anique de bruin erasmus university rotterdam metacognition and cognitive load the effect of...
Post on 16-Dec-2015
217 Views
Preview:
TRANSCRIPT
Anique de BruinErasmus University Rotterdam
Metacognition and cognitive load
The effect of self-explanation when learning to play chess
Goal present studies
• Initial framework:
Expertise development
• Deliberate practice (Ericsson et al., 1993):
metacognitive activities crucial for expertise development
Rehearsal and correction of errors
Metacognitive strategies
• Self-explanations:
– Process more deliberately
– Recognizing inconsistencies
– Stimulate integration new information
Enhances accuracy metacognition
Metacognition in skill acquisition
To what extent do metacognitive activities foster learning in skill domains (chess)?
- non-verbal nature of material
- no explicit information provided
- novices
Self-explanations in chess
• Three groups (N = 15 per group):
1. Observation only (O)
2. Predict next move (PO)
3. Predict and self-explain next move (PSE)
Procedure
• Three phases: – Basic rules
– Learning phase:
• Predict and self-explain
• Prediction only
• Observation
– Test phase: play against computer
Discover chess principles
• Chess rules: too little information to play endgame
Chess principles necessary:• King checkmated at the edge of the
board• Rook minimizes space of the King
What instruction fosters development of principled understanding most?
Results learning phase
Principles applied in learning phase
3040
506070
8090
1 2
Practice Session
Perc
en
tage
pri
nci
ple
s co
rrect
PSE condition
PO condition
Self-explanations
• Three categories:
1. Basic chess rules
2. Partial explanation of principles
3. Complete explanation of principles
Results self-explanations
• Median split on number of SEs:• High-explainers: >51 (mean=95.1)• Low-explainers: <51 (mean=32.4)
Compare differences in SE and chess performance between high- and low-explainers
Chess rule explanations
0
5
10
15
20
25
30
35
1 2
Session
Nu
mber
of
self-e
xpla
nati
on
s
High-explainers
Low-explainers
Partial principle explanations
0
5
10
15
20
25
30
35
40
1 2
Session
Num
ber
of s
elf-
expl
anat
ions
High- explainers
Low- explainers
Complete principle explanations
0
1
2
3
4
5
6
7
1 2
Session
Num
ber
of s
elf-
expl
anat
ions
High- explainers
Low- explainers
Results self-explanations
• Test exercises: high-expl more checkmate than low-expl
• However: No difference in time needed to self-explain
Results test phase
0
10
20
30
40
50
60
70
80
90
1 2 3 4 5
Test exercise
Mea
n p
erce
nta
ge
chec
kmat
e PSE condition
PO condition
Observationcondition
Cognitive load
• From CL perspective surprising:– Despite low prior knowledge, prediction
+ self-explanation foster learning better principled understanding
What (meta)cognitive mechanisms explain SE effect in novices?
Conclusions I
• More explanation of basic chess rules better discovery of principles
• Rehearsing basic rules frees up processing resources for principle discovery
Conclusions II
• Verbalization of self-explanations crucial: No effect in PO condition
– Meaningful self-explanations
• Wording of the SE instruction: Explain why the computer would make that move
• No re-reads (as in text learning) possible
– Verbalized (partial) discoveries of principles receive more activation in WM
Future research
• Examine covert self-explanation in PO condition
• Test effect SE only
• Manipulate rehearsal of basic chess rules to test effect on principle discovery
Thank you
Questions?
top related