EconomicsISC Class 11 · Collection, Organisation and Presentation of Data

Collecting Data: Primary and Secondary Data, Census and Sampling (ISC Class 11)

Updated 17 October 2026 · Fact-checked

Data collected afresh by the investigator for a specific purpose are primary data; data already collected by someone else and used again are secondary data. Primary data are collected by direct personal investigation, interviews, questionnaires and observation. A census studies every unit of the population; a sample studies a part, chosen by random, stratified or systematic sampling, to draw conclusions about the whole.

A census studies every unit; a sample studies only a few chosen units, which is cheaper and faster but carries sampling error.

Illustration drawn for QuestionBuddy.

Understand Collecting Data: Primary and Secondary Data, Methods, Census and Sample

Primary data. First-hand data collected by the investigator for the purpose in hand. Merits: reliable, specific, up to date. Demerits: costly, time-consuming.

Secondary data. Data collected earlier by others and now used. Sources: published (government reports, census, Economic Survey, RBI and NSO publications, journals, newspapers, international bodies) and unpublished (office records). Merits: cheap, quick. Demerits: may not fit the purpose, may be out of date or biased. Check suitability, reliability and adequacy before use.

Methods of collecting primary data.

*Direct personal investigation:* the investigator meets respondents; accurate but costly; suits a small area.
*Indirect oral investigation:* information from persons who know the matter (witnesses).
*Questionnaires (mailed or online):* a set of questions sent to respondents; cheap and wide; low response.
*Schedules filled by enumerators:* trained enumerators fill in the form.
*Observation:* watching behaviour.
*Telephone and online surveys.*

A good questionnaire has few, clear, simple, non-leading, ordered questions, and is tested on a small group (a pilot).

Census and sample. *Census (complete enumeration):* every unit of the population is studied; accurate but costly and slow; used for population counts. *Sample survey:* only a part; cheaper, faster, can be more careful, but sampling error exists.

Sampling methods.

*Simple random:* every unit has an equal chance (lottery, random numbers).
*Stratified:* population divided into homogeneous groups (strata) and a random sample taken from each, usually in proportion to the size of the stratum.
*Systematic:* every k-th unit, where k = N / n.
*Cluster, convenience, judgement* are other methods.

Errors. *Sampling error:* difference between the sample result and the true value; falls as the sample grows. *Non-sampling errors:* wrong recording, non-response, biased questions; present even in a census.

Sampling fraction = n / N. Sample size by strata (proportional) = Sample size x (Stratum size / Population).

Collecting Data: Primary and Secondary Data, Methods, Census and Sample: key facts

Systematic sampling: k = N / n | Every k-th unit after a random start; N population, n sample.
Proportional stratified sample: Sample from stratum = n x (Stratum size / N) | Rounded; Strata total equals n.
Sampling fraction: Fraction = n / N | Percent = n / N x 100; Share sampled.

How to answer Collecting Data: Primary and Secondary Data, Methods, Census and Sample questions

Classify the data first, then match the method to the situation.

1Is the data collected afresh or taken from existing records?
2Choose the method of collection suited to the size, cost and accuracy needed.
3For sampling, use the formulas with N and n.
4Mention one source of error.

Collecting Data: Primary and Secondary Data, Methods, Census and Sample in real life

The idea, met in an everyday situation. Answer the question in your own words, then compare with the model answer. The one-line hook is what to remember.

Stratified sample

Suppose a school has 1,200 students: 480 in Class 9, 420 in Class 10 and 300 in Class 11, and a survey needs 120 students.

How many students from each class are chosen for a proportional stratified sample?

Try it first, then show the solutionSolution

Class 9: 120 x 480 / 1,200 = 48.
Class 10: 120 x 420 / 1,200 = 42.
Class 11: 120 x 300 / 1,200 = 30.

Answer: 48, 42 and 30 students.

Memory hook: Each class sends its fair share.

Tricks to remember Collecting Data: Primary and Secondary Data, Methods, Census and Sample

Short memory aids. Say each one aloud once, then use it on a question the same day.

Primary = fresh, Secondary = second-hand
Primary is first-hand; secondary is borrowed.
Use it when: Data types.
k = N over n
Skip interval = population divided by sample.
Use it when: Systematic sampling.
Strata = layers
Take a share from each layer.
Use it when: Stratified.

Collecting Data: Primary and Secondary Data, Methods, Census and Sample: solved examples

Ordered from board level to HOTS. Try each one before you open the solution.

Example 1: Data type · 1 mark · MCQBoard level

Data taken from the Economic Survey for a research project are:
(a) primary data (b) secondary data (c) sample data only (d) census data only

Try it first, then show the solutionSolution

Already published by another body: secondary.

Answer: (b) secondary data

Example 2: Census vs sample · 4 marks · Short answerBoard level

Distinguish between a census and a sample survey on four points. (4 marks)

Try it first, then show the solutionSolution

Coverage: all units vs a part.
Cost and time: high vs low.
Accuracy: complete vs with sampling error.
Use: population counts vs large or destructive studies.

Answer: A census covers every unit, is costly and slow but free from sampling error; a sample covers part, is cheaper and faster but has sampling error.

Example 3: Sampling numericals · 6 marks · Long answerHarder board level

A population of 5,000 households has 2,000 urban and 3,000 rural households. A sample of 250 is needed. Find the sampling fraction, the proportional stratified sample for each stratum, and the interval k for a systematic sample, and say how many would be urban in a systematic sample if the households are listed randomly. (6 marks)

Try it first, then show the solutionSolution

Sampling fraction = 250 / 5,000 = 1/20 = 5%.
Urban: 250 x 2,000 / 5,000 = 100; rural: 250 x 3,000 / 5,000 = 150.
k = 5,000 / 250 = 20.
A random listing gives about 40% urban: roughly 100 urban households.

Answer: 5%; 100 urban and 150 rural; k = 20; about 100 urban households in a systematic sample from a randomly ordered list.

Example 4: Questionnaire design · 5 marks · Long answerHarder board level

A researcher plans a survey on household spending using a mailed questionnaire to 800 households and expects a 35% response. Find the number of replies, and the number of mailings needed to get at least 560 replies. List three features of a good questionnaire. (5 marks)

Try it first, then show the solutionSolution

Replies = 35% of 800 = 280.
Mailings needed for 560 replies = 560 / 0.35 = 1,600.
Features: few and clear questions, no leading questions, logical order.

Answer: 280 replies; 1,600 mailings for 560 replies; a good questionnaire is short, clear, unbiased and ordered.

Example 5: Choosing sources · 4 marks · Short answerHOTS / competency

A student wants to study how house prices in her city changed over 20 years. Decide whether to use primary or secondary data, justify, and name two checks on the data before use. (4 marks)

Try it first, then show the solutionSolution

Secondary data from registration records or published reports are practical because 20 years of past data cannot be collected afresh.
Check reliability (who collected it, how).
Check suitability and definitions (same areas, same property types).

Answer: Use secondary data, since past prices cannot be recollected, after checking reliability and consistency of definitions.

Where students lose marks in Collecting Data: Primary and Secondary Data, Methods, Census and Sample

The slips examiners see most often, and what to write instead.

Mistake: Calling published data primary
Used by the student for the first time.
Wrong working
Economic Survey data are primary
Right working
They are secondary
Do this: Primary means collected first-hand by the investigator.
Usually costs: 1 mark.
Mistake: Strata shares not adding to n
Rounding.
Wrong working
47 + 42 + 30 = 119
Right working
48 + 42 + 30 = 120
Do this: Check the total.
Usually costs: 1-2 marks.
Mistake: Believing a census has no errors
Complete coverage.
Wrong working
A census is free of all errors
Right working
Free of sampling error but not of non-sampling errors
Do this: State both.
Usually costs: 1 mark.

Exam tips

Always define the term and give one source or example.
Show k = N / n steps for systematic sampling.
Mention non-sampling errors in a census.

Practice questions on Collecting Data: Primary and Secondary Data, Methods, Census and Sample

Write each answer on paper first. Open the hint only if you are stuck, and the answer only when you have finished.

Q1 · 1 mark · MCQBoard level
In systematic sampling from 600 units for a sample of 30, the interval is:
(a) 10 (b) 20 (c) 30 (d) 60
Hint
N / n.
Answer
(b) 20
Q2 · 4 marks · Short answerBoard level
Explain the merits and demerits of primary and secondary data. (4 marks)
Hint
Reliability vs cost.
Answer
Primary data are reliable, specific and current but costly and slow; secondary data are cheap and quick but may be unsuitable, outdated or biased.
Q3 · 4 marks · Short answerHarder board level
A college has 900 students: 450 in arts, 300 in commerce and 150 in science. A sample of 60 is to be taken. Find the proportional stratified sample, the sampling fraction and the interval for a systematic sample. (4 marks)
Hint
60 / 900.
Answer
Arts 30, commerce 20, science 10; fraction 60 / 900 = 6.67%; k = 900 / 60 = 15.
Q4 · 5 marks · Long answerHarder board level
A market-research firm wants 400 usable responses. Past surveys give a response rate of 25% by post and 60% by phone. Find the contacts needed by each method and the number of contacts saved by phone. If a postal contact costs Rs 20 and a phone contact Rs 45, compare the total cost. (5 marks)
Hint
400 / 0.25 and 400 / 0.60.
Answer
Post: 400 / 0.25 = 1,600 contacts, cost Rs 32,000. Phone: 400 / 0.60 = 667 contacts (rounded up), cost 667 x 45 = Rs 30,015. Phone saves 933 contacts and Rs 1,985.
Q5 · 3 marks · Short answerHOTS / competency
The government plans a survey to find the average life of electric bulbs made by a company. Explain why a sample is better than a census here and what sampling method could be used. (3 marks)
Hint
Destructive testing.
Answer
Testing a bulb's life destroys it, so a census would destroy the whole stock. A random sample from each batch (stratified by batch) gives a good estimate at low cost.

Test your recall

Answer in your head or on paper first, then open the card. Pulling an answer out of memory makes it stick far better than re-reading.

1.Primary data means?Show answer
Data collected first-hand by the investigator.
2.Sampling fraction?Show answer
n / N.
3.Interval in systematic sampling?Show answer
N / n.
4.Census covers?Show answer
Every unit.
5.A secondary source?Show answer
Economic Survey or Census reports.

Collecting Data: Primary and Secondary Data, Methods, Census and Sample: frequently asked questions

Is a questionnaire primary data?

Yes, when you collect responses yourself.

Which sampling is unbiased?

Random sampling gives every unit an equal chance.

What is a pilot survey?

A trial run of the questionnaire.

Do larger samples reduce error?

Sampling error falls.