In this article, we will cover 7 sampling methods, which we are going to divide into 2 types: probability sampling methods, and non-probability sampling methods.

Probability sampling methods involve random selection of participants, and therefore tend to produce unbiased samples; Non-probability methods do not involve random selection of participants, and therefore are cheaper to apply, but produce samples that are more prone to bias.

We will start by discussing 4 probability sampling methods:

And then 3 non-probability sampling methods:

## 1. Simple random sampling

### 1.1. How it works:

Step 1: Make a list of all *N* individuals in the population.

Step 2: Randomly select *n* participants to be included in the study (selection can be with or without replacement).

### 1.2. Advantage:

- It provides a simple and fair way of selecting the sample, since every person will have the same probability of being included in the study.
- Since it is based on randomness, it will produce an unbiased sample that is representative of the population.

### 1.3. Limitation:

- Randomly choosing a sample from a list of all individuals in the population requires us to have such a list, which is in general difficult to obtain.
- It can be very expensive and time consuming, especially when participants are geographically dispersed.

## 2. Systematic sampling

### 2.1. How it works:

Step 1: Select a random starting point *i*, and include the i^{th} person in the sample.

Step 2: Pick an interval *k*, then include every k^{th} person in the sample until you reach the sample size *n*. Those included will be the set: {*i, i+k, i+2k, …, i + (n-1)k*}.

### 2.2. Advantage:

- The sample is evenly spread across the population..
- Sometimes it is easier to apply than simple random sampling.

### 2.3. Limitation:

- A patterns may be present in the population making the sample biased.

## 3. Stratified sampling

### 3.1. How it works:

Step 1: Divide the population into subgroups/strata of individuals who share certain characteristics (like age, gender, income level, etc.).

Step 2: Calculate the number of individuals you want to sample from each subgroup/stratum (so that the proportion of each subgroup in your sample will match that of the population), and then apply simple random sampling within each stratum to build your sample.

### 3.2. Advantage:

- It forces the sample to resemble the population regarding certain attributes that may influence the outcome, and therefore, increases the generalizability of our results.
- It is mostly useful when we want to make sure that minority groups are represented in the sample.

### 3.3. Limitation:

- Sometimes the population cannot be divided into different strata.
- Creating stratified lists of individuals may be expensive.

## 4. Cluster sampling

### 4.1. How it works:

Step 1: Divide the population into clusters (for example, according to their location on the map).

Step 2: Apply simple random sampling to select a subset of clusters.

Step 3: Include all individuals from the selected clusters in your sample.

### 4.2. Advantage:

- Cheaper and more efficient to apply when dealing with a geographically dispersed population.

### 4.3. Limitation:

- The sample will be biased if the clusters do not represent the population.

## 5. Convenience sampling

### 5.1. How it works:

Select individuals that are the easiest to locate or contact.

### 5.2. Advantage:

- Cheap, time saving, and simple to implement.

### 5.3. Limitation:

- Often leads to selection bias and thus the study results will not be generalizable.

## 6. Quota sampling

### 6.1. How it works

Step 1: Divide the population into subgroups/strata of individuals who share certain characteristics (like age, gender, income level, etc.).

Step 2: Calculate the number of individuals you want to sample from each subgroup/stratum (so that the proportion of each subgroup in your sample will match that of the population), and then apply convenience sampling within each stratum to build your sample.

### 6.2. Advantage:

- Compared to stratified sampling, quota sampling is cheaper, faster, and easier to apply because it does not require random sampling.
- Compared to other non-probability sampling methods (convenience sampling and network sampling), this method ensures representativeness of all subgroups (especially minority groups).

### 6.3. Limitation:

- Since this method is not based on random selection of participants, selection bias is possible and the sample representativeness is not guaranteed.

## 7. Network sampling

### 7.1. How it works

Step 1: Identify few initial participants to be included in your study.

Step 2: Ask these participants to refer others to be included in the study. (Repeat this step many times until you reach the target sample size).

### 7.2. Advantage:

- Useful when individuals in the population are difficult to identify (e.g. drug users).

### 7.3. Limitation:

- Selection bias is possible since the sample is not selected at random.
- May be slow because it relies on the referral of existing participants to build up the sample.

## References

- Berndt, A. E. (2020). Sampling Methods.
*Journal of Human Lactation*,*36*(2), 224–226. https://doi.org/10.1177/0890334420906850 - Elliott, M. R., & Valliant, R. (2017). Inference for Nonprobability Samples.
*Statistical Science*,*32*(2), 249–264. https://doi.org/10.1214/16-STS598 - Sharma, G. (2017). Pros and cons of different sampling techniques.
*International Journal of Applied Research*,*3*(7), 749–752.