CS 235: User Interface Design December 3 Class Meeting Department of Computer Science San Jose State...

45
CS 235: User Interface Design December 3 Class Meeting Department of Computer Science San Jose State University Fall 2014 Instructor: Ron Mak www.cs.sjsu.edu/~mak

Transcript of CS 235: User Interface Design December 3 Class Meeting Department of Computer Science San Jose State...

CS 235: User Interface DesignDecember 3 Class Meeting

Department of Computer ScienceSan Jose State University

Fall 2014Instructor: Ron Mak

www.cs.sjsu.edu/~mak

Computer Science Dept.Fall 2014: December 3

CS 235: User Interface Design© R. Mak

2

Developing an Effective Dashboard

Understand the application domain.

Clearly define the dashboard’s purpose. What are its objectives?

Get inside the heads of the dashboard viewers.

Focus on the objectives, not the means.

Most dashboard creation tools over-emphasize fancy artwork and cute skeuomorphs.

Computer Science Dept.Fall 2014: December 3

CS 235: User Interface Design© R. Mak

3

Matching a Dashboard to its Viewers

A dashboard should represent information that’s needed to update a viewer’s mental model …

… in a way that closely matches that mental model.

What information is displayed. How it’s displayed. The arrangement.

Computer Science Dept.Fall 2014: December 3

CS 235: User Interface Design© R. Mak

4

Dashboard Requirements

Understand the viewers’ mental models.

Elicit requirements from potential viewers.

How does this piece of information relate to that? If a problem occurs here,

what does it affect and how?

Novices to an application domain may not have formed clear mental models.

Experts may not be able to describe their mental models.

Computer Science Dept.Fall 2014: December 3

CS 235: User Interface Design© R. Mak

5

Information Complexity

Assess the degree of information complexity that viewers can easily process.

People benefit most from dashboardsthat are well-matched to their abilities.

Viewers may have different levels of abilities.

Computer Science Dept.Fall 2014: December 3

CS 235: User Interface Design© R. Mak

6

Questions the Dashboard Designer Should Ask

How frequently to update the information?

Who will use the dashboard? One person? One group? Different departments?

What will the dashboard monitor?

What objectives will it support?

Computer Science Dept.Fall 2014: December 3

CS 235: User Interface Design© R. Mak

7

Questions to Ask, cont’d

What questions should the dashboard answer?

What actions will be taken in response?

What items should the dashboard display? What does each item tell you? Why is it important?

Computer Science Dept.Fall 2014: December 3

CS 235: User Interface Design© R. Mak

8

Questions to Ask, cont’d

At what level of summary or detail should information be displayed for a quick overview?

Which information items are most important to achieve the dashboard’s objectives?

What are logical groupings of information? To which group does each item belong?

What are useful comparisons to provide meaningful context? Are there historical data or targets?

Computer Science Dept.Fall 2014: December 3

CS 235: User Interface Design© R. Mak

9

Include Only Useful Information

A dashboard should only include information that actually affects performance.

Questions to ask future viewers:

Can you describe a situation when this information would lead you to do something?

What is an example of the actual data that would appear and what action would you take in response?

Computer Science Dept.Fall 2014: December 3

CS 235: User Interface Design© R. Mak

10

Purpose of a Dashboard

Not just for keeping viewers up to date.

Also to maintain situational awareness:

Prevent problems. Take advantage of opportunities. Take steps to maintain or improve performance.

Computer Science Dept.Fall 2014: December 3

CS 235: User Interface Design© R. Mak

11

The Need for Context

Data values by themselves are useless. Compared to what?

Data only comes alive when compared to related values:

targets standards the norm the past

Computer Science Dept.Fall 2014: December 3

CS 235: User Interface Design© R. Mak

12

Fundamental Considerations

Update frequency User expertise Audience size Technology platform Screen type Data type

Computer Science Dept.Fall 2014: December 3

CS 235: User Interface Design© R. Mak

13

Update Frequency

Daily? Hourly? Real time? Include a timestamp if updated more than daily.

If updated less often than daily,then it’s not a dashboard. Not situation awareness at a glance.

If updated in real time, then the dashboard should be simpler and contain less information. Increased need for speed. There is a limit to the amount of information

a viewer can take in at a glance.

Computer Science Dept.Fall 2014: December 3

CS 235: User Interface Design© R. Mak

14

User Experience

A viewer with a high level of expertise in the application domain can handle a high degree of complexity in the information display.

A novice can handle relatively little complexity.

Provide a simpler dashboard display. Increase complexity as the viewer gains experience.

Computer Science Dept.Fall 2014: December 3

CS 235: User Interface Design© R. Mak

15

Audience Size

One person Multiple people with similar requirements Multiple people who need to monitor

different subsets of information

Designing a dashboard for a group of people is more challenging.

One-size-fits all may be a poor compromise. Instead, show different levels of information based

on the identity of the logged-in viewer.

Computer Science Dept.Fall 2014: December 3

CS 235: User Interface Design© R. Mak

16

Technology Platform

Desktop or laptop Web browser and server Mobile device

Web constraints Less functionality Less elegant interface

Mobile devices add touch and gesture input.

Computer Science Dept.Fall 2014: December 3

CS 235: User Interface Design© R. Mak

17

Screen Type

Screen characteristics size viewing distance aspect ratio resolution

We can only see clearly within a small angle. A large screen doesn’t mean we can show

more information in the dashboard.

Designers may have to accommodate multiple screen sizes, aspect ratios, and resolutions.

Computer Science Dept.Fall 2014: December 3

CS 235: User Interface Design© R. Mak

18

Data Type

Quantitative data

numerical values measures

Non-quantitative data

top 10 customers issues to be resolved people to contact etc.

Computer Science Dept.Fall 2014: December 3

CS 235: User Interface Design© R. Mak

19

Quantitative Data: Timing Variations

Time frames

This year to date This quarter to date This month to date This week to date Today so far

The appropriate time frame is determined by the nature of the dashboard’s objectives.

Computer Science Dept.Fall 2014: December 3

CS 235: User Interface Design© R. Mak

20

Quantitative Data: Comparisons

Understand data measures in context.

Context is often provided by comparisons:

same day last year end of last year budgeted amount percentage of the budget percentage of the forecast average, typical range, benchmark projection into the future competitor’s measure

Computer Science Dept.Fall 2014: December 3

CS 235: User Interface Design© R. Mak

21

Quantitative Data: Comparisons, cont’d

Multiple instances of a measure Each represents a subdivision of a whole Examples: regions, departments, subranges

Temporal instances of a measure Time series

Questions to ask a viewer:

Can you tell how well what you’re monitoring is performing?

Can you tell when something is abnormal?

Computer Science Dept.Fall 2014: December 3

CS 235: User Interface Design© R. Mak

22

Quantitative Data: Predictions

Develop good predictive models based on:

historical data an understanding of probability and statistics

If possible, incorporate predictions into a dashboard.

The dashboard can help shape the future.

Computer Science Dept.Fall 2014: December 3

CS 235: User Interface Design© R. Mak

23

A Well-Designed Dashboard

“Perfection is achieved, not when there is nothing more to add, but when there is nothing left to take away.” -- Antoine de Saint-Exupery

Exceptionally well organized. Condensed primarily in the form of summaries. Specific to the task at hand. Customized clearly to the viewers. Uses concise media that communicate in the

clearest and most direct way possible.

Computer Science Dept.Fall 2014: December 3

CS 235: User Interface Design© R. Mak

24

What is Wrong with this “Dashboard”?

Display a single traffic signal.

Light is green: All is well. Light is red: Drill down for more details.

Viewers rely on a predetermined threshold of what is or is not well.

Computer Science Dept.Fall 2014: December 3

CS 235: User Interface Design© R. Mak

25

Data-Ink Ratio

Maximize the data-ink ratio (within reason).

Every bit of ink on a graphic requires a reason. Nearly always, the reason should be that

the ink presents new information.

Keep it simple! Eliminate chart junk!

Computer Science Dept.Fall 2014: December 3

CS 235: User Interface Design© R. Mak

26

Dashboard Design: Goals and Steps

Reduce the non-data pixels.

Eliminate all unnecessary non-data pixels. De-emphasize and regularize

the non-data pixels that remain.

Enhance the data pixels.

Eliminate all unnecessary data pixels. Highlight the most important data pixels that remain.

Computer Science Dept.Fall 2014: December 3

CS 235: User Interface Design© R. Mak

27

Data-Ink Ratio, cont’d

Can you separate the data ink from the non-data ink?

The non-data inkis highlighted in red.

Information Dashboard Design, 2nd ed.by Stephen FewAnalytics Press, 2013

Computer Science Dept.Fall 2014: December 3

CS 235: User Interface Design© R. Mak

28

Data-Ink Ratio, cont’d

Too high percentage of non-data pixels:

Information Dashboard Design, 2nd ed.by Stephen FewAnalytics Press, 2013

Computer Science Dept.Fall 2014: December 3

CS 235: User Interface Design© R. Mak

29

Data-Ink Ratio, cont’d

Colors that don’t encode any meaning:

Information Dashboard Design, 2nd ed.by Stephen FewAnalytics Press, 2013

Computer Science Dept.Fall 2014: December 3

CS 235: User Interface Design© R. Mak

30

Data-Ink Ratio, cont’d

Unnecessary borders fragment the display.

Information Dashboard Design, 2nd ed.by Stephen FewAnalytics Press, 2013

Computer Science Dept.Fall 2014: December 3

CS 235: User Interface Design© R. Mak

31

Data-Ink Ratio, cont’d

Unnecessary and distracting fill colors.

Information Dashboard Design, 2nd ed.by Stephen FewAnalytics Press, 2013

Computer Science Dept.Fall 2014: December 3

CS 235: User Interface Design© R. Mak

32

Data-Ink Ratio, cont’d

Distracting use of color gradients.

Information Dashboard Design, 2nd ed.by Stephen FewAnalytics Press, 2013

Computer Science Dept.Fall 2014: December 3

CS 235: User Interface Design© R. Mak

33

Data-Ink Ratio, cont’d

Unnecessary grid lines in graphs.

Information Dashboard Design, 2nd ed.by Stephen FewAnalytics Press, 2013

Computer Science Dept.Fall 2014: December 3

CS 235: User Interface Design© R. Mak

34

Data-Ink Ratio, cont’d

Unnecessary grid lines in tables.

Information Dashboard Design, 2nd ed.by Stephen FewAnalytics Press, 2013

Computer Science Dept.Fall 2014: December 3

CS 235: User Interface Design© R. Mak

35

Data-Ink Ratio, cont’d

Complete borders instead of x- and y-axes.

Information Dashboard Design, 2nd ed.by Stephen FewAnalytics Press, 2013

Computer Science Dept.Fall 2014: December 3

CS 235: User Interface Design© R. Mak

36

Data-Ink Ratio, cont’d

Avoid 3-D

Information Dashboard Design, 2nd ed.by Stephen FewAnalytics Press, 2013

Computer Science Dept.Fall 2014: December 3

CS 235: User Interface Design© R. Mak

37

Data-Ink Ratio, cont’d

Make separator lines lighter.

Information Dashboard Design, 2nd ed.by Stephen FewAnalytics Press, 2013

Computer Science Dept.Fall 2014: December 3

CS 235: User Interface Design© R. Mak

38

Data-Ink Ratio, cont’d

Make separator lines lighter, cont’d.

Information Dashboard Design, 2nd ed.by Stephen FewAnalytics Press, 2013

Computer Science Dept.Fall 2014: December 3

CS 235: User Interface Design© R. Mak

39

Data-Ink Ratio, cont’d

Make grid lines lighter.

Information Dashboard Design, 2nd ed.by Stephen FewAnalytics Press, 2013

Computer Science Dept.Fall 2014: December 3

CS 235: User Interface Design© R. Mak

40

Data-Ink Ratio, cont’d

Make grid lines lighter, cont’d.

Information Dashboard Design, 2nd ed.by Stephen FewAnalytics Press, 2013

Computer Science Dept.Fall 2014: December 3

CS 235: User Interface Design© R. Mak

41

Data-Ink Ratio, cont’d

Mute the shading used to delineate table rows.

Information Dashboard Design, 2nd ed.by Stephen FewAnalytics Press, 2013

Computer Science Dept.Fall 2014: December 3

CS 235: User Interface Design© R. Mak

42

Data-Ink Ratio, cont’d

Dashboards should not need instructions.

Information Dashboard Design, 2nd ed.by Stephen FewAnalytics Press, 2013

Computer Science Dept.Fall 2014: December 3

CS 235: User Interface Design© R. Mak

43

Data-Ink Ratio, cont’d

Chart junk!

Information Dashboard Design, 2nd ed.by Stephen FewAnalytics Press, 2013

Computer Science Dept.Fall 2014: December 3

CS 235: User Interface Design© R. Mak

44

Data-Ink Ratio, cont’d

Overuse of color.

Information Dashboard Design, 2nd ed.by Stephen FewAnalytics Press, 2013

Computer Science Dept.Fall 2014: December 3

CS 235: User Interface Design© R. Mak

45

Data-Ink Ratio, cont’d

Overuse of color, cont’d.

Now the importantitems stand out.

Information Dashboard Design, 2nd ed.by Stephen FewAnalytics Press, 2013