Broadening the Study of Spatial Intelligence Mary Hegarty - - PowerPoint PPT Presentation

Broadening the Study of Spatial Intelligence

Mary Hegarty University of California, Santa Barbara

Terminology

Spatial Cognition Spatial Thinking Spatial Intelligence

What is Spatial “Thinking”?

Spatial Intelligence as Adaptive Spatial Thinking

• Thinking about space:

Thinking about spatial processes and structures at the scale of objects.

• Thinking in space:

Thinking about spatial processes and structures at the scale of environments

• Thinking with space:

Using spatial representations of non- spatial entities

Spatial Processes and Structures

Chemistry Problem Solving Mechanical Reasoning Imagining Cross Sections Geological Inference

Br Br H H H Br CH3 CH3 Br H

Please indicate the stereochemical relationship between these two molecules.

Using Spatial Representations

a b f d c g i e h

Spatial Entities Non-Spatial Entities

How Do Psychologists Study Spatial Thinking?

Laboratory Studies

Mental Rotation

Shepard & Metzler, 1971

Individual Differences (Spatial Abilities)

Mental Rotation Visualization of Views Embedded Figures

Spatial thinking “in the Wild” (Discipline Diving)

Mechanics Meteorology Medicine Chemistry Navigation Geology

What Have we Learned from these Approaches?

Laboratory Studies

Mental Rotation

• Analog Mental Processes
• Piecemeal vs holistic
• Motor Interference
• Parietal (sometimes premotor) activation

Limitation: Primary focus on one task. A single rigid transformation

Individual Differences Studies

• Spatial abilities separable from verbal, reasoning ability
• Different spatial abilities (e.g., spatial perception vs.

spatial visualization etc)

• Predict success in STEM education and careers
• Sex differences (especially mental rotation)
• Performance improves with practice, training

Limitation: Focused on small number of tests, development of tests not theoretically motivated

Insights from Examining Spatial Cognition “in the Wild”

• Small-Scale vs. Large Scale
• Mental Simulation vs. Analytic Thinking
• Internal vs. External Representations

Distinction 1: Small-Scale vs. Large Scale

• Small-Scale Spatial Ability
• Large-Scale Spatial Ability

– Learning the layout of a new environment from

• direct experience (walking through the environment),
• a video
• a virtual environment

Correlations

Learning from Direct Experience: Direction .30 .21 Distance .27 .16 Map .20 .11 Learning from Video Direction .45 .40 Distance .22 .08 Map .40 .24 Learning from VE Direction .45 .33 Distance .29 .17 Map .29 .22 Embedded Mental Figures Rotation

Hegarty, Montello, Richardson, Ishikawa & Lovelace, 2006

Learning from Visual Media .93* .72* .73* .77* .43* .54* .69* .34* .65* Spatial Ability Self-Reported Sense of Direction

Mental Rotation Test

• Embedded. Figures Test

Spatial Working Memory (Arrow Span) Santa Barbara Sense of Direction Scale

.50* .18* .39* .87* .48* .65* .53* .94* .30* Learning from Direct Experience

Pointing Distance Estimation Map Drawing Pointing Distance Estimation Map Drawing Pointing Distance Estimation Map Drawing After learning from Direct Experience After learning from A Video After learning from A Virtual Environment

Hegarty, Montello, Richardson, Ishikawa & Lovelace, 2006

Spatial Ability and Science Disciplines

• Do STEM disciplines depend on both large-

scale and small-scale spatial abilities?

• Do all STEM disciplines depend equally on

spatial abilities?

• Preliminary Study: Self reports of scientists

Hegarty, Crookes, Dara-Abrams & Shipley, 2010

Small-Scale Spatial Ability

Hegarty, Crookes, Dara-Abrams & Shipley, 2010

Large-Scale Spatial Ability

Hegarty, Crookes, Dara-Abrams & Shipley, 2010

Summary

• Large and small scale spatial abilities partially

dissociated

• Sciences may not depend equally on small and

large-scale spatial abilities

• Self-report data is suggestive but need to follow

up with more objective testing

Distinction 2 Visualization vs. Analytic Thinking

Visualization: Mental Images, Analog transformations, mental simulation Analytic Thinking: Task decomposition, mathematical/propositional representations, rule- based reasoning Not either or! Collaborative processes

Mechanical Reasoning

When the rope is pulled, in what direction will the lower pulley turn? Inferring Rigid Transformations of Multiple, Interconnected Objects

The middle pulley turns counterclockwise The lower pulley turns clockwise

1 2 3 4 5 2 1 3

The upper pulley turns counterclockwise

1

Not pure visualization (simulation)! Also task decomposition

Interplay between visualization and rule-based reasoning

Initial Strategy: Mental Simulation Infer Rule: “every other gear turns the same way”

If the large gear turns clockwise, Which direction will Gear A turn?

Schwartz & Black, 1996

Interplay also found in

• Inferring cross sections
• Spatial ability tests
• Stereochemistry

CH3 OH Cl H H3C HO Cl H

Questions

• What about imagining non-rigid transformations of
• bjects?
• What are the limits of visualization/mental

simulation?

• When does analytic thinking take over?
• How much emphasis should we put on

visualization vs analytic thinking in education?

Distinction 3 Internal vs. External Representations

• Internal:

– Mental images, propositional representations

• External:

– Diagrams, graphs, maps, models, interactive computer visualizations

Metarepresentational Competence (diSessa, 2004)

Ability to:

• Learn to use representations with minimal instruction
• Choose representations, e.g., critique and compare the

adequacy of representations and judge their suitability for different tasks

• Design representations

Task: Draw the Cross section from the Perspective of the Arrow

Using Representations

Use of Interactive Visualization Keehner, Hegarty, Cohen, Khooshabeh & Montello (2008)

Test Page View Arrow View

Use of External Visualization

Access of Arrow View

Proportion of trials where Arrow View was accessed

.88 .75 .63 .50 .38 .25 .13 0.00 10 8 6 4 2

• Std. Dev = .32

Mean = .40 N = 30.00

0.1 0.2 0.3 0.4 0.5 0.6 0.7 Low-access participants High-access participants

Proportion correct (duct relations)

Representation Translation in Chemistry

H H H H H CH3 Stull, Hegarty, Dixon & Stieff, 2012

Number of Trials on Which Participants Used the Model

Models vs. No Models

F (2,44) = 50.11, p < .001* (Max = 18) No Models Received Models

Meta-representational Competence: Using Representations

• Not everyone discovers how to use the

visualization productively without instruction

• Performance related to access of the most

informative view

• Independent of spatial ability

Meta-representational Competence Choosing Representations

• Interactive visualizations allow people to chose
• r create the representations that they will use to

accomplish a task

• Effectiveness of visual displays is relative to the

task to be performed

(e.g., Bertin, 1983; Cheng, 2002; Gattis & Holyoak, 1996; Larkin & Simon, 1987; Shah, Mayer & Hegarty, 1999)

• Do people choose the best external

representations for their task? (Hegarty & Smallman & Stull, 2012)

Choosing Representations

• Novices and experts perform simple read-
• ff and comparison tasks with maps with

different complexity

• Express their intuitions about display

effectiveness by choosing the display they would use to perform a task

4 4.5 5 5.5 6 6.5 7 7.5 1 2 3 1 2 3 Without Realism With Realism Number of Displayed Variables Response time (sec) Pressure Wind Temperature

Results:

• About third of novices and experts prefered realistic
• ver non-realistic maps
• Some experts also prefer maps with additional

meteorological variables

• However, both realism and complexity slowed

performance for both experts and novices

Conclusions and Questions

• People do not always discover how to use

an external visualization productively

• People make poor representation choices
• Can we teach people to use and critique

external representations

• What is the best way to do this?

Spatial Thinking in the Wild

• Thinking about spatial processes and structures

– Flexible interaction between mental simulation and analytic thinking

• External representations can enhance cognition,

but also depend on cognition for their use – Not as intuitive as we think!

• Spatial center as an opportunity to discover how

people think spatially in different disciplines, and hopefully to enhance spatial thinking across the discipleins.

Thank you!