Key points are not available for this paper at this time.
Intensive interviews were conducted with 36 older people, using think aloud and cognitive probes, to study response errors in measuring functional status in health surveys. Consistent with the original hypothesis, words or phrases, such as "aids," "help from another person," and "difficulty," were not interpreted by respondents to have the same meaning as intended. Similarly, evidence was obtained indicating that the subjects did not recall all the kinds of help they received, nor report their health as the basis for needing assistance. Moreover, replies to a variety of functional status items were often conditional upon the circumstances. Recommendations for questionnaire design and further research are discussed.
Keller et al. (Sun,) studied this question.