The study was directed towards the further exploration of the use of attribute ratings as the basis for establishing the job component validity of tests, in particular by using different methods of combining "attribute-based" data with "job analysis" data to form estimates of the aptitude requirements of jobs. The primary focus of the study related to the use of attribute data for deriving estimates of requirements for psychomotor tests, since the previous use of attribute data with such tests had proved to be ineffective. Twenty-one methods of combining these two sets of data were investigated. The findings generally confirm the results of the previous study using such attribute data in indicating reasonably satisfactory prediction with cognitive tests, moderate prediction with perceptual tests, and poor prediction in the case of psychomotor tests. There were however, some variations in the effectiveness of the different methods in predicting aptitude requirements, with some of the methods being differentially effective in the prediction of such requirements with different types of tests. (Author/RC)