Sampling Methods Final

7/31/2019 Sampling Methods Final

1/37


2/37

Sampling

0 Sampling is the strategy of selecting a smaller section of the population that will

accurately represent the patterns of the target population at large.

0 In research terms a sample is a group of people, objects, or items that are taken

from a larger population for measurement.

0 The sample should be representative of the population to ensure that we can

generalize the findings from the research sample to the population as a whole.


3/37

0 Sometimes, the entire population will be sufficiently small, and when the

researcher includes the entire population in the study. This is called a

census study

0 Researchers usually cannot make direct observations of every individual in

the population they are studying. Instead, they collect data from a subset of

individuals a sample and use those observations to make inferencesabout the entire population

0 The researcher's conclusions from

the sample are mostly applicableto the entire population


4/37

Purpose Of Sampling

There would also be difficulties measuring whole populations

because: -

0 The large size of many populations

0 Inaccessibility of some of the population

0 Destructiveness of the observation

0 Accuracy in sampling


5/37

Sampling Process

Define thepopulation

Specify thesampling

frame

Specify thesamplingunit

Specifysamplingmethod

Determinesamplingsize

Specifysamplingplan

Select thesample


6/37

Sampling Methods

SamplingMethods

Non-ProbabilitySampling

ProbabilitySampling


7/37

Non Probability Sampling

0

It relies on personal judgment of the researcher rather than the chance toselect the sample

0 The researcher can decide which element to include

0 Any statistical method cannot be used to draw inference from this sample

0 It becomes essential in some situations

0Eg. Suppose we have to take a small sample from a big heap of coal. Wecannot make a list of all the pieces of coal. We have to use our judgment

0 The nonprobability sampling is also called non-random sampling


8/37

Divided Into

Non-ProbabilitySampling

ConvenienceSampling

JudgmentalSampling

Quota SamplingSnowfallSampling

PurposiveSampling


9/37

Convenience Sampling

0It is a method of choosing subjects which are available or easyto find. As the name implies, the sample is selected becausethey are convenient

0 The selection of the sample is left to the researcher

0 It is used in exploratory research where the researcher isinterested in getting an inexpensive approximation of thetruth

0 This nonprobability method is often used during preliminaryresearch efforts to get a gross estimate of the results, withoutincurring the cost or time required to select a random sample


10/37

0 Advantages:

Least expensive and least time consuming

Sampling units are accessible and easy to measure

0 Disadvantages:

It contains sample errors and there is a sampling bias andthat the sample is not representative of the entire population

0 Examples:

Use of students, church groups and members of anorganization

Mall intercept interviews without qualifying the respondentsTear out questionnaire included in magazines

People on the street interviews

Choosing five people from a class or choosing the first five

names from the list of patients


11/37

Judgmental Sampling

0The elements are purposively selected based on the judgmentof the researcher

0 The sample is drawn by a judgmental selection process

0 The process involves nothing but purposely handpickingindividuals from the population based on the authoritys orthe researchers knowledge and judgment.

0 It is usually used when a limited number of individualspossess the trait of interest. It is also possible to usejudgmental sampling if the researcher knows a reliableprofessional or authority that he thinks is capable ofassembling a representative sample


12/37

0 Advantages:

It is low cost, convenient and quick

The value depends on the researchers judgment

Useful when broad population inferences are not required

0 Disadvantages:

Sampling bias is there

The judgment of the researcher might not be reliable

0 Examples:

When a research has to be done on bicycle riding habits. Ajudgment sample can be taken in which schools and parks canbe the locations for research.


13/37

Quota Sampling

0 First Control categories or quotas are developed

Second Elements are selected based on convenience or

judgment

0

The samples are selected in such a way that the interestparameters represented in the sample are in the same

proportions as they are in the universe / population

0 Such control measures are taken to reduce the sampling errors


14/37

0 Advantages:

Sampling error is reduced

Samples selected are on criteria relevant to the populationQuick and Easy to relate

0 Disadvantages:

Incorrect information might lead to incorrect selection of thesample

Other sampling errors might be there which might not be

determined

0 Examples:

Quota sample on the test of a flavored tea, the parameters are

ethnic background, income bracket, age group and area


15/37

Snowball Sampling0 A researcher identifies one member of some population of interest,

speaks to him/her, then asks that person to identify others in thepopulation that the researcher might speak to. This person is thenasked to refer the researcher to yet another person, and so on.

0 It is very good for cases where members of a special population aredifficult to locate.

0 Eg: If you want to look at pattern of recruitment to a communityorganization over time, you might begin by interviewing fairly recentrecruits, asking them who introduced themto the group.

Then interview the people named,asking them who recruited them to thegroup.

0 A researcher is not sure who is inthe sample


16/37

Purposive Sampling

0 Subjects selected on the basis of specific characteristics or

qualities

0 Subjects selected for a good reason tied to purposes of

research

0 Eg. Users of a particular technology, Young mothers with

small children, members of a fan club etc.


17/37

Probability Sampling

0 Sampling units are selected by chance

0 The human mind has no control on the selection or non-selection of the units for the sample

0 Every unit of the population has known non-zero probabilityof being selected for the sample

0 The probability of selection may be equal or unequal but it ispossible to specify the probability

0 The advantage of probability sampling is that sampling errorcan be calculated

0 The probability sampling is also called the random sampling


18/37

Divided Into.

ProbabilitySampling

Simple RandomSampling

SystematicSampling

StratifiedSampling

ClusterSampling


19/37

Simple Random Sampling

0 Random sampling is generally the best and simplest way todraw a sample from a population

0 With random sampling, every member of the population hasan equal opportunity to be included in the sample

0 Every element is selected independently

0 It is equivalent to a lottery system


20/37

0 Advantages:

Easy to understand

0 Disadvantages:

a) Such a sampling frame is difficult to be madeb) Can be spread over a large area which makes it costly

c) Larger errors

d) Sample may or may not be relevant

0 Examples:

Selecting any name from a list


21/37

Systematic Sampling0 This method of sampling is at first glance very different from SRS.

0 In practice, it is a variant of simple random sampling that involves some

listing of elements - every nth element of list is then drawn for inclusion in

the sample

0 It starts from a random point

0 The nth element = Population Size

---------------------Sample Size


22/37

0 Advantages:Less costly and easier than SRSIt can be used without the knowledge of the elements

0 Disadvantages:

0 Sampling errors can take place. A wrong sample might be selected

0 Examples:

Say you have a list of 10,000 people and you want a sample of 1,000.

Creating such a sample includes three steps:

Divide number of cases in the population by the desired sample size.

In this example, dividing 10,000 by 1,000 gives a value of 10.Select a random number between one and the value attained in Step

1. In this example, we choose a number between 1 and 10 - say wepick 7.

Starting with case number chosen in Step 2, take every tenth record

(7, 17, 27, etc.).


23/37

Stratified Sampling

0 Population is divided into homogenous subsets before sampling, thendrawing a random sample within each subset

0 A stratum is a subset of the population that share at least one common

characteristic. Examples of stratums might be males and females, ormanagers and non-managers

0 The researcher first identifies the relevant stratums and their actual

representation in the population. Random sampling is then used to select a

sufficient number of subjects from each stratum


24/37

0 Advantages:

0 Reduces sampling errors

0 Disadvantages:

0 Limits the area of approach

0 Examples:

Suppose a study has to be made on spending according to age. Then, the

population will be divided into 18-34, 35-49, 50-64, 65 and above and then

separate samples are drawn from each group


25/37

Cluster Sampling

0 The target population is divided into mutually exclusive and collectivelyexhaustive subpopulation called clusters

0 Then a random sample of clusters is selected. For each cluster, either all the

elements are included or a sample of elements is drawn

0 If all the elements in each selected cluster are included in the sample, the

procedure is called one stage cluster sampling

0 If a sample of element is drawn probabilistically from each selected cluster,the procedure is called two way cluster sampling


26/37

0 Advantages:

0 Low population heterogeneity/ high population homogeneity

0 Low cost

0Disadvantages:Sampling error

0 Examples:

Suppose we consider a sample of 100 households to be selected fora personal interview from a city. The city would be divided intoblocks and 10 households from 10 selected blocks would beinterviewed.


27/37


28/37

Non sampling error

0 Sampling error arises from inaccurate sampling frame, dataclarification or verification methods, reporting or coding ofdata, and/or specifications

0 It may also arise from poorly designed survey questionnaires,improper sample allocation and selection procedures, and/orerrors in estimation methodology

0 Even if we study the population units under ideal conditions,

there may still be the difference between the observed value ofthe population mean and the true value of the population mean


29/37

Contd..

0

1. The units of the population may not be definedproperly

2. There may be poor response on the part ofrespondents

3. The data may not be collected properly from thepopulation or from the sample

4. Another serious error is due to bias


30/37

Contd..

Non-Sampling

Errors

RandomErrors

SystematicErrors


31/37

Random errors

0 They are the unpredictable errors resulting from estimation

0 They are generally cancelled out if a large enough sample is use

0 However, when these errors do take effect, they often lead to

an increased variability in the characteristic of interest(i.e., the

greater the difference between the population units, the larger

the sample size required to achieve a specific level of reliability)


32/37

Systematic errors

0 These errors tend to accumulate over the entire sample

0

For example, if there is an error in the questionnairedesign, this could cause problems with the respondent's

answers, which in turn, can create processing errors, etc

0These types of errors often lead to a bias in the finalresults


33/37


34/37


35/37

T f E E l H i i i i


36/37

Type of Error Example How to minimize it

Random errorsYou measure the mass of a

ring three times using thesame balance and get

slightly different values:

17.46 g, 17.42 g, 17.44 g

Take more data. Random

errors can be evaluated

through statistical analysisand can be reduced by

averaging over a large

number of observations.

Systematic

errors

The cloth tape measurethat you use to measure

the length of an object had

been stretched out from

years of use. (As a result,

all of your lengthmeasurements were too

small.)

Systematic errors aredifficult to detect and

cannot be analyzed

statistically, because all of

the data is off in the same

direction (either to highor too low). Spotting and

correcting for systematic

error takes a lot of care.


37/37

Thank You..!!!

0 Aditya Rane 06

0 Ambreen Panjwani 12

0Chetan Tatia 18

0 Deepika Pathak 24

0 Hardik Dave 30

0 Imran Malik 36

0 Kashif Qureshi 42

Documents

Sampling Methods Final