Extreme Learning – Professor Mike Sharples’ Inaugural Lecture
28th September 2006

Reminder: good source of e-learning info from the Pew Learning Report 2004 (available online at www.pewinternet.org)

Often learning is perceived as difficult – and ways to overcome these difficulties to enable ‘easier’ learning are sought. Discussed the effectiveness of different types of learning – for example computer-assisted learning has been shown to be less effective than others (Parr et al meta-study). Widely perceived that best learning should be relaxed and in a non-confrontational environment. But sometimes learning under stressful or hazardous conditions is also successful – e.g. skiing (see Burton et al 1984). But how is learning measured as “effective” in these studies? Often quantitative data e.g. an intervention is only used if it produces Ð0.4 difference in learning outcomes (Pask, 1961). But teaching is the only control to learning – learning is easy it’s the TEACHING that is hard! This can be illustrated by studies showing most adults are involved in some kind of learning (95% according to Livingstone’s study). In this study learning episodes were self-defined and logged in personal diaries. The episodes occurred mainly at work (e.g. learning how to order a flower delivery) but could be separated into 3 distinct types of “everyday” learning :a)work; b)leisure (e.g. studying family history) –and c) “survival” e.g. tax returns on-line.

Argues that the ease of learning also illustrated by babies of only 9 months understanding speech patterns in that they attended more to previously heard words. They were learning by active repetition (see Morris & Fitz, 2000).

Described the MEDIATE study – a computer-based knowledge system where language is being developed to describe MRI scans – this is especially important in the case of abnormalities. Briefly, in this study 1200 MRI data where disease was known (through autopsy etc) were described as typical or atypical within a visual multidimensional pattern of an “80 dimensional” space. Importance of consistent, reliable language for each case – especially for trainee doctors. Word and image and repetition. This is echoed in successful models of learning, e.g. the managed, motivational and guided learning of “scaffolding” (see David Wood’s work, 2001).

So to maximise effectiveness of learning it must be supported and challenging – informal and incidental. But crucially it must contain an element of (perceived) risk whilst remaining challenging to the learner. In the past we have wrongly eliminated risk from the learning environment. But how to present “real” risk? E.g. consider the difference between a learner using a virtual calculator and a flight simulator. The item’s physical use is therefore important. Can simulated environments provide real risk? Transformational, empowering learning through positive trauma has been shown to lead to radical self-development (Woods, 1996). Extreme learning – you never know when it’s gonna happen!