Home > Education, Memory > “Practice makes Perfect”–but what type of practice?

“Practice makes Perfect”–but what type of practice?

Students spend an immense amount of their time studying—after all, have you ever taken a class in hopes of failing the final exam?  Most students study class material before a test to try and avoid this fate.  Despite making an effort to do this, almost anyone who has ever been in school can recall a time when they spent hours, days, or even weeks studying, only to arrive at a test and find that they are unable to answer any of the questions.  While this can be very frustrating, it also shows that the amount of time you spend studying is not the only thing that determines how well you will do on a test.  In order to determine which other factors might play a part, psychologist Andrew Butler conducted a study at Washington University in St. Louis that looked at different studying techniques and how they affect test performance.  More specifically, Butler compared test-enhanced learning, which involves studying by being tested on the material at hand (like testing yourself with flashcards) to repeated restudying of information (picture yourself reading a textbook page over and over again).

In order to compare these two strategies, Butler asked participants to read six different passages about a variety of topics.  These participants then repeatedly restudied two of the passages, took the same test over and over again on another two of the passages, and repeatedly took different tests on the last two passages.  While the studying technique differed across passages, all six of them were studied for the same amount of time.  Then, one week later, the same participants took a final test on the readings.  Butler administered two versions of this final test; in one version the questions asked about information that was explicitly stated in the passages, while in the other one the answers had to be inferred from them.  For example, if a the passage stated, “there was snow all over the ground” an inferential question might ask, “in which season did this passage take place?”

For both the explicit and inferential tests, participants answered more questions correctly about the readings that they had been tested on during studying, rather than the ones that they repeatedly read over.  However, it did not make a difference whether or not they had been tested with the same questions repeatedly, or with a variety of different questions.  Interestingly, if people answered a question correctly while they were studying, they were also more likely to answer it correctly on the final test.  This suggests that not only is it helpful to be asked question during studying, but studying until you can correctly answer all of these questions correctly is critical.

Although test-enhanced learning certainly helped people recall information from passages, Butler wanted to see if this phenomena extended to learning material in other contexts.  For example, what if a student only wanted to learn isolated facts, like terminology for a spelling quiz?  To determine if test-enhanced learning leads to better test performance in other situations, Butler modified his original experiment by giving people isolated facts and concepts to learn, instead of information embedded in passages.  Despite this change, participants still did better on the final test if they were previously tested on the information as opposed to if they simply restudied it.

One criticism of test-based learning is that it only teaches people to “spit back” a specific response: sure, it helps you to re-iterate facts, but are you really gaining a better understanding of the material?  Butler also aimed to explore this facet of studying.  To do so, he had people to read passages and study them either by testing or repeated restudying (just like in the previous experiments), but the final test that they took consisted of inferential questions from different areas of knowledge than those of the passages.  This required them to apply material that they studied to never-before-seen questions about a new (but still related) topic.  This condition differs from the previous condition in which participants were asked inferential questions because they tested people’s ability to transfer their knowledge to a completely new knowledge domain, whereas previously the test questions were still related to the same topic as the study material.  For example, participants may learn that bats and birds have very different wingspans, and then be asked a week later, “How would an aircraft modeled after batwings differ from other aircrafts?”  Again, people were better able to answer questions correctly if they were tested than if they restudied the material.  In my opinion, this is the most interesting of all Butler’s findings because it shows that the benefits of testing go beyond recalling specific facts, and can help you apply the material you’ve studied to questions and related topics that you have not previously encountered.

Ultimately, there is very strong evidence that you can boost your test grade with methods as simple as using flash cards to quiz yourself.  Students should utilize such strategies regardless of the type of test that they’ll be taking; Butler’s experiments show that the benefits of test-enhanced learning extend to various types of assessments, from ones that require you to reiterate facts to inferential questions you have never seen before.  Test-enhanced learning is also a valuable study strategy because it allows students to study in a more time-efficient manner.  Instead of spending an extraordinary amount of time re-reading the material, you can test yourself until you feel comfortable with it.  This will save time and, in all likelihood, result in a better grade.


Butler, A. C.  (2010).  Repeated Testing Produces Superior Transfer of Learning Relative to Repeated Studying.  Journal of Experimental Psychology: Learning, Memory, and Cognition, 36(5), 1118-1133.

Can be easily accessed at: http://people.duke.edu/~ab259/pubs/Butler(2010).pdf

Categories: Education, Memory Tags: ,
  1. Kacie Wrean
    October 18th, 2015 at 13:24 | #1

    Some of the findings of this study seem like pieces of information that teachers and students have kind of figured out themselves. For example, the study showed that studying until you can answer the question correctly seems obvious when you think about it because when quizzing yourself, if you know you got it wrong, you know you’ll get it wrong in the future unless you put in the effort to learn the correct answer. Similarly, through all my years of test-taking I have encountered teachers who love the questions that ask you to apply what you know to a new idea or topic, whether it was in a math class or a history class. Teachers, though they are not scientists necessarily, often have to think the way scientists do to discover how best to teach a certain subject or how to help a particular student grasp a concept. And I think it makes sense that teachers should know the best ways to test whether a student knows the information and understands it enough to apply it to something new or different because they understand that what they are teaching is not straight facts but information that should be used in students’ lives long after they have finished being tested in school on the facts.

You must be logged in to post a comment.