Chapter 2 Collection of Data
Practice questions
According to the textbook, which of the following does the Census of India record?
- Daily retail prices of oranges in every market
- Size, density, sex ratio, literacy and migration of the population
- Only the incomes of farmers
- The results of exit polls
Answer and explanation
Answer: B
The textbook lists size, density, sex ratio, literacy, migration and rural-urban distribution among the aspects of population that Census officials collect.
Assertion (A): A census is free from all errors. Reason (R): A census has no sampling error because every unit is covered.
- Both A and R are true, and R is the correct explanation of A
- Both A and R are true, but R is not the correct explanation of A
- A is true, but R is false
- A is false, but R is true
Answer and explanation
Answer: D
A is false: even a census can contain non-sampling errors such as non-response and recording mistakes. R is true: with every unit covered, there is no gap between estimate and parameter due to sampling.
Illustrative: a population of four farms has yields of 20, 24, 28 and 32 quintals. A sample of farms 1 and 2 is taken. The sampling error of the mean is:
- 2 quintals
- 26 quintals
- 4 quintals
- 22 quintals
Answer and explanation
Answer: C
Population mean = (20 + 24 + 28 + 32) ÷ 4 = 104 ÷ 4 = 26. Sample mean = (20 + 24) ÷ 2 = 22. Sampling error = 26 − 22 = 4 quintals.
Case: a made-up company surveys customer opinion by stopping only those people who walk into its biggest showroom. Which problem is most likely?
- Sampling bias, because customers who never visit that showroom cannot be included
- No problem, because it is a census
- Only a sampling error that a larger sample will remove
- A pilot survey error
Answer and explanation
Answer: A
Customers who shop elsewhere had no chance of being included, which is sampling bias, a non-sampling error that a larger sample at the same showroom would not cure.
Illustrative: a frame is numbered 01-40 and a sample of 3 is needed. A row of random numbers (made up) reads 52, 17, 08, 17, 63, 29, 35. The sample is:
- 52, 17, 08
- 17, 08, 17
- 17, 29, 35
- 17, 08, 29
Answer and explanation
Answer: D
Skip 52 and 63 (above 40) and the repeated 17. The first three distinct numbers within 01-40 are 17, 08 and 29.
Illustrative: how many different samples of 3 can be drawn from 7 units?
- 21
- 35
- 210
- 343
Answer and explanation
Answer: B
Number of samples = 7 × 6 × 5 ÷ (3 × 2 × 1) = 210 ÷ 6 = 35. 210 counts each group in all 6 orders, so it must be divided by 6.
Which is NOT a purpose of a pilot survey?
- Checking the clarity of instructions
- Estimating the cost and time of the main survey
- Replacing the main survey
- Assessing the performance of enumerators
Answer and explanation
Answer: C
A pilot survey is a small try-out before the main survey; it improves and plans the main survey but never replaces it.
A researcher must collect information quickly from respondents who are reluctant to answer some questions face to face, and most own telephones. The best mode is:
- telephone interview
- personal interview
- census
- pilot survey
Answer and explanation
Answer: A
Telephone interviews are quicker and cheaper than personal interviews and suit respondents who are reluctant in face-to-face meetings; the respondents can be reached because they have phones.
Illustrative: 800 questionnaires are mailed; 260 are returned, of which 20 are incomplete. The usable response rate is:
- 32.5%
- 30%
- 35%
- 2.5%
Answer and explanation
Answer: B
Usable returns = 260 − 20 = 240. Rate = (240 ÷ 800) × 100 = 30%. 32.5% wrongly counts the 20 incomplete forms as usable.
The option 'Any other (please specify)' is added to multiple-choice questions mainly to:
- make the questionnaire longer
- turn it into a two-way question
- lead the respondent
- capture answers that the researcher did not anticipate
Answer and explanation
Answer: D
A respondent's true answer may not be among the given options; 'Any other' lets him or her write it in.
Assertion (A): Open-ended questions are hard to score. Reason (R): Responses to open-ended questions vary a great deal from person to person.
- Both A and R are true, and R is the correct explanation of A
- Both A and R are true, but R is not the correct explanation of A
- A is true, but R is false
- A is false, but R is true
Answer and explanation
Answer: A
Both are true, and R explains A: because each respondent answers in his or her own words, the variety of answers makes them hard to interpret and score.
Which of these is the best-framed question?
- Don't you think fuel prices are too high?
- How many hours a day do you study at home? Less than 1 / 1-2 / 2-3 / more than 3
- How much do you love this excellent textbook?
- Do you study a lot?
Answer and explanation
Answer: B
It is precise and gives clear ranges to tick. The others use a 'Don't you' opening, are leading ('excellent') or vague ('a lot').
Which statement about primary data is correct?
- They are always cheaper than secondary data
- They are taken from published reports
- They are based on first-hand information collected by the investigator
- They cannot be used in economics
Answer and explanation
Answer: C
Primary data come from the investigator's own enquiry and so are first-hand. They usually cost more time and money than secondary data.
A student uses crop data from a government report for her project. For her, these data are:
- secondary, because another agency collected and processed them
- primary, because she is using them
- primary, because a government collected them
- neither primary nor secondary
Answer and explanation
Answer: A
Data are secondary for anyone who uses data already collected and processed by another agency; they are primary only for the agency that collected them first.
Illustrative: a variable Z takes the values 48, 52, 47, 55 and 58 in five successive years. In how many of the year-to-year changes did Z fall?
- None
- Two
- Three
- One
Answer and explanation
Answer: D
Changes: 48 to 52 rise, 52 to 47 fall, 47 to 55 rise, 55 to 58 rise. Only one of the four changes is a fall.
Assertion (A): Non-sampling errors are more serious than sampling errors. Reason (R): Non-sampling errors can occur even in a census and cannot be removed simply by increasing the sample size.
- Both A and R are true, and R is the correct explanation of A
- Both A and R are true, but R is not the correct explanation of A
- A is true, but R is false
- A is false, but R is true
Answer and explanation
Answer: A
Both are true, and R explains A: since a larger sample or even a full census does not remove them, non-sampling errors are harder to control and so more serious.
According to the textbook, the journal through which NSS data are released is:
- the Census Gazette
- the Labour Bulletin
- the Statistical Abstract
- Sarvekshana
Answer and explanation
Answer: D
The chapter states that NSS data are released through reports and its quarterly journal Sarvekshana.
An enumerator records a household's size as 13 when the true figure is 31. This is an example of:
- sampling error
- sampling bias
- error in data acquisition
- non-response
Answer and explanation
Answer: C
Recording or copying the wrong figure is a mistake in data acquisition, a kind of non-sampling error.
Which error can be reduced by taking a larger sample?
- Non-response error
- Sampling error
- Recording error
- Sampling bias
Answer and explanation
Answer: B
Sampling error arises from studying only part of the population and falls as the sample becomes larger. The other three are non-sampling errors, which a larger sample does not cure.
Illustrative: the true average income of a population is ₹750, and a sample gives an average of ₹712. The sampling error is:
- ₹38
- ₹712
- ₹1,462
- ₹62
Answer and explanation
Answer: A
Sampling error = parameter − estimate = 750 − 712 = ₹38.