A review of: Xia, Jingfeng. “Assessment of Self-Archiving in Institutional Repositories: Across Disciplines.” The Journal of Academic Librarianship 33.6 (Dec. 2007): 647-54. Objective – To test the assumption that authors familiar with subject-based repositories are more likely to self-archive to institutional repositories. Design – Comparative content analysis. Setting – Institutional repositories (IRs) from the following seven universities: Queensland University of Technology (QUT), University of Melbourne, University of Queensland, Lund University, University of Glasgow, University of Southampton, and University of Strathclyde. The IRs included in the study were selected on the basis of repository size and use of EPrints software. Faculty size data and IR deposit policies were drawn from universities’ Web sites. Methods – Each IR was searched to determine the number of deposits in the disciplines of chemistry, physics, economics and sociology. Physics and economics were selected because these disciplines have established internationally renowned subject-based repositories, in contrast to chemistry and sociology, which have not. Deposits from the disciplines were identified from subject terms, keywords and departmental names in metadata records. A “deposit rate” for the four disciplines in each IR was calculated. The metadata records were examined for name of the depositor, date of deposit, full-text availability, item type, and format. Information in the field “Deposited By” was used to identify the extent of self-archiving (that is, deposited by the author). Faculty size for the four disciplines at the seven universities was established from departmental Web site information. For the purposes of making comparisons between the IRs, these data were converted into “rates of faculty” size by dividing the number of faculty in the department by the total number of faculty at the institution. A weighted rate of deposits by discipline was calculated by dividing the rate of faculty size by the deposit rates. To take into account disciplinary differences in publication productivity, these rates were subjected to further analysis. Using an “average publications per year” calculation for each discipline (from a 1977 paper), a final weighted rate of depositing was calculated for the four disciplines in the seven IRs. Main Results – Without weighting for faculty size, deposit rates vary greatly between disciplines. In most institutions, deposit rates for chemistry and sociology were higher than rates for physics and economics. When faculty size is controlled for, the highest deposit rates in five IRs were for chemistry and sociology. Only two IRs were found to have the highest deposit rates for physics and economics. These results did not change overall when the weighting for publishing productivity was applied: the same five IRs had highest deposit rates for chemistry and sociology. Exceptions to these findings were the IRs at University of Melbourne and University of Queensland, where the highest deposit rates were for economics and physics. On examination of depositor information, it was found that only 2.3% of economics deposits in the Melbourne IR were self-archived. Administrative assistants and other staff were responsible for depositing 97.7% of the IR’s economics holdings. Self-archiving of physics items to the Melbourne IR was 90%; however, these deposits comprised student theses and dissertations only. Self-archiving practices were examined for: chemistry, physics and economics deposits at the University of Melbourne; chemistry and economics at the University of Queensland; and chemistry, physics and sociology at Queensland University of Technology (the only IR in the sample with a mandatory deposit policy). Like Melbourne, self-archiving of economics deposits at the University of Queensland was also low, at 17%. Of the remaining economics deposits, a librarian was responsible for depositing 68%. Chemistry deposits at both Melbourne and Queensland had much higher self-archiving rates, 76.2% and 100% respectively, than those found for physics and economics. At QUT, where deposit into the IR is mandatory, self-archiving rates are high for the three disciplines for which findings are reported. The self-archiving rate for chemistry was 68.3%, sociology 46.3%, and physics 42.9%. A librarian was responsible for the majority of the remaining deposits. Conclusion – This research tested the proposition that disciplines familiar with subject-based open access repositories, such as physics and economics, are more likely to contribute to IRs. Its findings did not support this view. Instead, the study found no particular pattern of deposit rate across the four disciplines of chemistry, physics, economics and sociology in the seven IRs. Operational aspects of IRs, such as assisted and mandated deposit, appear to have a more significant effect on deposit rates. Assisted deposit, either through departmental administrative staff or librarians, accounted for relatively high deposit rates for economics in the Queensland and Melbourne IRs. Deposit date information in the Queensland IR suggests administrative staff of the economics department deposit to the IR on an ongoing basis. Students showed a high rate of self-archiving for theses and dissertations. It might be speculated that a mandate policy at Queensland University of Technology is responsible for the high self-archiving rates seen for economics, chemistry and sociology. However, librarians have assisted in the process, depositing over half the items for physics and sociology. The author recognises the value of both assisted and mandated deposit, but raises questions about how this will affect faculty use of IRs. For example, in cases where faculty have no role in contributing to an IR and therefore no familiarity with it, will they in fact use it? Another important consideration is the policy approach taken to temporary faculty and a mobile academic workforce. In conclusion, the author states, “Institutional repositories need a mandate policy to ensure success”.