Collection of Data

Important Terms

• Investigator: is a person who conducts the statistical enquiry.
• Enumerator: is a person who actually collects the data for investigation from the field of enquiry.
• Respondent: is a person from whom data is collected (or one who responds to the enquiry)
• Statistical enquiry: is an investigation on a topic by an agency wherein relevant quantitative information is collected.
• Survey: is a method of gathering information from individuals on a topic. It is conducted by asking questions about a topic from a selected group of people.

Sources of Data

There are two sources of data: Primary data and Secondary data

Primary Data	Secondary Data
Data which is originally collected by an investigator or agency for the first time for some specific purpose directly from the field of enquiry.	Data which has already been collected and processed by some agency, other than the investigator, for a different purpose.
Data is more accurate and reliable as it is the original data collected by the investigator himself for a specific purpose.	Data is less reliable as it has been collected by someone else for a different purpose or is Second-hand data.
It requires more time and efforts for data collection.	It requires less time and efforts as data is already available.
It is more costly as more number of enumerators are required to collect the data.	It is less costly as it is taken from a published or unpublished data source.
It is raw data on which statistical tools need to be applied.	It is finished and processed data.
E.g: Census data collected by the government.	E.g: Data published in economic survey

Methods of Primary Data Collection   

• Direct Personal Investigation (or Direct Personal Interview) Under this method, the investigator collects data by having a direct face-to-face interview with the respondent.

The investigator goes to the field personally and conducts an on-the spot enquiry.

• Suitability: When the area of investigation is not very large and when maximum degree of accuracy is required.

Merits

1. The first hand information obtained by the investigator himself is bound to be more reliable and accurate.
2. The facial expressions or reactions of the respondent can be observed and some additional information can also be gathered.
3. The investigator has the flexibility of clarifying ambiguous questions and avoiding misinterpretation of questions.

Demerits

1. There is a possibility of influencing the respondents.
2. It is more time consuming and more expensive.
3. It is not suitable if the area of enquiry is large or when many respondents are to be interviewed.

• Telephonic Interview

Under this method, data is collected by the investigator through an interview with the respondent over the telephone.

Suitability: In case the respondent is reluctant to answer certain questions in a face to face interview and data is to be collected in a short period of time.

Merits

1. They are cheaper and can be conducted in short period of time.
2. The method can cover investigation over a large area.
3. It is possible to assist the respondent by clarifying the questions.

Demerits

1. It has limited scope as it is not useful in case the respondent has no phone connection.
2. Facial expressions or reactions of the respondents cannot be observed.

 

• Mailed questionnaire method (or Mailed Interview)

Under this method, the investigator makes a questionnaire pertaining to the field of investigation which is sent to the respondents along with a covering letter specifying the purpose of the enquiry and a request to complete and return the same by a given date. The respondents are also assured of secrecy of the information provided by them. A stamped self-addressed envelope is also enclosed for returning the questionnaire by post.

 

Suitability: When the area of investigation is large and the respondents are literate.

A questionnaire is a list of questions pertaining to the topic of investigation.

Merits

1. It is less expensive and can be used to access remote areas.
2. Every question is interpreted by the respondent himself hence is free from the personal bias of the investigator.
3. It is the best method when anonymity of the respondent needs to be maintained.

Demerits

1. It can only be used if the respondents are literate.
2. It does not allow the investigator to see the reactions of the respondent.
3. Accuracy or reliability of data cannot be testified.
4. There are chances of ‘No response’ or long response time because of mail delays.

 

Pilot Survey (Pre-testing of the questionnaire)

A try-out or trial-run of the questionnaire with a small group of respondents is known as Pilot Survey.

• The pilot survey helps in providing a preliminary idea about the survey and in pre-testing of the questionnaire, so as to know the shortcomings and drawbacks of the questions.
• It also helps in assessing the suitability of questions, clarity of instructions, performance of enumerators and the cost and time involved in the actual survey.

Essentials of a good questionnaire A good questionnaire should consist of:

1. Covering Letter: A polite letter explaining the purpose and scope of the survey should be sent to the respondents ensuring them of secrecy or confidentiality of their responses.

In case of mailed questionnaires, a stamped self-addressed envelope should also be enclosed.

(A) Type of Questions: While preparing the questionnaire/interview schedule, one should keep in mind the following points:

1. 1. The series of questions should move from general to specific. The questionnaire should start from general questions and proceed to more specific ones.
   2. The questions should be precise and clear and should not be ambiguous. They should enable the respondents to answer quickly, correctly and clearly.
   3. The question should not use double negatives or personal questions.
   4. The questionnaire may consist of closed-ended (or structured) questions or open-ended (or unstructured) questions. Closed-ended can either be a two-way question or a multiple choice question. Open-ended questions are subjective so are difficult to interpret and hard to score, since there are a lot of variations in the responses. Hence, preference should be given to closed ended questions.

(B) Layout: The questionnaire should be attractive and pleasing to the eyes of the respondent. Enough space should be provided for each answer. The questionnaire should not be too long. The number of questions should be as minimum as possible.

Sources of Secondary Data

There are two types of sources of secondary data:

(a) Published sources (b) Unpublished sources

 

Published sources include:

• 1. Official Government Publications like Annual Economic Survey, Census of India reports, NSSO (National Sample Survey Organization) reports published in Sarvekshana journal etc.
  2. Semi Government Publications like reports published by municipalities on births, deaths, education etc.
  3. Reports of committees and commissions like Pay Commission reports, Education Commission reports etc.
  4. Publications of research institutes like reports of ICAR (Indian Council of Agricultural Research), ISI (Indian Statistical Institute), reports of NCAER (National Council of Applied Economic Research, Institute of Economic Growth, reports of NCERT etc.
  5. International Publications like reports of UNO (United Nations Organisation), IMF (International Monetary Fund), World Bank etc.
  6. Newspapers and Magazines like data collected by Economic Times, Financial Express, Outlook Money etc.
• Unpublished sources include: unpublished statistical material maintained by research scholars, private investigation agencies, hospital administration, school administration and private firms etc.

Limitations of Secondary data

1. One may not be sure of the method or procedure adopted for collecting the data.
2. It may be influenced by the personal bias of the investigator.
3. It may lack accuracy as data was collected for a different purpose.
4. The data may be outdated or may not cover the full period of investigation.

 

Precautions while using secondary data

1. Suitability for the purpose: The investigator must ensure that the data is suitable for the purpose of enquiry by checking the nature and scope of data as well as the time period covered.
2. Reliability of the data: The reliability of the data can be checked by the experience of the agency source of information and the method of data collection used.
3. Adequacy and accuracy: It is necessary to use adequate data to avoid biases leading to erroneous or inaccurate results. One should keep in mind the degree of accuracy maintained by each investigator.

Primary Data Collection Techniques: Census and Sampling methods

• Population or Universe in statistics is always all the individuals/items who possess certain characteristics (or a set of characteristics), according to the purpose of the survey. It means totality of the items under study.
• A sample refers to a group or section of the population from which information is to be obtained.

Requisites of a good sample

• The selected elements should be representative of the characteristics of the population.
• The size of the sample should be adequate so that all characteristics of the population are represented.
• A good sample (representative sample) is generally smaller than the population and is capable of providing reasonably accurate information about the population at a much lower cost and shorter time.

Census method	Sampling method
It is a method of collecting data where data is collected from each and every element of the population or universe and there is 100% enumeration.	It is a method of collecting data where only some representative items of the population (part of the universe or population) are selected for the study.
It is more reliable and accurate since there is 100% enumeration.	It is less reliable and less accurate since only part of the universe is taken into account.
It is more costly and time consuming	It is less costly and less time consuming.
It requires more number of enumerators.	It requires less number of enumerators.
It is difficult to verify or crosscheck the data. May not be possible to verify in case of large amount of data.	It is easier to verify or cross-check since it involves less data.
E.g: Data collected by Census of India.	E.g: Sampling data collected by NSSO.

 

Suitability of Census over Sampling:

(i) Can be used either when the population is homogenous (similar characteristics) or heterogenous.

(ii) When the area of investigation is limited.

(iii) When high degree of accuracy is desirable.

Suitability of Sampling over Census:

(i) When the area of investigation is large and population is homogenous.

(ii) When there are time and cost constraints (less time and budget).

(iii) When less number of enumerators are available.

 

Merits of census method

(i) It provides an intensive and in-depth information covering many facets of the population. For Example: In population census, many characteristics or attributes like age, marital status, income levels etc. can be obtained at the same time.

(ii) There is high degree of accuracy since each and every item of the universe is taken into account.

(iii) It can be used even when the population is not homogenous.

Demerits of census method

(i) It is expensive since is to be collected from each and every item of the universe. Especially so when the population is very large.

(ii) More time and manpower is required to collect large volumes of data and for its its further analysis and interpretation.

(iii) It is difficult to cross-check and verify the data and if the data is huge, it may even be impossible.

Merits of sampling method

(i) It is more economical (less costly) than census method as only part of the population Is taken into account.

(ii) It takes less time and less number of enumerators to collect the data.

(iii) It can be cross checked and verified as the data involved is not very large.

Demerits of sampling method

(i) Since the results of sampling are based only on part of the population, it may not be 100% accurate.

(ii) The investigator’s bias may be involved in the selection of the sample.

(iii) It may not be as effective in case the population is heterogenous.

 

Types of sampling methods

Random Sampling (Probability Sampling)	Non- Random Sampling (Non-Probability Sampling)
In this kind of sampling each and every item of the universe has an equal chance of being selected in the sample.	In this kind of sampling each and every item of the universe does not have an equal chance of being selected in the sample and convenience or judgement of the investigator plays an important role in selection of the sample.
Selection is by chance not by choice of the investigator.	Selection is not by chance but by choice of the investigator.
Personal bias of the investigator is not involved.	Personal bias of the investigator is involved.
Example: Lottery method, stratified random sampling, systematic random sampling	Example: Judgement sampling, convenience sampling and quota sampling

 

 

Random Sampling (Probability Sampling)

Simple / Unrestricted Random Sampling

Restricted Random Sampling Stratified Systematic

• Simple Random Sampling

A simple random sampling is one in which every item of the population has an equal chance of being selected.

LOTTERY METHOD: In this method, all items of the population are numbered or named on separate paper slips which are then placed in a bowl and mixed thoroughly. The elements are then selected randomly from the bowl according to the required sample size.

Merits

(i) Personal bias of the investigator is not involved.

(ii) It is based on the rules of probability.

Demerits

(i) It is time consuming especially if the population is large as all the elements have to be numbered or named and then the sample is drawn.

(ii) For a small population, the sample may not be representative of the population.

 

• Stratified Random Sampling

In this method, the universe or the entire population is first divided into a number of homogenous groups or ‘strata’ and then the required number of items are selected randomly from each group as per the sample size.

It is suitable when the population is heterogenous. This ensures that all the characteristics of a heterogenous population are adequately represented in the sample.

For example: If a sample of 10 is to be drawn representing all the types of occupation in a locality, The population is first divided into homogenous strata of different types of occupations in the locality such as Teachers, Lawyers, Doctors, Businessmen, Engineers etc. After this, the sample is drawn randomly from each of these groups so that the resulting sample is representative of the heterogenous population.

 

Merits

(i) The sample is more representative of the population and hence is useful in case the population is heterogenous.

(ii) Personal bias of the investigator is not involved.

Demerits

(i) It requires complete knowledge regarding the diverse characteristics of the population.

(ii) It is difficult to ascertain the different strata or groups in the population.

(iii) The stratified samples, if widely distributed, may prove to be expensive and time consuming.

 

• Systematic Random Sampling (Quasi Random Sampling)

In this method, the elements of the population are first listed or ordered alphabetically or numerically and then the sample is selected by taking every K^th item from the list where K is the interval size.

K (interval Size) = Size of the population/ Size of the sample

The randomness lies in the choice of the first sample item.

For example: If a sample of 4 students is to be drawn from a population of a group of 12 students, The students are first ordered alphabetically and then the interval size is calculated as:

K (interval Size) = 12/4 i.e. K= 3

Hence, the sample is then drawn by taking every 3^rd student from the group. The randomness lies in the choice of the first student. E.g: If the 2^nd student is chosen, the next would be the 5^th, followed by the 8^th and so on.

Merits

(i) It is less time consuming and less effort is required.

(ii) Personal bias of the investigator is not involved.

Demerits

(i) Complete knowledge about the population from which the sample is to be drawn is required.

(ii) It is not suitable for a large population since it would be difficult to prepare the sampling frame.

Non- Random Sampling (Non-Probability Sampling)

Sampling

Convenience Sampling

Quota Sampling

Judgement/ Deliberate Sampling

In this method, the sample items are chosen exclusively by the judgement of the investigator. Hence, the chance of inclusion of some items in the sample is very high while that of the others would be very low.

For example: An investigator who wants to conduct a survey to ascertain the choice of streams in a school, may select 50 students who, according to his judgement or opinion, are representative of the population.

Merits

(i) It is an easy method as it does not involve complicated selection procedures.

(ii) It prevents unnecessary and irrelevant items from being selected in the sample.

Demerits

(i) Personal bias of the investigator is involved.

(ii) The investigator should have full knowledge about the population without which there could be error in his judgement.

 

• Quota Sampling

Under this method, the items of the population are first sub-divided into various groups and then a quota (number of items to be selected from each subgroup) is fixed. Within the given quota, the selection of the sample units depends upon the personal judgement of the investigator.

For example: In a product survey, it may be decided that out of the total population, 50% should be females who are housewives, 30 % should be school going girls while the remaining 20% should be working females. Within the quota, the investigator is free to select the people to be interviewed.

Such sampling is used for opinion polls and market surveys of new products.

Merits

(i) It is very economical and gives more reliable results.

(ii) It is easy to administer and the sample can be selected to suit the enquiry.

Demerits

(i) It involves personal bias of the investigator.

(ii) It is not possible to estimate the degree of accuracy achieved.

 

• Convenience Sampling

In this method, the sample items are selected according the convenience of the investigator.

For example: If a sample of 20 students is to be made to study the quality of education imparted in schools, the investigator may choose students from the schools near his residence or place of work for his convenience.

Merits

(i) It is less time consuming and less costly.

(ii) It is convenient and the sample can be easily located and contacted by the investigator.

Demerits

(i) Personal bias of the investigator is involved.

(ii) The results may be misleading or unsatisfactory as the sample may not represent the universe or given population.

Important government agencies of data:

Two important agencies at the national level which collect, process and tabulate data on important economic and social issues are:

• Census of India

(i) The Census of India provides the most complete and continuous demographic record of population.

(ii) The Census is being regularly conducted every ten years since 1881. The first Census after Independence was conducted in 1951.

(iii) The Census officials collect information on various aspects of population such as the size, density, sex ratio, literacy, migration, rural-urban distribution, etc.

(iv) Census data is interpreted and analysed to understand many economic and social issues in India.

• National Sample Survey Office (NSSO)

(i) The NSS or National Sample Survey came into existence in 1950 and was reorganised as NSSO in 1970.

(ii) The NSSO was established by the Government of India to conduct nationwide surveys on socioeconomic issues. It conducts continuous surveys in successive rounds.

(iii) The data collected by NSSO are released through reports and its quarterly journal Sarvekshana.

(iv) It provides periodic estimates of literacy, school enrolment utilisation of educational services, employment, unemployment, manufacturing and service sector enterprises etc.

(v) The NSSO also collects details of industrial activities and retail prices for various goods.

• Few other agencies which collect data at the national level are:

• Central Statistics Office (CSO), Registrar General of India (RGI), Directorate General of Commercial Intelligence and Statistics (DGCIS), Labour Bureau, etc.

 

NOTE : (reports of these organisations are important sources of secondary data)