What is External Validity? (+3 Common Threats)

External validity is determining how much your findings in research apply to the “real world”, or in practice.

This practice was born out of the idea that results should be able to be widely understood by the majority of the population.

A study that is externally valid will not only have followed all of the steps needed to produce an accurate study result, but it will also mention any limitations or biases found during the conduct of the study.

External validity ensures that the research we do is effective and applicable to a large population.

In order to better understand external validity, let’s consider an example where someone is conducting research.

Imagine that the research is not sure how valid the outcomes of the study are.

This would be an instance in which a study has very low external validity. People will often look to see if your research is trustworthy, has meaning and can be applied to a broader population.

If it does not have those aspects, then it will not have external validity. 

In psychology, external validity also means that your research can be understood and explained by people who work in a variety of fields.

There are three areas in which your study should be able to be generalized: across populations, settings and times.

If your study is generalizable across all three of these areas, then you will have an externally valid study.

When your study is generalizable across populations, this means that your study can be applied to a larger sample size and not just the specific group that participated in or is targeted by your experiment.

Generalizability across settings means that your study can be conducted in a variety of settings, and not just the setting in which you conducted your experiments.

A study can be viewed as generalizable across time if your research can be conducted during any time of day or year, and not just at the point in time when you conducted the study.

External validity is not possible unless you perform multiple studies to confirm that your results are in fact externally valid.

Repeating your experiment multiple times shows that the results you got the first time you did the experiment did not merely happen by chance.

This is common across several scientific disciplines-researchers often repeat their experiments to ensure that it is as accurate as possible and that no errors occurred during the first experiment.

If there are any uncertainties about the study, then repeating your experiments is the best way to make sure that your study has a chance at being externally valid.

When you discover that your findings are in fact accurate, you will be able to defend your own research when you present it to other people.  

External validity is such an important concept in research because it ensures that other research occurring simultaneously is also being done as accurately as possible.

Consider an example where a researcher could have many ideas that they would like to study and prove true.

However, over the life cycle of the study, the researcher slowly encounters more and more difficulties and realizes his study is only applicable to a very small group.

Additionally, the researcher now has a harder time than ever defending their work.

This study, which is not externally valid, immediately loses ground in the professional realm because the results do not apply to different populations, settings and points in time. 

Although it would be ideal to have an externally valid study from the beginning, this doesn’t mean that we can’t improve the external validity of the study.

Thankfully, there are several methods that can help improve the external validity of a research study.

For example, if your study has a large data set, then it might be helpful to make sure that the data you have corresponds to the population that you are studying.

Another possibility to consider is that you can break up the population into groups, which may provide external validity within the study that you are conducting.

Sometimes, we cannot achieve the sample size or population that we are looking for when conducting experiments.

Grouping your population is one way to make sure that your study could potentially have some generalizability.

It’s possible to group certain populations based on characteristics such as demographics, geographic location or religion.

Within these groups, we can create a scale that determines what similarities and differences exist in our data sets and in our smaller groups.

There is not just one but two types of validates; Internal validity and External validity, each having more sub-types.

An example of External validity can be ecological validity and for Internal, face validity.

Threats to External Validity

There are a few study limitations that can threaten the external validity of a study. These are often known as different types of biases.

A threat to external validity means that the generalizability of a study is not as great as we would like it to be and that the study results/outcomes are not as meaningful as researchers had hoped.

Your results may not be relevant to your study population, and therefore won’t be applicable to a larger study group.

The following biases are known to pose some of the greatest threats to the external validity of a study:

 Selection bias:

No two individuals are exactly alike in this world. This fact is especially important to keep in mind when you are selecting your study population for your research.

If all of your subjects do not represent the population that is going to be analyzed, then your study is likely to be a victim of selection bias.

Without accounting for selection bias, such as controlling for certain factors that may differ among individuals, the study will not render the most accurate results possible.

If people volunteer to participate in the study, it is extremely important the researchers stratify them in the most appropriate way possible so the study group does not experience selection bias. 


Another aspect of a study that can threaten its external validity is if confounding is present.

Confounding occurs when a third variable affects the relationship between the dependent and independent variable.

If a confounding variable is present, then the study results will be skewed and rendered inaccurate.

Confounding can threaten external validity because the results are not completely representative of the relationship between X and Y.

Instead, the study shows results of the relationship between X and Y when Z is involved.

This will threaten your study’s external validity – therefore, it’s important to keep in mind when confounding might be present in your experiment. 

The real world versus the experimental world:

At times, the environment that we set up for our experiment doesn’t mirror the world in which we would like our research results to be applicable.

It’s important to keep in mind the differences between the experimental world and the population at large.

It’s also important to make sure that individuals who can withstand experimental conditions will also be able to withstand study conditions outside of the bounds of the research being done.

Ensuring that the environment in which you conduct your experiment is as close as possible to the environment in which you would like your study results to be most generalizable is incredibly important. 

Time and external validity:

Time also plays a critical role in determining external validity. When was your experiment conducted?

Was it done at a specific point in time?

Understanding this information and how it might affect your research is extremely important in determining the results of the study.

If the timing of your study has an impact on your results, it’s important to keep this in mind when conducting your analysis and understanding your study limitations. 

In short, external validity plays a very important role not just in research but also in our world.

Several experiments that have been conducted have externally valid results.

Thanks to this, we’re able to understand parts of our world more clearly and fully. 

FAQs on External Validity:

How can I improve the external validity of my study?

A common method that researchers use to improve external validity is the usage of inclusion/exclusion criteria.

This helps ensure that individuals with certain characteristics or conditions are enrolled safely in your study.

Furthermore, this will ensure that there are no confounding variables when you conduct your experiments. 

How is external validity different from internal validity?

While external validity is focused on how applicable your study is to the outside world, internal validity is focused on the accuracy of your research methods.

Internal validity. These studies also control for external variables and minimizes the likelihood that there are alternate explanations for this study.

Interested in Learning More? Check out these books on External Validity:

External Validity (Leadership Analysis Book 4)

Population-Based Survey Experiments

Critical Appraisal from Papers to Patient: A Practical Guide


  • En.wikipedia.org. 2020. External Validity. [online] Available at: <https://en.wikipedia.org/wiki/External_validity> [Accessed 17 April 2020].
  • Verywell Mind. 2020. Understanding Internal And External Validity. [online] Available at: <https://www.verywellmind.com/internal-and-external-validity-4584479> [Accessed 17 April 2020].

Was this helpful?

Thanks for your feedback!