How To Improve Spatial Skill: Converging on Mechanism Alexandra Morrison Temple University
Dec 23, 2015
How To Improve Spatial Skill: Converging on Mechanism
Alexandra Morrison
Temple University
Acknowledgements
Nora Newcombe Tim Shipley
Alexandra Morrison Shannon Fitzhugh
Graduate Students
Research Staff
Danielle EaganMike DemersDominique Dumay Adrienne Villagomez
Jason Chein
Spatial Intelligence Learning Center & Temple University Neurocognition Lab
3 Questions
• Why Train Spatial Skills?
• How can we train spatial skills?
• What are the results of different training approaches?– In progress data
Why Spatial?
How did you get to San Diego for Aera?
Pack a Suitcase?
Navigate a busy airport?
Read a map?
Why Spatial? Scientific Thought and Communication
• STEM Disciplines– Science– Technology– Engineering – Math
Why Spatial?Predicts choice of STEM Major
– Higher spatial ability is associated with choice of STEM major (Hedges and Chung)
• Controlling for IQ & SES• Equal in gender
– SATM also related to choice of STEM discipline (Shae, Lubinski, & Benbow, 2001)
How can we train spatial skills?
• Two Approaches
– Strategy 1: Train Spatial Skills– Domain Specific Approach
– Strategy 2: Train shared mental resources– Domain general Approach
Approach 1: Train Domain Specific Skills• Can train spatial skills by
1. Familiarizing participants with spatial stimuli A. geometric shapesB. Shepard-Metzler Cubes
2. Practice spatial Transformations
Approach 1: Example of Domain Specific Spatial Training
• Wright, Thompson, Ganis, Newcombe, & Kosslyn, 2008
– Trained in either mental rotation or paper folding– Transferred to nontrained spatial but not verbal
task– Results domain specific but not task specific
Approach 2:Domain General Training
• Repetitive mental exercise improves general cognitive abilities
• Small scientific literature supports this
Approach 2: Domain General Cognitive Training
An emerging science of cognitive training …
Demonstrates1. Performance also relies on general cognitive abilities2. General cognitive abilities are both malleable &
trainable.
Approach 2: Working Memory Training
WM
Language Comp.
VerbalReasoning
SpatialReasoning
Problem Solving
FluidIntelligence
• WM capacity is a domain-general resource
• WM capacity predicts performance in a broad range of tasks(e.g., Kane et al., 2004)
• Successfully trained• Klingberg, 2002 & 2005
• Vaerhaghen, 2004
• Jaeggi, 2008
Working Memory vs. Short Term Memory
• Working Memory (WM)– Simultaneous storage and processing
• Short Term Memory (STM)– Storage only
What do we find when we compare types of training?
• Our goals1. Look at the malleability of spatial skills
2. Contrast different Cognitive Training ParadigmsDomain Specific
Spatial Visualization Training
Verbal Short Term Memory Training
Domain GeneralVerbal Working Memory Training
3. Determine Transfer Patterns
Training Methods Overview
Pre-test of cognitive skill level
Spatial Visualization Training
Working Memory TrainingPost-test of cognitive
skill level
Short Term Memory Training
Cognitive Assessment Battery
– MRT-A (Peters, 1995)– Solid Object Mental Rotation Test (Fitzhugh, Shipley, Newcombe & Dumay
(in prep.)– Spatial/Verbal STM
• Storage only
– Spatial/Verbal WM• Storage + processing
– Verbal Reasoning• Non-sense syllogisms, Inference
– Cognitive Control• Stroop
– Spatial Reasoning• Paper Folding, Surface Development
– Reading Comprehension• Nelson Denny
Training Regimes
• 5 days a week – 30 min per day• 4 weeks of training• Trained at home while guided by lab staff
Training Type 1:Spatial Visualization Training
• Domain Specific Spatial Training
• Introduction to 3D Visualization (Sorby, 1996)– Focus on transforming 3D objects into 2D
representations• Isomorphic – 2D perspective rendering• Orthographic – 2D flat projections• Surface Development • Rotation about one axis
Working Memory TrainingC
Glib
XFoot
TPrust
PMile
• Domain general task
•Storage + Processing
•Adaptive Difficulty
Short Term Memory Training
• Domain Specific Verbal
• Storage only– 4s interval
• Adaptive difficulty
C****
X****
T****
P****
Results of In Progress Study (n=91)
• Examine the following Questions: Do different types of training yield different
results?What is the impact of each training type on:
1. Spatial Skills2. STM (storage only) 3. WM (storage and information processing)
Differential Effects of Training
-40.00
-35.00
-30.00
-25.00
-20.00
-15.00
-10.00
-5.00
0.00
Selective Attention (Stroop)
SV WM STM
Redu
ction
in S
troo
p Ef
ect i
n M
S
Does the type of Training Matter?
*
d = .33
Which training types improve spatial skills?
-0.10
0.00
0.10
0.20
0.30
0.40
Mental Rotation (MRT-A)
SV WM STM
Perc
ent C
hang
e in
MRT
-A S
core
* * *
d = .9 d = 1.18 d = 1.22
Which Training types increase memory storage?
24 http://www.spatialintelligence.org-1.00
0.00
1.00
2.00
3.00
4.00
5.00
6.00
7.00
8.00Verbal STM
SV WM STM
Diff
eren
ce in
Sco
re P
re/P
ost
test
*
*
-1.00
0.00
1.00
2.00
3.00
4.00
5.00
6.00
7.00
8.00
Spatial STM
SV WM STM
Diff
eren
ce in
Sco
re P
re/P
ost
test
*
Which Training types improve storage and processing of information?
25
-1.00
0.00
1.00
2.00
3.00
4.00
5.00
6.00
7.00
8.00 Verbal WM
SV WM STM
Chan
ge in
Sco
re o
n O
pera
tion
Span
*
*d = .49
d = .39
-1.00
0.00
1.00
2.00
3.00
4.00
5.00
6.00
7.00
8.00 Spatial WM
SV WM STM
Chan
ge in
Sco
re o
n Sy
mm
etry
Spa
n
**
*d = .44
d = .35d = .48
Summary – Spatial Visualization Training
– SV training transfers to other spatial measures• Mental Rotation• Surface Development • Spatial Working Memory
– No transfer to verbal tasks– Domain specific but not task specific!
Summary Working Memory Training
– WM training transfers to non-trained domains • WM training spatial visualization measures• Also transferred to cognitive control measure
Summary Short Term Memory Training
– Transfer to verbal tasks– Some Transfer to spatial tasks
• May approximate WM training at longer list lengths (Unsworth & Engle, 2007)
Future Directions
• Trivia Training Control Group• Examine eye movement data
– Looks like Spatial training and WM may improve by different mechanisms
• 3 month follow-up– Are improvements stable?
Thank You! • Nora Newcombe• Jason Chein• Tim Shipley• Shannon Fitzhugh• Danielle Eagan• Dominique Dumay• Mike Demers• Adrienne Villagomez
**This research was supported by a National Science Foundation grant to support the Spatial Intelligence and Learning Center (No. SBE0541957)