ABSTRACT
The Atlantic Refining Company research was published originally in the Journal of Applied Psychology (Miner 1960a), and supplemented later in Chapters 3, 4, 7, and 8 of a book (Miner 1965a). The latter materials supersede what is presented in the original article on occasion, and are given primary attention here. For one thing the experimental group was increased from fifty-five to fifty-six with the addition of an individual whose posttest was considerably delayed. Second, the concurrent validity sample was increased from eighty-one to one hundred in computing relationships with job grade, as new data became available. Third, rare scores were calculated using a normative sample of 160 managers from the same company rather than the original 120, thus producing more stable results and a number of changes. Finally item score findings were incorporated in the presentation for the first time. In hindsight the original publication appears to have been somewhat premature, and that situation is corrected in what follows.
