M. Pariès, A. Robert and J. Rogalski recently published an article called “Analyses de séances en classe et stabilité des pratiques d’enseignants de mathématiques expérimentés du second degré” in Educational Studies in Mathematics. The article is in French, but here is the abstract in English:

In this paper we tackle the issue of an eventual stability of teachers’
activity in the classroom. First we explain what kind of stability is
searched and how we look for the chosen characteristics: we analyse the
mathematical activity the teacher organises for students during
classroom sessions and the way he manages the relationship between
students and mathematical tasks. We analyse three one-hour sessions for
different groups of 11 year old students on the same content and with
the same teacher, and two other sessions for 14 year old and 15 year
old students, on analogous contents, with the same teacher (another
one). Actually it appears in these two examples that the main
stabilities are tied with the precise management of the tasks, at a
scale of some minutes, and with some subtle characteristic touches of
the teacher’s discourse. We present then a discussion and suggest some
inferences of these results.