The intent of this investigation was to learn what factors may have contributed to low usage of a small set of reputable open access scholarly journals, as evidenced in graduate student reference lists. Inquiry into differences among open access sources referenced between 2005 and 2014 revealed that non-indexed journals were being used notably fewer times than their commercially indexed counterparts. A comparison of HTML source code between indexed and non-indexed journals illustrated that certain metadata and other coding features may have affected the visibility of non-indexed journal content.The study suggests to librarians the presence of less visible empirical literature on the web and to smaller, independent academic publishers, the need to apply better quality metadata and optimization strategies to their open access journal content.