The PDF/A standard is a component in many institutional repository file format policies. However, full conformance with the standard is often difficult to achieve for born-digital documents, such astheses and dissertations created by students. This study aims to contribute to the growing body ofresearch on the PDF/A standard, focusing particularly on areas which have previously had little academic attention, such as the use of Han (Chinese, Japanese, and Korean) characters in student theses and dissertations.
The research questions addressed by this study are:
- [ primary RQ ] Is PDF/A an adequate file format for creation or conformance of electronic theses and dissertations?
- [ secondary RQ ] What areas of non-conformance to the ISO 19005 standard are impractical to avoid for theses and dissertations deposited in a non-PDF/A file format?
- [ secondary RQ ] Do those areas of non-conformance precipitate considerable preservation risks?
These questions have been addressed through a literature review, interviews with repository staff, and practical application of the PDF/A standard to born-digital and digitized thesis and dissertation documents. The latter was achieved through trial migrations/normalization of over 100 document files held within the Oxford University Research Archive. The datasets consisted of theses and dissertations which were deposited by students between 2007and 2017. The study trialed seven different conformance and creation software currently in use by institutions, and used the PDF validation tool, veraPDF v.0.8, to assess the level of success of the different conformance and creation software.
In addition to assessing the success of migration/ normalization of documents to the PDF/A standard, significant properties of theses and dissertations were considered as a metric for determining success.
The study is significant as it demonstrates different types of PDF/A non-conformances which occurred within a diverse dataset. Results from the study have the potential of informing new risk-based validation policies for use with the veraPDF tool. The study has already informed recommendations for PDF creation at one of the repositories which participated in the study.