Collection of data refers to when data is gathered and measured in a systematic manner on the basis of variables of interest. It enables one to find a sound solution to any problem. Data is required to understand any inconsistencies in the output.
The process used in the collection of data in any field is generally the same. Every field of work is highly dependent on data. So, the accuracy of the data collected is crucial. It helps one to make informed decisions as well as changes. Data is basically information that is collected in the form of numbers.
Types of data
There are mainly two types of data-
1. Primary data
Primary data is based on firsthand information. For instance, let us assume you want to know the popularity of a specific genre of music amongst the students of a school. You will have to ask questions from a good number of students to collect the required information. This data collected is a form of primary data.
2. Secondary data
When the data collected is scrutinised and examined by any agency, it is secondary data. Government documents, newspapers, reports by economists, etc., are examples of secondary data.
Types of data collection
There are three ways of collecting data. They are discussed as follows-
1. Personal Interviews
Personal interviews enable the researchers to conduct interviews with the respondents in person. It also allows them to explain the questions and the requirements of the questionnaire to the respondents.
There is no scope for any miscommunication. Moreover, the way the respondents react to a particular question can give more information to the researchers. Though, this method of collection of data is expensive.
2. Mailing Questionnaire
In this method, the questionnaire is sent across in emails. The respondents are supposed to fill in the questionnaire and send it to the researchers before a specific date mentioned in the questionnaire. This method is effective and cheap.
3. Telephone interviews
Researchers gather information from the respondents over a telephone call, as the name suggests.
Once a questionnaire has been prepared, it is tested in a small group of people. This is called the pilot survey. This helps the researchers understand if there are any drawbacks to the questionnaire. It helps them identify areas of improvement to give more accuracy in the collection of data.
Census and Sample Surveys
Census is a survey that includes all the elements of the population. It is also called the method of complete enumeration. In India, a door-to-door inquiry is carried out in rural and urban households. The Registrar General of India collects this demographic data.
2. Population and Sample
A population is always all the individuals/items who possess specific characteristics (or a set of characteristics). If the researcher finds it difficult to conduct surveys of an entire population, they can select a representative sample.
A sample is a section of the population from which a researcher can obtain information. A good sample can provide accurate and sound information in a shorter period and much lower cost.
3. Random sampling
In random sampling, individual units of a population are randomly selected. Suppose the government has to determine the impact of a hike in prices of a specific commodity in a locality of 450 households. Researchers can take up to 45 households to conduct their survey. The names of all the 450 households are jotted down in paper and mixed. Only 45 papers can be selected out of them to be interviewed later. There is an equal chance for every household to be a part of the survey.
4. Non-Random Sampling
In this sampling method, the researchers cannot randomly choose the individuals of a population. Instead, they are handpicked. This selection depends upon the researcher’s preference, quota, or requirement. So, in this method, every individual in the population does not have equal chances to participate in the research.
Sampling and Non-Sampling Errors
1. Sampling error
The sampling error is the difference between the actual value of a parameter of the population and its estimate (from the sample). One can reduce such errors by taking larger chunks of the sample.
2. Non-sampling error
Non-sampling errors are more severe errors that one cannot eliminate by taking larger samples. They are of different types:
- Sampling Bias – Occurs when a particular section of the target population is left out from the sample.
- Non-response errors – Occurs when some respondents refuse to be a part of the research or when the researcher is unable to connect to a certain number of individuals mentioned in the sample list.
- Error in data acquisition – Occurs due to the recording of incorrect responses from the respondents while carrying out a survey.
Census of India and NSSO
India’s census provides a continuous and complete record of the nation’s demographic and population. Since 1881, the census has been conducted at a 10 year interval.
The National Sample Survey Organization is established to conduct surveys on socio-economic issues. Reports are published quarterly on employment, education, unemployment, morbidity, childcare, etc.
Facts, when expressed and collected in numbers, are called data. Its collection helps one to understand and find the solution to problems. Several planned steps are executed to conduct surveys to obtain primary data directly from the respondents. On the other hand, data collected and published by certain agencies and government bodies are secondary data.