Lecture 1 – An Introduction to Research

Lecture Objectives

• By the end of the session a student should be able to:

  • Define research in clear, operational terms.

  • Differentiate basic (academic) research from applied research.

  • Explain foundational concepts: phenomenon, paradigm, theory, hypothesis, theoretical framework.

  • Describe and contrast the inductive and deductive theory–building processes.

Defining Research

• General definition: a systematic, step-by-step, methodological process of inquiry that collects, records, analyses and interprets data to answer questions, solve problems and extend knowledge.

• Key attributes:

  • Systematic: follows an ordered sequence.

  • Methodological: guided by accepted methods and logic.

  • Evidence-based: conclusions rely on empirical data, not opinion.

  • Transformative: raw data → information → knowledge → understanding.

IT Research (Wilson, 2010)

• Purpose: gather information to aid IT-related decision making.

• Definition: “the systematic and objective process of collecting, recording, analysing and interpreting data for aid in solving managerial problems.”

• Emphasises managerial relevance, objectivity and measurability.

Research Problem

• Formal definition (Bryman, 2007): a clear statement about an area of concern, condition to improve, difficulty to eliminate, or troubling question in literature, theory or practice that requires deeper understanding and deliberate investigation.

• Alternative wording: gap in existing knowledge or a deviation from a norm/standard needing exploration.

• Crucial criterion: there must be potential solutions; if no solution exists (e.g., “how can humans live forever?”) research resources are wasted.

Illustrative Examples

• “Everybody wants to go to heaven but nobody wants to die.” – death has no realistic solution; poor research topic.

• Artificial Super-Intelligence (ASI): originally sought to build systems “smarter than the best human brains in every field.” Realisation of futility shifted focus to machine learning (machines mimicking human learning rather than surpassing all human cognition).

• Equifinality (Systems Theory): many paths can reach the same end; hence multiple methodological options exist—researchers must justify their chosen approach.

Characteristics of a Good Research Problem

• Addresses a knowledge gap and is significant to the field.

• Triggers further research and permits empirical investigation.

• Fits researcher’s interests, skills, time and resources.

• Ethically approachable.

• Clearly stated, grounded in objective reality, hints at possible solutions, advances existing knowledge.

• Goes beyond mere description (e.g., simple "yes/no" questions or trivial correlations) and asks “what” and “why.”

Example Computing/IT Research Problems

• Algorithmic (Theory/Complexity):

  • Fastest multiplication algorithm for two $n$-digit numbers.

  • Fastest matrix multiplication algorithm.

  • Can the Schwartz–Zippel lemma be derandomised?

  • Depth-first search tree in NC (poly-log time, polynomial processors).

  • Existence of a strongly polynomial algorithm for linear programming.

• Natural Language Processing (local languages):

  • Perfect syllabification algorithm for Eswatini language.

  • Perfect stemming algorithm for Sepedi.

  • Perfect POS-tagging algorithm for Zulu.

• Domain-based knowledge acquisition & text mining, health surveillance, semantic integration, humanoid robotics, bioinformatics, multi-agent systems, etc.

• Cloud Computing Security: assuring confidentiality, auditability, tamper-evidence via technologies like TPM-based remote attestation under provider-infrastructure split.

• Machine Learning: cross-disciplinary emphasis on statistical learning, optimisation, control, complexity; aims to mimic human learning by generalising patterns: “If this pattern has recurred, it will recur.”

Broader IT-Related Research Topics

• Data science / Big Data; Internet of Things (IoT); VR/AR; mobile computing; cloud computing; business intelligence & predictive analytics; agile development; MOOCs; smart cities & vehicles; knowledge management; socio-technical & behavioural science angles.

Digital Business & IoT Context

• Transition from siloed departmental IT → ERP → SOA → today’s big data, IoT, social media & cloud.

• Internet’s dual nature: great success in connectivity; failure in security, privacy, performance, resilience, etc.

• IoT will be ubiquitous, embedding online services into “every activity in life” – thus secure, dependable infrastructure becomes critical.

• Enabling buzzwords/technologies in Southern Africa: energy storage, fuel cells, gas beneficiation, renewable inputs, medical devices, natural products, 3-D printing, nanotechnology.

Types of Research

1. Basic (Academic) Research

• Conducted “for knowledge’s sake.”

• Seeks to develop or test theory; outcome is theoretical contribution.

• Often sparks further investigations.

• Goal: produce theory supported by empirical evidence (or not refuted by it).

2. Applied Research

• Seeks solutions to real-world problems; grounded in existing theory but oriented toward application.

• Produces findings that shape or evaluate policies, programmes, products, interventions.

Key Difference

• Objective: contribute to academic knowledge (basic) vs. solve practical problem (applied).

• Activities may appear similar; intent must be stated by researcher.

Foundational Concepts

Phenomenon

• Observed fact/event whose cause/explanation is questioned.

Paradigm

• Framework or worldview for thinking about a domain; guides the “how” of doing or thinking.

Theory

• Structured set of assumptions, propositions, facts explaining cause-and-effect among phenomena.

• Built on hypotheses and evidence; logically coherent; remains valid only while all new data support it.

• Examples (IS discipline): Delone & McLean IS Success Model; Design Theory; Diffusion of Innovations; Technology Acceptance Model (TAM); TOE framework; Theory of Planned Behaviour; Unified Theory of Acceptance & Use of Technology (UTAUT).

Hypothesis

• Tentative, testable statement about relationships or effects; a researcher’s educated guess.

• Null hypothesis ($H_0$): asserts no relationship/effect.

  • e.g., “Information Systems do not speed up data processing.”

• Alternative hypothesis ($H<em>a$): contradicts $H</em>0$.

  • e.g., “Information Systems do speed up data processing.”

• Null hypotheses are never proven, only rejected or fail-to-reject.

Theoretical Framework

• Organised set of concepts & relationships underpinning a study; derived from literature; defines study boundaries; keeps researcher focused.

Inductive vs Deductive Theory Processes

Inductive (Bottom-up)

1. Research question → observation → hypothesis → test/analysis → theory.

2. Generalisations are derived from data; theory is the outcome.

3. Subsequent work may apply deduction to test conditions under which the inductively-built theory holds.

Deductive (Top-down)

1. Existing theory → hypothesis → data collection → findings → confirm/reject hypothesis → revise theory.

2. Theory is springboard; research seeks empirical verification/falsification.

Report Writing

• Writing is an art/skill refined by practice and feedback.

• Always audience-centred:

  • Who will read? What is their technical level? Purpose: inform or persuade?

Core Qualities

• Adherence to objectives: show study goals were met.

• Clarity & logical flow: easy to follow, well-structured, lucid prose.

• Objectivity: report methodology, results, conclusions without bias.

• Selectivity & conciseness: include what matters; brevity without sacrificing completeness.

• Professional presentation: quality paper, good typing, attractive binding.

• Visual aids: numbered, titled tables/figures with units, sources, explanatory notes.

Brief Writing Style Tips

• Omit needless words; combine sentences; rewrite for brevity.

  • Weak: “Campus Jewelers’ main objective is to increase sales. Specifically, the objective is to double sales in the next five years by becoming a more successful business.”

  • Strong: “Campus Jewelers aims to double sales within five years.”

Standard Report Structure

1. Preliminary Material

   • Title, Table of Contents, Abstract/Synopsis.

2. Body of Report

   • Introduction (may include literature review), Methodology, Results, Discussion.

3. Supplementary Material

   • Conclusion, Recommendations, References/Bibliography, Appendices (interview transcripts, questionnaires, raw data, previous reports).

Expanded Section List

• Purpose & Scope; Limitations, Assumptions & Methods; Background/History; Conclusions & Recommendations.

• Types of specialised reports: proposals, user manuals, technical manuals, white papers, business, research, feasibility, annual reports, etc.

Oral Presentation

• Complements the written report; should employ visuals to illustrate concepts quickly; audience-focused.

Tools & Techniques

• Software: word processors, dictionaries/thesauri, citation managers, spreadsheets.

• Hardware: computer, printer.

• Information sources: books, journals, Internet, newspapers, research papers.

Malpractice & Academic Dishonesty

• Plagiarism, collusion, falsification, impersonation, copying are unethical and violate academic integrity.

Steps in Report Writing (Process View)

1. Define the research problem.

2. Gather information (literature, data).

3. Analyse information.

4. Organise findings.

5. Write the report following chosen structure.

Ethical, Philosophical & Practical Implications

• Ethics: selecting viable problems, treating participants/data honestly, choosing methods responsibly, adhering to academic integrity.

• Philosophical: understanding paradigms shapes methodology (positivist → deductive, interpretivist → inductive, etc.).

• Practical: quality research underpins sound IT-related decisions, policy formation, technological innovation (e.g., secure IoT, ML applications in indigenous languages, cloud data protection).

Connections & Real-World Relevance

• Inductive findings in one study become deductive starting points for another, forming a feedback loop advancing knowledge.

• Basic research (e.g., complexity of matrix multiplication) eventually influences applied solutions (faster graphics, cryptography, ML efficiency).

• IoT security failures illustrate the cost of neglecting rigorous applied research; emphasises need for multidisciplinary studies combining IT, policy and human factors.

Numerical / Statistical Elements (Samples)

• Nick’s Class (NC): decision problems solvable in $O((\log n)^k)$ time on a parallel computer using $n^{O(1)}$ processors.

• Goal example: “double sales in five years” (quantified target).

End of comprehensive lecture-note bullet points.