Introduction
Chapter 1
Dr. Andrew Besmer
Types of HCI Contributions
- Empirical
- Qualitative
- Quantitative
- Surveys, Focus Groups, Time Diaries, Sensors, Instrumentation, Ethnography, etc…
- Artifact
- Often accompanied by empirical
Types of HCI Contributions
- Methodological
- New approach or process to carry out research
- Instrument for measurement
- Theoretical
- Predictive or descriptive concepts/models
- Dataset
- Opinion
Research
- Research generally needs to be
- Relevant/rigorous
- Influence
- Interface design
- Development processes
- User training
- Public policy
- Traditionally focuses on practical results that improve quality of life
Measurement
- Early measurement based on human performance
- Completion Time
- Success Rate/Error Rate
- Time to Learn
- Retention Over Time
- Satisfaction
- Typically Task Based Study
- Separated - Discrete/make sense
- Quantified/Measured
Micro/Macro HCI
- Micro-HCI
- Macro-HCI
- User Motivation, Collaboration, Trust, etc…
- Societal Impact
- Consider something enjoyable like a video game
Experimental
- Experimental Model
- Hypothesis testing with statistics
- Good for how often, long, successful, etc…
- Not always suited to why
- Well understood
- Other Models
- Ethnography
- Focus Group
- Interview, etc…
Early Research
- Researching a new topic?
- Consider exploratory research using:
- Surveys
- Interviews
- Ethnography
- etc…
Types of Research
- Certain domains may lend different types of research
- Consider mobile “in the wild” vs laboratory
- Issues to consider such as consent to data collection
Data Collection and Replication
- No such thing as perfect data collection
- Rarely is one study or finding the definitive answer
- Replication is important
- Studies should include sufficient details for replication
- Results in “triangulation”
Replication
- Even better when verified using different research methods
- Multiple studies
- Broad shallow navs are better than narrow deep
- Findings can change over time
- Frustrations of early web users vs today
Design Nuance
- Design is not simple and not an optimization problem
- Different stakeholders can be… different
- Understand conflicts/tradeoffs between groups
- Even if something is better, may not be adopted
- Sometimes people desire it… iPhone