Subjects in the experimental group received a series of culture-based multicomponent cognitive training activities via research assistants (RAs) and caregivers of the centres for 8 weeks, then provided by caregivers of the day care centre for 8 weeks with RAs providing assistance with the activity process.

Participants in the comparison group received these activities via two RAs and day care caregivers for 16 weeks.

Activities of the program were structured, cognitive, leisure activities based on Chinese and Taiwanese culture for cognitively impaired older people. Activities contained structured brain-exercise activities and musical activities. The structured brain-exercise activities included playing musical instruments or singing, playing board games, and classes in Chinese, sciences, sociology, physical education, cooking, and math. Chinese and Taiwanese cultures were included within the structured brain-exercise activities curriculum, such as food, clothing, architecture, places of interest, folklore, etiquette, celebrations, religion, art, etc. Participants began with a musical exercise, then had 20-30 minutes of physical exercise with 5 minutes of warm-up at the beginning and 5 minutes of cool down. The program was provided for at least 40-60 minutes, twice a week, for 8 weeks via RAs and day centre caregivers.

The Mini-Mental State Examination (MMSE) was used to measure the cognitive status of the participants. It was collected through individual interviews. The MMSE is a criterion-referenced instrument that contains 11 items. It can be used to examine a person's orientation, registration, recall, attention, calculation ability, understanding and use of language, and praxis. The score ranges from 0 to 30. The sum of the score is used to stratify cognitive function (Crun, Anthony, Bassett, & Folstein, 1993).

The Clifton Assessment Procedures for the Elderly - Behaviour Rating Scale (CAPE-BRS) was used to measure behavioural problems, including activities of daily living, apathy, communication difficulties, and social disturbance. It was collected by observations. The CAPE-BRS includes 18 items; each item is scored from 0 to 2. The total score ranges from 0 to 36. Higher scores indicate greater disability (Pattiem & Gilleard, 1979).

The Cornell Scale for Depression in Dementia (CSDD) was used to examine symptoms of depression. It was collected by observations. It includes 19 items and measures mood-related signs, behavioural disturbances, physical signs, cyclic functioning, and ideational disturbance. The total score ranges from 0 to 38. A score from 0 to 6 indicates no depression, 7 to 9 indicates mild depression, 10 to 19 indicates possible severe depression, and 18 and over indicates severe depression. The CSDD has been translated into Chinese, and the validity and reliability have been examined (Sharp & Lipsky, 2002).

The Refined ADL Assessment Scale (RADL) was used in this study to measure the performance of activities of daily living (ADLs). It was collected by observations. The RADL is designed to measure 5 tasks of basic ADLs for older adults with Alzheimer's and related disorders. It includes tasks of feeding, washing, grooming, dressing, and toileting. The toileting task was not assessed in this study because it was likely to disturb the privacy of older adults. Each task in the RADL is broken down into 2 to 5 subtasks. Each subtask is further broken down into the sequence of steps needed for the completion of the activity. The scores for the feeding activity range from 0 to 126, for the washing activity range from 0 to 204, for the grooming activity range from 0 to 174, and for the dressing activity range from 0 to 150. A higher score indicates greater independence in ADL performance.