phd-defense

44
Maps, how do users see them? Kristien Ooms Department of Geography – Ghent University An in depth investigation of the map users’ cognitive processes

Upload: oomskristien

Post on 25-May-2015

76 views

Category:

Education


0 download

DESCRIPTION

slides used for my PhD defense 'Maps, how do users see them?' December 2012

TRANSCRIPT

Page 1: PhD-defense

Maps, how do users see them?

Kristien OomsDepartment of Geography – Ghent University

An in depth investigation of the map users’ cognitive processes

Page 2: PhD-defense

Maps, how do users see them?

Improve the effectiveness of (screen) map designs based on the users’ characteristics.

After Kolácný (1969)

Purpose of maps Communication

Page 3: PhD-defense

Maps, how do users see them?

Contribute to the understanding of how map users read, interpret, store, and use the presented visual information on screen

maps.

AttentionRead

Store

Use

Understand

InfluenceImprovedesign

Retrieve

Interpret

Page 4: PhD-defense

Maps, how do users see them?

Contribute to the understanding of how map users read, interpret, store, and use the presented visual information on screen

maps.

After van Elzakker and Wealands (2007)

Page 5: PhD-defense

Maps, how do users see them?

Investigate the influence of (cartographic) expertise on the map users’ cognitive processes and their limitations while

processing the visual information presented on screen maps.

After van Elzakker and Wealands (2007)

After Kolácný (1969)

Page 6: PhD-defense

Maps, how do users see them?

• Research Questions1. How do map users read and interpret the visual

information presented on screen maps?

2. How do map users store and retrieve (use) the information that was previously gathered from screen maps?

3. How are the map users’ cognitive processes influenced by deviations in the map image?

4. How does (cartographic) expertise influence the cognitive processes investigated in the previous research questions?

Page 7: PhD-defense

Maps, how do users see them?

Ch 5-6: Complex Map Design

Ch 7: General Discussion

Ch 8: General Conclusion

Ch 5: Ch 6:

eye tracking

thinking aloud

sketch map

RQ 2

RQ 4

RQ 1 RQ 3expert

novice

questionnaire

Ch 1: Introduction

Ch 2-4: Basic Map Design

RQ 2

Ch 2: Ch 3: Ch 4:

RQ 4

RQ 1 RQ 3expert

novice

eye tracking

questionnaire

reaction times

Page 8: PhD-defense

Part I - Basic Map Design

• Task:– Visual search

• Techniques:– Eye tracking– Reaction times– Questionnaire

• Analyses:– Statistical– Visual

Page 9: PhD-defense

Part I - Basic Map Design

• Eye tracking?

...

Page 10: PhD-defense

Part I - Basic Map Design

• Eye tracking?• Fixations:

Insights in...- Attentive behaviour- Interpretation- Cognitive processes

Metrics:- Location of the

fixations- Fixation duration- Number of fixations

• Saccades:

Page 11: PhD-defense

Part I - Basic Map Design

• Reaction time measurements?– Time needed to locate label– Recorded through button actions– Time intervals are compared

• Questionnaires?– Background information!

• User characteristics• Familiarity with stimuli• Feedback

Page 12: PhD-defense

Part I – Experts vs. novices

• Aims:– Study cognitive processes– Difference experts vs. novices?– Explain by Cognitive Load Theory

• Structure WM: limited Cognitive load• Influence of map design

– Content– Symbolisation

• Room for learning

Chapter 2

Page 13: PhD-defense

Chapter 2

Part I – Experts vs. novices

• Results: ‘experts’ vs. ‘novices’

– Reaction time measurements• Experts vs. novices• Before vs. after

Page 14: PhD-defense

Chapter 2

Part I – Experts vs. novices

• Results: ‘experts’ vs. ‘novices’

Fixation duration (s)

Fixation count (fix/s)

Page 15: PhD-defense

Chapter 2

Part I – Experts vs. novices

• Results:– Fixation distribution

Low number of fixations

High number of fixations

Page 16: PhD-defense

Part I – Experts vs. novices

• Conclusion– Similar trend in both user groups: CLT– Experts significantly faster at locating the names – Explained by eye movement metrics

Shorter fixations

More fixations per second

Locates the names faster

Can interpret a larger part of the map in the same amount of time

Interprets map more efficiently

Can interpret the map’s content more efficiently

Chapter 2

Page 17: PhD-defense

Chapter 3

Part I – Visual Analytics

• Aims:– Extend statistical analyses

• Maps: communicate spatial information• Study spatial dimension• Influence of map layout

– Visual Analytics Toolkit• Filter data: time & attributes• Aggregate data

1 participant | 10 seconds

Page 18: PhD-defense

Part I – Visual Analytics

• Time series• Aggregation• Simplification• Selection

Chapter 3

Page 19: PhD-defense

Part I – Visual Analytics

• Conclusion– Selection, aggregation, simplifation

• Tools are indispensable

– Patterns: search behaviour• Time series: evolution search behaviour• Influence of map layout (labels)• Individual differences

Chapter 3

Page 20: PhD-defense

Chapter 4

Part I – Efficient and effective labels?

• Aims:– Evaluate different map designs

• Label placement algorithm– Improved efficiency faster– Lower map quality?

– Influence on (novice) users?• Effectiveness of the map?

bord

er-d

esig

nto

tal-d

esig

nor

igin

al v

iew

Page 21: PhD-defense

Chapter 4

Part I – Efficient and effective labels?

• Results: ‘border’ vs. ‘total’

– Reaction time measurements

– Eye movements• Fixation duration• Fixation count• Visualisation scanpaths

– Questionnaires

Page 22: PhD-defense

Part I – Efficient and effective labels?

• Conclusion– Improved (algorithmic) efficiency

– No influence on effectiveness• Consiously: user statements

– “no difference was seen”

• Unconsiously: measurements– No deviations in

» Reaction time measurements» Eye movement metrics

Chapter 4

Page 23: PhD-defense

Part II – Complex Map Design

• Task:– Study & draw

• Techniques:– Eye tracking– Thinking aloud– Sketch maps– Questionnaire

• Analyses:– Statistical– Visual

Page 24: PhD-defense

Part II – Complex Map Design

• Aims:– Communication process:

– Expertise?– Influence of deviations

SensoryInput

WorkingMemory

TransferredLong Term

Memory

Transferred

Retrieved

Virtually unlimited

Limited in -size -time (debate)

Using links, pointers with previous knowledge

Page 25: PhD-defense

Part II – Reading and Interpretation

• Eye movement:– Metrics

• Average fixation duration:Experts significantly shorter

• Number of fixations per secondExperts significantly more

Confirm previous findings

Chapter 5

Page 26: PhD-defense

Part II – Reading and Interpretation

• Eye movements– Gridded visualisations

• Fixation count• Total dwell time• Average fixation duration

• Average per user group• Maximum per user group

Chapter 5

Page 27: PhD-defense

Part II – Reading and Interpretation

• Eye movements– 2D gridded visualisations

Average fixation count

Maximum fixation count

Chapter 5

Page 28: PhD-defense

Part II – Reading and Interpretation

• Eye movements– 3D gridded visualisation

Average total dwell time

Average fixation duration

Chapter 5

Page 29: PhD-defense

Part II – Reading and Interpretation

• Eye movements– Gridded visualisation: statistical

comparisonStatistical comparison (ANOVA)

Chapter 5

Page 30: PhD-defense

Part II – Reading and Interpretation

• Eye movements– Scanpaths

Chapter 5

Page 31: PhD-defense

Part II – Reading and Interpretation

• Eye movements– Conclusion

• Focus on general structuring elements• Fixate more on top and left side

– Experts: more pronounced

• Influence of deviations– No influence for less important elements– Confusion for structuring elements

» Colour water bodies» Mirrored map elements

– Novices: more pronounced

Chapter 5

Page 32: PhD-defense

Chapter 6

Part II – Cognition and Memory

• Thinking aloud?– Say out loud every thought– Audio & video recordings– Insights in...

• Working Memory– How is information stored– How much information

• Information retrieval– Links with Long Term Memory

Square...rectangle...euhm...house? ..what do you guys

think?

Page 33: PhD-defense

Chapter 6

Part II – Cognition and Memory

• Thinking aloud– Recordings– Transcriptions– Segmentation

• Words– Word count

• Full thought (sentences)– Protocol analyses

Page 34: PhD-defense

Part II – Cognition and Memory

• Thinking aloudChapter 6

Page 35: PhD-defense

Part II – Cognition and Memory

• Thinking aloud– Word segmentation (count in ‰)

Based on theme Based on frequency

Chapter 6

Page 36: PhD-defense

Chapter 6

Part II – Cognition and Memory

• Thinking aloud– ‘Full thought’

• 4 Levels of codes:Level 1: Map Level

Orientate – Execute - EvaluateLevel 2: Item Level

Gather Thougts – Draw – Correct - EvaluateLevel 3: Confidence

Confident – Neutral – Not ConfidentLevel 4: Actions

Check – Correct – Draw – Erase – Fill Colour – Talk – Take Pencil

• Time ratio for each code: [0-1]

PsychologicalTheories

Task Analysis

Psychological Model

Proposed Codes

CodingScheme

Segmented Protocols

Transcriptions(Raw Protocols)

Coded Protocols

T H E O R Y

USERDATA

Page 37: PhD-defense

Part II – Cognition and Memory

• Sketch maps?– Draw map from memory– Paper & pencil– Insights in...

• What is remembered– Shape, location, relations, ...

• How much is remembered• Importance of objects

– Order of drawing

Chapter 6

Page 38: PhD-defense

Part II – Cognition and Memory

• Sketch mapsChapter 6

Page 39: PhD-defense

Part II – Cognition and Memory

• Sketch maps– Order of drawing

– Scores on maps

• Questionnaire– Stated confidence

Chapter 6

Page 40: PhD-defense

Part II – Cognition and Memory

• Conclusion– General structures: similar

• Novices: store more information– Descriptions, locations, etc.– No extra knowledge– Not derive extra information

• Experts: can retrieve more information– Know objects’ names – Background information

» Derive information– Larger chunks in WM

Chapter 6

Page 41: PhD-defense

General Conclusion1. How do map users read and interpret the visual

information presented on screen maps?• Guided by main structuring map elements• Influence on cognitive load Experts: more efficient interpretation

2. How do map users store and retrieve (use) the information that was previously gathered from screen maps?

• Limited information in WM• Retrieve information from LTM through links (chunking) Important influence of background knowledge (carto & geo)

3. How are the map users’ cognitive processes influenced by deviations in the map image?

• Subtle differences: almost no influence• Deviations to main structures (mirror, colour): confusion Novices are most influenced

4. How does (cartographic) expertise influence the cognitive processes investigated in the previous research questions?

Page 42: PhD-defense

General Conclusion

• Effective screen map design?– Guide attentive behaviour

• Reference frame: clearly visible• Other elements: neutral

– Structure/organise information• Group objects, e.g. through colour use

– Optimal map?• Less essential for non-structuring elements• Important for structuring elements

Page 43: PhD-defense

More [email protected]

1. Ooms, K., De Maeyer, P., Fack, V., Van Assche, E., & Witlox, F. (2012). Interpreting maps through the eyes of expert and novice users. International Journal of Geographical Information Science, 26 (10), 1773-1788.

2. Ooms, K., Andrienko, G., Andrienko, N., De Maeyer, P., & Fack, V. (2012). Analysing the spatial dimension of eye movement data using a visual analytic approach. Expert Systems with Applications, 39(1), 1324-1332.

3. Ooms, K., De Maeyer, P., Fack, V., Van Assche, E., & Witlox, F. (2012). Investigation the effectiveness of an efficient label placement method using eye movement data. The Cartographic Journal, 49(3), 234-246.

4. Ooms, K., De Maeyer, P., & Fack, V. (under revision). Understanding expert and novice map users: Reading and interpretation. Cartography and Geographic Information Science.

5. Ooms, K., De Maeyer, P., & Fack, V. (under revision). Listen to the map user: Cognition, memory, and expertise. The Cartographic Journal.

Page 44: PhD-defense

Thank you for your attention!

Questions?