The document presents the findings of a study that compared scientific journal articles to their pre-print versions. The study gathered a corpus of over 11,000 matched pre-print and post-print article pairs. It analyzed and compared sections of the texts, such as titles, abstracts and bodies, using multiple text similarity comparison methods. The preliminary findings showed very high levels of similarity between pre-print and post-print versions across all sections. This suggests that publishers may not be substantively improving the content of articles to warrant their high costs.