Design Science Research (DSR)

A scientific methodology for creating and evaluating IT artifacts that solve real problems.

Quick Reference

Activity	Question	Output
1. Problem	What problem needs solving?	Problem statement, motivation
2. Objectives	What would a solution look like?	Requirements, success criteria
3. Design	How do we build it?	Artifact architecture
4. Demonstration	Does it work?	Working prototype
5. Evaluation	How well does it work?	Metrics, comparison
6. Communication	Who needs to know?	Documentation, papers

---

The DSRM Process Model

┌──────────┐   ┌──────────┐   ┌──────────┐   ┌──────────┐   ┌──────────┐   ┌──────────┐
│ 1.       │   │ 2.       │   │ 3.       │   │ 4.       │   │ 5.       │   │ 6.       │
│ Problem  │──►│Objectives│──►│ Design & │──►│ Demon-   │──►│ Evaluate │──►│Communi-  │
│ Identify │   │          │   │ Develop  │   │ strate   │   │          │   │ cate     │
└──────────┘   └──────────┘   └──────────┘   └──────────┘   └──────────┘   └──────────┘
     │                             ▲              │              │
     │                             │              └──────────────┘
     │                             │                 (iterate)
     └─────────────────────────────┘
              (refine problem understanding)

---

Activity 1: Problem Identification & Motivation

Purpose

Define the research problem and justify the value of a solution.

Questions to Answer

• What is the problem?
• Why is it important?
• Who is affected?
• What is the current state?
• What are the consequences of not solving it?

Template

<problem_identification>
  <problem_statement>
    [Clear, concise statement of the problem]
  </problem_statement>

  <importance>
    [Why does this matter? Quantify if possible]
  </importance>

  <stakeholders>
    [Who is affected by this problem?]
  </stakeholders>

  <current_state>
    [How is this currently handled? What are the limitations?]
  </current_state>

  <consequences>
    [What happens if we don't solve this?]
  </consequences>
</problem_identification>

Example

<problem_identification>
  <problem_statement>
    Human-AI collaboration lacks personalization, leading to repetitive
    context-setting and suboptimal interactions.
  </problem_statement>

  <importance>
    Developers spend 15-30% of AI interaction time re-establishing context.
    This reduces productivity and leads to frustration.
  </importance>

  <stakeholders>
    - Developers using AI assistants
    - Teams adopting AI-augmented workflows
    - Organizations investing in AI tools
  </stakeholders>

  <current_state>
    - Generic system prompts used for all users
    - No persistent memory of user preferences
    - Context lost between sessions
  </current_state>

  <consequences>
    - Continued inefficiency in human-AI collaboration
    - Slower AI adoption due to friction
    - Missed potential of personalized AI assistance
  </consequences>
</problem_identification>

---

Activity 2: Define Objectives of a Solution

Purpose

Infer objectives from the problem definition and knowledge of what is possible.

Questions to Answer

• What would an ideal solution do?
• What are the success criteria?
• How will we know if we've succeeded?
• What constraints must we respect?

Template

<objectives>
  <ideal_solution>
    [Description of what success looks like]
  </ideal_solution>

  <success_criteria>
    <criterion metric="[metric_name]">
      [Specific, measurable criterion]
    </criterion>
    <!-- Add more criteria -->
  </success_criteria>

  <constraints>
    <constraint type="technical">[constraint]</constraint>
    <constraint type="resource">[constraint]</constraint>
    <constraint type="time">[constraint]</constraint>
  </constraints>

  <non_goals>
    [What are we explicitly NOT trying to achieve?]
  </non_goals>
</objectives>

Objective Types

Type	Description	Example
Quantitative	Measurable numbers	"Reduce context-setting time by 50%"
Qualitative	Observable improvements	"Users report higher satisfaction"
Comparative	Better than alternatives	"Outperform baseline approach"
Satisficing	Meet minimum threshold	"Achieve 80% accuracy"

---

Activity 3: Design & Development

Purpose

Create the artifact - constructs, models, methods, or instantiations.

Artifact Types

Type	Description	Examples
Constructs	Vocabulary, concepts	DMMF dimensions, pattern names
Models	Abstractions, representations	Architecture diagrams, frameworks
Methods	Algorithms, processes	Workflow patterns, evaluation procedures
Instantiations	Working systems	Implemented software, prototypes

Template

<design>
  <artifact_type>[construct|model|method|instantiation]</artifact_type>

  <architecture>
    [High-level design of the artifact]
  </architecture>

  <components>
    <component name="[name]">
      <purpose>[what it does]</purpose>
      <interface>[how to interact with it]</interface>
    </component>
  </components>

  <design_decisions>
    <decision>
      <choice>[what was decided]</choice>
      <rationale>[why this choice]</rationale>
      <alternatives>[what else was considered]</alternatives>
    </decision>
  </design_decisions>

  <theoretical_foundation>
    [What theories/prior work inform this design?]
  </theoretical_foundation>
</design>

Design Principles (Hevner's Guidelines)

1. Artifact Focus - Produce something concrete and usable
2. Problem Relevance - Solve a real problem that matters
3. Rigor - Apply appropriate theoretical foundations
4. Search Process - Iterate through solution space systematically

---

Activity 4: Demonstration

Purpose

Demonstrate use of the artifact to solve the problem.

Demonstration Methods

Method	Use Case	Example
Proof of Concept	Show feasibility	Minimal working implementation
Case Study	Real-world application	Apply to actual project
Simulation	Controlled testing	Run with synthetic data
Experiment	Comparative evaluation	A/B test with alternatives

Template

<demonstration>
  <method>[proof_of_concept|case_study|simulation|experiment]</method>

  <context>
    [Where/how is the artifact being demonstrated?]
  </context>

  <scenario>
    <input>[what goes in]</input>
    <process>[what happens]</process>
    <output>[what comes out]</output>
  </scenario>

  <observations>
    [What did we observe during demonstration?]
  </observations>

  <artifacts_produced>
    [What tangible outputs were created?]
  </artifacts_produced>
</demonstration>

---

Activity 5: Evaluation

Purpose

Observe and measure how well the artifact supports a solution.

Evaluation Methods

Category	Method	Description
Observational	Case Study	In-depth analysis of real use
Observational	Field Study	Natural environment observation
Analytical	Static Analysis	Examine structure/properties
Analytical	Architecture Analysis	Evaluate design qualities
Experimental	Controlled Experiment	Isolate variables
Experimental	Simulation	Model-based testing
Testing	Functional Testing	Does it work correctly?
Testing	Structural Testing	Code coverage, paths
Descriptive	Informed Argument	Logical reasoning
Descriptive	Scenarios	Detailed use cases

Template

<evaluation>
  <method>[evaluation method]</method>

  <criteria>
    <criterion name="[name]" weight="[0-1]">
      <definition>[what does this measure?]</definition>
      <measurement>[how do we measure it?]</measurement>
      <threshold>[what counts as success?]</threshold>
    </criterion>
  </criteria>

  <results>
    <result criterion="[name]">
      <score>[actual measurement]</score>
      <evidence>[supporting data]</evidence>
    </result>
  </results>

  <analysis>
    [Interpretation of results]
  </analysis>

  <limitations>
    [What are the limitations of this evaluation?]
  </limitations>

  <improvements>
    [What could be improved based on evaluation?]
  </improvements>
</evaluation>

Evaluation Checklist

• Are criteria aligned with objectives?
• Is evidence quantifiable where possible?
• Are limitations acknowledged?
• Is comparison to alternatives included?
• Are threats to validity addressed?

---

Activity 6: Communication

Purpose

Communicate the problem, artifact, and results to relevant audiences.

Audiences

Audience	Focus	Format
Researchers	Methodology, rigor, contribution	Academic paper
Practitioners	How to use, benefits	Documentation, tutorials
Management	Business value, ROI	Executive summary
Community	Accessibility, adoption	Blog posts, talks

Template

<communication>
  <audience>[target audience]</audience>

  <key_messages>
    <message priority="1">[most important takeaway]</message>
    <message priority="2">[second takeaway]</message>
    <message priority="3">[third takeaway]</message>
  </key_messages>

  <contribution_type>
    [artifact|foundation|methodology]
  </contribution_type>

  <artifacts>
    <artifact type="[type]" location="[where]">
      [description]
    </artifact>
  </artifacts>

  <reproducibility>
    [How can others reproduce/build on this work?]
  </reproducibility>
</communication>

---

Hevner's 7 Guidelines Checklist

Use this to validate your DSR project:

#	Guideline	Question	Status
1	Design as Artifact	Did we create a concrete artifact?	[ ]
2	Problem Relevance	Does it solve a real problem?	[ ]
3	Design Evaluation	Did we rigorously evaluate it?	[ ]
4	Research Contributions	What's new/novel?	[ ]
5	Research Rigor	Did we apply appropriate methods?	[ ]
6	Design as Search	Did we explore the solution space?	[ ]
7	Communication	Did we share with relevant audiences?	[ ]

---

Entry Points

You can enter the DSRM process at different points:

Entry Point	Trigger	Start At
Problem-Centered	New problem identified	Activity 1
Objective-Centered	Industry/research need	Activity 2
Design-Centered	Existing artifact to extend	Activity 3
Client-Centered	Consulting engagement	Activity 4

---

Iteration Patterns

Pattern 1: Linear

1 → 2 → 3 → 4 → 5 → 6

Use when: Problem is well-understood, single iteration sufficient.

Pattern 2: Evaluation Loop

1 → 2 → 3 → 4 → 5 → (improvements) → 3 → 4 → 5 → 6

Use when: Iterative refinement needed based on evaluation.

Pattern 3: Problem Refinement

1 → 2 → 3 → 4 → (new insights) → 1 → 2 → 3 → 4 → 5 → 6

Use when: Demonstration reveals problem was misunderstood.

Pattern 4: Continuous

1 → 2 → 3 → 4 → 5 → 6 → (feedback) → 1 ...

Use when: Ongoing research program with multiple cycles.

---

Integration with Cognitive Workflows

DSR activities can use cognitive workflow patterns:

DSR Activity	Useful Patterns
Problem ID	Parallel - Multiple stakeholder perspectives
Objectives	Chain - Problem → Constraints → Criteria
Design	Orchestrator - Break down into design tasks
Demonstration	Route - Different demo paths per audience
Evaluation	Evaluator-Optimizer - Iterative refinement
Communication	Parallel - Different audience versions

---

References

• Hevner, A. R., March, S. T., Park, J., & Ram, S. (2004). Design Science in Information Systems Research. MIS Quarterly, 28(1), 75-105.
• Peffers, K., Tuunanen, T., Rothenberger, M. A., & Chatterjee, S. (2007). A Design Science Research Methodology. JMIS, 24(3), 45-77.

---

Evidence: evidence/patterns/004-design-science-research-methodology.md