Abstract

An attempt was made to develop a newly integrated version of CES-D (Center for Epidemiologic Studies Depression Scale, Radloff, 1977) using items contained in its earlier three versions currently available in Korea. A series of three studies were preformed. Study 1 was to select the most powerful items from earlier versions. A preliminary scale with 30 items taken from earlier three versions was tested with 594 college students and 228 adults sample. 20 best items were identified on the basis of item-total correlation in 9 different analyses. In Study 2, an intermediate new scale consisting of the selected 20 items was validated with 235 adults. The internal consistency of the intermediate scale was .91, higher than or equal to earlier versions. However, one item (i.e., item 4) in the intermediate scale was shown unsatisfactory revealing low item-total correlation of .20. In Study 3, four new items as tentative candidates alternative to the poor item (i.e., item 4) were developed and tested in terms of psychometric and translation criteria using data from 189 college students. One among the candidates was finally selected and incorporated into the final scale. The final integrated adaptation of CES-D revealed .91 for its internal consistency and four distinctive factor structure similar to original English scale. However, the new scale needs further validation test with diverse samples in both normal and clinical settings in future studies.