Unit 1: Foundation of research

Research is the systematic, objective investigation of a problem to discover, interpret, or revise facts, theories, or applications. The word derives from the French recherche ("to search again"). Redman and Mory define it as a "systematized effort to gain new knowledge." Everything in this unit rests on the following defining properties.

• Systematic: follows a defined sequence of steps rather than guesswork.
• Objective: conclusions rest on evidence, not on the researcher's bias.
• Empirical: grounded in observable, measurable data.
• Logical: guided by valid reasoning, inductive or deductive.
• Replicable: methods are documented so others can repeat and verify them.
• Cyclical: ends often generate new questions, restarting the process.

II. Types of Research

The classification of enquiry by purpose, data, and time.

A. Basic vs. applied research

Research is first split by whether it seeks knowledge for its own sake or for use.

1. Basic (fundamental/pure): aims to add to the body of theory; e.g., establishing a law of motion or a theory of learning. No immediate application intended.
2. Applied: aims to solve a concrete problem; e.g., testing which fertiliser raises crop yield, or improving a manufacturing defect rate.

B. Qualitative vs. quantitative research

The split rests on the nature of the data collected.

1. Qualitative: deals with non-numeric data — words, images, meanings. Uses interviews, focus groups, ethnography; asks why and how. Analysis is thematic.
2. Quantitative: deals with numeric data measured on scales. Uses surveys and experiments; asks how many/how much. Analysis is statistical (means, correlations, significance tests).

C. Longitudinal research

Studies the same subjects repeatedly over an extended period to track change.

• Design: repeated measurement of one cohort at intervals (e.g., health of 500 children measured yearly for 10 years).
• Strength: reveals developmental trends and cause-over-time; weakness: slow, costly, subject attrition.

D. Cross-sectional research

Studies many subjects at a single point in time to give a snapshot.

• Design: one measurement across groups (e.g., surveying age groups 10, 20, 30 today).
• Strength: fast and cheap; weakness: cannot show change in individuals, only differences between groups.

E. Correlation research

Measures the strength and direction of association between two or more variables without manipulating them.

• Statistic: Pearson's coefficient r, ranging from −1 to +1.

r = +1  perfect positive   r = 0  no linear relation
r = −1  perfect negative

• Caution: correlation is not causation; a confounding variable may drive both.

F. Historical research

Investigates past events by systematically gathering and evaluating evidence.

• Sources: primary (original documents, artefacts) and secondary (later accounts).
• Method: external criticism (is the source authentic?) and internal criticism (is its content accurate?).

G. Inductive vs. deductive research

Two opposite directions of reasoning.

1. Inductive: moves from specific observations to a general theory (bottom-up). Observe many white swans → propose "all swans are white." Typical of qualitative, exploratory work.
2. Deductive: moves from a general theory to a specific tested prediction (top-down). Theory → hypothesis → data → confirm/reject. Typical of quantitative, confirmatory work.

III. Research Approaches

The methodological stances that operationalise the types above.

• Quantitative approach: generation of data in quantitative form for rigorous statistical analysis. Sub-forms:
  • Inferential: survey a sample to infer characteristics of a population.
  • Experimental: manipulate an independent variable under control (treatment vs. control group).
  • Simulation: build a mathematical model of a system to observe its behaviour.
• Qualitative approach: concerned with subjective assessment of attitudes, opinions, and behaviour, using techniques such as in-depth interviews and projective methods.

IV. Significance of Research

Research matters because it converts curiosity and problems into reliable knowledge and action.

• Fuels policy: government economic and social planning relies on survey data (e.g., census-based welfare allocation).
• Solves operational problems: operations research optimises inventory, logistics, scheduling.
• Advances theory: basic research widens the intellectual frontier for later application.
• Aids decision-making: business forecasts demand and market trends before investing.
• Trains the mind: develops disciplined, logical habits and a spirit of enquiry in students.

V. Research Process

The ordered sequence from problem to report.

The process is a chain of interdependent steps; a fault in an early step propagates forward.

• 1. Define the problem: state the question precisely and delimit its scope.
• 2. Review of literature: survey existing work to avoid duplication and locate gaps.
• 3. Formulate hypotheses: state testable, tentative relationships (e.g., "training raises output").
• 4. Prepare the research design: choose exploratory, descriptive, or experimental design and plan data collection.
• 5. Determine sample design: select probability (random, stratified) or non-probability (quota, convenience) sampling.
• 6. Collect data: via observation, interview, questionnaire, schedules, or secondary sources.
• 7. Analyse data: coding, tabulation, and statistical or thematic analysis; test hypotheses for significance.
• 8. Interpret and report: draw generalisations and write the report in a standard structure (introduction, method, findings, conclusion).

VI. Criteria of Good Research

Sound research displays fixed hallmarks regardless of field.

• Clearly defined purpose: the aim and terms are unambiguous.
• Detailed, replicable procedure: enough method detail for another to repeat the study.
• Objectivity: findings free of the researcher's bias.
• Adequate analysis: methods suited to the data; error revealed, not hidden.
• Valid and reliable data: validity = measures what it claims; reliability = gives consistent results.
• Justified conclusions: claims confined to what the data support.
• Ethical integrity: informed consent, honesty, no fabrication or plagiarism.

VII. Types of Journals

The outlets through which research is published, ranked by rigour.

• Peer-reviewed (refereed) journals: manuscripts vetted by expert referees; highest credibility (e.g., Nature, The Lancet).
• Non-refereed journals: editor accepts without external review; lower authority.
• Open access journals: freely readable online, often funded by article-processing charges (e.g., PLOS ONE).
• Subscription journals: access paid by readers or libraries.
• Predatory journals: charge fees but skip genuine review; a mark of low quality to be avoided.

VIII. Journal Indexing

Metrics that quantify a journal's or scholar's influence.

Indexing databases (Scopus, Web of Science) list vetted journals; from their citation data several metrics are computed.

A. Impact factor of journals

Measures the average citations a journal's recent articles receive.

IF (year) = citations in year Y to articles from Y-1 and Y-2
            -------------------------------------------------
            number of citable articles published in Y-1 and Y-2

• Interpretation: IF of 5 means each article was cited about five times on average.

B. h index

A joint measure of productivity and citation impact for an author or journal.

• Definition: h is the largest number such that h papers each have at least h citations.
• Example: an author with h = 20 has 20 papers cited ≥ 20 times each.

C. SJR value

The SCImago Journal Rank weights citations by the prestige of the citing journal.

• Principle: a citation from a high-ranked journal counts more than one from an obscure one (a Google-PageRank-style algorithm applied to Scopus data).

D. Journal citation indices

Broader ranking tools drawn from citation databases.

• JCR (Journal Citation Reports): publishes IF and quartile rank (Q1–Q4) per subject.
• CiteScore: Scopus's counterpart, averaging citations over a four-year window.
• Eigenfactor / Immediacy index: measure total influence and speed of citation respectively.

IX. High-valued / Reputed Publishing Houses

Established publishers whose imprint signals quality control.

• Elsevier: ScienceDirect platform; publishes The Lancet, Cell.
• Springer Nature: publishes Nature and thousands of SpringerLink titles.
• Wiley (Wiley-Blackwell): broad science and humanities catalogue.
• Taylor & Francis: strong in social sciences and engineering.
• IEEE: flagship for electrical, electronic, and computing research.
• University presses: Oxford, Cambridge — rigorous scholarly monographs and journals.

X. Type of Publication: Review Process of Publication

How a manuscript travels from submission to print.

The review process is the gatekeeping mechanism that certifies quality before publication.

• Submission: author uploads manuscript to the journal's system.
• Editorial screening: editor checks scope, format, and plagiarism; may desk-reject.
• Peer review: sent to two or more expert referees. Modes contrast:
  1. Single-blind: referees know the author's identity, author does not know referees.
  2. Double-blind: neither side knows the other, reducing bias.
• Referee decision: accept, minor revision, major revision, or reject.
• Revision: author addresses comments and resubmits, often through several rounds.
• Final acceptance and production: copy-editing, proofing, assignment of a DOI, and publication.
• Types of publication produced: research articles, review articles, short communications, case studies, conference papers, and book chapters — each with its own length and review depth.