The day started with **Jane White** and **Paola Iannone** on *Assessment practices and learning mathematics*. There was some discussion of the rationale for different forms of assessment, and then Paola reported on an interesting survey that asked students to rate different forms of assessment by (a) their personal preference; (b) their perceived ability to discriminate. For mathematics students there was some correlation. The traditional closed book exam was the most preferred option, and also regarded as having the greatest discriminatory power.

From interviews with students several other themes emerged: students dislike coursework, knowing how easy it is plagiarise (it was quickly pointed out that plagiarising *successfully* is not so easy), and group assessment, perhaps because of the ‘free-rider’ problem. They value fairness highly, and were surprisingly open to unusual forms of assessment. Several suggested short oral exams would give them a fair chance to show what they could do. Paola reported that vivas are used instead of one problem sheet in a course at UEA. The workload is roughly equivalent, but shifts from Ph.D markers to the lecturer.

The next talk was **Kevin Houston** and **Mike Robinson** on *The flipped classroom*. Kevin talked about flipping a first term ‘Number Systems’ course at Leeds. I lectured a similar course at Royal Holloway for three years, with Kevin’s book as required additional reading, and most recently gave a ‘partially flipped’ follow on course on Matrix Algebra, so naturally my ears pricked up. Before flipping, Kevin’s course had 10 weeks of 1 hour lectures, with a 1 hour weekly tutorial in groups of 10 to 13. After flipping, it has two 1 hour lectures, on Monday and Fridays. On Monday he lectured in his usual style; on Friday students were expected to have worked through written notes in advance, and the hour was spent solving problems and reviewing the key ideas in proofs. (I’m not sure if Kevin fully adopted Eric Mazur’s peer instruction cycle.) He used socrative to run quizzes in lectures, relying on students to own a smartphone: they all did. I’d be worried about students getting distracted, but there are clearly advantages over clickers: for instance, students can submit short text responses, so questions do not have to be purely multiple choice.

Students were encouraged to do the preliminary work by an assessed online quiz, available before each Friday lecture. Full exam credit was available for any result above 50%, with multiple attempts permitted — this seems a neat way to provide some additional motivation, while keeping an essentially formative character to the assessment. Exam results were ‘comparable to previous years on a slightly harder course’. Student engagement was significantly improved. Kevin commented on the high workload in preparing the new course for the first time, and the need for very accurate timing to get students to the right point on Monday evening.

Mike then talked about his experience flipping a second year course ‘Dynamical systems and Fourier analysis’. Before flipping, this course had a 1 hour lecture, and 1 hour practical workshop in a PC Lab. Post flipping, it had a single 2 hour class, held in a room with plentiful laptop computers. The preliminary work was to watch short online videos prepared by Mike. The advantage of online videos over lectures, that the viewer can pause at any moment, was appreciated by the students. They also observed, however, that there was no immediate opportunity to ask questions to the lecturer or their peers. Mike reported on a survey, addressed only to serial non-attenders at this course: a typical non-attender missed classes for reasons unrelated to the courses, but a few, who admitted not doing the preliminary work, felt their lack of preparation would be exposed. Another survey showed the new course was very popular with students, with many seeing only advantages to the new format.

I would like to know whether strong or weak students benefit more from the flipped classroom. My guess is that flipping can help weak but hard-working students to get somewhere, rather than nowhere, but is maybe of less benefit to stronger students. And under any system there will be a core of students who’ll fail no matter what we do. Kevin made the interesting point that many students adopt a pragmatic attitude and aim to do just enough in any given course. So it is unrealistic to expect huge improvements in exam performance, even from successful interventions.

All speakers were enthusiastic about their experience; both Kevin and Mike felt they ‘hadn’t flipped enough’. A recurring comment was that, post flipping, they had more meaningful interaction with the students. Paola commented that it was often surprising how much students turned out to know. I agree. Going the other way, I’m now only rarely surprised by the many essential things (definitions, for example) that pass them by, causing endless stumbling blocks. The flipped classroom may create — at least in the short term — more work, and is probably not suited to everyone, but it offers a way to tackle these problems head on.

The final talk was by **Peter Giblin** and **Alice Rogers** on *The Teaching Excellence Framework, HE White Paper, and Quality Assurance*. Peter gave an excellent summary of the white paper (he said slides would be available from the LMS website), highlighting the somewhat strange way in which the TEF is expected to be introduced, at the institutional level very soon (but with everyone who agrees to play guaranteed at least the minimum ‘Meets Expectation’ rating), and at the discipline level only by 2019/2020. Peter quoted one particularly discouraging requirement, buried in the additional technical documentation, that submissions should ‘avoid focusing on successful but highly localised practices’. The whole game seems to be a bureaucrat’s dream and an academic’s nightmare.

There was much uncertainty about the way departments would be assessed, and considerable debate about what, if any, benchmarks could fairly be used. On this note, the day ended.