A Direct-Observation Coding Instrument for Doctoral Viva Questions: Retrieval, Extension, Adjacency, and the Examination of Claimed Contribution

Research on doctoral examination has developed over twenty-five years, but it has relied almost entirely on indirect methods: interviews with examiners, analysis of examiners' written reports, candidate recollection, and policy review. The reason is structural rather than a matter of research priorities. As Morley, Leonard and David (2002) noted, the viva is rarely witnessed by anyone outside the examination process, which constrains both monitoring and research into the assessment itself. In the systems that produced this literature, the doctoral defence is closed. In Jordanian universities, and in a number of other systems, the doctoral defence is a public session. The methodological barrier does not apply, and direct observation becomes possible. This instrument supports the systematic observation and coding of doctoral defences along two axes. The first classifies every question posed on a single formal criterion — whether the answer is available within the thesis itself — across four categories: direct retrieval, extension, adjacent, and formal. The criterion is deliberately formal rather than evaluative, so that coding does not depend on the observer's judgement of question quality. The second axis records whether the contribution to knowledge required by the institution's own regulations is explicitly examined, or is instead assumed from the existence of the document. The instrument additionally captures response latency, questioner role, and the verbatim wording of the contribution question where one is posed. The instrument is released for reuse in any system where doctoral defences are held in public. It includes coding rules with worked examples, boundary-case resolutions, an inter-rater reliability protocol, and ethical constraints prohibiti