Types of Probability sampling

1. Simple random sampling:
2. Systematic sampling
3. Stratified sampling
4. Cluster sampling
5. Area sampling
6. Multistage sampling @ MuCAS

a. Simple random sampling

• Complete random method of selection of selection.
• Is an easy as assigning number to the individuals.
• Applicable when population is small, homogenous and readily available.
• Each element has equal probability of selection.
• Lottery system is used to determine which units are to be selected.
• Estimates are easy to calculate.

Disadvantages of simple random sampling:

• If sampling frame large, this method is impracticable,

b. Systematic sampling

• Here, you choose nth individual to be part of the sample. i.e 3^rd , 6^th, 9^th or 2^nd , 4^th, 6^th , 8^th , etc.
• Sample is selected at regular periods.
• Equal opportunity for every sample to be selected.
• Relies on arranging the target population into ordering scheme.
• Involves a random start and continues with K^th

Advantages:

• Easy to select.
• Suitable sampling frame can be identified easily.

Disadvantages:

• Sample may be biased.
• Difficult to assess precision to estimate from one survey.

c. Stratified sampling

• Larger population can be divided into smaller groups.
• Usually don’t overlap and represent entire population.
• Eg: Male-Male, Female-Female, teenager-teenager, Old-Old, etc.
• Every unit has same chance of being selected.

Advantages:

• Every unit has same chance of being selected.
• Proportionate representation of sample.
• Adequate representation of minority subgroups.

Disadvantages:

• Sampling frame of entire population has to be prepared separately.
• Complicate design to examine multiple criteria.
• Requires larger sample than others.

d. Cluster sampling:

• Way of randomly selecting participants when they are geographically out.
• Clusters selected by dividing the greater population into various small sections.

Advantages:

• Cuts down the cost of preparing a sampling frame.
• Reduce travel and administration cost.

Disadvantages:

• Sampling error is high.

Types of cluster sampling

a. One-stage sampling: All of the elements within selected clusters are selected.

b. Two-stage sampling: Subsets are randomly selected for inclusion.

i) Simple cluster sampling

ii) Probability proportionate to size (PPS) sampling

c. multi-stage sampling:

• A complex form of cluster sampling.
• Random number of samples are taken of various niche ( Districts, Municipalities, Village, House).
• All ultimate units selected at last step are then surveyed.

Steps of cluster sampling

• An example of ‘two-stage sampling’.
• Area is chosen.
• Respondent within the area is selected.
• Population divided into clusters of homogenous units.
• Sample divided into groups rather than individual.
• Sample of clusters then selected.
• All units from selected clusters selected.

Post-stratification

• Done at the end of sampling phase.
• Implemented due to lack of knowledge at the beginning of the stratifying variable.
• Helps improve the precision of the sample’s estimate.

Over sampling

• Model is built on biased sample.
• Estimation is done with more precision.

PPS ( Probability proportionate to size) sampling

• Larger clusters have more probability of getting selected.
• We can use both simple random sampling and systematic sampling to draw clusters from the population.
• Sample size for each cluster is same.
• PPS approach can improve accuracy for a given sample size by concentrating sample on large elements that have greatest impact on population estimates.