Post on 08-Jul-2020
transcript
Measuring and improving self-‐regulated learning From cogni6ve psychology to educa6onal prac6ce Anique de Bruin, PhD
Department of Educa6onal Development and Research Faculty of Health, Medicine, and Life Sciences
> 50% 0
10
20
30
40
50
60
70
80
90
1st 2nd 3rd 4th
Percen
tages
Performance quar0les
Test performance
Es6ma6on
Why are poor self-‐judgments in educa0on troublesome?
Overconfidence decreases
performance and causes
underachievement
Poor monitoring
Poor Self-‐regulated
learning
Lower academic performance
Why and when are we poor at monitoring our learning?
Why don’t we learn from our experiences?
Why and when are we good at monitoring our learning?
How well will you do on the test?
Metacognitive intervention
The typical paradigm
Text Problem Skill
Who’s good at it?
Gender
Intelligence
Personality
Knowledge
0
20
40
60
80
100
1st 2nd 3rd 4th
Does it maJer?
Knowledge
0
20
40
60
80
100
1st 2nd 3rd 4th
YES, DEFINITELY Experts excel in self-‐reflec0on (Ericsson et al., 1993) Students with more knowledge are beJer at self-‐judging
“DOUBLE CURSE”
Does it maJer?
Intelligence
NOT REALLY It’s more WM, if at all Veenman (2013): highest IQ à poorer metacogni0ve skills.
Does it maJer?
Gender
Intelligence
Personality
Does it maJer?
Personality
POSSIBLY SOMEWHAT Narcissism & op0mism (De Bruin, Kok, Lobbestael, & De Grip, in prep)
Higher narcissism à more overes0ma0on Posi0vism: ? Not directly
Does it maJer?
Gender
YES, SOMEWHAT Males are usually somewhat more overconfident, but: Females are also overconfident And females usually score higher
Who’s good at it?
Gender
Intelligence
Personality
Knowledge
0
20
40
60
80
100
1st 2nd 3rd 4th
Individual differences maQer in a small to moderate extent, but never do they preclude the
possible effect of instruc6onal interven6ons
When are we good at it?
Task complexity & WM
Implicit rela6on
to judgment
Genera6on Rela6ve vs absolute accuracy
Timing of judgment
When are we good at it?
Timing of judgment
YES, DEFINITELY, BUT (D-‐JOL effect, Nelson & Dunlosky, 1991)
Delay ensures LTM retrieval. If it is in WM now, it doesn’t have to be at the test. No (or reversed?) effect of delay in problem solving tasks (Baars et al., in press)
Memory
Judgment
Shared areas
(Falbo et al., submiJed)
When are we good at it?
Genera6on
YES, DEFINITELY Ac0vely genera0ng informa0on from LTM vs passive reading Diagrams, keywords, sentences (Van Loon et al., 2013; De Bruin et al., 2011) à Usually, only aier a delay Self-‐tes0ng
(Van Loon et al., revision resubmiJed)
When are we good at it?
Task complexity & WM
Implicit rela6on
to judgment
Genera6on Rela6ve vs absolute accuracy
Timing
Cogni6ve mechanism behind adequate monitoring
CUE UTILIZATION MODEL
(Koriat, 1997)
People use informa6on (‘cues’) that is most available (comes to mind easily) when making a self-‐judgment.
These cues are oeen invalid and superficial (e.g.,
accessibility, familiarity).
By improving the cues people use, we can improve monitoring.
METACOGNITIVE INTERVENTIONS
If we iden6fy valid cues for a learning task (i.e., the ones that accurately predict
learning performance),
And we design an interven6on that requires students to generate these cues,
and only these cues,
We can improve monitoring of learning.
To conclude
How well will you do on the test?
Metacognitive intervention
How about real life?
Text Problem Skill
How well will you do on the test?
Metacognitive intervention
Cavalcanti et al. (in prep)
Diagnose on a simulator
How confident are you in your diagnosis?
How well will you do on the test?
Metacognitive intervention
Wagner et al. (in prep)
Simulated pa0ent contact
Describe 5 moments where you
were (dis)sa0sfied
How well will you do on the test?
Metacognitive intervention
Kok et al. (in prep)
Learn CXR
Judge accuracy of diagnosis
Take test on CXR
Use this framework
1. Determine VALID CUES for your learning task 2. Design instruc0on that allows students to self-‐generate these cues 3. Have them judge their learning 4. Determine accuracy of judgments
Poor monitoring
Self-‐regulated
learning
Lower academic performance
For future research
Real life self-‐regulated learning: Avoiding the delay Self-‐tes0ng Selec0ng valid cues when both are available
Poor monitoring
Self-‐regulated
learning
Lower academic performance
Thanks to: Ellen Kok MarieQe van Loon Mar6ne Baars Luciana Falbo Gino Camp
Thank you for your aJen0on
Tamara van Gog Peter S6ers Jeroen van Merrienboer Fred Paas Jill Lobbestael Andries de Grip
anique.debruin@maastrichtuniversity.nl
There are three things extremely hard: steel, a diamond, and to know one’s self -‐
Benjamin Franklin (1750)
Use this framework
1. Determine VALID CUES for your learning task 2. Design instruc0on that allows students to self-‐generate these cues 3. Have them judge their learning 4. Determine accuracy of judgments
Poor monitoring
Self-‐regulated
learning
Lower academic performance
Textbook learning Skills training
Planning and goal seong