Monday 26 November 2012

Expert Vs Novice: Place Your Bets Now

So what's your selection? You know, to even contemplate such a task takes considerable knowledge, let alone successfully finding the winner! But wait a minute. How come my mate Joe seems to be ahead of the game? In fact, he seems to have the bookies on the back foot. He's been banned by many. I guess you could call him a professional gambler. But is there a difference between how an expert and novice solve problems?  While it is obvious experts know more than novices until recently the lay person's view of the expert might presume their skills were due to a superior mental capacity rather than a vast body of specialist knowledge. 

However, there has been a shift in emphasis on ground-breaking research regarding chess skills. The chess analogy is interesting because not only does it investigate problem-solving strategies but it has the focus of the adversary opponent. 

De Groot (1946/65) conducted a series of chess studies which conflicted with the assumption that skilled problem solvers must have superior information processing skills.  He asked five grandmasters & five skilled chess players to think aloud as they studied a chessboard and choose a move. If grandmasters used such superior information processing they would be expected to make broader searches for their next move. Interestingly, evidence illustrated there was no qualitative difference between the expert and novice. The difference between the two groups was unremarkable - the grandmasters simply made the better moves.  Players were shown chessboards with pieces arranged from actual games. The boards were presented to players for a short time and then removed. They were asked to construct the board positions from memory. The grandmasters constructed the board almost without error while the novice faltered (91% - 41%). Skill level was linked to the amount of information remembered about the chessboard positions. Further research from Chase and Simon (1973) suggest experts not only possess more knowledge but it is organized in more meaningful and readily accessible ways.    

Larkin et al (1980) were interested in the possible strategic differences between experts and novices. They asked expert and novice physicists to solve a range of physics problems. They found that experts tended to use a working forwards strategy. Using the information to derive a solution. Novices use a working backwards strategy starting with the goal. In gambling terms, this would amount to thinking ''I must find the winner''.

It appears experts use their knowledge to generate good problem representations which support working forward strategies while novices rely on trial and error.

It is often said 'practice makes perfect'. But what researchers noticed many years ago that performance improves with practice in a very systematic and predictable way. The 'power law of practise' has been known for a long time. Practice seems to be a factor in the development of skills over a range of activities.  Performance improves with practise because individual task components are executed more efficiently; sequences of task components are executed more efficiently & qualitative changes occur in representations of task structure.    

Performance improves with practise because the time to recovery memory is reduced and importantly sequences of units or chunks. In addition, performance improves because the task is restructured.

But how much practice is needed to achieve excellence? Ericsson et al. (1993) have given ten years as a ballpark figure for attaining high levels of performance in a variety of areas (chess, mathematics, violin playing). Ericsson (1991) suggests that it takes at least ten years to reach the international level of performance in sport, the arts and sciences. Simon and Chase (1973) estimated it took 3,000 hours of practice to become an expert and around 30,000 hours to become a chess master. Many of those who achieve excellence start at a very young age simply because it takes such a long time to acquire the necessary knowledge. 

However, it is possible to train participants to improve on their best performance. Ericsson and Harris (1990) trained an individual who was not a chess player over a period of 50 hours to recognise chess positions almost as accurately as some chess masters. Although Ericsson and Polson (1988) found, the practice itself is not a guarantee of superior performance. In their study, the waiter most skilled in remembering orders used more effective encoding strategies compared to equally experienced counterparts. The critical point is not how much practice individuals have, but what they actually do while they are practising the skill. (This point will be explored in our next article.)