ASSESSMENT OF CHILDHOOD DEPRESSION - SAMPLING MULTIPLE DATA SOURCES WITH ONE INSTRUMENT

The low correspondence among reporting sources in the measurement of c hildhood depression is well documented. However, no effort has been ma de to quantify the effects of using different measures with different sources. In this research the administration of the Peer Nomination In ventory of Depression (PNID) was modified to obtain three scores: (1) each child's rating of depression as given by his/her peers; (2) each child's self-rating of depression and (3) the number of times children assigned ratings of depression to others. The Children's Depression I nventory (CDI) was administered for comparative purposes. Results sugg est that differences in test items and formats can explain some, but n ot all, of the low correlations among reporting sources. Gender differ ences in the three modified PNID scores also were investigated.