Description
The GDS was developed from an initial pool of 100 items chosen by clinicians and researchers for their ability to distinguish elderly depressive people from unaffected people. Questions that might elicit defensiveness were avoided and, for simplicity, a yes/no answer format was used for all items. The 30 items in the GDS were empirically selected from the pool, on the basis of high item-total correlations when administered to a mixed sample of 47 community residents and people hospitalized for depression, all of whom were older than 55 (1, p41).