Abstract
This study utilizes generalizability theory to examine the construct-related validity of a carefully implemented assessment center. Previous research has indicated that assessment centers do not exhibit construct-related validity. This study suggests that the lack of evidence demonstrating construct-related validity may be due to implementational differences in assessment centers. Results indicate that the person as well as person by dimension effects contribute large proportions of variance to the total variance in assessment center ratings. Rater and person by exercise effects, however, contribute little to the total variance. This study is consistent with assessment center theory, suggesting that performance does vary across dimensions while remaining relatively more stable across exercises.
Maldegen, Robyn Marie (1996). Further examination of multiple systematic sources of variance in assessment center ratings. Master's thesis, Texas A&M University. Available electronically from
https : / /hdl .handle .net /1969 .1 /ETD -TAMU -1996 -THESIS -M35.