Chapter 2 Collection Of Data

1. Introduction

In the previous chapter, we established that economics relies heavily on statistics. To analyze economic issues and formulate effective policies, we need factual evidence. This evidence comes in the form of data. Data serves as the primary tool that provides information, allowing us to understand problems and arrive at clear, well-supported solutions.

Consider an economic statement like, "The production of food grains in India rose from 108 million tonnes in 1970-71 to 272 million tonnes in 2016–17." The figures in this statement—108 million and 272 million—are data points. These values change over time, so they are referred to as variables. Each specific value of a variable (e.g., the production in a particular year) is called an observation.

To study any economic phenomenon, such as the fluctuations in food grain production, we first need to collect the relevant data. This chapter explores the fundamental aspects of data collection: where data comes from (sources) and how it is gathered (methods).

2. What Are The Sources Of Data?

Statistical data can be sourced in two primary ways, which determines whether the data is classified as primary or secondary.

• Primary Data: This is data collected for the first time by the researcher or investigator for their specific purpose. It is original, first-hand information gathered directly from the source. For example, if you conduct a survey among your classmates to find out their favorite brand of smartphone, the information you gather is primary data.
• Secondary Data: This refers to data that has already been collected and processed by some other person or agency and is then used by a different researcher for their study. This data is second-hand. It can be sourced from published materials like government reports, research papers, newspapers, or websites. For example, if a business analyst uses the data from your smartphone survey to study market trends, your primary data becomes secondary data for the analyst.

Using secondary data can be very efficient, as it saves a significant amount of time, money, and effort. However, it's crucial to ensure that the secondary data is reliable, suitable for the current research objective, and was collected using proper methods.

3. How Do We Collect The Data?

The collection of primary data is often done through a survey. A survey is a method of gathering information by asking questions to a group of individuals (respondents). The purpose is to describe certain characteristics of the group, such as their opinions, preferences, or behaviors.

Preparation of Instrument

The most common tool used in surveys is the questionnaire or interview schedule. This is a list of questions designed to collect specific information. Designing a good questionnaire is a critical step and requires careful planning. Here are some key principles:

• Keep it Short: The questionnaire should not be too long. A shorter questionnaire encourages a higher response rate.
• Use Clear Language: Questions should be simple, easy to understand, and avoid difficult or ambiguous words.
• Logical Order: Questions should be arranged logically, moving from general to more specific topics to make the respondent feel comfortable.
  • Good Example: Start with, "Is the electricity supply in your locality regular?" and then ask, "Is the increase in electricity charges justified?"
  • Poor Example: Asking about justifying the price increase before establishing if the service is regular.
• Be Precise: Avoid vague questions. Instead of "Do you spend a lot on books?", ask "How much do you spend on books per month? (Less than $\textsf{₹}$200, $\textsf{₹}$200-300, etc.)"
• Avoid Double Negatives: Questions like "Don't you think smoking should be prohibited?" can be confusing and lead to biased answers. A better phrasing is, "Do you think smoking should be prohibited?"
• Avoid Leading Questions: Do not phrase a question in a way that suggests a particular answer. Instead of "How do you like the flavour of this high-quality tea?", simply ask, "How do you like the flavour of this tea?"
• Avoid Indicating Alternatives: Let the respondent provide their own answer. Instead of "Would you like to do a job or be a housewife?", ask "What would you like to do after college?"

Types of Questions:

• Closed-ended (Structured) Questions: These offer a limited set of pre-defined answers. They can be two-way (e.g., Yes/No) or multiple-choice. They are easy to analyze but may not capture the respondent's true opinion if the right option isn't available.
• Open-ended (Unstructured) Questions: These allow respondents to answer in their own words. They provide rich, detailed information but are more difficult to interpret and analyze. Example: "What is your view about globalisation?"

Mode of Data Collection

There are three basic ways to administer a survey and collect data:

Pilot Survey

Before launching a full-scale survey, it is essential to conduct a pilot survey, which is a try-out of the questionnaire on a small group. This pre-testing is crucial because it helps to:

• Get a preliminary idea about the survey's potential outcomes.
• Identify shortcomings in the questions (e.g., ambiguity, confusion).
• Assess the clarity of instructions for the respondents and enumerators.
• Estimate the time and cost required for the actual survey.

4. Census And Sample Surveys

Once the data collection instrument is ready, the researcher must decide whether to survey the entire group of interest or just a part of it.

Census Or Complete Enumeration

A Census is a survey that includes every single element of the population under study. It is also known as the method of complete enumeration. The most well-known example is the Census of India, conducted every ten years. It involves a house-to-house enquiry to collect demographic data (like population size, literacy, employment) from every household in the country.

Population And Sample

• In statistics, a Population or Universe refers to the entire group of individuals or items that a researcher is interested in studying.
• A Sample is a smaller, representative group or section selected from the population, from which information is obtained.

Most surveys are sample surveys because they are more practical than a census. A well-selected sample can provide reasonably accurate information about the entire population at a much lower cost and in a shorter time. Because the scale is smaller, it's possible to collect more detailed information and better train and supervise the investigators.

Random Sampling

Random sampling is a method where every individual unit in the population has an equal and independent chance of being selected for the sample. This ensures that the sample is representative of the population and free from investigator bias. A common method is the lottery method, where names are written on slips of paper, mixed, and then drawn randomly. Today, computer programs are widely used to generate random samples.

Non-random Sampling

In non-random sampling, not all units of the population have an equal chance of being selected. The investigator uses their judgment, convenience, or a quota to select the sample. While this method can be easier to implement, it is prone to bias and may not yield a sample that is truly representative of the population.

5. Sampling And Non-sampling Errors

Errors in statistical surveys are broadly classified into two types: sampling errors and non-sampling errors.

Sampling Errors

A sampling error is the difference between the result obtained from a sample (the sample estimate) and the true value that would have been obtained from a census of the entire population (the population parameter). This error arises simply because a sample is only a part of the population and may not perfectly reflect it.

The magnitude of the sampling error can be reduced by taking a larger sample size. A larger sample is more likely to be representative of the population, thus reducing the error.

Answer:

Non-sampling Errors

Non-sampling errors are much more serious because they are not related to the act of sampling and can occur even in a census. Increasing the sample size does not reduce these errors. They include:

• Sampling Bias: This occurs when the sampling plan is flawed in such a way that certain members of the population have no chance of being included in the sample.
• Non-response Errors: This happens when an investigator is unable to contact a selected person, or the person refuses to answer. This can make the sample unrepresentative.
• Errors in Data Acquisition: These are mistakes made during the data collection process, such as:
  • Recording incorrect responses.
  • Errors in measurement due to faulty instruments.
  • Carelessness by the investigator or respondent.
  • Transcription errors (e.g., writing down 13 instead of 31).

6. Census Of India And Nsso

In India, several government agencies are responsible for collecting, processing, and publishing statistical data. Two of the most important agencies at the national level are:

1. Census of India: Conducted by the Registrar General of India (RGI), the Census provides the most complete and continuous record of India's population. It has been conducted every ten years since 1881. It collects comprehensive data on population size, density, sex ratio, literacy, migration, and other demographic and socio-economic characteristics.
2. National Sample Survey (NSS): Formerly the National Sample Survey Organisation (NSSO), the NSS was established by the government to conduct regular nationwide surveys on various socio-economic issues. It operates in successive "rounds," each focusing on different topics. The NSS provides crucial data on literacy, employment, unemployment, household consumption, healthcare, and the public distribution system. Its findings are released through reports and its quarterly journal, Sarvekshana, and are vital for government planning.

7. Conclusion

Data, which are economic facts expressed in numbers, are essential for understanding and analyzing economic problems. The process of data collection is a foundational step in any statistical inquiry and must be planned carefully.

Data can be collected first-hand (primary data) through surveys, or it can be sourced from existing publications (secondary data). The choice between a full census and a smaller sample, the method of sampling (e.g., random sampling), and the mode of data collection (personal interview, mail, telephone) all depend on the specific objectives, budget, and timeline of the study.

Recap

• Data is a fundamental tool for drawing sound conclusions about economic problems.
• Primary data is first-hand information collected by the researcher, often through surveys using methods like personal interviews, mailing questionnaires, or telephone interviews.
• A Census includes every member of the population, while a Sample is a smaller, representative group selected from it.
• In random sampling, every member of the population has an equal chance of being selected.
• Sampling errors arise from the difference between a sample estimate and the true population value. They can be reduced by increasing the sample size.
• Non-sampling errors are mistakes in data acquisition, non-response, or biased selection, and are more serious as they are not fixed by larger samples.
• The Census of India and the National Sample Survey (NSS) are key national agencies that provide vital statistical data for planning and analysis in India.

Exercises

This section contains questions for practice and self-assessment, designed to test the learner's understanding of the concepts discussed in the chapter, such as framing questionnaire questions, distinguishing between census and sample, and identifying different types of statistical errors.