Measuring the User Experience
(Tullis and Albert:2013)
4. Performance metrics
Task
success
Binary success
Levels of success
Issues in measuring success
Time on
task
Importance
Collecting and measuring
Analyzing and presenting
Issues to consider
Errors
When to measure
What is
Collecting and measuring
Analyzing and presenting
Issues to consider
Efficiency
Collecting and measuring
Analyzing and presenting
As a combination of task success and time
Learnability
Collecting and measuring
Analyzing and presenting
Issues to consider
5. Issue-based metrics
What is an issue?
Real vs. false
How to identify
In-person study
Automated
study
Analyzing and
reporting metrics
for usability issues
Freq of unique issues
Freq of issues per participant
Freq of participants
Issues by category
Issues by task
Consistency and bias
Number of participants
83% of issues by 1st 5 people (31% avg prob of any one issue)
Other study only 35% of issues
Tullis: Catch big issues; Homogenous user group; Limited design
More people when all issues must be caught; various user groups; Large design
Severity ratings
Based on UX
Based on
combination of
factors
Using rating
system
Caveats
6. Self-reported metrics
Importance: User perception; Insight into feelings
Rating scales
Likert
Semantic differential scales
How and when:
Analyzing
Post-task rating
Ease of use
After-Scenario Questionnaire (ASQ)
Expectation measure
Post-session ratings
Aggregating individual task ratings
System Usability Scale
Computer System Usability Questionnaire
Questionnaire for UI satisfaction
Usefulness, Satisfaction, and Ease-of-Use Questionnaire (USE)
Product reaction cards
Other types
Specific attributes
Specific elements
Open-ended questions
Awareness and comprehension
Awareness and usefulness gaps
7. Behavioural and
Physiological
Metrics
Unprompted verbal expressions
Eye tracking
Scan path
Heat map
Areas of Interest (AOIs)
Binning chart
Metrics
Dwell time
# of fixations
Fixation duration
Sequence
Time to First Fixation
Revisits
Hit ratio
Analysis
Pupillary response
Emotion
Affectiva; Q-Sensor;
Blue Bubble Lab;
Emovision; Seren;
Emotiv
Stress and
physiological
measures
Heart rate variance
Skin conductance
PressureMouse; Posture Analysis Seat
8. Combined and
comparative metrics
Single Usability Scores
Based on target goals
Based on Percentages
Based on Z scores
Using single usability metric
Usability scorecards
Comparisons against
Goals
Expert performance
The basics
1. Introduction
Definitions
UX: User involved, interact with interface, measurable and interesting
Value of UX metrics
Structure, extent, evaluation of changes, ROI, patterms
Ten myths
Anmerkungen:
1. Take too much time
2. Too expensive
3. Not useful for small improvements
4. Don't help understand causes
5. Too noisy
6. Trust your gut
7. Don't apply to new products
8. No metrics exist for this
9. Management doesn't care
10. Can't collect data with small sample
1. Take too much time
2. Too expensive
3. Not useful for small improvements
4. Don't help understand causes
5. Too noisy
6. Trust your gut
7. Don't apply to new products
8. No metrics exist for this
9. Management doesn't care
10. Can't collect data with small sample
2. Background / Statistics
Independent and
dependent
variables
Dependent: Things you measure (e.g. # of errors)
Independent: Things you control (e.g. Age, designs)