Teaching statistics at any level might be challenging with traditional teaching methods. Formative assessment may be a viable candidate to achieve some difficulties faced in classrooms.
Formative assessment is vital if you want your students to be able to actually apply their knowledge! However, you need to ensure alignment between the formative assessment, and your learning outcomes. Too often the formative questions only cover the calculation and recall aspects because it is easier to devise computer marked question banks for these. So they can calculate a statistic and give a definition of a simple random sample for example ( and know they can because of the formative element!) but still have no conceptual understanding so do not know when to use a particular sampling method, nor recognize when a particular statistic is the appropriate one to use, nor are they able to interpret the results and even if they can they are not confident in doing so (because they have not practiced these skills nor received formative feedback on then)
Formative assessment is a way of avoiding the assembly-line model of education in which assessment is employed to accept or reject a product (student) according to pre-determined quality criteria. Everything students do in the teaching-learning process should be conducive to their learning, assessment inclusive. That said, given the way universities and schools are structured and accredited, we cannot do away with summative assessment. However, I believe the former should be employed only at the end of the course.
The nature of Statistics as a course, in my view, makes formative assessment an integral part of the methods of teaching in order for students to master concepts under non-threatening conditions. Many students find statistics challenging because of the many abstract and complicated formulas involved. It helps the students to work things out on their own if formative assessment is applied throughout the course. They also gain confidence in handling a lot of conceptual issues in statistical methods. By the time students have to deal with summative assessment they are in a better position to do well.