External Validity: Definition, Types, Threats & Examples

What Is External Validity: Definition

First and foremost, what is external validity in research? When you understand its fundamental principles, it will be easier to get externally valid data for study results in the future. It helps you define whether you can apply your research results to real-world examples. 

External validity is the extent to which you can expand your results for other target audiences, circumstances, or settings. Unlike internal validity, it deals with generalizability of study outcomes. In other words, the external validity of a study is the ability to use your insights from quantitative or qualitative research to make a conclusion for the broader public. 

To achieve high external validity, researchers must meet 2 criteria:

1. Sample should be representative of the population.
2. Methods and measures are appropriate for the research question.

Why Is External Validity Important?

Another essential question is why you even need to care about it. You have research questions, and you have your answers. But this is not a goal for almost all academics. Your aim is to change this situation and help essential institutions to be more productive. 

You need to ensure external validity for valuable research to achieve such goals:

• Ensure that findings can be applied to similar populations or circumstances
• Make your study more valuable
• Extend future research
• Make a connection between scientific findings and real-world issues
• Complete a stronger study.

External Validity Example

To understand the transferability of your research results better, and go deeper into different types of validity, let’s start with a basic example. 

Here you can see a common example of external validity, and next, we will focus on its types and specifics.

Example 

Types of External Validity 

There are 2 main types of external validity that you may face while working on research:

• Population validity
• Ecological validity.

In the following paragraphs, we will briefly discuss each type and share examples of external validity.

Population Validity

Population validity means that you can generalize your findings from your smaller samplings to a larger group of people with the same characteristics. Your results may be limited to the population you are researching in case of applying a non-probability sampling method. However, in some situations, you can easily extend your results to bigger groups of people.

Example of Low Population Validity

Analysis

Ecological Validity

Another type of external validity is ecological validity – the extent to which your result can be generalized for other circumstances or environmental conditions. In other words, can we make a conclusion and apply insights to the same sampling but in other situations? This type of external validity is very similar to population validity, but instead of examining different groups of people, it examines different settings.

In some cases, your study results can not be applied to all settings, as the experiment was conducted in quite specific circumstances (e.g.,using a computer instead of a simulation). We would call it low ecological validity.

Below, you will find an exact example of ecological validity. 

Example of Low Ecological Validity

Analysis

How to Establish External Validity

It can be challenging for some research to ensure its external validity. Scientists need to understand all details of assessing external validity and examine all options for improving it. The only way to ensure that your results can be used for solving real-life problems is to repeat your study for other target audiences or use other settings. You may increase the validity by expanding the criteria for your samplings. However, before you try to enhance validity, you need to figure it out and define if you have any problems.

How to Increase External Validity?

As we have already mentioned, increasing external validity can be challenging and require a lot of additional research work. However, if the purpose of the study is to expand the insight for a broad audience, you need to focus on how to improve the external validity of the research. 

Here’s how to improve external validity and make it more applicable for real-live:

1. Use random selection for your samplings if it’s possible.
2. Bring more focus on researching the group for your samples, and outline similarities of different groups that are a part of your work.
3. Use conceptual framework to understand the degree of similarities between target groups.
4. Conduct your study with different groups of people in various circumstances and time slots.

Threats to External Validity

Finally, we came to the most important part of your research – identifying threats to external validity 

External validity can be compromised by various extraneous variables that may influence the causal relationship. These variables may include external factors such as the research setting, participant characteristics, and the method of data collection.

Your opponents can use external validity threats to make your work weak. That’s why you need to know all possible threats to be able to avoid them.

1. Sampling Bias

One of the common threats to external validity is when research samples are not representative and do not illustrate the whole population you want to generalize the study. In other words, your sampling group includes only one type of people but not the average data about the population you are looking for. 

Example

2. History Threat

Other external threats to validity are some events that occur during your tests or gathering samples that affect study results but are not directly related to your study. It can be natural disasters or political rallies that make your target audience answer in a different way. As a result, your data can be compromised. 

Example

3. Observer Bias

One common threat to external validity is how the behavior or other characteristics of the person who is conducting research influence your results. For instance, people often react to an instructor instead of research questions during focus groups.

Example

4. Pre- and Post-Test Effects

Pre and post-test communication also threatens the external validity of a study. In other words, how your research team or instructor organizes pre and post-test directly influences how people behave during the survey. For instance, how stressed they will be about tests or experiments. 

Example

5. Hawthorne Effect

One more threat that affects external validity is the Hawthorne effect. It stands on the point that people can change their behavior if they know that they are part of an experiment or study. People usually try to behave better than usual in case they know that researchers observe this behavior.  

Example

6. Novelty and Disruption Effect

In some cases, researchers can’t apply external validation of their study. For example, they are not sure if the results depend on the novelty of the treatment. Or if any study disruptions affect the insights of the study. This is often a threat in medicine and ecology studies. 

Example

7. Aptitude Treatment

This is one more threat to external validity! It is also possible that group characteristics, together with some individual variables, influence your dependent variable. That is why the insights from the study can be criticized and cannot be expanded to a broader audience or different circumstances. 

Example

8. Situational Factors

There are a lot of factors, like day time, weather, specific season, or other circumstances, that can undermine your study outcomes. In this case, the scientists should ensure that all independent variables are analyzed and nothing can influence the final results. However, this thread is often used to manipulate data. 

Example

How to Minimize Threats to External Validity?

After you are aware of key threats to your research, you also need to know how to counter threats to external validity. We would say that the essential part is to understand all the threats and try to prevent them. However, there are a few things to consider for minimizing threats:

• Random sampling By selecting participants randomly from the target population, you can ensure that no participants are underrepresented.
• Increase sample size The larger your sample size, the more likely it is that your outcomes will be accurate.
• Diverse samples Try to recruit participants from various backgrounds to investigate multiple perspectives.
• Naturalistic settings Use natural context for experiments you are going to conduct.
• Replication By conducting the same study several times with different samples, researchers can determine if findings remain consistent.

Bottom Line on External Validity

External validity is referred to the ability to extend the research results to a broad audience or various circumstances. It helps scientists to understand how their work can be applied to solving real-life problems. This is why every researcher needs to understand its importance and know how to improve the data and results. In this text, we also outlined key threats to external validity and focused on some steps that can help you to minimize them. 

However, every time you are launching research, you need to start with your aims. After you understand it and have a clear vision of what you are going to study, you can easily navigate external validity and make your findings applicable to broader cases.