# What is a sample in statistics? (Uses, methods and examples)

By Indeed Editorial Team

Published 15 June 2022

The Indeed Editorial Team comprises a diverse and talented team of writers, researchers and subject matter experts equipped with Indeed's data and insights to deliver useful tips to help guide your career journey.

In statistics, a sample is important for determining relevant information about groups of people. Trying to collect data on each individual in a study can be time-consuming and samples allow you to create more manageable data sets. Learning about the uses of sampling can improve your knowledge of the field and strengthen your skills as a statistician. In this article, we discuss a sample in statistics, explore key sampling methods and provide examples that can help you better understand this concept.

## What is a sample in statistics?

A sample in statistics is a small data set that you obtain from a larger set of data to represent a whole, for example, the entire population. Using samples is common among statisticians because it makes it easier to gather and analyse information. For example, if they'd like to determine how many people over the age of 60 in the country have pets, it would be difficult to ask every single senior about this. Instead, it's possible to gather data from a smaller number of people over that age and use it to make a reasonable assumption about the whole.

Related: 19 jobs that use statistics (Plus duties and salaries)

## Uses for statistical sampling

Statistical sampling is a process that both statisticians and other professionals use to back up their studies and learn more about a specific group of people, such as their clients. Here are some common uses for sampling:

Science: scientists use sampling regularly, as it allows them to gather information from small groups of people to assume what the entire population thinks about concepts such as global warming or water quality. Typically, they invest in more time-consuming sampling methods to make sure their results are as accurate as possible.

Marketing: businesses of all sizes can use sampling techniques to learn more about their customers or potential target markets. They can then use this knowledge to create more effective marketing campaigns.

Medicine: physicians and other medical professionals use sampling to conduct clinical trials. For example, it's common that they use it to determine the side effects of a new medication.

Related: What does a social researcher do? (Top duties explained)

## Common sampling methods

Depending on the purpose of your project and what you're analysing, there are several sampling methods you can choose. There are two key types of sampling methods, which are probability and non-probability sampling. Each has several subtypes that define what technique you'd use to approach gathering information, including:

### Four probability sampling methods

Probability sampling is a sampling approach where you, as the researcher, choose samples from a larger data set using the theory of probability. To consider a probabilistic sampling method, it's necessary to incorporate random selection into the process. The four key types of probability sampling are:

#### 1. Simple random sampling (SRS)

SRS is one of the most popular sampling methods which allows you to choose random selections. To use this method, you can use a lottery system or a number generating software. For example, if you're responsible for sampling from a large group of people, you can start by assigning each person a different number and inputting the range into the software. Then, the app would automatically select a few numbers randomly. People with those numbers are now the members that you can include in the sample.

Related: How to become a data analyst

#### 2. Systematic sampling

Systematic sampling is another probabilistic sampling method that involves selecting a sample group starting at a random point and then following fixed, periodic intervals. To determine your interval, you can simply divide the population, or your large data set, by the size of the sample you wish to create. For example, if the size of the population is 6000 and the size of your sample is 600, then you can calculate your sample interval like this:

Interval = size of the population / size of the sample

Interval = 6000 / 600

Interval = 10

#### 3. Stratified sampling

To perform stratified sampling, or random quota sampling, you'd start by dividing a large group of people into smaller groups. Your primary goal is to select groups that represent the entire population as accurately as possible without overlapping. For example, you can have several groups that represent people of certain ages, such as 0-18, 19-25, 26-39, 40-59 and 60-100. Then, you'd simply take a sample from each group separately.

Related: 10 common types of variables in research and statistics

#### 4. Cluster sampling

Cluster sampling requires you to take the entire population and divide it into smaller groups, often referred to as clusters. Then, you randomly select individuals from your clusters to create a sample. This probability sampling method is great for analysing large populations, as it allows you to use real-life concepts as clusters, such as cities or facilities.

### Four non-probability sampling methods

Non-probability sampling relies on subjective methods, rather than selecting samples randomly. It's a time-efficient and cheaper way of obtaining data to analyse, which means these methods can be great when you want to access data quickly or even in real time. Common non-probability sampling methods are:

#### 1. Convenience sampling

In convenience sampling, also called haphazard or accidental sampling, you select your sample based on what's convenient at the moment. This method involves no prior speculation or assessment of the population. The only criteria are that you create the sample based on the ease of obtaining participants. For instance, you might even ask your friends to participate in the study because they're available and wouldn't ask for a payment, which makes selecting them convenient.

#### 2. Quota sampling

Quota sampling involves selecting your sample non-randomly, which means all members of the population have different chances of participating. For example, researchers might decide to choose their sample based on specific traits or personal qualities that only some members of the population have. There are two kinds of quota sampling:

Controlled sampling: controlled quota sampling is when there are specific limitations that researchers follow when selecting samples. For example, their employer may require that they only select people above the age of 60.

Uncontrolled sampling: in uncontrolled quota sampling, there are no limitations or restrictions that researchers follow. It means they can create the sample however they want.

#### 3. Judgement sampling

Judgement sampling, also known as purposive or authoritative sampling, requires you to select your sample based on your knowledge or personal judgement. It's an easy to conduct and quick non-probability sampling method that allows you to approach participants directly and obtain almost real time results, for example, as a result of a poll or survey. In many instances, this method can be highly accurate, which usually happens when there's a restricted number of individuals in a population who own qualities that you expect from the entire population.

Related: What does a research scientist do? (With responsibilities)

#### 4. Snowball sampling

Snowball sampling is a unique non-probability method in which you find some participants and they help recruit more people for the study. This is a great method for researching specific industries, markets or fields. For example, if the aim of your study is to determine how many career coaches experience stress before public appearances, it's only necessary that you contact a few. Then, your existing participants can use their professional networks to spread the word about the study because it's likely that they know more professional career coaches.

## Examples of statistical samples

Reviewing examples of some sampling methods can help you understand how you can use these techniques for your projects. Here are examples of how two common types of sampling work:

### Cluster random

A cinema wants to gather information about its customers' experiences. Using randomisation software, they select three random showings throughout the day of different films. They then ask the guests from these films to complete a survey as they exit. This method of sample collection is usually a reliable sampling technique that provides good enough results for marketing or business studies.

### Stratified sampling

A university wants to gather information about their students' preferences for upcoming holiday celebrations. To gather a fair representation of information from each course, university representatives survey 10% of students from each department. For example, because there are 5,000 business students, they survey 500 of them and do the same thing within the remaining departments. This ensures fair representation for all courses.

## Explore more articles

- Independent software vendor: definition and what they do
- Job sharing pros and cons: A guide to flexible working
- What is a burn up chart? (Definition, uses and benefits)
- How to write a template for a statement of work (with tips)
- What is paid search? (With benefits for SEO professionals)
- How to communicate a meeting's purpose: steps and tips
- What is a logical flow diagram? Definition, uses and example
- What is vector graphic software? With features and benefits
- A guide to organic traffic (and how to increase it)
- Why do peer relationships at work matter? (With tips)
- Dealer markets and how they differ from other markets
- How to do video advertising and what makes it important