Design and natural science research on information technology

1
Design and natural science research on
information technology
Salvatore T. March, Gerald F. Smith
Decision Support Systems 15 (1995)
Discussion for INSY 6306 Fall 2002 David Weltman
2
Contents

• Introduction
• Background
• Research Framework 4X4
• Research Outputs
• Research Activities
• Research Area Illustrations (cells)
• Data management
• Further Research Areas/Discussion

3
Introduction

• This paper discusses two types of research
• Design Science Research (DS)
• Natural Science Research (NS)
• Proposes a 4X4 framework for IS research
• Design Natural Science Research
• Outputs Produced by Design Research
• Argument Both Design Natural Science
  Activities Needed in an IT research framework

4
Survey of Information Systems

• Information
• Technology, practical nature
• IT practice, design activities
• Artifacts can major have or impacts,
  pervasiveness
• Two IT Research Arguments/Divisions? Dual Nature.
• Understanding the nature of IT (NS) knowledge
  producing
• Improving IT performance (DS) knowledge using
• 4X4 Framework Reconciles These 2 Schools of
  Thought
• And suggests a further IT research study agenda

5
Natural Science

• How and why things are. Understanding of
  reality.
• Theories (ultimate products of NS)
• Explanations
• Progress new theories with better/deeper
  explanations
• Discovery Justification
• D is generating/proposing claims (creative)
• J is testing the claims for validity
  (logic/empirical)

6
Design Science

• Designing artifacts (systems) to attain goals
  (new and/or better systems)
• Practical the creation of useful things
  (models, methods, implementations that are
  innovative and useful).
• Technology oriented
• The technology products are assessed against a
  criteria-Evaluation
• Does it work?
• Is it an improvement?
• Is the evaluation criteria good (considering the
  environment)?
• Used extensively in engineering, architecture,
• DS products are constructs, models, methods, and
  implementations
• DS activities are build and evaluate
• Progress ? when technologies are improved

7
DS and NS Interactions/Synergies

• DS ? Artifacts
• NS ? Artifacts
• Past Example
• DS? GDSS
• NS studies the behavior the GDSS fosters
• Current/Future Example
• DS? Instant Messaging Systems
• Certainly this has changed organizational
  behavior and is a significant area for NS
  investigations
• The converse is true also
• NS? Knowledge (explanations of behaviors/reactions
  )
• This knowledge can be used to develop better/more
  useful technologies
• DS can substantiate NS Claims
• NS theories receive support through actual tests
  (things that were designed). Engineering
  examples. Medical Examples. Build or Make and
  it works or not.

8
Problems with Prior Research IT Frameworks

• Characterized by
• Id sets of variables to be studied
• Hypothesis and positing interactions among
  variables??
• 1. Fails to provide prioritization among
  variables
• 2. Does not account for large body of DS research
• 3. Fails to recognize artificialness phenomena
• 4. Does not handle changing phenomena/environments

9
Research 4X4 Framework Overview
Research Activities
Design Science
Natural Science
Evaluate
Theorize
Justify
Build
Constructs
Research Outputs
Model
Method
Instantiation
Increased Performance Knowledge Application
How and Why, Claims Knowledge Creation
Design Science Products
10
Research Output - Constructs

• Basic language used to characterize phenomena.
• Concepts to describe problems and solutions in
  the domain.

11
Research Output - Models

• Description/representation of how things are.
• Statements expressing relationships among
  constructs.
• Representations of the problem as solution
  statements.
• Examples
• E-R constructs to represent data requirements
• Tree Structure
• Prototype
• Simulations
• Expert System
• Describe tasks, situations, or artifacts.

12
Research Output - Methods

• Set of steps (algorithms or guidelines) used to
  perform a task.
• Ways of performing goal-directed activities.
• Examples
• Data Structures new ways to store and retrieve
  data
• OR- model with objective function and constraints
  that leads to new optimization algorithms

13
Research Output Instantiations/Implementations

• Specific products and implementations that
  perform certain tasks.
• Realization of an artifact in its environment.
• Provides working artifacts (for further study).
• They demonstrate the feasibility and
  effectiveness of the models and methods upon
  which they are based.
• Leads to improved instantiations.
• Example Linux

14
Research Activities-Design Science

• Build
• Evaluate

15
Research Activities-Natural Science

• Theorize
• Explain why the observed performance happen in
  its environment.
• Can explain things about both the artifact and
  the environment.
• Example When employees are involved in
  non-targeted activities (unit conference calls,
  meetings), they access applications (e-mail, IM,
  web) for engaging activities. Theory ?
  Traditional/Non-customized activities need to
  change (use engaging systems) to be effective in
  todays environment.
• Justify
• Example Use a GDSS or other engaging system to
  conduct meetings. Then gather evidence to
  support (or refute) effectiveness of the meetings
  (compare with traditional meetings).
• The metrics/measures of success must be
  justified.

16
Application of Framework to IT Research

• There can be valid research activities about each
  of the outputs.
• Each cell has different objectives and methods.
• Research can cover multiple cells.
• Evaluation of research should be based on the
  cell(s) in which the research lies.

17
Build Research

• Should be evaluated based on utility.

18
Evaluative Research

• Develops metrics and compares the performance of
  artifacts.
• Metrics define what the research area is trying
  to accomplish.
• Are subsequent artifacts significant performance
  improvements?
• Evaluation of Constructs
• Completeness, simplicity, elegance
• Evaluation of Models
• Match real world, completeness, detail
• Evaluation of Methods
• Usefulness, efficient, quality results
• Evaluation of Instantiations
• Efficiency and effectiveness of the
  implementation and its impacts on the
  environment/users.
• Goal Determination of how well the artifact
  works (not how or why).

19
Theorize and Justify

• Theorize explanation of how and why the
  artifacts work
• Justify testing of theories, can involve
  empirical research. Can provide future direction
  for technologies.

20
Illustrations

• 1. Evaluating a Model
• Simulation of creative brain activity during
  video game play
• How well does the computer simulation match
  actual measured results in a lab?
• Better/improved models
• Theories about brain activity
• 2. Theorizing about an Instantiation
• Explain how IM affects communication
• New theories about communication
• New artifacts related to IT communication
  systems

21
Conclusions/Discussions/Future Research

• (3) Implications
• 1. Our theories need to be based on Natural
  Phenomena.
• 2. Our artifacts are perishable.
• 3. Our artifacts are being produced at an
  increasing rate. Many areas for research. The
  4X4 will help with the catagorization, can reduce
  wasted efforts and provide further direction.
• Example There has been only moderate efforts in
  Evaluation Activity.
• Example Very little research has been done
  theorizing why and how artifacts work (NS area).