Material Phase 6: Research Methodologies & Design Science Research (DSR)

Complete study guide for Course Unit 6 (Direct Dec 2024 Exam Q4 & Q5 focus [40 Marks]). Master Research vs. Software Development methodologies, Hevner's 3-Cycle Design Science Framework, and Mixed-Methods research defenses.

1. Research Methodologies vs. Software Development Methodologies

A critical distinction frequently tested in NUST final examinations is the difference between software development execution models and scientific research methodologies.

Dimension	Software Development Methodology (Agile, Scrum, Waterfall)	Scientific CS Research Methodology (Design Science, Benchmarking, Action Research)
Primary Objective	Deliver a working software product meeting specific client/user requirements on time and within budget.	Generate novel, generalizable scientific knowledge, models, or algorithms that solve a broad research gap.
Evaluation Focus	Feature completeness, user acceptance testing (UAT), software release cycles.	Methodological rigor, empirical benchmarking against baselines, validity, and statistical evaluation.
Target Output	Executable application / commercial system.	Peer-reviewed thesis, theoretical artifact, benchmark logs, and generalizable architecture design.

⚠️ Examiner Trap Alert (Dec 2024 Exam Q4 Defense):
Never claim that "Agile" or "Waterfall" is your research methodology! Agile is a software project management framework. Your research methodology is Design Science Research (DSR) or Experimental Benchmarking, while Agile may only be the internal software development cycle used to build the evaluation prototype.

2. Design Science Research (DSR) & Hevner's 3-Cycle Framework

Design Science Research (DSR) is a computer science methodology that creates and evaluates novel technical IT artifacts (algorithms, software frameworks, database schemas, network protocols) to solve real-world problems.

Hevner's 3-Cycle DSR Model

1. The Relevance Cycle

Bridges the application environment (people, business tasks, technical problems) with the research. Defines problem requirements and accepts the artifact for field testing.

2. The Rigor Cycle

Connects research activities with the existing Knowledge Base (foundational theories, scientific algorithms, design methodologies, state-of-the-art benchmarks).

3. The Design Cycle

The core iterative feedback loop between construction (building the artifact) and evaluation (testing against performance metrics).

The 6 Steps of DSR Execution

Problem Identification & Motivation: Defining the exact computing gap.
Define Objectives of a Solution: Quantitative performance targets (e.g., reduce throughput overhead below 3%).
Design & Development: Creating the artifact (architecture, code, schema).
Demonstration: Deploying the artifact in a simulated or real environment.
Evaluation: Empirically measuring performance against objectives using controlled benchmarks or user trials.
Communication: Documenting results in scientific publications and theses.

3. Mixed Methods Research & Defense Strategy

Mixed Methods Research combines both quantitative (benchmarks, statistical telemetry) and qualitative (developer interviews, usability focus groups) approaches to provide a complete understanding of a research problem.

Sequential Explanatory

Quantitative benchmark collection $\rightarrow$ Followed by qualitative developer interviews to explain unexpected quantitative anomalies.

Sequential Exploratory

Qualitative interviews to discover developer needs $\rightarrow$ Followed by quantitative prototype benchmarking to validate design solutions.

Concurrent Triangulation

Quantitative performance logs and qualitative user feedback collected simultaneously and compared to cross-validate findings.