Objective referenced tests were scaled using the Rasch latent trait model to measure Student Learning Objectives (SLOs) in grades one through eight. The tests, in the areas of reading, language and math, were scaled and linked through students, blocks of items, and test item banks. The study tests latent trait theory in a public school setting. General assumptions are substantiated. Limitations that should be considered by developers and users of test item banks are clarified, and some questions are raised about the appropriateness of expecting large increments of measured growth among students, particularly at higher grade levels. (Author/GK)