Complete study guide for Course Unit 1. Master the 6 core characteristics of scientific research, problem statement architecture, aims vs. objectives, and hypothesis vs. research question formulation to achieve distinction grade in SCS2110.
Scientific research in Computer Science is a controlled, empirical, critical, and systematic inquiry designed to solve an identified computing problem and contribute new, generalizable knowledge to the discipline.
A frequent exam question asks why building a web portal, mobile app, or CRUD software system is NOT research by itself. Remember:
Extraneous variables are isolated so observed effects are strictly attributable to the independent variable. CS Example: When comparing two database engines, CPU core count, memory allocation, storage disk I/O, and concurrent query loads must be held strictly constant.
Every methodological choice (algorithm selection, sample size, metrics like throughput or memory footprint) is meticulously justified using domain literature and theoretical principles.
Conclusions are anchored strictly in observable, measurable data collected from real-world execution, execution traces, or user trials, never subjective claims.
Follows a structured, logical sequence: Problem Definition → Literature Review → Methodology Design → Data Collection & Execution → Data Analysis → Critical Conclusion.
Conclusions follow logically from evidence, and independent researchers can replicate the experimental setup to achieve identical results.
The researcher actively audits potential flaws, edge cases, algorithmic biases, and hardware limitations rather than hiding negative results.
A research problem statement articulates a specific gap, inefficiency, contradiction, or limitation in existing computing solutions.
A NUST-compliant problem statement consists of 3 distinct components:
| Feature | Research Aim | Research Objectives |
|---|---|---|
| Definition | Broad, high-level statement of what the research intends to achieve overall (The Ultimate Destination). | Specific, measurable operational steps required to achieve the overall aim (The Roadmap). |
| Quantity | Single overarching sentence. | Typically 3 to 5 sequential bullet points. |
| Verbs Used | To investigate, To design, To develop, To evaluate... | Must start with SMART action verbs: To analyze, To design, To implement, To measure, To evaluate, To compare. |
| CS Example | "The aim of this research is to design and evaluate a lightweight anomaly detection algorithm for resource-constrained IoT edge devices." |
1. To analyze existing IoT anomaly detection algorithms for memory overhead. 2. To design a quantized feature extraction model for microcontrollers. 3. To evaluate the latency and detection accuracy of the model against baseline datasets. |
In CS research, deciding whether to formulate a Research Question or a Hypothesis depends on the nature of the research paradigm and prior knowledge available.
A tentative, testable prediction about the directional relationship between independent and dependent variables. Used in quantitative experimental research.
Formula: "If [Independent Variable], then [Dependent Variable]."
Example: H1: Implementing dynamic memory page compression in Linux kernel 6.x reduces cache miss rates by at least 20% under high concurrency workloads.
An open-ended inquiry used when exploring a novel area where existing literature is insufficient to predict a directional outcome.
Formula: "How does [X] impact [Y]?" or "What are the primary factors affecting [Z]?"
Example: RQ1: How do software developers perceive the maintainability challenges of microservice architectures during serverless migration?
If you encounter Phase 1 drill questions or exam questions asking you to critique or write hypotheses vs RQs: Always verify that Hypotheses specify measurable variables (Independent vs Dependent) and clear directional metrics (e.g., latency reduction by 25%, energy savings by 15%).