Communicating Data Certainty on Maps cisa

Abstract

How does it arise in data

Surveys being conducted

ConclusionsConclusions

Abstract

The work being done

A visual example:

Symbols fall into two categories

Symbols will be placed at point locations

Currently, there is no clear conclusion as to how data quality should be depicted, and this has been recognized as an important challenge to the visualization field (Wittenbrink et al. 1996). There have been numerous suggestions, but there has been little testing of these proposed methods (MacEachren 1997). In order to address this need, fifteen sets of symbols have been designed that aim to communicate a data value as well as its corresponding degree of certainty. These symbols were developed based upon ideas posed in the cartographic literature from the authors MacEachren, Schweizer and Goodchild, Leitner and Buttenfield, Drecki, Wittenbrink, Pang, and Lodha, Deitrick and Edsall, and Cliburn et al..

Intrinsic symbols modify a characteristic of the symbol such as opacity, shape, or texture to portray uncertainty, while extrinsic symbols use additional geometry.

Moderately Dry,High Certainty

Moderately Dry, Low Certainty

Moderately Dry,Medium Certainty

Extremely Dry,High Certainty

Extremely Dry, Low Certainty

Extremely Dry,Medium Certainty

Moderately Wet,High Certainty

Moderately Wet, Low Certainty

Moderately Wet,Medium Certainty

Informed Decisions

Predictable Outcomes

Understanding

Background informationSubject

Actions

Uncertainty

Data reliability

Enhanced Deci-sion Making

DataMaps

+

Decisions

Decisions

Decisions

Decisions

Uncertainty occurs when you lack a complete understanding and background information about a subject. It makes it difficult to make informed decisions and judge outcomes.

What is uncertainty?

E F

1/2

E F

1/2

Why communicate it?Why communicate it?

Intuitive abilities of symbols

All of the symbol sets (left) and symbols (right) that appeared in the

survey

Intuitive abilities of symbol sets

Feedback regarding participant’s perceived effectiveness of the symbols

Maybe the inaccuracy was not disclosed to you Maybe it was and you prefer just to ignore it

Uncertainty info is disclosed and you take it into accountYour fueling habits are altered

You borrow a car with an inaccurate gas gauge. Its true to within a quarter tank of the needle.

Dipping below a quarter tank is risky business. Are you keen on running out of gas?

Knowing about the uncertainty leads to informed decisionsand helps you avoid running out of gas.

Information always carries some degree of uncertainty. Therefore, if a decision support systems uses a map to communicate, should it not convey the data’s reliability to allow for optimal decisions?

An example:

Decision Support ToolsDecision Makers

Aid in decision making

BUT... Experience difficulty communicating data uncertainties

Individuals participating in various forms of management seek information to guide their learning, understanding, and decision-making. Web based decision support tools facilitate this, but often fail to provide any measure of the presented data’s reliability. Therefore, decision support systems (D'S) put managers in contact with data of varying degrees of reliability, as uncertainty is unavoidable and inherent in information.

Decision support systems (DSS’s) increasingly communicate through information visuals such as maps. When viewing a map, it is often assumed that all the data presented is truthful and accurate. This is never quite the case, however, as maps are just simplified representations of reality. Cartography faces the challenge of communicating data’s reliability in order to enhance decision making.

Interpolation Example

Natural Phenomenon

Synthesized Understanding of

Natural Phenomenon

Collection

Examination

Presentation

Communicating Data Certainty on Maps cisaPoster by Jay Fowler

Special thanks to Dr. Sarah Battersby for all of the insight, knowledge and encouragement

Data that is being analyzed varies in reliability. This may be due to human error such as incorrectly measuring a phenomenon, or due to instrument error if a certain tool is not working correctly.

Data is often manipulated introducing error. Examples of this are interpolation and extrapolation, two methods producing results that are not completely accurate.

It is impossible to perfectly capture and represent the complexities of reality on a map.

As data is collected, examined, and presented, uncertainties compound.

A cluster of data points with medium level certainty

Two clusters of data points with low to medium level certainty

Data Certainty shown Through Symbol Opacity

High Certainty

Low Certainty

q00 rb_lowq01 rb_lowq02 rb_lowq03 rb_lowq04 rb_lowq05 rb_mediumq06 rb_mediumq07 rb_mediumq08 rb_mediumq09 rb_mediumq10 rb_highq11 rb_highq12 rb_highq13 rb_highq14 rb_highq15 rb_lowq16 rb_lowq17 rb_lowq18 rb_lowq19 rb_lowq20 rb_mediumq21 rb_mediumq22 rb_mediumq23 rb_mediumq24 rb_mediumq25 rb_lowq26 rb_lowq27 rb_lowq28 rb_lowq29 rb_lowq30 rb_lowq31 rb_lowq32 rb_lowq33 rb_lowq34 rb_lowq35 rb_mediumq36 rb_mediumq37 rb_mediumq38 rb_mediumq39 rb_mediumq40 rb_mediumq41 rb_mediumq42 rb_mediumq43 rb_mediumq44 rb_mediumq45 rb_highq46 rb_highq47 rb_highq48 rb_highq49 rb_highq50 rb_lowq51 rb_lowq52 rb_lowq53 rb_lowq54 rb_lowq55 rb_mediumq56 rb_mediumq57 rb_mediumq58 rb_mediumq59 rb_mediumq60 rb_lowq61 rb_lowq62 rb_lowq63 rb_lowq64 rb_lowq65 rb_mediumq66 rb_mediumq67 rb_mediumq68 rb_mediumq69 rb_mediumq70 rb_lowq71 rb_lowq72 rb_lowq73 rb_lowq74 rb_lowq75 rb_lowq76 rb_lowq77 rb_lowq78 rb_lowq79 rb_lowq80 rb_mediumq81 rb_mediumq82 rb_mediumq83 rb_mediumq84 rb_mediumq85 rb_lowq86 rb_lowq87 rb_lowq88 rb_lowq89 rb_lowq90 rb_mediumq91 rb_mediumq92 rb_mediumq93 rb_mediumq94 rb_mediumq95 rb_lowq96 rb_lowq97 rb_lowq98 rb_lowq99 rb_low

q100 rb_mediumq101 rb_mediumq102 rb_mediumq103 rb_mediumq104 rb_mediumq105 rb_lowq106 rb_lowq107 rb_lowq108 rb_lowq109 rb_lowq110 rb_mediumq111 rb_mediumq112 rb_mediumq113 rb_mediumq114 rb_mediumq115 rb_lowq116 rb_lowq117 rb_lowq118 rb_lowq119 rb_lowq120 rb_mediumq121 rb_mediumq122 rb_mediumq123 rb_mediumq124 rb_mediumq125 rb_mediumq126 rb_mediumq127 rb_mediumq128 rb_mediumq129 rb_mediumq130 rb_lowq131 rb_lowq132 rb_lowq133 rb_lowq134 rb_lowq135 rb_mediumq136 rb_mediumq137 rb_mediumq138 rb_mediumq139 rb_mediumq140 rb_highq141 rb_highq142 rb_highq143 rb_mediumq144 rb_mediumq145 rb_mediumq146 rb_lowq147 rb_lowq148 rb_low

q01

q02

q03

q04

q05

q06

q07

q08

q09

q10

q11

q12

q13

q14

q15

q16

q17

q18

q19

q20

q21

q22

q23

q24

q25

q26

q27

q28

q29

q30

q31

q32

q33

q34

q35

q36

q37

q38

q39

q40

q41

q42

q43

q44

q45

q46

q47

q48

q49

q50

q51

q52

q53

q54

q55

q56

q57

q58

q59

q60

q61

q62

q63

q64

q65

q66

q67

q67q68

The second survey will place the most successful symbol sets from study one on maps for re-evaluation. Generating results from a testing environment similar to what decision-makers actually experience is desired (Hope and Hunter 2007).

Two separate human-subject surveys are being conducted evaluating symbol performance. The first is a comprehensive evaluation of all fifteen sets of symbols. It examines:

Study Two

High Certainty

Low Certainty

LegendExtremely Dry

Moderately Dry

Regular

Moderately Wet

Extremely Wet

Characterize the drought within the circle

Characterize the related data certainty within the circle

Extremely Dry

Moderately Dry

Regular

Moderately Wet

Extremely Wet

High Certainty

Medium Certainty

Low Certainty

Communicating uncertainty is important, but there is limited knowledge and studies examining comprehensively the best way to do this on a map. The proposed effectiveness testing will provide valuable information to the uncertainty visualization community and allow for better communication with individuals using decision support tools and maps.

Cliburn, D. C., J. J. Feddema, et al. (2002). "Design and evaluation of a decision support system in a water balance application." Computers & Graphics-Uk 26(6): 931-949.Deitrick, S. and R. Edsall (2008). "Making Uncertainty Usable: Approaches for Visualizing Uncertainty Information." Geographic Visualization: Concepts, Tools and Applications.Drecki, I. (2002). "Visualization of Uncertainty in Geographical Data." Spatial Data Quality(W. Shi, P. Fisher, and M. Goodchild (eds)): 140-159.Hope, S. and G. J. Hunter (2007). "Testing the effects of positional uncertainty on spatial decision-making." International Journal of Geographical Information Science 21(6): 645-665.Leitner, M. and B. P. Buttenfield (2000). "Guidelines for the display of attribute certainty." Cartography and Geographic Information Science 27(1): 3-14.MacEachren, A. M. and M. J. Kraak (1997). "Exploratory cartographic visualization: Advancing the agenda." Computers & Geosciences 23(4): 335-343.Pang, A. T., C. M. Wittenbrink, et al. (1997). "Approaches to uncertainty visualization." Visual Computer 13(8): 370-390.Penrod, J. (2001). "Refinement of the concept of uncertainty." Journal of Advanced Nursing 34(2): 238-245.Schweizer, D. M. and M. F. Goodchild (1992). "Data quality and choropleth maps: An experiment with the use of color." Proceedings, GIS/LIS '92, San Jose, California. ACSM and ASPRS, Washington, D.C. : 686-699.Wittenbrink, C. M., A. T. Pang, et al. (1996). "Glyphs for visualizing uncertainty in vector fields." Ieee Transactions on Visualization and Computer Graphics 2(3): 266-279.