating the performance of learning environments.Although the use of AI-supported methods has increased in recent years, the behavioural dimensions addressed or overlooked in AI-based and conventional research have not been systematically compared across educational settings.This study evaluates spatial behaviour research in educational spaces from the perspective of the representation gap and compares the behavioural and environmental dimensions examined by AIsupported and traditional approaches.A systematic literature review yielded an analysis matrix containing 181 coded publications, including 121 non-AI and 60 AI-supported records.The publications were coded according to educational level, behavioural parameters, environmental variables, data collection tools, and methodological approach.The findings show that AI-supported studies more frequently examine duration, route, and headcount, while user interaction and density are more frequently considered in non-AI studies.Environmental parameters, including temperature, acoustics, air quality, humidity, and lighting, are infrequently represented in both groups.The results indicate that increased data collection and analytical capacity does not necessarily result in a broader examination of social and environmental dimensions.The differences identified concern the variables addressed in the reviewed publications rather than the inherent capabilities of individual methods.The study therefore emphasises the value of analysis frameworks that integrate behavioural, social, and environmental components in the evaluation of educational buildings.
Source: Online Journal of Art and Design
Authors: Beyza Çilli, Kutlu SEVİNÇ KAYIHAN
Language: English
Subfield: Education
Field: Social Sciences
Domain: Social Sciences
SDG: Quality education
Open Access status: bronze
OpenAlex ID: W7225996977
