I'd suggest that it wouldn't tell us anything about why tutors behaved in the way they did, or of the extent to which tutors were responding to the behaviour of the students themselves, in that the tutor's gendered actions may have been a response to their gendered behaviour. Also, the apparent objectivity afforded by ticking boxes disguises the judgements made beforehand about what are the significant categories and during about how to interpret the meaning of different utterances. On the other hand, the quantifiable data would be an interesting measure once supplemented by other sorts of information and the categories and the coding devices could be refined so as to produce a degree of standardisation and reliability.