In 1994 and 1995, this variable is part of the Disability Supplement (NHIS-D) that collects information on disability. NHIS-D included two household interviews: in Phase I all members of sample families were screened for any indication of disability; in Phase II persons with any indication of disability were followed-up for additional information. Phase I includes sensory, communication, and mobility problems; health conditions; activities of daily living and independent activities of daily living; functional limitations; mental health; services and benefits; special health needs of children; early child development; education; relationship to respondent; and perceived disability. Phase II includes four Disability Followback Survey (DFS) questionnaires: one for children, one for adults, one for elderly persons (69 years of age and over) without any indication of disability (also called the Supplement on Aging or SOA; only in 1994), and one for persons with a history of polio.

The term "difficult" was respondent-defined. A brief definition of "special equipment" based on examples (i.e., "such as a cane, a wheelchair, a special bed, or a special telephone") was part of the preceding question. The Field Representative's Manual for 1997 forward defined "special equipment" as "any device, tool, utensil, instrument, implement, etc., used as an aid in performing an activity because of a physical, mental, or emotional problem." The Manual for 1997-2000 continued, "This includes the use of adult 'diapers' for incontinence. However, ordinary eyeglasses and hearing aids should not be considered 'special equipment.' For example: a spoon is not normally considered as 'special equipment'; however, a uniquely designed or functioning one used for eating by a person because of physical, mental, or emotional problems is considered 'special equipment.'"

Beginning in 2001, the Field Representative's Manual formally defined the words "by yourself." Specifically, "By yourself is considered to be without the help from another person or without hands-on assistance with performing an activity. Another person may be a friend, relative, paid helper, volunteer from an agency or organization or anyone else who helps the family member in doing the activities mentioned. He or she may be a household member or a non-household member."

A formal definition of "health problem" was also included in the Field Representative's Manual, with only slight changes in wording over time. The Manual for 1997 stated:

No definition of terms except the aforementioned definition of "health problem" as "any physical, mental, or emotional problem or illness (not including pregnancy) was routinely supplied to respondents during the interview.

Other questions in the series collected information about how difficult the sample adult found the following activities, if performed "by yourself and without using any special equipment":

• Walking about 3 blocks (FLWALK3BL)

• Standing for about 2 hours (FLSTAND2HR)

• Walking up 10 steps without resting (FLCLIMB)

• Sitting for about 2 hours (FLSIT2HR)

• Stooping, bending, or kneeling (FLSTOOP)

• Reaching above the head (FLREACH)

• Carrying 10 pounds (FLCARRY)

• Pushing large objects (FLPUSHLAR)

• Going out for shopping (FLGOOUT)

• Participating in social activities (FLSOCIAL)

• Doing things to relax at home (FLRELAX)

For years prior to 2000, a small number of cases were coded as "Do not do this activity" in the NHIS public use data files. The Field Representative's Manual for 2000 explained,

Although, technically, "Do not do this activity" was a category for FLGRASP for 1997 forward in the NHIS public use files, this category is not actually comparable across the two periods 1997-1999 and 2000 forward. Very few cases (0 in 1997) were coded as "Do not do this activity" via data editing, compared to the number of respondents who self-selected this category for 2000 forward (33 in 2000). To increase comparability across years, researchers are advised to combine the categories "Can't do at all" and "Do not do this activity" when analyzing data collected both before and after 2000 for this variable.