Optimisation of Maslow's hierarchy of needs-based survey form for nursing home residents using machine learning

Abstract

The global increase in the elderly population necessitates accurate assessment of the needs of individuals residing in nursing homes. This study aims to optimise a 30-item questionnaire developed on the basis of Maslow's Hierarchy of Needs by employing machine learning algorithms. As one of the pioneering applications in Turkey, the research was conducted with 310 participants from four nursing homes. Data were analysed using Artificial Neural Network (ANN), Gaussian Process Regression (GPR), Linear Regression (LR), and Support Vector Machine (SVM). F-Test results identified the most significant variables, reducing the number of items from 30 to 10. Within the comparative analyses, the GPR model outperformed the others by yielding the lowest mean error metrics (RMSE = 0.252, MSE = 0.064, MAE = 0.195) and the highest predictive accuracy (R2 = 0.86). Findings indicate that the physiological, social, and psychological needs of older adults can be assessed through shorter, more reliable questionnaires. This study offers academic and practical contributions to elderly care and interior design.