Upload
shashankgulati30
View
226
Download
0
Embed Size (px)
Citation preview
8/10/2019 3.2 Types of Sampling
1/16
8/10/2019 3.2 Types of Sampling
2/16
PROBABILITYSAMPLING DESIGN
1. Simple probability
sampling2. Stratified random
sampling
3. Systematic sampling
4. Cluster sampling/multi- stage sampling
NON-PROBABILITYSAMPLING
1. Deliberate sampling
2. Quota sampling3. Convenience
sampling
8/10/2019 3.2 Types of Sampling
3/16
Probability or random sampling gives allmembers of the population a known chanceof being selected for inclusion in the sample.
Selection of the specific units in the sampledepends entirely on chance.
It is possible to measure the sampling errorand hence we know the degree of precision.
8/10/2019 3.2 Types of Sampling
4/16
1. Simple random Sampling :This is the ideal
choice as it is a perfectrandom method.
Using this method, individuals are
randomly selected from a list of thepopulation and every single individual has
an equal chance of selection.
This method is ideal, but if it cannot beadopted, one of the following alternatives
may be chosen if any shortfall in accuracy.
8/10/2019 3.2 Types of Sampling
5/16
8/10/2019 3.2 Types of Sampling
6/16
Stratified sampling methods also come intwo types proportionate anddisproportionate.
In proportionate sampling, the stratasample sizes are made proportional to the
strata population sizes.For example if thefirst strata is made up of males, then asthere are around 50% of males in the UKpopulation, the male strata will need torepresent around 50% of the total sample.
In disproportionate methods, the strataare not sampled according to thepopulation sizes, but higher proportionsare selected from some groups and notothers. This technique is typically used in anumber of distinct situations
8/10/2019 3.2 Types of Sampling
7/16
Systematic sampling is a frequently usedvariant of simple random sampling. Whenperforming systematic sampling, every kthelement from the list is selected (this is
referred to as the sample interval) from arandomly selected starting point. Forexample, if we have a listed population of6000 members and wish to draw a sampleof 200, we would select every 30th (6000
divided by 200) person from the list. Inpractice, we would randomly select anumber between 1 and 30 to act as ourstarting point.
8/10/2019 3.2 Types of Sampling
8/16
The one potential problem with this
method of sampling concerns the
arrangement of elements in the list.? If
the list is arranged in any kind of ordere.g. if every 30th house is smaller than
the others from which the sample is being
recruited, there is a possibility that the
sample produced could be seriouslybiased.
8/10/2019 3.2 Types of Sampling
9/16
Cluster sampling is a frequently-used, andusually more practical, random samplingmethod. It is particularly useful insituations for which no list of the elementswithin a population is available and
therefore cannot be selected directly. Asthis form of sampling is conducted byrandomly selecting subgroups of thepopulation, possibly in several stages, itshould produce results equivalent to asimple random sample.
8/10/2019 3.2 Types of Sampling
10/16
The sample is generally done by firstsampling at the higher level(s) e.g.randomly sampled countries, then samplingfrom subsequent levels in turn e.g. within
the selected countries sample counties,then within these postcodes, the withinthese households, until the final stage isreached, at which point the sampling is
done in a simple random manner e.g.sampling people within the selectedhouseholds. The levels in question aredefined by subgroups into which it isappropriate to subdivide your population.
8/10/2019 3.2 Types of Sampling
11/16
Cluster samples are generally used if: - No list of the population exists.
- Well-defined clusters, which will oftenbe geographic areas exist.
- A reasonable estimate of the numberof elements in each level of clusteringcan be made.- Often the total sample size must befairly large to enable cluster sampling tobe used effectively.
8/10/2019 3.2 Types of Sampling
12/16
The probability of in-decision of any
unit (of population) in a sample is
not known.
The selection of units within a
sample involves judgment rather
than pure chance.
The degree of accuracy is notknown.
8/10/2019 3.2 Types of Sampling
13/16
1. Deliberate/ purposive sampling: this is also called
judgement sampling. The investigator excercises hisdiscretion in selecting sample observations from theuniverse. As a result, there is an element of bias in theselection.
8/10/2019 3.2 Types of Sampling
14/16
2. Quota sampling: In a Market Research context, themost frequently-adopted form of non-probabilitysampling is known as quota sampling.? In some waysthis is similar to cluster sampling in that it requiresthe definition of key subgroups. The main difference
lies in the fact that quotas (i.e. the amount of peopleto be surveyed) within subgroups are set beforehand(e.g. 25% 16-24 yr olds, 30% 25-34 yr olds, 20% 35-55yr olds, and 25% 56+ yr olds) usually proportions areset to match known population distributions.Interviewers then select respondents according to
these criteria rather than at random. The subjectivenature of this selection means that only about aproportion of the population has a chance of beingselected in a typical quota sampling strategy.
8/10/2019 3.2 Types of Sampling
15/16
3. Convenience sampling: a sample
obtained from readily available
lists such as telephone directories
or other prepared populationstatements is called convenience
sample.
8/10/2019 3.2 Types of Sampling
16/16
Technique
Strengths
Weaknesses
Nonprobability SamplingConvenience sampling
Least expensive, leasttime-consuming, mostconvenient
Selection bias, sample notrepresentative, not recommended fordescriptive or causal research
Judgmental sampling Low cost, convenient,not time-consuming
Does not allow generalization,subjective
Quota sampling Sample can be controlledfor certain characteristics
Selection bias, no assurance ofrepresentativeness
Snowball sampling Can estimate rarecharacteristics
Time-consuming
Probability samplingSimple random sampling
(SRS)
Easily understood,results projectable
Difficult to construct samplingframe, expensive, lower precision,
no assurance of representativeness.Systematic sampling Can increase
representativeness,easier to implement than
SRS, sampling frame notnecessary
Can decrease representativeness
Stratified sampling Include all importantsubpopulations,
precision
Difficult to select relevantstratification variables, not feasible tostratify on many variables, expensive
Cluster sampling Easy to implement, cost
effective
Imprecise, difficult to compute and
interpret results