Mukt Shabd Journal ISSN NO : 2347-3150

Techniques for Sampling

Dr.Ambarish Kalikar : Bharatiya Mahavidyalaya,Amravati
---------------------------------------------------------------------------------------------------------------------

Abstract -

Research benefits greatly from sampling. It is among the key elements that influences how
accurate the results of your survey or research will be. Should there be any issues with your
sample, the outcome will be directly affected. Numerous methods exist to assist us in gathering
samples based on the circumstances and needs.

Selecting a small group of individuals from a larger group is known as sampling. This is
frequently discovered when population data collection is required for statistical analysis,
demographic surveys, and economic research.

Sampling, as used in statistics, quality control, and survey technique, is the process of choosing a
subset or statistical sample abbreviated "sample" for short of people from a statistical population
in order to estimate the characteristics of the entire population. The goal of the subset is to
represent the entire population, and statisticians make an effort to gather representative samples.
Sampling can be useful when it is not practical to measure the entire population because it is less
expensive and allows for faster data collection than capturing data from the full population.

Every observation quantifies one or more characteristics of distinct items or people, such as
mass, weight, location, colour, or other attributes. Weights can be used in survey sampling,
especially in stratified sampling, to adjust the data for the sample design. The technique is guided
by findings from statistics and probability theory. Sampling is a common method used in
business and medical research to obtain data about a community. To find out if a production
batch of material satisfies the relevant requirements, acceptance sampling is utilised.

Key word- Sampling, Observation , probability ,

Introduction -

The sampling technique studies only a portion of the universe and draws conclusions for the
entire universe based on that analysis, rather than looking at every unit in the universe.

The concept of sampling is not very new, even if the theory of sampling has evolved
significantly in recent years. People have inspected a handful of grains to determine the quality
of the entire bunch since ancient times. To determine whether the rice is ready, a homemaker
looks at two or three grains of boiling rice. A physician analyses a few blood droplets to
determine the blood composition of the entire body. After reviewing a small sample of the
substance, a businessman places an order for it. To determine whether the class as a whole is

Volume XIII, Issue IX, SEPTEMBER/2024 Page No : 177

Mukt Shabd Journal ISSN NO : 2347-3150

understanding the lesson, a teacher may quiz one or two pupils. Actually, there are very few
fields where the technique of sampling a not used either consciously or unconsciously.

Selecting a Sampling Technique: It should be remembered that every technique has a unique
set of applications. The choice of a specific sampling technique would be influenced,
consciously or unconsciously, by a variety of factors, including the nature of the problem, the
size of the universe, the size of the sample, the availability of money, time, etc.

Sample size: Selecting an appropriate sampling strategy requires careful consideration of the
sample size. Experts on this topic have voiced differing perspectives. While some suggest the
sample size should be at least 10%, others have proposed that it should be 4% of the population
size. These opinions are, however, largely useless because, in actuality, the right sample size
depends on a number of factors related to the topic being studied, such as the desired level of
accuracy, the time and cost considerations, etc. The idea of sampling is not very helpful in
determining the appropriate sample size in any given circumstance. When choosing the right
sample size, it's advisable to take into account the next two factors.

* As the variety in the particular item rises, the size of the simple should also increase.
* The larger the sample size, the higher the desired level of accuracy.

Positive aspects -

1.Quickness, or less time

2. The Economy
3. Convenience for Administration:
4. Trustworthy
5. Expanded Range
6. The population is infinite or hypothetical.
7. In-depth Examinations

ERROR - It is the discrepancy between the estimated or approximated value and the true value.
Numerous factors contribute to statistical errors, including

1. Measurement approximation -

2) Rounding numbers to the closest hundreds, thousands, or millions or rounding decimals to the
nearest

3) Biases caused by incorrect data collection, incorrect data analysis, and prejudiced results
presentation and interpretation.

Volume XIII, Issue IX, SEPTEMBER/2024 Page No : 178

Mukt Shabd Journal ISSN NO : 2347-3150

4) Personal biases of the investigators and so on

Sample Error – Sampling error is the mistake incurred when approximating population
characteristics primarily only on a sample, and it occurs because only a small percentage of the
population is used in the calculation.

Reasons of sampling errors -

1. Sample selection facility: If a sample is chosen using an inadequate sampling approach, bias is
introduced.

2. Substitution : In the event that the investigators are unable to count a specific sampling unit
that is part of the random sample, they often replace it with a practical member of the population.
Due to the fact that the attributes held by the replacement unit will typically differ from those of
the unit that was initially included in the sample, this causes bias..

3. Improper sampling unit demarcation: In area surveys, such as field agricultural trials or crop
cutting surveys, bias resulting from improper sampling unit delineation is especially important.

4. Error generated by bias in the forecasting process: An incorrect choice of estimation methods
can lead to an error.

5. Variability of the population- Sampling error also depends on the variability or heterogeneity
of the population to be sampled.

Significant elements that contribute to non-sampling errors in surveys include:

1. Inadequate planning, such as imprecise and incorrect definitions of the population or the
statistical units to be employed; a deficient population membership list.

2. A vague and poor questionnaire that could provide inaccurate or incomplete data
3. Ineffective techniques for conducting interviews and posing queries:
4. Lack of clarity regarding the nature of the data to be gathered.
5. Exaggerated or incorrect responses to the questions that play to the respondents' egos, pride, or
self-interest.
6. The investigator's personal prejudice
7. A shortage of supervisory staff as well as a dearth of experienced and trained investigators.
8. The respondent's inability to recollect past events or occurrences.
9. Failure to respond and insufficient or lacking in information
10. Inadequate insurance

Volume XIII, Issue IX, SEPTEMBER/2024 Page No : 179

Mukt Shabd Journal ISSN NO : 2347-3150

11. Gathering Mistakes

12.Publication mistake.

Errors that are prejudiced or cumulative - Errors with bias emerge as the outcome of

1) Bias on the part of the enumerator or investigator, whose preconceptions and ideas could
influence the investigation's finding

2) Bias in the equipment or gauge used to record the observations.

3) Bias resulting from faulty data gathering methods, statistical methods, and formulas applied to
the data analysis.
4) Respondent's bias: Appealing to someone's pride or prestige can lead to a bias known as
prestige bias, which causes them to inflate their age, understate their wealth, improve their
education, or work status, among other things, in order to protect their own interests.

5) Bias in the technique of approximations: If, while rounding off, each individual value is either
approximated to next highest or lowest number so that all the errors move in the same direction,
there is bias for overstatement or understatement respectively.

Fair Mistakes (Compensating Mistakes) - if the likelihood of overestimating something is

almost equal to the likelihood of underestimating something. Because these faults travel in both
directions, the errors in one disruption are mostly offset by the errors in the other, so that the
final outcome is not much impacted.

References -

1. Aitkin, M. (2008). Applications of the Bayesian bootstrap in finite population inference.

Journal of Official Statistics, 24, 21–51.
2. Alexander, C. H. (1987). A class of methods for using person controls in household
weighting. Survey Methodology, 13, 183–198.
3. Asok, C., & Sukhatme, B. V. (1976). On Sampford’s procedure of unequal probability
sampling without replacement. Journal of the American Statistical Association, 71, 912–918.
4 .Bahadur, R. R. (1966). A note on quantiles in large samples. Annals of Mathematical
Statistics, 37, 577–580.
5 .Baker, R., Brick, J. M., Bates, N. A., Battaglia, M., Couper, M. P., Dever, J. A., et al.
(2013). Report of the AAPOR task force on non-probability sampling. Journal of Survey
Statistics and Methodology, 1, 90–143.

6.Ben-Shlomo Y, Brookes S, Hickman M. 2013. Lecture Notes: Epidemiology, Evidence-

based Medicine and Public Health (6th ed.), Wiley-Blackwell, Oxford.
7.http://www.stats.gla.ac.uk/steps/glossary/sampling.html - Accessed 8/04/17

Volume XIII, Issue IX, SEPTEMBER/2024 Page No : 180