HCI research methods by Sanjay Tripathi, ABB

Dr. Sanjay Tripathi, ABB Corporate Research, August 28th, 2010

ISS
HCI Research /Methods

© ABB Group
September 15, 2010 | Slide 1

Computing Era 1960-1980

Source: Being Human: Human Computer Interaction in the year 2020

© ABB Group

Computing Era 2000-2020


© ABB Group

Interfaces Gestures

The Reactable
A multi touch interface used for playing
music. Interaction is done by moving
and rotating physical objects on its
surface.

The HotHand Device
A ring worn by electric guitar players
that uses motion sensors and a wireless
transmitter to create different kinds of
sound effects by various hand gestures.


© ABB Group

Prediction

Computing will move to the background,
will extremely merge in our everyday living spaces

Human user  into the foreground.

Interfaces will have the capacity to understand and
emulate human communicative intentions

© ABB Group

Technology Reliant World

© ABB Group

Phenomena

© ABB Group

Prologue
 Computer has become a mediator between the human operator
and the physical system being controlled.
 This occurred first in aviation
 Then in process control, manufacturing and military systems
 More recently it happened in hospitals, trains, automobiles and home
appliances

Example: Power and automation
 New form of control is often called supervisory control,
 Human operator = Manager
 planning
 setting goals and constraints
 diagnosing failures
 intervening in control as necessary, and
 learning from experience

The hope has been to relieve the operator of tedious work and
make the system both more efficient and more reliable

© ABB Group

Introduction to HCI Research
 HCI is a young discipline (about 30 years)
 Has developed from computer science, with influences from
cognitive psychology
 Other influences: sociology, anthropology, communication, design,
ergonomics
 Related area: Interaction design

 We will Cover 
 Overview of Science and research
 Introduction
 Fundamentals of research (in general)
 Research approaches and methods
 Their limitations in application

© ABB Group

Definition:
 Human-computer interaction is a discipline concerned with
the design, evaluation and implementation of interactive
computing systems for human use and with the study of
major phenomena surrounding them.

 ACM SIGCHI Curricula for HCI (Hewett et al. 1992)
 http://sigchi.org/cdg/cdg2.html

© ABB Group


HCI as a ”user science”
 All HCI research includes studies of the use of technology
 Empirical study methods are essential
 But HCI is also closely related to the design and
development of technology
 Important to discuss the relationship between research and
development

© ABB Group


Is HCI inter- or multidisciplinary?
 One discipline = one research community
 Shared values, methods and concepts
 Criteria for ”good” and ”bad” research
 Disciplines may have difficulties understanding each other
(e.g. due to cultural differences)
 HCI demands cooperation among researchers
 How does real interdisciplinary arise?

© ABB Group


Inter- and multidisciplinary research
 Interdisciplinary research transcends the borders
between knowledge areas and organisations … The
goal is an integration, gradually developing into a
science of its own
 Multidisciplinary research has a clear division of
labour among the sciences involved

© ABB Group


Research approaches and perspectives
 Experimental psychology
 Cognitive modeling, task analysis
 Situated and distributed cognition
 Ethnography
 Conversation analysis
 Activity theory
 Organizational approaches

© ABB Group


HCI heritage from Computer Science
 Computer science was itself interdisciplinary
 Mathematics + engineering
 HCI  Computer Science + Psychology
 A model for all kinds of engineering:
 Identify requirements
 Specify a design
 Build an implementation
 Waterfall model (top-down)
 More realistic: Iterative model!

© ABB Group


HCI heritage from psychology
 Research object in cognitive psychology: mental processes
(memory, thinking, learning)
 Approach
 modeling mental processes
 influences from information processing in computers
 oriented towards experimental methods and controlled data
management
 testing of hypotheses
 Limitations for HCI
 great individual variations; many variables interact
 de-contextualization
 strong focus on causal relations

© ABB Group


HCI heritage from psychology
 Work psychology: originates from ergonomics
 Task analysis
 Informal description of tasks (e.g. as a scenario)
 Systems for formal description of work tasks
 Purpose: to match the system design to users’ tasks
 Problem: many tasks are open-ended and cannot be
described in a simple way with rules or tree structures
 Situated action: more recent approach
 Socially oriented psychological theory
 Individuals’ cognitive processes interacting with a social
context
 Method: interpretative, more field studies

© ABB Group

Perspectives on Human-Computer Interaction

 By John Kammersgaard
 The Systems perspective
 The system is looked upon from a bird's eye view
 The relevance of a task can only be expressed on the
organisational level
 All interaction is seen as transmission of data between
human and automatic components
 The goal is to make transmission as fast and correct as
possible.
 The dialogue partner perspective
 The tool perspective
 The media perspective

© ABB Group


 brought to attention through the artificial intelligence
research
 focuses the use of computers within an individual
context
 the computer always be seen as acting like a human
being in a communication process
 The goal: to make communication as similar as possible
to human-human interaction
 Caution: only be used for certain special purposes and
always in combination with other perspectives

© ABB Group


 the computer becomes a tool box providing utensils that can help the
user in accomplishing a task
 The user possesses all the knowledge and have full control over the
tools
 The user knows which tools are needed and the designer knows how
to make them  the user leads the development process with the
designer as an expert resource
 The purpose of the system is not to take over some part of work but
to function as a powerful tool for the user
 The tool ideally disappear from the user's conscience in the same
way a hammer is used without conscious reflection by a carpenter.
 strength: the knowledge of the user is properly utilized
 Weakness: no generalized conclusions


© ABB Group


 the computer is seen as a medium through which
humans communicate with each other
 The focus is on use within a collective context
 Two types of communication is of interest
 The communication within groups of users
 the one-way communication from the designer to the
users

© ABB Group

Science ? Research?

 Scientific knowledge is different from everyday knowledge in that
“ the conditions, methods and reliability of the knowledge
are reviewed systematically “ (Wallèn)
 What is NOT science?
 Knowledge acquired through intuition, or skills (?)
 Is an enquiry a kind of science?
 Is construction of computer systems a science?

 Research is a systematic and methodical search for new
knowledge and new ideas

© ABB Group

Scientific traditions

 Positivism
 Empirically verifiable knowledge
 Measurement and measurement methods
 Replicating results by repeated experiments
 Explanations by cause and effect
 Looking for general facts
 The researcher should be objective
 Critique of positivism
 The human is viewed as an object
 Interpretation cannot be avoided when investigating
 human and social affairs!

© ABB Group

Theories and Models

 Theory
 A conceptual description of a phenomenon
 Basic concepts and their internal structure
 Model
 A simplified description of how different entities are related
 Example: Norman’s model of the interaction process

© ABB Group

Measurements

 How can variations in a phenomenon be measured?
 Subjective measures
 Ratings, question-answer studies
 Objective measures
 Direct observation
 E.g. Time measurements, or process logs
 Validity: do you measure what you intended to measure?
 Reliability: do different measurements give the same result?

© ABB Group

Scientific studies

 Explorative study
 May initiate new research in a field
 Contributes to discovery and understanding of problems
 Identify typical cases and relevant variables
 Descriptive study
 Characterization of a research object and its properties
 Can be rather free from theoretical assumptions
 Explanatory study
 Causal connections are looked for
 Identifying the basic mechanisms behind a certain effect

© ABB Group

Induction and deduction

 Induction
 going from collected data, trying to reach general
conclusions
 E.g. In Grounded theory: collection of data, analysis and
classification, conceptual analysis
 Deduction
 empirical testing of a hypothesis, generated from a theory
 Usually in experimental study Different factors (variables)
are varied systematically and the effects are measured

© ABB Group

In HCI Research

 What we do:
 Systematic study of users in relevant tasks
 Construction of a prototype (”hypothesis”)
 Evaluation with users (”testing the hypothesis)
 Usually an iterative process!

Is this science?

© ABB Group

Acknowledgment
Erick Stoltermann
Kerstin Severinson Eklundh
Ann Lantz
Wynekoop and Conger

Thank You
Dr. Sanjay Tripathi
Sanjay.tripathi@in.abb.com
Mob. 9740215857

© ABB Group

HCI research methods by Sanjay Tripathi, ABB

More Related Content

What's hot (12)

Viewers also liked (8)

Similar to HCI research methods by Sanjay Tripathi, ABB (20)

More from STC India UX SIG (20)

Recently uploaded (20)

HCI research methods by Sanjay Tripathi, ABB