Factor analysis is used frequently by researchers as a data reduction and summarization technique. Many analysts use exploratory factor analysis to search for underlying dimensions in attitudinal studies. Concern arises when novice researchers rely solely on information derived from computer printouts to factor analyze data, dismissing theoretical consideration of concepts underlying this analytical procedure. A primer on principal components exploratory factor analysis is presented, and five decision rules for selecting the number of principal components to retain are discussed: (1) the K1 rule; (2) the Scree test; (3) Bartlett's test; (4) the minimum average partial method; and (5) parallel analysis. A small data set, obtained in an actual exploratory study, was used to illustrate the discussion. The study addressed effects of preemployment tests on attitudes toward a firm formed by individuals outside that firm. In a pilot study, responses of more than 400 graduating seniors to three different preemployment tests were analyzed. In a second study, 249 graduating seniors and master's candidates responded to preemployment test scenarios. Dimensions of applicants' attitudes were examined through exploratory factor analysis. It is concluded that the results of different decision rules must be used when determining the number of principal components, and that factor analyses should be run with one or two components above and below those suggested with the five methods in order to avoid underextraction or overextraction. Analysts are cautioned to not rely on computer programs and preset default outputs as the "last word." Three figures and four tables supplement the discussion. A 35-item list of references is included. (Author/SLD)