CampusLearn
← CSC 316 Research Methodology in Computing

Introduction to Computing Research and Problem Formulation

Notes

Introduction to Research in Computing Research in computer science is a systematic inquiry aimed at creating new knowledge, validating existing models, or solving complex computational problems. Unlike ordinary software engineering where the primary goal is delivering a product to a client, computing research focuses on generalizable algorithms, architectural novelty, empirical performance boundaries, and formal verification. A computer science researcher seeks to understand why an algorithm behaves in a particular manner under constraints and how computational theories apply to real-world phenomena. Defining a Computable Research Problem A valid research problem in computing must be precise, testable, and grounded in a verifiable gap within existing scientific literature. Novice researchers often mistake commercial software requirements for research questions. A genuine research problem addresses computational limitations such as algorithmic complexity, throughput bottlenecks, security vulnerabilities, or poor model convergence under noisy input. Formulating the problem requires defining the scope, input parameters, computational constraints, and expected scientific contributions. Research Questions and Hypotheses Research questions guide the investigation by specifying what the researcher aims to uncover. In quantitative computing research, questions are accompanied by formal null and alternative hypotheses. For example, a null hypothesis might state that an optimized routing algorithm offers no statistically significant reduction in packet drop rates compared to standard protocols. The alternative hypothesis claims the presence of an effect, which must be tested using empirical metrics. Literature Review and Knowledge Gap The literature review serves as the foundation for positioning any computing research. Researchers systematically evaluate peer-reviewed conference proceedings and journals to establish the state of the art. Through comparative analysis of related works, researchers identify methodology gaps, datasets that lack diversity, or algorithmic assumptions that fail under non-ideal operating conditions. Practical example A third-year student at a Kenyan university investigates network latency issues in USSD-based financial services used by tea farmers in Nyeri County. Commercial aggregators report transaction timeouts during peak harvest seasons. Problem Statement: Existing USSD gateway load-balancing algorithms use static round-robin scheduling, resulting in queue overflow and an average transaction latency exceeding 8.5 seconds during peak agricultural pay periods. Research Question: Can a dynamic reinforcement-learning-based queuing algorithm reduce USSD transaction drop rates while keeping latency below 3 seconds on constrained telecommunications infrastructure? Hypothesis: The proposed dynamic queuing model reduces average end-to-end transaction latency by at least 40 percent compared to static round-robin scheduling under simulated high-volume concurrent USSD requests.

Quiz

1. What fundamentally distinguishes computer science research from standard commercial software engineering?

2. Why is static round-robin scheduling insufficient in the Nyeri USSD practical example?

3. What is the primary role of a null hypothesis in quantitative computing experiments?

Sign in to submit and securely grade your quiz.