Sampling Introduction

Sampling

Purpose of sample:

            It is difficult to conduct research by using all the members of population. So due to this difficulty, researchers select some members from targeted population who will represent their population, e.g. if researcher want to conduct research on university adolescents, it is difficult for him to collect data from all students of university. So researcher will select some of the students from university population by using different method of sampling (will discuss later). Those students who selected by the researcher from university population are called sample. In another example, if researcher want to assess the relationship between self esteem and emotional intelligence in psychology class. He will not administer the questionnaire on whole class, but will select some students from the class. Those students will be called sample.

            Before discussing sampling techniques, we will try to know the basic terms regarding sampling. (Population, sampling frame, Sampling strategy, Sampling bias, sample)

Population:

            All the people of interest in a study are called population e.g. if researcher want to investigate the effect of emotional intelligence on self esteem in university students, all the students of university be called population

Sampling frame:

A list of all the individual of target population is called sampling frame e.g. list of all the students enrolled in university is called sampling frame.

Sampling strategy:

            Techniques, procedures or ways used for selection of sample is called sampling strategy.

Sampling bias:

            When individuals selected for the sample will be not representative of target population.

Sample:

            Individuals, selected from the population by using sampling techniques is called sample.

Advantages of sampling:  

1.       Low cost         (low cost is required as compare to population)

2.       Less time consuming   (time saving as compare to whole population)

3.       Easy in process            (very easy to get information through sample as compare to population)

4.       Suitable in case of large population    (suitable method for studying population)

5.       Detailed information   (researcher collect detailed and comprehensive data or information about phenomenon through sample)

Disadvantages of sampling:

1.      Accurate and careful sampling is difficult

2.      Experts are required for sampling

3.      If information is required from every individual then sample is not best and suitable choice

4.      Difficult in selection of truly representative sample.

Sampling techniques:

            There are two sampling techniques:

1.      Probability sampling

2.      Non-probability

Probability sampling

            A sampling technique, in which every individual has equal chances of  being selected from the population. There some methods include in probability sampling"

Simple random sampling:

            It is not specific method of sample selection. In this technique all the members of population has an equal chances of selection in sample.

 In simple random sample:

1.      Write down the names of members of population on separate slips

2.      Put all the slips in a box and shuffle them

3.      Blindly draw the slips from the box

For example, if the members of population are 100 and researcher wants to select 10 members as a sample, above mentioned steps will be used. First, write the names of 100 members on separate slips one by one. Put all the slips in a box and shuffle them, then draw 10 slips blindly from the box. Those 10 members will be sample.

This technique of sampling can be used with small population. i.e. students in a class or workers in a factory. For large population it is not applicable or too much difficult.

Systematic random sampling:

            In this sampling technique researcher use the following steps for sample selection

1.      Prepare a list of population members. List them all as 1, 2, 3, 4, 5, 6, 7……1000

2.      Determine interval size through dividing the numbers of subjects in population by numbers of subjects in sample. I.e. if population size is 1000 and sample size is 100 then . Ten will be the interval size.

3.      Start from anywhere and select every 10^th member in the list e.g. if researcher start from 3 then next members will be 13 then next 23 and so on.

Stratified random sampling:

      In stratified random sampling population members are divided in different strata on the basis of homogeneity and then use sample and systematic random sampling technique for selection of individual from each stratum. For example if you have 100 students of three different classes like B.sc, M.sc and M.phil and you want to select 30 students for your sample. If you use simple random sampling and systematic random sampling techniques, then it is possible that you cannot give a representative sample. In order to avoid from this unconvinced you divide the population in their respective groups or strata then select the sample from each stratum. There may be two situations in stratified sampling technique:

1.      Equal allocation

2.      Proportional sampling

Equal allocation will present when each stratum has equal members. For example in classes B.sc, M.sc and M.phil have equal students. Supposedly, each class has 60 students. If sample is 30 then 10 students will be selected from each class.

Proportional sampling situation will be when the strata's are not equal, when each stratum has different population size e.g. if B.sc class has 60, M.sc has 80 and M.phil class has 40 students then we select members as proportionally.

Sample size=30           population=180           layer size (size of each class: b.sc=60, m.sc=80 & m.phil=40)

Students from B.sc class will be:

Students from M.sc class will be:

Students from M.phil class will be:               

Cluster sampling:

Cluster sampling technique is used in the case of very large population when it is impossible for the researcher to access or meet every element in the population e.g. whole city or country.

For example members of our population are the students of philosophy in Pakistani universities. So each university is a cluster and after location of these cluster, researchers use random sampling in order to select the sample.