Codes and Frequencies
BEDAYR2 reports, in broad intervals, the number of days during the preceding 12 months that illness or injury kept the person in bed for more than half the day, including days while hospitalized.
For 1976-1981 versus 1982 forward, there were subtle differences in the initial question used to elicit the information for BEDAYR2.
For the earlier period, interviewers asked, "During the past 12 months (that is, since [date] a year ago), ABOUT how many days did illness or injury keep [person] in bed all or most of the day? (Include the days in the past 2 weeks.) (Include the days while a patient in a hospital.)" For the later period, interviewers asked, "During the past 12 months, (that is, since 12-month date a year ago), ABOUT how many days did illness or injury keep [person] in bed more than half of the day? (Include days while an overnight patient in a hospital.)" Comparability may be slightly compromised by the specification of "all or most of the day" versus "more than half the day" and "while a patient in a hospital" versus "while an overnight patient in a hospital."
For 1976-1981, follow-up questions encouraged respondents to supply answers falling into broad intervals.
Specifically, interviewers asked, "Was it more than 7 days or less than 7 days? Was it more than 30 days or less than 30 days? Was it more than half the year or less than half the year?" The values for BEDAYR2-- None, 1-7 days, 8-30 days, 31-180 days, 181-365--mirror the intervals mentioned in these follow-up questions.
Beginning in 1982, the follow-up questions were dropped, and respondents had to supply a single number falling between 0 and 365 (for 1982-1996) or 0 and 366 (for 1997 forward).
To facilitate comparisons over time, BEDAYR2 recodes the more detailed responses for 1982 forward into the broad intervals that are the only information available in the public use files for 1976-1981. Researchers who are only interested in data from 1982 or later may wish to analyze the non-intervalled data on bed days during the preceding 12 months, included in BEDAYR.
There was a significant change in the variable universe, with all persons included up through 1996 and only sample adults included for 1997 forward.
Prior to 1997, one member of the household could answer this question for other household members (i.e., proxy reporting). Beginning in 1997, when the information was collected only for sample adults, individuals answered the question for themselves, except in rare cases when disability barred sample adults from responding. In general, self-reporting is more accurate than proxy reporting. Thus, researchers should be cautious about comparing results from BEDAYR from prior to 1997 to results from BEDAYR for 1997 forward, even if the analysis is limited to adults.
- 1976-1996: All persons.
- 1997-2018: Sample adults age 18+.