Unit 2: An Introduction to Research

I. Foundations of Research

Research is a systematic and objective investigation undertaken to answer a question, solve a problem, test an explanation, or generate knowledge. It differs from casual inquiry because it follows an explicit process, uses appropriate evidence, and permits others to evaluate how conclusions were reached.

• Systematic character: Research proceeds through connected stages—problem formulation, literature review, design, data collection, analysis, and interpretation—rather than through random information gathering.
• Empirical basis: Conclusions are supported by observable evidence, such as survey responses, laboratory measurements, historical documents, or interview transcripts.
• Logical reasoning: Researchers use deduction to derive predictions from theory and induction to develop broader explanations from observed patterns.
• Objectivity: Procedures are designed to reduce personal bias; for example, a standardized questionnaire presents the same questions to every respondent.
• Verifiability: Methods, measures, and sources are documented so that findings can be checked, replicated, or critically examined.
• Ethical conduct: Research protects participants through informed consent, confidentiality, voluntary participation, and avoidance of unnecessary harm.
• Purpose orientation: Basic research expands theoretical knowledge, whereas applied research addresses practical problems such as employee turnover or declining crop yields.
• Provisional knowledge: A conclusion is accepted according to the available evidence and may be revised when stronger evidence appears.

II. Research and Its Process — From Inquiry to Evidence

Research converts a broad area of curiosity into a clearly framed inquiry and then uses a disciplined sequence of decisions to produce defensible findings.

A. Meaning of research

Research is the organized search for reliable knowledge through the collection, analysis, and interpretation of relevant evidence.

• Etymological sense: “Research” suggests searching carefully or repeatedly, but academically it means more than locating facts; it involves examining relationships, explanations, or solutions.
• Central objectives: A study may explore an unfamiliar issue, describe a condition, explain causal relationships, predict outcomes, or evaluate an intervention.
  • Exploration: Interviews may identify reasons why first-year students leave college.
  • Description: A survey may estimate the percentage of students using online learning platforms.
  • Explanation: A study may test whether platform use affects academic performance.
• Basic research: It develops or tests concepts and theories without requiring an immediate practical result; for example, studying how memory retrieval operates.
• Applied research: It seeks an actionable answer in a particular setting; for example, comparing two training programmes to reduce workplace accidents.
• Quantitative research: It represents variables numerically and commonly uses statistical analysis, as when income is recorded in currency units and related to household expenditure.
• Qualitative research: It studies meanings, experiences, and processes through material such as interviews, observations, and documents.
• Mixed-methods research: It combines numerical and qualitative evidence; survey results might show the scale of job dissatisfaction, while interviews explain its causes.
• Essential distinction: Data collection alone is not research. A customer list becomes research evidence only when collected and analysed in relation to a defined question.

B. Research process

The research process is an interrelated sequence in which each stage shapes the quality and meaning of later stages.

• Problem identification: The researcher recognizes a knowledge gap or practical difficulty, such as a sustained fall in employee retention.
• Preliminary investigation: Background reading and informal discussion clarify the context, terminology, affected population, and available evidence.
• Literature review: Existing theories, methods, and findings are examined to prevent unnecessary duplication and reveal unresolved questions.
• Problem formulation: The broad concern is converted into a precise statement, such as: “What is the relationship between flexible working hours and annual turnover among nurses in urban private hospitals?”
• Objectives and questions: The main purpose is divided into answerable tasks; an objective may be “to compare turnover rates under fixed and flexible schedules.”
• Hypothesis development: Where appropriate, a testable prediction is stated. A simple relational hypothesis is:

H₁: X is associated with Y.
H₀: X is not associated with Y.

• H₁ is the alternative hypothesis, H₀ the null hypothesis, X flexible scheduling, and Y employee turnover.
  • Research design: The researcher selects the overall plan, including study setting, time horizon, sampling method, measurement tools, and analytical procedures.
  • Sampling: A subset is selected from the target population; for example, 300 nurses may be sampled from all nurses employed by selected hospitals.
  • Data collection: Evidence is gathered through questionnaires, interviews, experiments, observation, records, or other suitable instruments.
  • Data analysis: Quantitative data may be analysed through percentages, correlations, or regression, while qualitative data may be coded into recurring themes.
  • Interpretation and reporting: Findings are related to the questions and literature, limitations are disclosed, and conclusions are restricted to what the evidence supports.
  • Iterative nature: The process is not always strictly linear; pilot testing may expose an unclear measure and require revision of the instrument or problem statement.

III. The Research Problem — Focusing the Investigation

A research problem is a clearly expressed issue, contradiction, knowledge gap, or practical difficulty that can be investigated through evidence. It determines what information is relevant and what the study is expected to accomplish.

A. Defining the research problem

Defining the research problem means converting a broad topic into a precise, feasible, and researchable statement.

• Broad-to-specific movement: “Social media” is only a topic; “the relationship between daily social-media use and sleep duration among university students” identifies variables and a population.
• Context specification: A complete definition may delimit the setting, participants, and period—for example, undergraduate students at public universities during one academic year.
• Variable identification: The researcher identifies the main concepts and their possible relationship; “daily use” may be measured in hours and “sleep duration” in average hours per night.
• Operational definition: Abstract ideas are translated into observable indicators. Academic achievement, for instance, may be operationalized as semester grade-point average.
• Problem statement: It explains the existing condition, the gap in knowledge, and the consequence of leaving that gap unresolved.
• Research questions: Questions specify the required evidence; “Does daily use predict sleep duration?” is more investigable than “Is social media harmful?”
• Objectives: Objectives use action verbs such as measure, compare, determine, or explain and must correspond directly to the problem.
• Delimitation: Boundaries deliberately narrow the study; excluding school students may improve manageability but limits generalization to that excluded group.

B. Research problem selection

Research problem selection is the evaluation of possible problems to identify one that is significant, researchable, ethical, and manageable.

• Researcher interest: Sustained interest supports careful work, but personal preference alone cannot establish scholarly value.
• Significance: The problem should contribute to theory, policy, professional practice, or social understanding; measuring hospital waiting time may guide resource allocation.
• Originality: A study may be original by examining a new population, period, variable, method, or context rather than by addressing an entirely unknown subject.
• Feasibility: The scope must fit available time, finance, skills, equipment, and access. A national census may be unsuitable for a short student project.
• Data availability: Necessary participants, records, or measurements must be accessible; confidential medical files cannot be assumed to be obtainable.
• Researcher competence: The chosen problem should match methodological and subject knowledge, or permit access to needed expertise.
• Ethical acceptability: Potential benefits must justify risks, and vulnerable participants require additional safeguards.
• Scope control: A problem should be neither so broad that it becomes superficial nor so narrow that it yields little meaningful evidence.
• Decision criterion: A feasible question with modest scope is preferable to an ambitious question that cannot be answered credibly.

C. Understanding the research problem

Understanding the research problem requires examining its background, concepts, assumptions, causes, stakeholders, and boundaries before fixing the method.

• Situational background: The researcher determines where and when the problem occurs and whether it is persistent, emerging, or context-specific.
• Literature mapping: Prior studies reveal accepted definitions, established findings, disputed explanations, and under-researched areas.
• Conceptual clarity: Similar terms must be distinguished; employee turnover intention is a stated intention, while actual turnover is recorded departure.
• Stakeholder analysis: Those affected may perceive the issue differently; patients, physicians, and administrators may use different criteria for healthcare quality.
• Causal caution: A visible association does not automatically establish cause. Low attendance and low grades may both be influenced by employment commitments.
• Assumption testing: The researcher should examine assumptions such as honest self-reporting or accurate institutional records.
• Preliminary evidence: Pilot interviews, field observation, or small-scale record analysis can reveal whether the proposed variables and questions are suitable.
• Conceptual framework: A diagram or written model organizes expected relationships, such as workload influencing stress, which then influences turnover intention.

D. Necessity of a defined problem

A defined problem is necessary because it provides direction and creates the standard against which every research decision is judged.

• Focus: It separates relevant evidence from interesting but unrelated information.
• Design alignment: A causal question may require an experiment or strong quasi-experimental design, whereas a descriptive question may require a survey.
• Measurement guidance: Clearly named variables determine which indicators and instruments must be used.
• Resource economy: Defined boundaries prevent unnecessary sampling, reading, and data collection.
• Analytical coherence: The problem links questions, hypotheses, evidence, analysis, and conclusions into one logical chain.
• Evaluation: Readers can judge whether the research has answered what it originally set out to investigate.
• Error prevention: An undefined problem encourages “data fishing,” in which patterns are sought without a prior rationale and chance findings may be overstated.

IV. Research Design — The Blueprint of Investigation

Research design is the overall plan connecting the research problem to the evidence needed for an answer. It specifies what will be studied, from whom data will be obtained, how variables will be measured, and how results will be analysed.

A. Research design

Research design translates questions and objectives into an operational structure for data collection and interpretation.

• Core components: A design identifies the unit of analysis, population, sample, setting, time horizon, variables, instruments, procedures, and analysis plan.
• Control of variation: It helps distinguish the relationship of interest from alternative explanations through randomization, matching, statistical control, or careful comparison.
• Validity: Internal validity concerns whether the observed effect has the proposed explanation; external validity concerns whether findings generalize beyond the study.
• Reliability: Consistent procedures improve repeatability; a standardized scale should produce reasonably stable results under similar conditions.
• Time dimension: A cross-sectional design examines one period, while a longitudinal design follows change across multiple periods.
• Ethical integration: Consent, privacy, secure data storage, and withdrawal procedures must be built into the plan rather than added after collection begins.

B. Need for research design

Research design is needed to obtain relevant evidence efficiently while reducing bias, ambiguity, and avoidable error.

• Advance planning: Decisions made before data collection reduce arbitrary changes prompted by early results.
• Question–method fit: The design ensures that the method can answer the question; an opinion survey cannot by itself establish that a medicine caused recovery.
• Bias reduction: Random sampling reduces selection bias, while blinding can reduce observer or participant expectations.
• Precision: Appropriate sample size and reliable measurement improve the accuracy of estimates.
• Coordination: A written plan aligns fieldworkers, instruments, schedules, budgets, and data-management procedures.
• Transparency: Explicit procedures allow supervisors, ethics committees, and readers to assess the study’s credibility.
• Economy: Design prevents collecting excessive data while ensuring that essential variables are not omitted.

C. Types of research design

Types of research design differ according to the study’s purpose, level of control, and kind of evidence required.

1. Exploratory design: Used when the problem is insufficiently understood; flexible interviews, focus groups, case studies, and literature searches help generate concepts or hypotheses.
2. Descriptive design: Portrays characteristics, frequencies, or trends; a cross-sectional survey may estimate unemployment within a defined population.
3. Diagnostic or analytical design: Examines associations and factors connected with a condition, such as whether training, experience, and workload predict productivity.
4. Experimental design: Manipulates an independent variable, uses a comparison or control group, and ideally assigns participants randomly to support causal inference.
5. Quasi-experimental design: Evaluates an intervention without full random assignment; one school adopting a programme may be compared with a similar school that does not.
6. Correlational design: Measures the direction and strength of relationships without manipulation; correlation alone does not prove causation.
7. Case-study design: Investigates a bounded case—an individual, institution, event, or community—in depth using multiple evidence sources.
8. Cross-sectional and longitudinal designs: Cross-sectional studies provide a one-time snapshot, whereas longitudinal studies track the same or comparable units over time.
9. Qualitative, quantitative, and mixed-methods designs: Selection depends on whether the study requires numerical estimation, contextual interpretation, or an integrated use of both forms of evidence.