9 Steps of Research

The research process follows a logical sequence of nine steps. The sequence is iterative — findings at later steps may force the researcher to revisit earlier ones — but the order is the standard reference frame for examination questions.

Nine Steps of the Research Process

Identification of the research problem.
Review of related literature.
Formulation of objectives and hypotheses.
Research design.
Sampling — defining population and selecting sample.
Selection / construction of tools and techniques.
Data collection.
Data analysis and interpretation.
Reporting — thesis, article, presentation.

flowchart TB
  P[1. Identify problem] --> L[2. Review literature]
  L --> H[3. Formulate objectives<br/>and hypotheses]
  H --> D[4. Research design]
  D --> S[5. Sampling]
  S --> T[6. Tools and techniques]
  T --> C[7. Data collection]
  C --> A[8. Data analysis<br/>and interpretation]
  A --> R[9. Reporting]
  R -. Feedback / future research .-> P
    classDef default fill:#003366,color:#ffffff,stroke:#ffcc00,stroke-width:3px,rx:10px,ry:10px;

9.1 Step 1 — Identification of the Research Problem

A research problem is a clearly stated question that research can answer. A good problem statement is specific, feasible, novel, ethical, and significant.

Sources of Research Problems

Source	What it gives the researcher
Personal experience	Felt difficulty in the field
Literature gap	Question raised but not answered in published work
Theory	Predictions of theory not yet tested
Practice / policy need	Real problems faced by practitioners
Replication	Verifying earlier studies in new contexts
Conferences and discussions	Open questions of the discipline

A problem becomes a researchable problem when it is narrowed enough to be answerable with available time, resources and methods. Example. “Education in India” is a topic, not a problem. “Effect of NEP-2020 multidisciplinary curriculum on first-year engagement in two undergraduate colleges in Tamil Nadu” is a researchable problem.

9.2 Step 2 — Review of Related Literature

The literature review accomplishes four things.

Four Functions of the Literature Review

Function	What it does
Establishes what is known	Summarises prior findings on the question
Identifies the gap	Shows what remains unanswered
Justifies method	Explains why this design is suitable
Provides a conceptual framework	Names variables and their hypothesised relationships

Types of Literature Review

Type	What it does
Narrative review	Critical summary of literature in a topic
Systematic review	Comprehensive search using pre-stated inclusion criteria
Meta-analysis	Statistical pooling of effects across multiple studies
Scoping review	Maps the breadth of literature on a broad question

9.3 Step 3 — Formulation of Objectives and Hypotheses

Objectives state what the study will do; hypotheses state what the study predicts.

General objective — overall purpose, one sentence.
Specific objectives — observable, measurable sub-goals (often 3–5).
Hypotheses — testable predictions about variable relationships.

Example.

General objective. To assess the effect of the flipped-classroom method on undergraduate engagement.
Specific objective 1. To compare engagement scores in flipped and lecture-based sections.
Specific objective 2. To compare end-semester achievement in the two sections.
Hypothesis (H₁). Flipped-classroom students show higher engagement than lecture-based students.

9.4 Step 4 — Research Design

The research design is the overall blueprint of the study — how data will be collected, from whom, and how analysed.

Three Broad Design Families

Family	When	Examples
Exploratory	When little is known	Pilot, focus group, in-depth interview
Descriptive	When a phenomenon is to be portrayed	Survey, observation, case study
Causal / experimental	When testing cause and effect	True experiment, quasi-experiment

A good design specifies: variables, sample, sampling method, tools, procedure, statistical analysis, and ethical safeguards.

9.5 Step 5 — Sampling

The researcher must define:

Sampling — Working Vocabulary

Term	Meaning
Population (universe)	All units to which findings will apply
Target population	The subset accessible to the researcher
Sampling frame	The list from which the sample is drawn
Sample	The units actually observed
Sampling unit	The entity sampled (individual, household, school)
Sampling error	Difference between sample statistic and population parameter, due to chance
Non-sampling error	Errors due to non-response, measurement, or processing

Probability and non-probability methods were summarised in the previous topic; the choice depends on whether generalisability (probability) or information richness (purposive) is the priority.

9.6 Step 6 — Selection or Construction of Tools and Techniques

The data-collection tool must be valid, reliable, practicable, and appropriate to the question.

Common Research Tools

Tool	What it captures	Best for
Questionnaire	Self-report — opinion, knowledge, behaviour	Surveys with literate populations
Interview schedule	Spoken responses; allows probing	Sensitive topics, low-literacy contexts
Observation schedule	Behaviour as it occurs	Classroom interactions, child development
Test / scale	Achievement, intelligence, attitude, personality	Measurable constructs
Rating scale	Judgments along a continuum (Likert, semantic differential)	Attitude, satisfaction, perception
Document analysis	Existing records	Historical and policy work
Checklist	Presence / absence of items	Audits, compliance
Sociometric tools	Group structure	Class peer relationships

Likert Scale — A Frequent NTA Stem

A Likert scale typically has 5 (or 7) ordered response options ranging from “Strongly disagree” to “Strongly agree”. It is named after Rensis Likert (1932). It is an ordinal measurement scale, though in practice often treated as interval.

9.6.1 Reliability and Validity of Tools

Forms of Reliability and Validity

Property	Form	What it asks
Reliability	Test-retest	Are scores stable across administrations?
	Parallel-forms	Do equivalent forms give similar scores?
	Internal consistency (Cronbach’s α)	Do items in the test cohere?
	Inter-rater	Do different graders agree?
Validity	Face	Does it look right on the surface?
	Content	Does it cover the construct fully?
	Construct	Does it measure the underlying construct?
	Criterion (concurrent / predictive)	Does it correlate with an external criterion?

9.7 Step 7 — Data Collection

Data are collected from primary or secondary sources.

Primary data — collected first-hand by the researcher (survey responses, observation logs, experimental measurements).
Secondary data — already collected by someone else (census, NSSO, AISHE, published studies, archival records).

Levels of Measurement (Stevens 1946)

Level	What it permits	Example
Nominal	Naming and counting categories	Gender, marital status
Ordinal	Rank order	Educational attainment (school / graduate / postgraduate)
Interval	Equal intervals; arbitrary zero	Temperature in Celsius
Ratio	Equal intervals; absolute zero	Income, height, age

The level of measurement determines which statistical tools are valid for analysis.

9.8 Step 8 — Data Analysis and Interpretation

Analysis depends on whether the data are quantitative or qualitative.

Quantitative Data Analysis

Stage	What it does	Examples
Descriptive statistics	Summarise the sample	Mean, median, mode, standard deviation, frequency, percentage
Inferential statistics	Generalise to the population	t-test, ANOVA, χ² test, correlation, regression
Effect size	Quantify magnitude of effect	Cohen’s d, η², r

Qualitative Data Analysis

Stage	What it does
Transcription	Render audio/video as text
Coding	Assign labels to chunks of data
Categorisation	Group codes into categories
Theme generation	Identify patterns across categories
Theoretical interpretation	Link themes to existing theory or build new theory

Interpretation is the conversion of statistical or thematic findings into answers to the research question, with attention to limitations.

9.9 Step 9 — Reporting

Reporting may take the form of a thesis, dissertation, journal article, conference paper, or policy report. The next topic deals with thesis and article writing in detail.

Standard Sections of a Research Report

Section	What it contains
Title page	Title, author, affiliation, date
Abstract	150–300 word summary
Introduction	Problem, significance, objectives, hypotheses
Review of literature	Prior work, gap, framework
Methodology	Design, sample, tools, procedure
Results / Findings	Data presentation, statistical results
Discussion	Interpretation, comparison with prior work, limitations
Conclusion	Key findings, implications, future work
References	Citation list
Appendices	Tools, raw data, ethical clearance

9.10 Practice Questions

Q 01 Research Process Easy

The first step in the research process is:

ASampling
BIdentification of the research problem
CData analysis
DReporting

View solution

Correct Option: B

Identification of the research problem is the first step; everything else flows from it.

Q 02 Step Functions Medium

Match the step with its activity:

(i)	Review of literature	(a)	Decide who to study
(ii)	Sampling	(b)	Identify gaps in prior work
(iii)	Data analysis	(c)	Apply statistical or thematic procedures
(iv)	Reporting	(d)	Write thesis or article

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

View solution

Correct Option: A

Literature → identify gaps; Sampling → who to study; Analysis → procedures; Reporting → write up.

Q 03 Literature Review Medium

Which is not a function of the literature review?

AEstablishing what is already known
BIdentifying the research gap
CReplacing the need for original data collection
DJustifying the chosen method

View solution

Correct Option: C

Literature review complements original data collection; it does not replace it.

Q 04 Sampling Frame Easy

The list from which a sample is actually drawn is called the:

APopulation
BSampling frame
CSampling unit
DSampling error

View solution

Correct Option: B

The sampling frame is the operational list of the population from which the sample is drawn.

Q 05 Cronbach's Alpha Easy

Cronbach's alpha is a measure of:

AConstruct validity
BInternal consistency reliability
CContent validity
DSampling adequacy

View solution

Correct Option: B

Cronbach's α measures internal consistency reliability — whether items in a test cohere.

Q 06 Levels of Measurement Easy

The level of measurement that has equal intervals and an absolute zero point is:

ANominal
BOrdinal
CInterval
DRatio

View solution

Correct Option: D

Ratio scales (income, height, age) have equal intervals AND an absolute zero.

Q 07 Likert Scale Medium

A 5-point Likert scale is an example of which level of measurement (most strictly)?

ANominal
BOrdinal
CInterval
DRatio

View solution

Correct Option: B

Strictly ordinal, although often treated as interval in practice.

Q 08 Order of Steps Easy

Which of the following is the correct order of the last three steps of the research process?

AReporting → Data analysis → Data collection
BData collection → Data analysis → Reporting
CData analysis → Reporting → Data collection
DReporting → Data collection → Data analysis

View solution

Correct Option: B

The correct order is Data collection → Data analysis → Reporting.

Quick recall

Nine steps: Problem → Literature → Objectives/Hypotheses → Design → Sampling → Tools → Data collection → Analysis → Reporting.
Sampling vocabulary: Population, Target population, Sampling frame, Sample, Sampling unit, Sampling error, Non-sampling error.
Tool properties: Validity (face, content, construct, criterion); Reliability (test-retest, parallel-forms, internal consistency = Cronbach’s α, inter-rater).
Levels of measurement: Nominal, Ordinal, Interval, Ratio (mnemonic “NOIR”).
Likert scale = ordinal (5 or 7 points), named after Rensis Likert (1932).
Standard report sections: Title, Abstract, Introduction, Literature, Method, Results, Discussion, Conclusion, References, Appendices.