A study examined factor structure invariance among the writing skills of graduating High School Seniors and Adult High School Non-Completers. The study had three purposes: (1) to use LISREL confirmatory factor analysis (CFA) procedures to specify and test a series of factor models based on the test specifications content of the General Educational Development (GED) Writing Skills Test; (2) to evaluate the fit of these models to a set of test data obtained from a national sample of graduating high school seniors; and (3) to test the invariance of the best fitting factor structure model in both the seniors and the adult high school non-completers. Subjects were 2,532 high school seniors who took the anchor form of the Writing Skills Test in the spring of 1987 and 699 adult high school non-completers between the ages of 17 and 19. Results indicated the nature of writing skill measured by this test to be a single construct reflecting generalized proofreading/editing skills. No support was found for the view that separate item-type methods factors accounted for variability of performance on the multiple-choice portion of the GED test. This study provided an empirical field-based illustration of the use of CFA procedures to evaluate the factor structure of multiple-trait, multiple-method data and to test for the invariance of plausible measurement models over multiple groups. (Two tables of data are included, and 24 references and a list of variables are attached.) (RAE)