Edited by Kate Findley and proofread by Angela Shoemaker, Wondrium Daily
Most people assume that impressive levels of performance are associated with talent. However, new data suggests that aptitude tests such as the SAT are not as accurate at predicting success as practice. Dr. Vishton explains.
Talent or Practice?
Is talent or practice a better indicator of success? That is the age-old question.
Indeed, some people are innately good at certain things. It seems that they were born that way, or at least born with the physical and/or mental capabilities that enabled them to quickly reach an exceptional level of performance.
If they’re children, we sometimes refer to these super-talented people as prodigies. Certainly, congenital abilities matter. It’s not a mistake that the average NBA player is much taller than the average person, or the average world-class marathon runner tends to be relatively smaller than the average person and possess a high percentage of a particular type of muscle called slow twitch muscle fiber.
It may be that there are physical, genetically specified characteristics of neurons that are different among very talented people. Thus, there might be such a thing as brain talent, in the same way that we think about physical talent.
Power of Practice
Therefore, talent does matter to some degree, but a wide range of research suggests that talent matters much less than we think it does. Most people’s overall level of achievement and expertise in any particular domain is not predicted by any innate characteristic, physical or mental. It is predicted by the type of practice and the level of experience and training that they have.
If there’s something that you want to excel at, studies of expertise suggest that you can. You should pick activities to pursue based on how much you like them, not on how good you are at them the first few times that you try them.
To develop expertise, be persistent. If you try something and it doesn’t go well the first few times, you shouldn’t use that to predict your potential ability in that realm.
Designing Aptitude Tests
Many studies are aimed at predicting how good someone will eventually be at some set of tasks. We usually call them aptitude tests. An IQ test is an aptitude test for later academic performance.
If you wanted to make your own IQ test from scratch, you would start by developing a list of questions that you think require intelligence in order to answer them well. You would then give that test to a large group of people.
Based on the data you collect, you would determine which of your questions correlate positively with academic performance—with grade point average (GPA), for example. If some of your questions aren’t correlated with school grades, then you get rid of those questions and come up with more. Eventually, you’ll have a test where you can take someone’s score and make a prediction about their ultimate academic GPA.
“I should say that, while it gets a bit more complicated, this is how every IQ test ever produced was developed—how they continue to be developed today, in fact,” Dr. Vishton said.
The SAT and ACT tests that high school students take every year are generated in exactly this fashion. For these tests, instead of high school GPA, the grades from the freshman year of college are used as the “to be predicted” variable. This basic method has been used frequently to try to assess people’s talent and aptitude for various skills: football success, and basketball performance.
In one study, the goal was to predict which people would become the best typists. We humans have a strong desire to predict the future, and this provides a tool for trying to do exactly that.
Do Aptitude Tests Work?
Inherent in the frequent use of this method is an implicit belief that’s not supported by modern research on expertise. The implicit belief is that the things that will make you an expert at something—or not an expert—are already present in you at the time of the initial aptitude test.
This implicit theory about talent and performance explains why we tend to use aptitude tests so often. However, in the multi-billion dollar industry that is aptitude testing, the correlations that are found are consistently very small.
The SAT predicts about nine percent of the variance in freshman college grades. That is, if you know a student’s SAT score, you’ll be about nine percent more accurate than if you just make a rote guess. The best versions of tests that attempt to predict even things as simple as typing aptitude only predict about four percent of the data.
Thus, when deciding whether talent or practice is a better indicator of success, these results demonstrate that talent is great, but talent does not provide a good predictor of overall performance.
This article was edited by Kate Findley, Writer for Wondrium Daily, and proofread by Angela Shoemaker, Proofreader and Copy Editor for Wondrium Daily.
Peter M. Vishton is an Associate Professor of Psychology at William & Mary. He earned his PhD in Psychology and Cognitive Science from Cornell University. Before joining the faculty of William & Mary, he taught at Northwestern University and served as the program director for developmental and learning sciences at the National Science Foundation.