This experimentation uses dependent interviewing to investigate the measurement of change in self-assessed measures of long-standing illness or disability. The current UKHLS method of obtaining information about disability is to use a yes/no question about long-standing conditions which is followed-up by a check-list of 11 areas of everyday life where people may have difficulties (i.e., “Activities of Daily Living”, or ADL indicators). The Wave 6 sample was split into two experimental treatments and a control group. The first experimental treatment compared the response to the initial question about long-standing conditions to the previous wave’s response and followed-up any responses which were different from previous with a “Why?” question. The ADL check-list was administered at a later point in the questionnaire for this set of respondents. The second experimental condition asked only the ADL check-list without filtering on whether the respondent reported that they had a long-standing condition. Finally, the control group repeated the original UKHLS approach.
Randomisation was of households within PSUs into the experimental treatments and control as follows: one-half of the sample into the first experimental treatment, one-quarter of the sample into the second experimental treatment, and the remaining quarter of the sample into the control group. All new entrants received the control version by default.
This experiment is controlled by variable ff_disabilityw6 on record hhsamp_ip. It takes the values:
Group 1 – edit check/explain inconsistencies, move DISDIF to annual events
Group 2 – only DISDIF, are not asked HEALTH
Group 3 – original design
The experiment was repeated in Wave 7 with the same allocations to treatments. The IP7 refreshment sample was allocated to Group 3.
Affected variables on the indresp records for Wave 6 and Wave 7 are: health, hthrdia11, hthrdia12, hthrdia13, hthrdia14, hthrdia15, hthrdia16, hthrdia21, hthrdia22, hthrdia23, hthrdia24, hthrdia25, hthrdia26, disdif1, disdif2, disdif3, disdif4, disdif5, disdif6, disdif7, disdif8, disdif9, disdif10, disdif11, disdif12, disdif96, disdifb1, disdifb2, disdifb3, disdifb4, disdifb5, disdifb6, disdifb7, disdifb8, disdifb9, disdifb10, disdifb11, disdifb12, disdifb96, disdifa1, disdifa2, disdifa3, disdifa4, disdifa5, disdifa6, disdifa7, disdifa8, disdifa9, disdifa10, disdifa11, disdifa12, disdifa96
