Abstract

The study aimed to examine the influence of personal characteristics on activity preferences using decision tree analysis and examine the effects of the variables using conventional approaches. A descriptive study was conducted with 251 nursing home residents with dementia in Korea to examine the relationship between their personal characteristics and activity preferences. Decision tree analysis was used to classify participants’ activity preferences, and preference levels were examined using logistic regression analysis. Activities were classified as either physical and social activities or cognitive and affective activities. This model showed an accuracy rate of 85.7% for positively predicting physical and social activity preference and 30.3% for positively predicting cognitive and affective activity preference. Gender was the strongest determinant of activity preference. The odds of preferring physical and social activities were 3.179 times higher among women, while the odds for preferring cognitive and affective activities were 0.412 times higher among men. Notably, cognitive and affective activity preference increased to 58.8% for married male participants. This study’s findings can contribute to the development of programs to decrease behavioral and psychological symptoms among older people with dementia residing in nursing homes and provide scientific evidence for integrating these activities into long-term services for this population.