I have 2 groups of students, both groups have pre-post qual and quant questions following a teaching intervention.
The students have 4 clinical skills sessions taught by academic staff, 2 sessions taught by students (but observed by lecturers, for university purposes), and 2 sessions taught solely by students.
The sessions will all follow a teaching plan (with specific outcomes)
The outcome is to see whether the student taught session are comparable/better/worse than the staff taught sessions.
Quantitative questions are Pre-post tests and and approved Likert Scale relating to teaching preferences.
Qual questions are provided for students to expand their answers.
What is the best option for analysing the quantitative data.
This is a feasibility study consisting of 40 students being taught, there are 9 student teachers, and 2 academic teachers.
They are split into 2 groups for timetabling purposes. Group 1 have their student teaching first then their academic teaching. Group 2 have their academic teaching then their student teaching.
Normal academic teaching is 1 academic to 20 students, while the student teaching consisted of 1 student teacher to 5 students
Thanks for any help.
Paired sample t test is most commonly used in the scenario.
You may also apply chi square phi or Cramer's test if you want to know the difference in both results.
Actually, test are suggested after having a look at the data.
A possible way of implementing it is just using a simple difference-in-differences. For instance, you can look at the 6th chapter of Angrist and Pischke's Mastering metrics.
In your case, it would be a regression like that
yi = b0 + b1*post_dummy + bk*individual_dummies + b2*treated_group_dummy*post_dummy
https://www.researchgate.net/post/When_one_should_use_the_likert_scale_for_a_questionnaire_and_how_do_you_interpret_the_data_collected_from_such_a_questionnaire
- Badges
- Science topic