Design Science Research Skill

Purpose

Supports DSR papers by structuring research per Hevner et al. (2004), enforcing the distinction between existence proofs and causal claims, and managing process evidence (Git commits, AI session logs, artifact versioning). Tracks co-evolution of artifact and manuscript, and keeps modality (claim strength) consistent.

When to Use

• When designing or writing a paper that adopts the DSR paradigm
• When developing an artifact (tool, skill, system) alongside the paper
• When you need to distinguish "this was constructible" from "this caused improvement"
• When reviewers flag conflation of existence proof and empirical demonstration

Prerequisites

• research-plan-workshop completed
• scholarly-concept-modeling completed

Procedure

Step 1: DSR Guidelines Check

Map the study to Hevner et al. (2004)'s seven guidelines:

Guideline	Check
1. Design as an Artifact	What is the artifact (skill, tool, method)?
2. Problem Relevance	What practical/theoretical problem does it solve?
3. Design Evaluation	How is it evaluated (existence proof, case study, experiment)?
4. Research Contributions	Contribution type: design proposition, existence proof, or empirical demonstration?
5. Research Rigor	How is rigor ensured in construction and evaluation?
6. Design as a Search Process	How is the search process recorded?
7. Communication of Research	How is it communicated to both practitioners and researchers?

Step 2: Existence Proof vs Causal Claim

Classify each claim in the manuscript:

Type	Definition	Example wording
Design proposition	A proposal for how to design	"should be...", "it is desirable to..."
Existence proof	That it was constructible/observable	"was constructible", "was observed"
Causal claim	That it caused improvement	"improved by...", "led to..."
Generalization	Applicability to other contexts	"generalizes to...", "universal"

Invariant: Do not use causal wording for existence proofs. Do not claim statistical generalization from n=1 or single-case studies.

Step 3: Process Evidence Management

In DSR, the design and construction process itself is evidence. Structure and record:

Evidence	Management
Git commit history	Audit index table (timestamp, milestone, hash)
AI session logs	Dialogue logs with redaction policy (local paths, PII)
Execution environment	Verify Agent:/Model: trailers
Negative cases (rejections, failures)	Structured NC-XX records

Step 4: Artifact ↔ Manuscript Mapping

Track artifact versions against manuscript descriptions:

| Artifact version | Manuscript section | Commit |
| :--- | :--- | :--- |
| v2.0.0 | Ch.4 §4.2 reference implementation | `abc1234` |
| v2.4.0 | Ch.4 §4.3 co-adaptation | `def5678` |

Step 5: Modality Calibration Report

Output to docs/design/dsr-structure.md:

• DSR guidelines mapping
• Existence proof / causal claim classification
• Process evidence index
• Artifact-manuscript mapping table

Outputs

• docs/design/dsr-structure.md (DSR structuring report)