Basic concepts in Skilled Behaviour

Question 1

Three stages of Skill acquisition (Fitts & Posner, 1967)

Answer 1

1. Cognitive Stage: Declarative encoding of skill (from self/coach)
   - Understanding the steps
   - Rehearsal as skill is performed (e.g. juggling)
   - Inadequate for highly skilled perf
2. Associative Stage (Knowledge Compilation)
   A. Errors in initial understanding are gradually detected and eliminated (e.g. coordination between clutch and accelerator)
   B. Connections between elements are strengthened (no longer need to think about how to do things)
   Result: Successful procedure for performing skill
3. Autonomous stage (Procedural): Procedures become more automatic
   - Better execution of skill - reflexive (change sin neural activation: less cerebellum, more temporal activ)
   - Declarative knowledge of skill may be lost
   - Can still have problems in production (may not want automatisation if want to improve skill)
   No sharp distinction between associative and autonomous stages - continuum?

Question 2

How does the morse code experiment conducted by Bryan and Harter (1899) support the “Three stages of skill acquisition” theory?

Answer 2

Learning progresses from recognising letters (smaller components) to recognising words (larger components) - different stages of proceduralisation.
Improvement was continuous, but became slower as performance improved
- Rapid improvement at start, then learning slows down

Question 3

What does Snoddy’s (1926) Power law of practice suggest as the trend of performance, as time increases?

Answer 3

T = A + FP^(-b) (classic positive hyperbola trend)
T = Time for task
A = Asymptotic performance
F = Initial perf (constant)
P = Amount of practice
b = Learning rate 

Theory suggests that, as practice time increases, time taken to complete the task gets exponentially smaller, and reaches plateau.

“Improvement can keep occurring forever, though at a diminishing rate.

Question 4

What evidence is there to support Snoddy’s (1926) Power law of practice?

Answer 4

Blackburn (1936): Continuing improvement observed in two subjects performing 10,000 addition problems
Law applies to many skills, even complex ones
E.g. Neves and Anderson (1981): Participants showed log-like improvement when learning justifications for geometry-like proofs

Question 5

Are there any constraints on Snoddy’s Power law of practice?

Answer 5

There are physical constraints - it limits improvement
Crossman (1959) studied Cuban cigar rollers:
- Speed followed power law for 10 years
- Little improvement after a while, because the cigars could not be physically rolled faster.

Question 6

What are the criticisms of Snoddy’s Power of practice law?

Answer 6

Hard to test whether improvement can/not go on forever
Heathcote et al (2000): Learning as exponential function
Aggregate data may follow law, but maybe not for individuals (Haider & Frensch, 2002)
Logan (1988): Not due to better procedures, but better memory processes

Question 7

EXPERT VS NOVICES: DIFF REPRESENTATIONS

What does Ericsson, Chase and Faloon’s (1980) study on recalling digits reveal about expertise and memory training?

Answer 7

One undergrad had to learn and recall sequences of random digits for over 20 months
- Incremental change in training: Improve –> +1 length, -1 otherwise
FOUND: Digit span increased from 7 to 79 digits
- End of session: participants could recall 80% of all the digits from the hour session. As good as any memory expert.

How was this achieved?
Strategies: chunking & association with previous knowledge - meaningful
- When sequences were deliberately not fitting with running times –> perf dropped (but then he supplemented this with ages and dates)

Mneumonics: could only get him to 28 digits, so needed retrieval structures
- shows hierarchical structuring of memory, as experts do
Not sure if his basic memory capacity increased:
- Chunking never more than 4 digits, 6 chunks, and 3 groups
- Switched to letters –> dropped to 6 consonants

Implications:
Practice did not change memory capacities, but improved use of meaningful chunks and structure, which form basis of memory
Practice –> Develop better strategies and structures
(anyone can do it, given enough time!)

Question 8

What does De Groot’s (1965) study on chess grandmasters tell us about Are experts qualitatively different from

Answer 8

Chessmasters: while chessmasters did not have better memories, search deeper or examine more moves, they looked at the right moves from the beginning, within set of available candidates in each turn
- Had better intuition - but what is intuition?
Had better memory for chess positions

Question 9

What does Chase and Simon’s (1973) study on normal and random placement of chess pieces reveal about experts?

Answer 9

Compared chess master and novice recall for normal and random placement of chess pieces
- Scrambled: chess piece positions that would not appear in a normal game
- also examined how the players reconstructed the board.
FOUND:
Experts only showed superior recall of chess pieces for normal chess boards: when pieces were scrambled, they were no better than novices.
Moreover, in reconstructing boards, masters used pawns to position the pieces, while novices used the position of the board to place the pieces.

Also examined what chess masters did when they had to reconstruct the chess positions of an adjacent game.
FOUND: After each glance, they created a meaningful chunk of that position.
- Masters acquire vocab of 50,000 chunks (takes ~10 years)