Writing researchers must be concerned with whether the instruments they are using are constructed as soundly as possible and, if so, whether they are performing the necessary analytic procedures to interpret the results correctly. To examine the logic in the construction of J. A. Daly and M. D. Miller's Writing Apprehension Test and the incomplete analytic procedures performed in K. Hunt's T-unit analyses, two experiments were conducted with college students. Results of the first experiment showed that the inclusion of the "uncertain" response choice in the writing apprehension test can lead to faulty results, thus lessening the validity of the instrument. Findings from the second experiment indicated that the assumption of a positive, linear relationship between syntactic complexity (T-unit analyses) and writing quality is inappropriate, especially when comparing diverse groups. The two experiments suggested that (1) either writers may be misassigned to writing apprehension groups or scores may not accurately reflect their degree of writing apprehension, thus confounding results, inferences, and theories on which other studies are subsequently based; and (2) essays may be evaluated for syntactic complexity with the false assumption that a positive, linear relationship exists between syntactic complexity and quality. (HTH)