Revision summary
Reliability is consistency, not truth. Test-retest, parallel forms, split-half, Cronbach’s alpha, and inter-rater agreement are the main tests. Durkheim’s suicide files already needed stable coding. Census definition changes can mimic social change. A reliable scale can still miss Weber’s intended meaning.
Model answer
Copper italics in this answer — like this — are the key facts. Each one is unpacked in the Facts & figures rail.
Introduction
Reliability is the consistency of a measure. If the same instrument, applied to the same kind of object under the same conditions, jumps about, the scores are unreliable. Validity is a later question: a clock can tick reliably and still tell the wrong time.
Body
What reliability is
- In survey and test work, reliability is the share of variance that is not random error.
- Emile Durkheim already needed stable classification of deaths before suicide rates could be compared. Unstable coding is an reliability failure in official data.
- Qualitative work uses a cousin idea: inter-coder agreement and stable use of a codebook, not a claim that meanings never shift.
Test-retest
- The same test is given twice to the same persons after an interval.
- Use when the trait should be stable, such as a literacy item, not a mood item after a riot.
- Memory and real change both threaten the correlation.
Parallel or alternate forms
- Two versions of the scale are built to the same blueprint and both are given.
- Use in examinations and long batteries to reduce item memory.
- Costly to write well.
Split-half and internal consistency
- The test is split, scores are correlated, and Spearman-Brown may correct the half-length.
- Cronbach’s alpha estimates consistency among items of a Likert battery.
- Use for attitude scales on, say, gender-role items in a student sample. Alpha is not a test of unidimensional truth, only of hanging together.
Inter-rater or inter-coder reliability
- Two observers code the same focus-group transcript or Census occupation string.
- Cohen’s kappa or percent agreement is reported.
- Essential for content analysis, courtroom files, and open-ended NFHS-type comments.
Conditions and cautions
- High reliability does not prove that “caste” or “work” was well conceptualised. That is validity, and Weber’s meaning problem.
- Official series such as the Census need stable definitions across decades; a definition change looks like social change if reliability of the series is ignored.
- Hidden populations reached by snowballing cannot borrow NFHS reliability claims without a design argument.
Flow diagram
flowchart TD REL[Reliability consistency] --> TT[Test-retest] REL --> PF[Parallel forms] REL --> SH[Split-half alpha] REL --> IR[Inter-rater] REL --> VAL[Validity is separate]
Conclusion
Reliability is consistency of a measure across time, forms, items, or raters. Test-retest, parallel forms, split-half, alpha, and inter-rater tests are the standard kit. Use them, then still ask whether the consistent score measures the sociological object.
Quick related
Students also ask
-
How has the idea of Work From Home forced us to redefine the formal and informal organisation of work ?
Next question on this syllabus topic (2021 · Q4(b)). View answer →
-
If alpha is high, is the study good?
It is internally consistent. It may still measure the wrong construct or a narrow one.
-
Does reliability apply only to numbers?
No. Coding of talk and occupations needs agreed rules, or the “finding” is one coder’s day.
Same topic · past papers
UPSC has asked this before
These previous-year questions sit on the same topic. Open one to practise the earlier ask.
-
2025 · Q2(c) · Sociology GS 1 · 10 marks
What do you mean by reliability? Discuss the importance of reliability in social science research. -
2026 · Q1(d) · Sociology GS 1 · 10 marks
Although sociology, as a subject of study, overarches all disciplines in social sciences yet it maintains its distinctiveness. Comment. -
2023 · Q6(c) · Sociology GS 1 · 10 marks
What are the ethical issues that a researcher faces in making use of participant observation as a method of collecting data? Explain -
2022 · Q1(a) · Sociology GS 1 · 10 marks
Delimit the scope of sociology in relation to other social sciences. -
2022 · Q1(b) · Sociology GS 1 · 10 marks
How does a researcher achieve objectivity in interpretative research?
More from this topic
Q1(c) · UPSC Mains 2025 · Sociology GS 1 · 10 marks · Solution
What is a variable in social research? What are their different types? Elaborate.
Works and Economic Life
A variable is a property with more than one value across cases. Independent, dependent, intervening and control names mark their role in a design. Stevens’s nominal, ordinal, interval and ratio levels mark how they are measured. Qualitative categories differ from quantitative counts and scales. A worked example is women’s education, fertility, age at marriage and wealth as control. Indian large surveys such as NFHS use this same grammar.
Q5(c) · UPSC Mains 2024 · Sociology GS 1 · 10 marks · Solution
Describe the differing principles of work organization in feudal and capitalist societies.
Works and Economic Life
Feudal work is tied to land, lord and custom rather than to a free labour market. Surplus is taken through extra-economic power and personal dependence. Capitalist work sells labour power as a commodity under wage and clock. The factory and office separate production from the household. Marx’s modes of production name this contrast of principles. Unfree labour can survive inside capitalism without erasing the type.
Q4(b) · UPSC Mains 2021 · Sociology GS 1 · 20 marks · Solution
How has the idea of Work From Home forced us to redefine the formal and informal organisation of work ?
Works and Economic Life
Formal organisation is the official chart; informal organisation is unwritten practice. WFH moves Weber’s office into dashboards and Mayo’s group into the household. Unpaid care and gig ratings intensify informal extraction. Lockdown split salaried WFH from reverse-migrant informal labour. The split is relocated, not abolished.