Understanding Society User Support

Hello,

I am attempting track the number of health conditions participants across waves, to examine any correlation between changes in health conditions and economic security. When doing a cross-sectional analysis of W10 data, I used a count function in SPSS to determine how many conditions each participant has. To do so, I used the variables j_hconds01 to j_hconds97 for new entrants to the survey and j_hcondns1 to j_hcondns13 for newly diagnosed health conditions of previous participants. Using this, I found 7260 cases with one health condition and 4741 with more than one condition. This is a similar figure to the 12296 who answered yes to having a long-standing illness or impairment (j_health).

I used the same syntax as the basis for creating a similar health condition flag in W9. It seems that between W9 and W10, there was a change in how the health condition questions were asked and far more conditions were inquired about. As such, I have used variables i_hconds01 to i_hconds17 for new entrants and i_hcondns1 to i_hcondns8. Using this, I found 3016 cases with one health condition and 708 with more than one condition. This is around a third of the total (12992) who answered yes to having a long-standing illness or impairment (j_health).

Were the changes to the way that health conditions were recorded between w9 and w10 so great as to cause this disparity? I imagine not, and would greatly appreciate any advice on how I can capture the remaining participants with health conditions in w9 and earlier.

Thanks,
James