CBSE Class 11 Economics Notes — Collection of Data

Chapter Summary

This chapter explains why data are collected and where they come from. It defines variables, observations, primary and secondary data, shows how to design a good questionnaire, compares personal, mail and telephone interviews, distinguishes census from sample surveys, introduces random and non-random sampling along with sampling and non-sampling errors, and lists important Indian sources of secondary data along with cautions for using them.

Key Concepts

• Variable and observation
• Data
• Primary data
• Secondary data
• Survey and questionnaire
• Closed-ended and open-ended questions
• Personal, mail and telephone interviews
• Population and sample
• Census (complete enumeration)
• Random and non-random sampling
• Sampling and non-sampling errors
• Sources of secondary data

Why We Collect Data: Variables and Observations

The purpose of collecting data is to supply evidence for reaching a sound and clear solution to a problem. The chapter starts with a statement about Indian food grain output: it rose to 132 million tonnes in 1978-79 from 108 million tonnes in 1970-71, then fell back to 108 million tonnes in 1979-80, and later climbed to 252 million tonnes in 2015-16 and 272 million tonnes in 2016-17.

Notice that production is not the same in every year. Whatever takes different values is called a variable, and variables are usually written as X, Y or Z. Each single value of a variable is an observation. In Table 2.1 of the chapter, the years are the variable X and food grain production in million tonnes is the variable Y. The pairs are 1970-71 with 108, 1978-79 with 132, 1990-91 with 176, 1997-98 with 194, 2001-02 with 212, 2015-16 with 252 and 2016-17 with 272.

These values together are the data. A little arithmetic shows what the data reveal: from 1970-71 to 2016-17 output grew by 272 minus 108, which is 164 million tonnes, or about 152 per cent. Between 1970-71 and 1978-79 the rise was 24 million tonnes, about 22 per cent. Without data we could not describe these fluctuations. Data are a tool that helps us understand a problem by giving information about it.

Key Points

• Data are collected to give evidence for solving a problem.
• A variable is a quantity that takes different values and is written X, Y or Z.
• Each value of a variable is an observation.
• In Table 2.1 the year is X and food grain output is Y.
• Data help us understand a problem by providing information.

Examples

• Food grain output rose from 108 to 272 million tonnes between 1970-71 and 2016-17, a rise of 164 million tonnes.
• The marks of students in a test form a variable with one observation for each student.

Primary Data and Secondary Data

Statistical data come from two kinds of sources. When a researcher collects the data herself by conducting an enquiry, the data are called primary data, because they rely on first-hand information. Suppose you want to know how popular a film star is among school students. You would have to ask a large number of students and record their replies. What you collect is primary data.

When the data have already been collected and processed, that is scrutinised and tabulated, by some other agency, they are called secondary data. These can be taken from published sources such as government reports, documents, newspapers and books by economists, or from a website. The chapter makes a subtle point: data are primary only to the source that collects and processes them for the first time. For every later user they are secondary. If you publish a report on the film star's popularity and another person uses your figures for a similar study, your primary data become his secondary data.

Using secondary data saves time and cost, which is why it is common in economic research. The price is that the user did not control how the data were gathered, so the definitions and quality must be checked. When answering a difference question, compare the two on source, cost, time, and control over quality.

Key Points

• Primary data are collected first-hand by the investigator through an enquiry.
• Secondary data are collected and processed by another agency and then used by others.
• Government reports, newspapers, books and websites are common secondary sources.
• The same data are primary for the collector and secondary for later users.
• Secondary data save time and cost.

Examples

• Surveying 200 students about a film star yields primary data for the surveyor.
• Using food grain figures printed in a government report is using secondary data.

Surveys and Designing a Good Questionnaire

A manufacturer deciding about a product, or a political party choosing a candidate, usually conducts a survey, which is a method of gathering information from individuals. The aim is to describe characteristics such as price, quality and usefulness of a product, or popularity, honesty and loyalty of a candidate.

The most common instrument is a questionnaire or interview schedule. It may be filled in by the respondent himself, or by a researcher called an enumerator or a trained investigator. The chapter gives clear rules for preparing it. It should not be too long, and the number of questions should be as small as possible. Language should be easy, avoiding ambiguous or difficult words. Questions should move from general to specific, so the respondent feels comfortable. For example, ask first whether electricity supply in the locality is regular, and only then whether the increase in electricity charges is justified.

Questions should also be precise. Instead of asking what percentage of income is spent on clothing, ask how much is spent on clothing. They should be unambiguous: rather than asking whether you spend a lot on books, offer options such as less than Rs 200, Rs 200-300, Rs 300-400 and more than Rs 400. Avoid double negatives such as asking whether one does not think smoking should be prohibited. Avoid leading questions like asking about the flavour of this high-quality tea, and avoid suggesting alternatives, such as job or housewife after college.

Key Points

• A survey is a method of gathering information from individuals.
• A questionnaire or interview schedule is the common instrument.
• Keep it short, simple and arranged from general to specific questions.
• Ask for precise amounts or give clear options instead of vague wording.
• Avoid double negatives, leading questions and questions that suggest the answer.

Examples

• Poor: Do you spend a lot of money on books in a month? Good: How much do you spend on books in a month, with four expense ranges to tick.
• Poor: How do you like the flavour of this high-quality tea? Good: How do you like the flavour of this tea?

Closed-Ended and Open-Ended Questions

A questionnaire may contain closed-ended or structured questions, or open-ended or unstructured questions. In an open-ended question the respondent writes the answer in his own words. The question about what a student wants to do after college is open-ended, and so is asking for one's view about globalisation.

Closed-ended questions offer fixed answers. When only two answers are possible, yes or no, it is a two-way question. When more than two answers are possible, a multiple choice question is more suitable, such as asking why a person sold his land, with options such as to pay off debts, to finance children's education, to invest in another property, or any other, please specify.

Closed questions are easy to use, score and codify because everybody chooses from the same options. Their drawbacks are that good alternatives, representing both sides of an issue, are hard to write, and the respondent's true answer may be missing. The option any other is provided to take care of this. Multiple choice questions also tend to restrict answers, since people might have answered differently without the listed options.

Open-ended questions give more individual responses but are hard to interpret and score because replies vary so widely. In practice a researcher chooses between them depending on whether easy analysis or rich detail matters more.

Key Points

• Closed-ended questions give fixed options; open-ended ones allow free answers.
• A two-way question has only yes or no as answers.
• Multiple choice questions suit situations with more than two possible answers.
• Closed questions are easy to score but may miss the true response; any other helps.
• Open-ended questions are individualised but hard to interpret and score.

Examples

• Why did you sell your land? with four options including any other, please specify.
• What is your view about globalisation? is an open-ended question.

Modes of Collecting Data: Personal, Mail and Telephone Interviews

There are three basic ways of collecting survey data: personal interviews, mailing questionnaire surveys and telephone interviews.

Personal interviews are used when the researcher can reach all the members and conducts face-to-face interviews. Their strengths are personal contact, the chance to explain the study and answer queries, the ability to ask the respondent to expand on important answers, fewer misunderstandings, and watching reactions that give extra information. The demerits are that they are expensive since trained interviewers are needed, they take long to finish, and the presence of the researcher may stop people from saying what they truly think.

In a mail survey the questionnaire is posted to each person with a request to return it by a given date. It is cheaper, reaches people in remote areas, does not allow the interviewer to influence the respondent, and gives time to think. Today online surveys and SMS surveys work in the same way. The weaknesses are that there is little help to explain instructions, so questions may be misunderstood, and response rates are low because forms are returned incomplete, not returned, or lost.

Telephone interviews are quick and cost less than personal visits, as the investigator can ring many people in a short time. Their limits are that they reach only those with phones, the questions must be short, and the respondent can end the call easily.

Key Points

• The three modes are personal interviews, mail surveys and telephone interviews.
• Personal interviews allow explanation and observation but are costly and slow.
• Mail surveys are cheap and reach remote areas but have low response rates.
• Online and SMS surveys work like mail surveys.
• Telephone interviews are quick but limited to people with phones and to short questions.

Examples

• A researcher visiting 50 households to ask about food expenditure is using personal interviews.
• A college sending a feedback form by email to former students uses an online survey, and many forms may never come back.

Census and Sample Surveys: Population and Sample

In statistics the word population means the whole group of items about which information is wanted, not only people. It could be all farms in a state, all households in a city, or all students in a school. A sample is a part of that population selected to represent it.

When information is collected from every unit of the population, the method is called a census or complete enumeration. India's Population Census, held every ten years, is the standard example. A census gives detailed and reliable information about every unit, but it is costly, takes much time, and needs a large organisation. It is also not possible where the population is very large or where testing destroys the item, such as testing every bulb for its life.

A sample survey collects data from only a part of the population and uses the result to understand the whole. It saves money and time, can be done quickly with smaller staff, and allows more careful checking of each answer.

A worked example makes this concrete. Suppose a school has 500 students and you wish to know how much time they spend on homework daily. Asking all 500 is a census. Choosing 50 of them is a sample, so the sample covers 50 out of 500, or 10 per cent, of the population. If the 50 average 2 hours of homework, you would estimate that the school's students also spend about 2 hours, with some margin of error.

Key Points

• Population is the whole group under study; a sample is a selected part of it.
• Census, or complete enumeration, collects data from every unit of the population.
• Census gives detailed information but is costly, slow and often impractical.
• A sample survey is cheaper and quicker but gives only an estimate.
• Sampling is essential when testing destroys the unit studied.

Examples

• India's Population Census is held every ten years and covers all households.
• Selecting 50 of 500 students to estimate average homework time gives a 10 per cent sample.

Sampling Techniques and Errors

How the sample is chosen decides how trustworthy the results are. Sampling methods are broadly of two kinds: random and non-random.

In random or probability sampling every unit of the population has a known chance of being selected, and the choice is made by chance, not by the researcher's preference. The simplest form is simple random sampling, where each unit has an equal chance. It can be done by the lottery method, writing the units on slips, mixing them and drawing. If a class has 60 students and 10 are to be selected, each student has a chance of 10 out of 60, which is 1 in 6. Another way is to use a table of random numbers to pick units.

In non-random sampling the researcher decides which units go in. Deliberate or judgement sampling selects units the investigator considers typical. Convenience sampling selects whatever units are easy to reach, such as the first ten shoppers outside a store. These are quick but can be biased, because the sample may not represent the population.

Results from a sample will differ from the true population value. The difference arising because only a part is studied is the sampling error, and it generally falls as the sample size grows. Errors that arise from other causes, such as non-response, wrong answers, faulty measurement or mistakes in recording, are non-sampling errors. They can occur even in a census.

Key Points

• Sampling methods are random (probability) or non-random.
• In simple random sampling every unit has an equal chance, using lottery or random number tables.
• Deliberate and convenience sampling leave the choice to the investigator and may be biased.
• Sampling error is the difference between the sample estimate and the true population value.
• Non-sampling errors arise from non-response, measurement or recording faults and occur in a census too.

Examples

• Drawing 10 names from 60 slips in a box gives each student a 1 in 6 chance.
• Interviewing only the first 20 people who walk past a market stall is convenience sampling.

Sources of Secondary Data and Cautions in Using Them

The chapter closes by describing important sources of secondary data in India, which students should know by name. The Census of India, conducted by the Registrar General and Census Commissioner every ten years, provides data on population, literacy, occupation, housing and more. The National Sample Survey Organisation (NSSO), now part of the national statistical system, carries out large sample surveys on household consumption, employment and unemployment. Government publications such as the annual Economic Survey, Reserve Bank of India reports and handbooks, and ministry and planning documents give data on production, prices, banking and trade. Other sources include international agencies, research institutes, newspapers, journals and websites.

Secondary data are cheap and quick to obtain, which is why a school project often starts there. However, the user must examine them before using. Check who collected the data and why, whether the definitions of terms match your purpose, the period and area covered, the method used to collect them, and whether the figures are reasonably accurate.

The food grain data in Table 2.1 are an example of secondary data, since they were collected by government agencies and are only being used here. Before using such figures, an investigator would confirm the unit of measurement, which is million tonnes in this table, and the years covered.

Key Points

• Census of India is conducted every ten years and gives population data.
• NSSO conducts sample surveys on household consumption and employment.
• Economic Survey and RBI publications give data on the economy.
• Secondary data are cheap but need checks on source, definition, period and accuracy.
• Food grain output in Table 2.1 is secondary data for the textbook reader.

Examples

• Using NSSO figures on household spending for a project on consumption patterns.
• Reading literacy figures from Census tables instead of surveying a district yourself.

Frequently Asked Questions

What is the difference between primary and secondary data?

Primary data are collected first-hand by the investigator for her own enquiry. Secondary data have already been collected and processed by some other agency, and are obtained from sources such as government reports, newspapers, books or websites. The same data are primary for the collector and secondary for every later user. Secondary data save time and cost.

What is a variable and what is an observation?

A variable is a quantity that takes different values, for example the year or the production of food grains, and is generally denoted by X, Y or Z. Each individual value taken by the variable is an observation. A set of observations on one or more variables forms the data used in a study.

What are the points to remember while designing a questionnaire (4 marks)?

It should be short, with as few questions as possible, and in simple language. Questions should be ordered from general to specific, and be precise and unambiguous. Double negatives, leading questions and suggested alternatives must be avoided. Closed or open-ended questions may be used, and any other should be provided in multiple choice questions.

Distinguish between closed-ended and open-ended questions.

Closed-ended questions give a fixed set of answers, either two-way yes or no, or multiple choice. They are easy to score and codify but may not include the respondent's true answer. Open-ended questions let respondents answer freely, giving individual and rich replies, but they are hard to interpret and score.

Compare personal interviews, mail surveys and telephone interviews (6 marks).

Personal interviews allow explanation, probing and observation, but are costly, slow and may inhibit honest replies. Mail surveys, including online and SMS ones, are cheap, reach remote areas and avoid interviewer influence, but have low response rates and little help with doubts. Telephone interviews are quick and cheaper than visits, but reach only phone users and permit only short questions.

What is the difference between a census and a sample survey?

In a census, information is collected from every unit of the population, as in India's decennial Population Census. It is detailed but costly and slow. In a sample survey only a selected part of the population is studied and the result is used to estimate the whole. It is cheaper and faster but contains sampling error.

What are random and non-random sampling?

In random sampling, units are chosen by chance and each has a known probability of selection, for instance in simple random sampling by lottery or random number tables. In non-random sampling the investigator decides which units to include, as in deliberate or convenience sampling. Random sampling reduces bias; non-random sampling is quicker but may not represent the population.

What is the difference between sampling error and non-sampling error?

Sampling error is the gap between the sample estimate and the true population value, arising because only a part of the population is studied. It usually shrinks with a larger sample. Non-sampling errors arise from non-response, faulty measurement, wrong answers or recording mistakes, and can occur even in a census.

Name some important sources of secondary data in India.

Important sources include the Census of India, conducted every ten years; the National Sample Survey Organisation, which runs household surveys on consumption and employment; the annual Economic Survey; and publications of the Reserve Bank of India and government ministries. Newspapers, journals, research institutes and websites are also used, but their reliability should be checked.

Why should secondary data be used with care?

The user did not control how they were collected. Definitions, units, time period, area covered and methods may differ from what the present study needs, and errors may exist. Before use, one should check the source, the purpose of collection, the definitions, coverage and accuracy of the data.