8 Methods of Research

The official syllabus names five methods: Experimental, Descriptive, Historical, Qualitative, and Quantitative. Each method has a different question form, a different data type, and a different inferential strength.

Five Methods at a Glance

Method	Asks	Data	Strength
Experimental	“Does X cause Y?”	Numerical (manipulated + measured)	Strong causal inference
Descriptive	“What is happening?”	Numerical or categorical	Accurate portrayal
Historical	“What happened in the past, and why?”	Documents, artefacts, oral records	Narrative reconstruction
Qualitative	“Why?” “How does it feel?” “What does it mean?”	Textual, observational	Depth, meaning
Quantitative	“How much?” “How many?” “Is X related to Y?”	Numerical	Generalisation, statistical inference

8.1 Experimental Method

The experimental method tests a causal hypothesis by manipulating an independent variable and measuring a dependent variable while controlling extraneous variables.

Three Defining Conditions of an Experiment

Condition	What it requires
Manipulation	Researcher creates the levels of the independent variable
Control	Extraneous variables are held constant or randomised
Comparison	Performance of treated group is compared with a control group

8.1.1 Common Experimental Designs

Working Experimental Designs

Design	Structure	Strength
One-shot case study	One group → treatment → post-test	Weakest — no control
One-group pre-test / post-test	One group → pre-test → treatment → post-test	Better but no control group
Static-group comparison	Treated group vs untreated group, post-test only	No randomisation
Pre-test / post-test control group	Random assignment → both groups pre-tested → only one treated → both post-tested	Strong — gold standard
Solomon four-group	Four groups; controls for testing effect	Strongest
Factorial design	Two or more independent variables manipulated together	Reveals interaction effects

flowchart LR
  P[Population] --> R[Random<br/>assignment]
  R --> T[Treatment group<br/>Pre-test → Treatment → Post-test]
  R --> C[Control group<br/>Pre-test → No treatment → Post-test]
  T --> A[Compare<br/>post-test means]
  C --> A
    classDef default fill:#003366,color:#ffffff,stroke:#ffcc00,stroke-width:3px,rx:10px,ry:10px;

8.1.2 Quasi-Experimental Designs

When random assignment is impossible (intact classrooms, naturally occurring groups), the design is quasi-experimental. Examples include the non-equivalent control group design and the interrupted time-series design. Quasi-experiments support weaker causal claims than true experiments.

8.1.3 Threats to Validity

Internal and External Validity

Validity	What it asks	Threats
Internal validity	Did the treatment cause the change?	History, maturation, testing, instrumentation, regression to mean, selection, mortality
External validity	Will the result generalise?	Sample-population mismatch, ecological validity, treatment-context interaction

8.2 Descriptive Method

The descriptive method portrays the characteristics of a phenomenon as it is, without manipulation. The researcher observes and reports — what is, not what causes what.

Sub-types of Descriptive Method

Sub-type	What it does	Tools
Survey	Gathers data from a sample by questionnaire or interview	Questionnaire, interview schedule
Case study	In-depth investigation of one or a few cases	Multiple sources of evidence
Observational	Records behaviour as it occurs	Observation schedule, field notes
Correlational	Measures relationships between variables without manipulation	Statistical tests of association
Developmental	Describes change across age or time	Cross-sectional / longitudinal
Comparative / Causal-comparative (ex-post-facto)	Compares groups that already differ	Statistical group comparison

The causal-comparative or ex-post-facto design is examined frequently. It looks like an experiment but the “treatment” has already happened naturally — the researcher cannot manipulate it. Examples: comparing reading scores of children who watch much TV vs little TV.

8.3 Historical Method

Historical research reconstructs past events by examining primary and secondary sources, evaluating the authenticity and credibility of the records, and synthesising the evidence into a coherent narrative or interpretation.

Sources in Historical Research

Source	Definition	Examples
Primary source	Original records produced at the time	Diaries, letters, official documents, photographs, interviews with eyewitnesses
Secondary source	Accounts produced later that interpret primary sources	Textbooks, review articles, encyclopaedia entries

External and Internal Criticism

Criticism	What it asks	Question
External criticism	Is the document authentic?	“Is this letter genuinely from 1857, and by the person it claims?”
Internal criticism	Is the content credible?	“Is the writer telling the truth, and were they in a position to know?”

flowchart TB
  S[Source<br/>identification] --> EC[External criticism<br/>Authenticity]
  EC --> IC[Internal criticism<br/>Credibility]
  IC --> SY[Synthesis<br/>Narrative / Interpretation]
    classDef default fill:#003366,color:#ffffff,stroke:#ffcc00,stroke-width:3px,rx:10px,ry:10px;

8.4 Qualitative Method

Qualitative research aims to understand the meaning people attach to actions, events and contexts. Data are textual or observational rather than numerical; sample sizes are small but rich.

Major Qualitative Approaches

Approach	Goal	Typical method
Ethnography	Understand a culture or community from the inside	Participant observation; long fieldwork
Phenomenology	Understand the lived experience of a phenomenon	In-depth interviews
Grounded theory	Build theory from data	Iterative coding; constant comparison
Case study	In-depth analysis of a bounded system	Multiple data sources
Narrative inquiry	Understand experience as story	Life-history interviews
Action research	Improve a local practice through cycles	Plan-Act-Observe-Reflect

8.4.1 Qualitative Data Collection Tools

Interview — structured, semi-structured, or unstructured.
Focus group discussion — 6 to 10 participants discussing a topic.
Participant observation — researcher embedded in the setting.
Document analysis — letters, reports, diaries, social media.
Field notes — researcher’s reflections and observations.

8.4.2 Trustworthiness in Qualitative Research

The qualitative parallel to validity and reliability is trustworthiness, with four criteria:

Four Criteria of Trustworthiness (Lincoln & Guba)

Criterion	Quantitative parallel	What it asks
Credibility	Internal validity	Are the findings believable?
Transferability	External validity	Can findings transfer to other contexts?
Dependability	Reliability	Are findings consistent?
Confirmability	Objectivity	Are findings shaped by participants, not by researcher bias?

8.5 Quantitative Method

Quantitative research asks questions that can be answered with numbers. Data are collected through measurement, summarised with descriptive statistics, and analysed with inferential statistics to draw generalisations from sample to population.

Quantitative Designs

Design	Question form	Example
Survey	“What is the prevalence?”	NSSO household survey
Correlational	“Is X associated with Y?”	Income and life expectancy
Experimental	“Does X cause Y?”	Drug trial
Quasi-experimental	“Does X cause Y, given non-random groups?”	Policy change in two states
Causal-comparative	“Why do these groups differ?”	Boys vs girls in maths

8.5.1 Sampling

Sampling Methods

Family	Method	When to use
Probability	Simple random	Every unit has equal chance
Probability	Systematic	Every kᵗʰ unit
Probability	Stratified	Sub-groups of interest must be represented
Probability	Cluster	Population naturally clustered (schools, villages)
Probability	Multi-stage	Combination of the above for very large populations
Non-probability	Convenience	Whoever is available
Non-probability	Purposive / Judgmental	Researcher selects information-rich cases
Non-probability	Quota	Pre-set numbers from each sub-group, not randomly chosen
Non-probability	Snowball	Each participant refers the next

flowchart TB
  S[Sampling] --> P[Probability]
  S --> N[Non-probability]
  P --> P1[Simple random]
  P --> P2[Systematic]
  P --> P3[Stratified]
  P --> P4[Cluster]
  P --> P5[Multi-stage]
  N --> N1[Convenience]
  N --> N2[Purposive]
  N --> N3[Quota]
  N --> N4[Snowball]
    classDef default fill:#003366,color:#ffffff,stroke:#ffcc00,stroke-width:3px,rx:10px,ry:10px;

8.6 Mixed Methods

Mixed-methods research combines quantitative and qualitative data in a single study.

Three Common Mixed-Methods Designs

Design	Sequence	Use
Convergent / parallel	Quantitative + Qualitative collected together; then merged	Triangulate findings
Explanatory sequential	Quantitative first → Qualitative to explain	When numbers raise the “why” question
Exploratory sequential	Qualitative first → Quantitative to test	When the field is new and constructs need development

8.7 Practice Questions

Q 01 Experimental Method Easy

Which of the following is a defining condition of an experiment?

ARandom sampling of population
BManipulation of an independent variable with a control group
CUse of a questionnaire only
DReliance on secondary data

View solution

Correct Option: B

Manipulation, control and comparison are the three defining conditions of an experiment.

Q 02 Causal-Comparative Medium

A study comparing the reading scores of children who watch a lot of television with those who watch little, where the researcher cannot assign children to viewing levels, is best described as:

ATrue experiment
BCausal-comparative (ex-post-facto)
CAction research
DPhenomenology

View solution

Correct Option: B

The "treatment" has already happened naturally; the researcher only compares groups — this is causal-comparative or ex-post-facto.

Q 03 Historical Research Medium

External criticism in historical research asks whether:

AThe conclusions are generalisable
BThe document is authentic — i.e., is it what it claims to be?
CThe writer is credible and well-positioned to know the facts
DThe evidence is statistically significant

View solution

Correct Option: B

External criticism = authenticity; internal criticism = credibility.

Q 04 Qualitative Approaches Medium

Match the qualitative approach with its goal:

(i)	Ethnography	(a)	Understand the lived experience of a phenomenon
(ii)	Phenomenology	(b)	Build theory from data
(iii)	Grounded theory	(c)	Understand a culture from the inside

A(i)-(c), (ii)-(a), (iii)-(b)
B(i)-(a), (ii)-(c), (iii)-(b)
C(i)-(b), (ii)-(c), (iii)-(a)
D(i)-(c), (ii)-(b), (iii)-(a)

View solution

Correct Option: A

Ethnography → culture; Phenomenology → lived experience; Grounded theory → build theory from data.

Q 05 Sampling Easy

Which sampling method ensures every unit in the population has an equal chance of selection?

AQuota sampling
BConvenience sampling
CSimple random sampling
DSnowball sampling

View solution

Correct Option: C

Simple random sampling gives every unit an equal probability of selection.

Q 06 Trustworthiness Hard

In Lincoln and Guba's framework, the qualitative parallel to external validity is:

ACredibility
BTransferability
CDependability
DConfirmability

View solution

Correct Option: B

Transferability = external validity; Credibility = internal validity; Dependability = reliability; Confirmability = objectivity.

Q 07 Factorial Design Medium

A factorial design is most useful when the researcher wants to study:

AThe effect of a single independent variable
BThe interaction between two or more independent variables
CA historical event
DA culture in depth

View solution

Correct Option: B

Factorial designs reveal interaction effects between two or more independent variables.

Q 08 Mixed Methods Hard

A researcher first conducts in-depth interviews to develop a measurement scale, then uses that scale in a large survey. This is an example of:

AConvergent / parallel mixed-methods design
BExplanatory sequential design
CExploratory sequential design
DPure quantitative design

View solution

Correct Option: C

Qualitative first (to develop), quantitative next (to test) = exploratory sequential.

Quick recall

Five methods: Experimental, Descriptive, Historical, Qualitative, Quantitative.
Experiment defining conditions: Manipulation, Control, Comparison.
Causal-comparative / ex-post-facto — looks like an experiment but treatment is naturally occurring.
Historical research: Primary source (original) vs Secondary source (later interpretation); External criticism (authenticity) vs Internal criticism (credibility).
Qualitative approaches: Ethnography, Phenomenology, Grounded theory, Case study, Narrative, Action research.
Trustworthiness: Credibility · Transferability · Dependability · Confirmability.
Sampling: Probability (random, systematic, stratified, cluster, multi-stage) vs Non-probability (convenience, purposive, quota, snowball).
Mixed methods: Convergent · Explanatory sequential · Exploratory sequential.