This study, a replication of Carbol's (1986), sought to compare GED test performance of first-time GED examinees in Virginia according to the method of testing administration experienced, one day or two half days. Carbol's study was extended to include an examination of how testing administration method affected the relationship of test performance and selected characteristics of the GED examinees (age, gender, race, highest grade completed, geographical location, preparation for the GED and motivations for taking the GED). These selected characteristics were also analyzed for their predictive usage for successful completion of the GED. No significant differences were found with test performance between the two treatment groups either with mean total test scores, pass/fail rate, first and second half of testing, or according to the order in which the five subtests were taken (sequence). Of the selected characteristics of GED examinees examined, testing administration method only affected the relationship of test performance and race of the GED examinees. The method marginally affected the relationship of test performance and age, geographical location and preparation for the GED. The method did not affect the relationship of test performance and any other of the GED examinee characteristics. Those characteristics deemed appropriate predictors of success included race, highest grade completed and preparation for the GED.

Two conclusions were drawn from this study. First, the initial testing policy currently utilized in Virginia with respect to method of testing administration utilized does not warrant further review due to the very small difference found between test scores earned in one and two half days of testing. Secondly, counselors should be aware of the various characteristics which might affect test performance and utilize this information in evaluating and updating current screening procedures to assist in advising prospective GED candidates.