Icwl2014 foredrag

ICWL 2014, 13-17 August 2014

How a Flipped Learning Environment Affects Learning in a Course on Theoretical Computer

ScienceDorina Gnaur!Department of Learning and Philosophy, Aalborg University, Aalborg, Denmark!!Hans Hüttel !Department of Computer Science, Aalborg University, Aalborg, Denmark

A well-known ritual

What’s the use of lectures?

?

Do lectures help students learn?

As early as 1971, Donald Bligh found that lectures might be good for transmitting information as effective as other media, but not good for!promoting thought or for changing for student attitudes

Deep learning vs. surface learning

Ferenc Marton and Roger Säljö (1976) !

✤ Surface learning: Learning in order to reproduce; learning the “signs”!

✤ Deep learning: Learning in order to understand; learning “what signs signify”!

(This has nothing to do with learning styles.)

Strategic learning

✤ Many students use strategic learning: learning (only) what they perceive as “necessary”. Sadly, strategic learning tends to be surface learning.!

✤ How can we ensure that the learning goals of the curriculum and deep learning can live together? Lectures may not be the best solution.

“The flipped classroom”

The “flipped classroom” is a recent, web-based strategy for learning:!

✤ Present new material to students in the form of downloadable podcasts!

✤ Devote the time spent together in plenary sessions to practice

An ideal “flipped class” cycle

Watch podcast Read text

Plenary session

Our hypothesis

Use a “flipped classroom” approach that is appropriate for the learning goals together with web-based peer reviews.!

This can be a good strategy for encouraging deep learning.

…but what does “appropriate for the learning goals” mean?

Signature pedagogies

✤ Signature pedagogies (Shulman, 2005) are the characteristic forms of teaching and learning associated with particular professions.!

✤ They define what counts as knowledge in a field and how things become known. !

✤ Cf. the chalk talk approach to teaching mathematics.

Threshold concepts

✤ Threshold concepts are the concepts one will need to master in order to “cross the threshold” to a subject area. (Meyer and Land, 2003)!

✤ Threshold concepts tend to be difficult and counter-intuitive.

A case study

A computer science course

✤ The course Computability and Complexity!

✤ 5 ECTS, 5th semester of the computer science and software undergraduate programmes!

✤ Written exam, external examiner!

✤ Textbook-based

This course is central to most undergraduate programmes in computer science.!!

Michael Sipser: !Introduction !to the Theory of Computation !is a widely used text.

The structure of the course

✤ 7 sessions about computability theory!

✤ 1 session devoted entirely to problem solving!

✤ 7 sessions about computational complexity theory!

✤ All materials available from Moodle.

Threshold concepts of the course

✤ In Computability and Complexity the threshold concepts are!

1. Precise definitions and their use in theorems and proofs of theorems!

2. Reducibility (many-to-one and polynomial time) as a way of relating decision problems wrt. their computational difficulty

Signature pedagogies of the course

✤ Going through definitions, theorems, proofs and examples!

✤ Lecturing at the blackboard (chalk talk)!

✤ Problem solving

Why text-related questions?

✤ Text-related questions are not a necessary component of a flipped classroom. The text-related questions were thought of as an additional strategy for deep learning and for putting the emphasis on threshold concepts

Text-related questions

✤ A small number (5 to 8) of questions about central aspects of the text associated with a session!

✤ Answers were to be written in LaTeX and uploaded to Moodle.!

✤ Answers were to be commented by a fellow student.!

✤ The collection of answers constituted the portfolio; this was the only aid allowed at the exam.

Our “flipped class” cycle

Watch pencastRead text

Plenary problem!solving session

Answer text-related!questions

Peer review answers to!

text-related!questions. Next

revise your own answers

Problem solving

✤ Problem sets. Their structure and content must be similar to what students would see at the exam (constructive alignment)!

✤ Solutions to problem sets. All solutions to all problem sets eventually became available via Moodle as the course proceeded.

Pencasts

✤ 1 pencast = a collection of video clips (10 to 15 minutes each), each presenting a part of the content of the session!

✤ Pencast = handwritten text + figures + voiceover!

✤ A pencast should be in the spirit of a blackboard lecture (writing while talking slows down the pace, showing slides tends to do the opposite)

Recording and editing

Wacom Bamboo stylus

Klipsch headsetiPad 2

Doceri app

…but how can we check if the strategy achieved what it set

out to do?

A survey

✤ A questionnaire for all students!

✤ Qualitative interviews

A survey

To which extent did you experience that the pencasts of Computability and Complexity supported your learning?

A survey

Which combination of activities was the most helpful for your learning?

What the students said

✤ Honest about their intent to use strategic learning!

✤ Pencasts helpful when preparing for the exam!

✤ Miss interaction with teacher

What the teacher learned

✤ Video editing takes an awful lot of time if one is not careful (the victory of the latent perfectionist?)!

✤ A challenge to keep up a good interaction with all students!

✤ Students focus too much on text-related questions as an exam aid (the note-taking fallacy = shallow learning)

What did we learn?

Is this the philosopher’s stone?

Some challenges

✤ How can the teacher ensure good interaction with students in a flipped classroom setting?!

✤ We are all born into a tradition of lectures and lecturing and tend to act accordingly.!

✤ When should we abandon signature pedagogies?!

✤ Is being able to produce good video material simply yet another demand imposed upon us as university teachers?