Conversation Analytic innovation for Teacher Education (CAiTE)


How can we train student teachers to be effective in their assessment practices? The CAiTE project aims to enhance the quality of teacher education and improve teachers' assessment practices.

How can we train student teachers to be effective in their assessment practices?

The CAiTE project (Conversation Analytic innovation for Teacher Education) aims to enhance the quality of teacher education and improve teachers' assessment practices by developing, evaluating and implementing a new research-based instruction method, based on the principles of CARM (Conversation Analytic Role-play Method), to train student teachers' feedback skills.

Based on conversation analysis of oral exams and teachers' supervision of group work in the classroom, the project has developed, piloted, executed and evaluated a CARM workshop for student teachers. Preliminary findings show that guidelines for feedback and supervision are experience-based and lack theoretical and empirical grounding. Empirical analyzes of 36 oral exams show that there is considerable variation in time spent on the various parts of oral exams. The exams consist of five main activity phases sequenced chronologically. The five phases are greeting and preparing, student presenting, subject conversation, grading, and examiners giving feedback and announcing the grade to the student. There is considerable variation in how much time is spent on the different phases. The shortest student presentation lasts 5 minutes and 16 seconds, while the longest lasts 16 minutes and 18 seconds. The shortest subject conversation lasts 7 minutes and 10 seconds, while the longest lasts 23 minutes and 2 seconds. There is also variation in how large a part of the curriculum is covered during student presentation and subject conversation. There is considerable variation in the number of questions asked and topics covered in the professional discussions between examiners and candidates. The examiners’ open, leading, and multiple questions during the oral examination put constraints for what it is possible to answer, and these questions can have both a scaffolding-building and inhibiting effect on the candidates' opportunities to demonstrate competence. Sometimes candidates give an answer that is marked as incorrect. The examiners then correct the candidate and provide the correct answer for them, or they attempt to get the candidate to self-correct by implicitly or explicitly initiating repair. That gives the candidate a new opportunity to answer correctly, but often leads to long side-sequences where the candidate does not necessarily have opportunity to show knowledge. 

The grading discussions are strongly consensus oriented and hold a high level of agreement. Disagreement is displayed implicitly. In 2 of 36 grading discussions, the internal and external examiners’ display diverging positions. The external examiner has the final say, but the internal examiner may influence the grade in benefit of the candidate. Proposing a grade is not a straightforward action; rather, it involves a stepwise calibration process where the assessors use different implicit resources to adjust their individual positions towards the candidate’s performance and secure alignment, before proposing a candidate grade. Scoring rubrics and explicit assessment criteria are to a small extent referred to in the discussions, but to a larger extent in VG3 compared with 10th grade.  

Analyzes of teachers' guidance in the classroom show how teachers initiate, pursue, and solve problems via verbal, prosodic and physical resources. Groupwork is a regular activity in Norwegian classrooms and provides opportunities for one-to-few interactions between students and their teacher. How do teachers decide when and how to offer help when not explicitly asked to do so? Teachers and students rely heavily on bodily conduct to display whether assistance may be offered or needed. Instant problem solving can thereby be achieved without disturbing students’ progress and without taking too much time. The teacher's presence and movement in the room are important for the students' opportunities to receive guidance.

The CARM workshop was piloted as an intervention in a randomized controlled study during 2019-. The first pilot of the intervention, autumn 2019, showed that the teacher students were mostly satisfied with the content and relevance of the course and ranked relatively high on self-awareness regarding spoken interaction in teacher-student conversations. A revised version of the intervention and evaluation will be carried out in October 2020. The results from this intervention showed statistically significant differences between the intervention group (study participants who took part in a CARM workshop ahead of the evaluation) and the control group (study participants who did not take part in CARM workshop ahead of the evaluation) on self-awareness. The results also showed a high level of acceptability of the CARM workshop: the participants found the workshop useful for their development as teachers, and they found its content and level of difficulty appropriate for teacher students.


Trening av samtaleferdigheter i lærerutdanningen (CAiTE)

Core research team

The project is financed by The Research Council of Norway (NFR) through the FINNUT-programme and consists of 5 researchers and two PhD.


University College of Southeast Norway, Faculty of Humanities, Sports and Educational Science

Loughborough University

Loughborough University, Department of Social Sciences

Collaborating projects

LOaPP: Learning Outcome across Policy and Practice (Læringsutbytte i politikk og praksis)