Studies on access: a review

Research on the accessibility of the scientific literature follows two main methodological approaches: the first is based on surveying researchers on their recollections, perceptions and desires of the journal publishing system; the second is based on unobtrusive studies of what scientists read and cite. Both approaches have their strengths and weaknesses. Surveys can gather the responses of thousands of individuals and allow a researcher to generalize the results over a target population. In-depth interviews, while limited in their generalizability, can explore a topic in more detail and draw out values and motivations from a respondent. Poorly constructed questionnaires, however, can mislead respondents and result in biased results. Similarly, interviewees may be prompted to provide what researchers want to hear or what scientists ought to believe, leading to significant response bias. For example, since one of the central values of science is openness (Merton, 1942), scientists may be supportive of the phrase " open-access " in spite of the ambiguity in how the term is used. Moreover, researchers have different – and often competing – interests when responding as authors or as readers (Mabe & Amin, 2002). Authors want to publish more: readers want to read less. This poses a problem for understanding the needs of researchers and makes the context of the study immensely important. Unobtrusive measures (such as counting article downloads or measuring citations) are a more direct approach to measuring what scientists actually do and not they say they do. These studies, however, are unable to answer questions such as why an article was downloaded or cited. Clearly, both types of studies are required to develop a more complete picture of the state of access to the scientific literature.