Flipped classroom and online learning: Personal FAQ

I spent most of June 2 and 3 attending two meetings about online learning: Flexible education (organized by Royal Holloway), at which a College wide teaching model was unveiled, and Teaching and Learning Mathematics Online (organized by Michael Grove, Rachel Hilliam and Kevin Houston). A recurring theme was ‘Active Blended Learning’ or the ‘Flipped Classroom’.  The purpose of this post is to record some observations made by speakers at these talks and my reflections on the research literature on these pedagogic models. Please may I emphasis that this post has absolutely no official status.

Does the College model require the flipped classroom?

This isn’t said explicitly, but I think it is implicit throughout. For example in the seven step programme that we are supposed to follow for each ‘teaching week’, we read

• 3. Engage with learning material. Provides time for students to independently or in groups to engage with learning materials.
• 4. Learning activity. Enables the practical and critical application of learning through individual or group activities.
• 5. Learning Check. Allows the student and lecturer(s) to check that the student learned the content and achieved the learning outcome(s) for the week through an activity or formative/summative assessment.

All this seems to sit fairly happily in a framework where students are expected to work off-line and then come together for synchronous problem solving sessions, maybe checking their understanding before/after by an online quiz. It does not, as as I see it, fit with the traditional model of three live lectures per week.  

What is the research evidence for the flipped classroom and active blended learning?

Freeman et al Active learning increases student performance in science, engineering, and mathematics, PNAS 111 (2014) 410–8415 is a meta-analysis of 225 studies that compared student performance in courses lectured in the traditional style and student performance courses with ‘at least some active learning’. The conclusion is clear: active learning improved student performance by about 1/2 of a standard deviation, and, more strikingly, the average failure rate was 21.8% under active learning compared to 33.8% under traditional lecturing. Measures of student engagement and satisfaction were not considered in this meta-analysis.

A survey talk by Robert Talbot, updating the conclusions of his 2017 book Flipped Learning: A Guide for Higher Education, also has clear conclusions. From the linked slides (emphasis preserved):

• Students in FL courses typically show either greater gains in measures of learning than students in traditional courses or else the differences are not statistically significant (slide 16).
• Students show higher satisfaction with FL and the active learning techniques once FL is in place (slide 18).
• Students often highly negative about FL when first introduced … even while acknowledging benefits of increased group work, more instructor attention, and better grades (slide 18).

In Jacqueline O’Flaherty and Craig Phlips The use of flipped classrooms in higher education, Internet and Higher Education 25 (2015) 85–95 the authors conclude in Section 4.4

Our review indicates a number of positive student learning outcomes from this pedagogical approach,

but adds:

This review found very few studies that actually demonstrated robust evidence to support that the flipped learning approach is more effective than conventional teaching methods. Only one study used empirical validation to show that a structured flipped classroom in comparison to the traditional one could effectively engage students in deep learning [Hung, Flipping the classroom for English language learners to foster active learning]. Whilst some studies referred to a modest improvement of academic performance, through outcomes of increased examination scores or improved student satisfaction, further research is required in this area …

The cited study was a randomised controlled trial comparing two flipped models with the traditional lecture model on the same content on 75 students.

What is ‘flipped learning’ anyway?

Talbot’s suggested definition (see about half way down) is

Flipped Learning is a pedagogical approach in which first contact with new concepts moves from the group learning space to the individual learning space in the form of structured activity, and the resulting group space is transformed into a dynamic, interactive learning environment where the educator guides students as they apply concepts and engage creatively in the subject matter.

He is clear that flipped learning does not necessarily require videos, and that a methodological flaw in some metastudies is that they exclude teaching models where videos were not used. Neither does flipped learning necessarily require lecturing. Instead a key feature is that the first contact with new concepts comes from a structured activity, done on one’s own. On my reading, the instructors’s duty is to plan this activity, and then enable students to apply the ideas creatively in individual and group work. Reassuringly (see slide 20), he reports

Implementation matters but not as much as simply offloading direct instruction into structured student activity and using the class time for active learning.

Finally I’ll mention a study by James McEvoy, Interactive problem-solving sessions in an introductory bioscience course engaged students and gave them feedback, but did not increase their exam scores. James ‘partially-flipped’ a Royal Holloway biology course by replacing one of two weekly lectures with an interactive problem solving class (see page 5 for details of how this was run). After the change there was a significant improvement in student responses to the two questions ‘The teaching style was engaging’ and ‘I received feedback on my progress during the course’. However other measures of student engagement, and (as the title makes clear) exam performance were not significantly affected; there was however a reduction in failure rate.

McEvoy’s findings are consistent with the Freeman et al study and my belief that the flipped classroom may benefit weakest students the most, by helping them to get somewhere, rather than nowhere (as alas, is too often the case in mathematics courses).

How does one make the flipped classroom work?

Here are some points that I jotted down from the various talks. Sue Pawley’s talk at TALMO Is there anyone out there? A guide to interactive activities in the online environment was especially useful.

• Make all live sessions worth attending. Don’t deliver the lecture (again), but instead get students to work in groups, learning from each other and you.
• Getting feedback from students is difficult: do not expect many students to speak in front of their peers. They will never speak unless recording is switched off. Quizzes are good and technologically easy. I want a system where students can collaborate in small groups on a ‘digital whiteboard’; at the moment this seems to require high-end tablets.
• Give students a ‘scaffold’. Don’t just say ‘read this’. Instead say ‘read this, watch that, do this quiz to check your (basic) understanding’.
• Do not use peer review since this just adds to the fear of being wrong. But critiquing anonymous wrong answers (we could even fabricate them) can be useful.
• One thought from me: in all the online talks, the convener collated questions in the chat and then selected the most useful (or most upvoted) to ask. This seems a clear improvement on the traditional model where the most assertive person gets to set the initial subject and tone of the questions.

What ideas are there for quiz questions?

George Kinnear’s talk at TALMO Using quizzes to deliver a course online had many useful ideas, going beyond the usual multiple choice model.

• Faded worked examples: leave gaps in a model proof or solution for students to fill in.
• Invite students to create examples (I think they will need a lot of hand-holding).
• The ‘test effect’: recalling information from memory is beneficial (in the context of the always-connected internet generation, I think this deserves saying). For instance, before a section on integration, he got students to make a table of important functions (left column) and their derivatives (right column).

I very much hope we will be able to use the STACK computer algebra system (which integrates with Moodle) to set questions. It can validate even quite complicated algebraic answers and give immediate feedback. For instance, it was used to mark the integration quiz above, and even suggest important functions the students had not used.

My own view is that quiz questions used to review material should be very basic, while questions in interactive workshops should be much tougher and focus on potential misconceptions and common errors.

Is this consistent with academic freedom?

I hope so, but at a recent EPMS meeting, the Head of School admitted some freedom would be lost. Personally I’m keen on moving to a flipped model, but I will defend to the (academic) death the freedom of colleagues to teach as they see fit, even when I firmly believe what they are proposing will be ineffective. As the slide below from James McEvoy’s talk shows, the lecturer’s preference for the flipped/traditional model is correlated with student performance.