Greetings from the Research School of Computer Science at the Australian National University in Canberra, where I am taking part in an examiner's meeting for the Master of Computing (Honours) sub-thesis. This is my first masters meeting, although I have taken part in the coursework examiners meeting previously. The students do some coursework, a presentation and a report of about 25,000 words. Two examiners, who did not supervise the students, each assess the thesis document separately, then try to come to a consensus.
This can be a difficult process as the examiners come from different part of the computing discipline (some are academics and other from industry). After the two examiners have separately reported their proposed mark with a few paragraphs of comments, the academic in charge looks to see if there is consensus and where there is not, encourages the two to discuss their differences. There is then a meeting of the examiners to provide their input.
Unlike the course-work process there are fewer statistics to help guide the process. The examiners need to avoid the temptation to base their mark on what the students got for their course-work, and just base it on the written thesis. The surprising part of this process is that even where there were divergent views at the start, a consensus between the examiners is usually achieved.
One part of the process which works remarkably well is that all examiners are treated based on their discipline knowledge, regardless of position in the university hierarchy. As an adjunct lecturer, I can happily disagree with the Dean, provided I can back up my position.