Sampling Process and Data Collection
Sampling Process and Data Collection
Sampling
In an experimental study covering the entire population is not possible
because it is costly, time consuming, laborious and complication for
handling data. Hence it is preferable to use an appropriate sampling
technique.
(
i) Population
A statistical population/universe is the aggregate (totality) of
measurements on characters under investigation.
For example
Population of doctors in a hospital, women of child bearing period (1549) in a village, smokers in a city, cancer patients in a district etc.
sample
(ii) Sample A small part (finite sub set) of statistical
Individuals chosen from the population for
studying its properties is called a sample.
The number of units in a sample is known
as sample size.
Random sampling
Non-random sampling
Snowball sampling
Snowball sampling : if the sampling frame
and population is not well defined then non of
sampling starategy works. In such case we
first locate any sampling unit at any cost and
take help of this sampling unit to find another
units and so on until we get appropriate no of
samples. As the size of sample gets larger
and larger each time like that of the size of
the snowball this method of sampling is
known as snowball sampling. This type of
sampling is done for cases like taking
responses where the respondent do not like
to reveal their identity.
Quota Sampling
Quota Sampling
A quota sampling is a special form of stratified
sampling. In this method, the population is first
divided into several strata and the sample units
usually known as quota is selected from
different strata according to the convenience of
the sampler so as to meet the predetermined
sample size.
For example Public opinion poll from n individuals consisting
n1 technicians, n2 administrators, n3 laborers,
etc. regarding in favor or against nuclear power
plant.
Systematic sampling
Cluster sampling
Definition
A cluster sampling is a SRS in which sampling
unit is a collection or cluster of elements.
In a certain district having 200 schools if 10
schools are chosen under SRS technique and
enumerate all the units within the selected
schools, then, the schools so selected are
known as clusters and the technique is known
as cluster sampling.
A cluster sampling wherein geographical
areas are used as clusters is known as area
sampling.
Multistage sampling
Definition
In a multi stage sampling mechanism in which sampling
procedures are carried on in several stages using random
sampling techniques. This technique is recommendable for large
country surveys.
Mechanism
Suppose we are handling a stage wise survey of status of eye
defective patients:
Let us consider the 75 districts of Nepal.
1st stage: Select 40 districts of Nepal SRS technique.
2nd stage: Select 20 VDC/ Nagar Palika from the selected 40
districts by SRS technique.
3rd stage : Select 80 households out of 20 VDC/ Nagar Palika by
SRS technique.
4th stage : Finally enumerate all eye defective patients in the
selected households. This is a four stage sampling.
Sampling Error
Error incurred by selection of sample that
are not representative of the population
Sources of Sampling Errors
-Faulty selection of sample
-Substitution
-Faulty Demarcation of Sampling Units
-Constant error due to improper choice of
the statist tics for Estimating population
parameter
Less time
Reduced cost of survey
Greater accuracy of results
Greater scope
If the population is too large or
If the test is destructive or
If the population is hypothetical then only
possible method of data collection is
sampling
Interview
It is a technique of primary data collection. It is an oral method in which one
person asks another person questions designed to obtain answer pertinent to
the Research problem. It is most commonly used direct method in the study
of human behavior. The interview is a face to face interpersonal role situation
in which one person asks another person being interviewed. The respondent
question designed to obtain answer pertinent to the purpose of the Research
problem. The interview may be regarded as a systematic method by which
one person enters more or less imaginatively with the inner life of another
who is
generally a comparative stranger to him. The purpose of interview is to find
out whats in or on someone elses mind.
According to Kerlinger The interview is a face to face interpersonal
role situation in which one person, the interviewer, asks a person
being interviewed, the respondent, questions designed to obtain
answers pertinent to the purpose of the Research problem.
Types of interview
Structured or standardized interview
It this interview, the questions, their sequence and their wordings are fixed. The
same question is asked to everybody in the same manner. in general, the
interviewer doesnt have flexibility to change the question, their format or order.
Sometimes an interviewer may be allowed some liberty in asking question but
relatively little. That liberty is specified in advance. Standardized interviews are
interview schedule that have been carefully prepared to obtained information
pertinent to the Research problem.
Unstructured or unstandardized interview
In the interview, the sequence and wording of question are not fixed. The
interviewer doesnt follow pre-planed list of the question. The interviewer writes
the response
of the interview during the interview or after interview is completed. It is more
flexible and open. Ordinary no schedule is used. They dont mean that
unstructured interview is casual. The interviewer will have a tentative list of
question to be covered during the interview.
Questionnaire
A questionnaire is a format list of questions designed to gather respondent from
respondent on a given topic. It is a formalized schedule for data collection. It is
used when fractional information is designed. When opinion rather than facts are
desired, opinionnaire or attitude scale is used. Questionnaire is a term used for
almost any kind of instrument that has questions or items to which individuals
respond. A questionnaire is a list of questions to number of persons for them to
answer. It secures standardized results that can be tabulated created statistically.
Type of questionnaire
Open or open-ended question:
Close or close-ended question:
restricted questionnaire
Types of close- questionnaire
Dichotomous questionnaire:
It offers only two alternatives; positive and negative. The response would be yes or
no; agree or disagree; and so on. It is also known as two way questionnaire.
Multi chotomous questionnaire:
It is also called multiple choice questionnaires.
Questionnaire Design
A questionnaire is designed to collect right type of primary data of right quality. The design of questionnaire is
more than art from than a scientific undertaking. Questionnaire design comes from the experience of the
Researchers who specialized in this area. The only way to develop this skill is to write a questionnaire, use it in a
series of interviews, analyze its weakness and revise it. There are no series of step, principle or guidelines with
guarantee on effective and efficient questionnaire. An effective questionnaire is one that
Draws out accurate information
Cab be completed easily by interviewers
Flow well
Leaves the respondent feel satisfied for their worth while participation in the Research
There are series of seven steps in questionnaire design as
started below:
It is versatile. Almost every problem of social Research can be approached from the questionnaire stand point.
Every social problem involves people. Therefore, ideas relative to the problem and solution can be obtained by
asking these people about the problem
Many people can be studied only by questioning. Knowledge, opinions, motivation, and intentions are usually
not open to observation. Similarly, it is not feasible to observe personnel activities such as burning teeth. So, it is
feasible.
Questioning is usually faster and efficient than observation. Some events that take place over a time period
would require lengthy observation, but a question on this behavior can be answered in a few seconds.
It is cheaper then observing. The Researcher has not to stand time to observe the behavior of the respondent.
A decrease in time usually lead to decrease in cost.
The person administrating the questionnaire has an opportunity to establish report, explain the purpose of the
study and explain the meaning of items that may not be clear.
Disadvantages of questionnaire
Respondents would be unwilling to provide information. Questions about
income or very personnel subject frequently meet refusal by respondent. The
show little or no interest to the question.
Despite a willingness to cooperate, many people are unable to give accurate
information of questions.
There can be influence on questioning process. Often respondent attempt to
give answers that they think will please the Researchers.
Filling out lengthy questions take a great deal of time and money.
Other disadvantages
It cannot be used in illiterate person.
Incomplete responses are possible
There is a possibility of wrong answer
Useless in depth problem
Uniform questions are not applicable to other people of different ethnicity,
culture etc.
Observation
It is a method of gathering primary data physically or
mechanically recording events or aspect of the phenomenon
under investigation. It involves recording of the respondents
behavior. It is the process of recognizing noting people, objects
and occurrence of events rather than asking for information. It can
supplement the information collected through
questionnaire and interview. Observation is the process of
recognizing and recording
behavior of people, objects and events. Observation is systematic
and deliberate study through eye, of spontaneous occurrence at
the time they occur. Example: instead of asking consumer what
brand they buy or what television program they watch, the
Researcher arrange to observe what product are brought and what
program they watch.
Sample Size
Too Big:
Requires
too much
resource
s
Too
small:
Wont
do the
job
Example: Sample
Size
for Mean
What sample size is needed to be 90%
confident of being correct within 5?
A pilot study suggested that the
standard deviation is 45
Z
1.645 45
n
219.2 220
2
2
Error
5
2
Round Up
Example: Sample
Size
for Proportion
What sample size is needed to be within
5 with 90% confidence? Out of a
population of 1,000, we randomly
selected 100 of which 30 were defective
2
2
Error
0.05
227.3 228
Round
2