While randomized trials and intent-to-treat analyses have often been considered the gold standards for evaluation of interventions, some psychologists have challenged this conventional wisdom. For example, in studies of respite care, even after randomly assigning caregivers to respite versus no respite, some caregivers in the respite condition will not use the service anyway while some caregivers in the non-respite condition will find a way to receive respite.
Zarit and colleagues (1998) noted the difficulty in conducting truly randomized trials of services such as adult day care. They argued strongly for the use of quasi-experimental designs in such circumstances.
Zarit, S. H., et al. 1998. Stress reduction for family caregivers: Effects of adult day care use. Journal of Gerontology 53B(5): S267-77.