Last Updated on: 29th August 2022, 08:16 am
Everybody knows external validity is important. With it, you’ll soar to the heights of research applicability. Without it, you’ll barely make it off the ground. But… what exactly is it?
Fear not: if you’re confused about what external validity is and how to achieve it, you’re in the right place. Or, if you already know what it is but haven’t the slightest clue how to actually put it into practice, we’ve got you covered too.
What Is External Validity?
When we consider external validity, we are asking whether the results of our study can be generalized beyond the scope of the study itself. Can the claim be realistically applied to larger populations, as well as to other times or situations?
When and How Is External Validity Important?
External validity is extremely important with frequency claims — studies that conclude how frequent or common something is. For example, “14% of College Students Consider Suicide” is a frequency claim. For us to take this claim seriously, we would need to know how they chose their study participants — did they ask a few students on the sidewalk? Did they ask 100 randomly chosen students across a variety of colleges?
Association claims — studies that claim that two things often occur together — also require interrogation of external validity. For example, “People Who Talk with Their Hands Are Often Warmer and Friendlier Than Those Who Don’t” is an example of an association claim. Do these results always hold true? Can a study of middle school girls in Connecticut that showed these results also hold true for a group of middle-aged men in California?
When saying one variable causes another (causal claims), we ask: to what populations, settings, and times can we generalize? For a study that claims that music lessons increase IQ, we would need to ask whether the results would apply in all cultures, for people of all socio-economic backgrounds, and at all ages. When making causal claims, however, interrogation is usually more rigorously focused on internal validity.
What’s the Difference Between External and Internal Validity?
Internal validity refers to the construction of the study and means that conclusions are warranted, extraneous variables are controlled, alternative explanations are eliminated, and accurate research methods were used.
External validity means the degree to which findings can be generalized beyond the sample, the outcomes apply to practical situations, and that the results can be translated into another context.
An Oregon State University researcher talks about the difference in practice:
In some more recent work, I was looking specifically at physical activity during pregnancy as the only exposure; thus, my advertisement to recruit women into the study mentioned that I was studying exercise in pregnancy (rather than pregnancy in general).[1] In this more recent study, I had very few sedentary people—indeed, I have a few who reported running half marathons while pregnant! Since this is not normal, my study—though it does have reasonable internal validity—cannot be generalized to all pregnant women but only to the subpopulation of them who get a fair bit of physical activity. It lacks external validity. Because it has good internal validity, I can generalize the results to highly active pregnant women—just not to all pregnant women.
Threats to External Validity
In order for your research to apply to other contexts (and thus be of use to anyone), it’s important to manage or eliminate threats to external validity. Here are some of the most common threats.
Sampling Errors
Polling during political campaign season offers a vivid example of the importance of sampling for predicting the behavior of a population: “Can we predict the results of the presidential election based on this sample of 1200 people?”
You’ll want to know if the sample comes from the population of interest, so you’ll want to define that first. Pollsters don’t ask children who they’d vote for, because children are not allowed to vote — they are not representative of the population of interest.
You’ll also need to be sure that the sample is representative of the population. Polls that sample only adult white males would be unlikely to accurately predict the election because they do not adequately represent the diversity of the voting population.
Special Circumstances
You also want to make sure major historical events do not influence the results of your study. For example, doing your study during a global pandemic. Does the pandemic have any influence on your results?
Or, if you planned to survey people over a time period (say, three months) and two months into the process, there’s a big change in society or within the profession, that might make your results have less external validity. Because some of the responses came before the major event and some came after, it’s hard to tell the impact the event had on the results.
How to Achieve External Validity
Larger, Representative Sample
Sample size is always a trade-off. Dedicating time and money needed to accumulate a large sample increases external validity, but a smaller sample allows for faster completion with fewer resources. Most people do a G*Power test, when doing a quantitative study, to determine the minimum sample size that’s needed to allow for external validity.
You want the sample to be large enough to sufficiently limit the influence of outliers. In a larger sample, someone offering unrepresentative views/behavior will not skew the results. For example, in a study where 4 people are asked how many doughnuts they can eat and 3 eat 2 doughnuts and 1 eats 34, the average is 10 doughnuts per person. However, if 1,000 people are surveyed and 999 eat 2 doughnuts and 1 eats 34 the average is 2.032. The outlier does not skew the results much.
Random samples are considered more valid than purposive samples, because you have a better chance of representing the population randomly than if you select who will be part of the study, or if they volunteer.
Replicability
If your study is set up so that others can repeat your study, then results have the potential to become much more externally valid. If others repeat the study and come up with similar results, that means your findings might actually be the case generally. If you replicate your own study under different circumstances or with different populations, you have more results to make your conclusion stronger.
Final Thoughts
External validity is an important concept to understand. A lack of understanding may doom your study, while strong external validity makes getting your dissertation accepted and/or article much easier to get published.