# Sampling

**It is not possible to include each element (variable) of the population in an experimental study. To overcome this limitation, sampling techniques are used.**- The results or conclusions obtained by studying a sample are considered representative or applicable to whole population.
- The sampling techniques used are:
- Simple random sampling
- Systematic sampling
- Stratified sampling
- Multistage sampling
- Cluster sampling

- â€‹Different sampling techniques are used according to their appropriateness in the study and the population.

â€‹

**Sample: Essential characteristics****â€‹**Adequate size (30 or more)- Random selection

- â€‹
**Simple random sampling****Every unit of population has an equal chance of being selected****(AIâ€™09).**- It is applicable when the population is: Small, Homogenous, and readily available.
- Examples:
- â€‹Patients coming to hospital
- Ward patients

- â€‹â€‹
**Systematic sampling**- It involves choosing elements in a systematic way â€“ such as every fifth patient admitted to a hospital or every third infant born in an area after having selected the first number randomly. This type of sampling provides the equivalent of a simple random sample without actually using randomization.
- It is used when the population is:
- â€‹Large
- Scattered
- Not homogenous

- â€‹Popularly used in those cases when a complete list of population, from which sample is to be drawn, is available.

**Stratified sampling****â€‹**It is used when the population is: Not homogenous**Merit:**Greater representation of each strata of sample.**Method:**Population under study

**â**

**Division into homogenous groups**

**â†“**

**Sampling drawn from each stratum at random in proportion to its size.**- Example: in order to determine prevalence of hemoglobinopathies in India, the equal number of children from Hindus, Sikhs, Christians and Muslims need to be selected.

- â€‹
**Multistage sampling****â€‹**It is employed in large country surveys.**Advantages****â€‹**Greater flexibility- Saves extra labor

- â€‹â€‹
**Cluster sampling [most frequently asked and difficult to answer] (AIIMS Novâ€™08)****â€‹**It is used when units of population are natural groups or clusters such as villages, wards, blocks, slums, and school.- Cluster samples may be used when it is too expensive or laborious to draw a simple or stratified random sample; eg â€“ in a survey in medical students in United states, an investigator may start by selecting a random set of groups or clusters â€“ such as a random set of 10 medical schools in the USA.
- In cluster sampling the clusters are selected randomly, and all members in the cluster are sampled.
- This method is much more economical and practical than trying to take a random sample directly from the widely scattered population of all medical students in the United states.
- Cluster sampling gives a higher standard error but the data collection in this method is simpler and less expensive. Another disadvantage of this method of sampling is that inter-cluster comparison is not possible.
**Design effect**is a phenomenon seen with cluster sampling and is the reason why sample size in cluster sampling is larger than that in simple random sampling.**As per module approved by WHO, it is most often used to evaluate vaccination coverage in expanded programme of immunization or to find out missed cases in pulse polio immunization, a 30 / 7 (30 clusters with 7 subjects in each) cluster technique is used.**^{Q}

**â€‹**

__Sample size calculation__**In prevalence / Cross sectional studies**

= 4

__PQ.__

L

^{2}

Where

P = Prevalence of disease

Q = 1 - P

ZÎ±

^{2}= Level of significance

â†’ at 95% level

ZÎ± = 1.96

ZÎ±

^{2}= (1.96)

^{2}

= 3.84 â‰… 4.

L = Precision