Are go/no-go tasks preferable to two-choice tasks in response time experiments with older adults?