Written by Dr. Althea Bauernschmidt
Now that the Spring 2018 semester is in full swing, CALM is back with a new series of blog posts exploring topics in attention, learning, and memory! This month we’ll be focusing on testing. Specifically we’ll be looking at retrieval practice, sometimes called the testing effect, as well as issues surrounding what makes a good test good (or a bad test bad). In this first introductory post I’ll be explaining retrieval practice.
We often get frustrated by everyday lapses in memory. Where did I park my car? What’s the name of the book my friend recommended me? When people find out that I study memory they almost always express that they wish they weren’t so forgetful. We get frustrated when we don’t have perfect recall of every aspect of our lives and assume that our memories are not working properly. As a memory researcher I can assure you that forgetting where you parked your car is not only normal – it’s actually a good thing. While it can be annoying, forgetting is actually a sign that our memories are working properly. Let me explain.
Our memories work by forming complex associations between cues in our environments and the information in our brains (Anderson & Schooler, 1991; Raaijmakers & Shiffrin, 1981). Cues are anything that triggers a memory. For example, seeing the words “peanut butter” might make you think of “jelly”*, or hearing a certain song might make you think of a movie. If we are studying for a test then we hope that the cues on the test (the questions) will make us remember the correct information. So what makes some cues better than others? Why does some information seem to be easier for us to remember than other information?
*For a certain set of older millennials you may have immediately thought of “peanut butter jelly time”. I’m sorry.
These questions can be better understood when we consider how memory actually works and what our memories are really designed to do. It can be tempting to think that memory is for remembering the past. On the surface the certainly seems reasonable. We use memory to remember the names of people we met at a work function, where we put our keys, and passages from our favorite books. But if we think of memory as primarily for remembering the past then we will be a bit let down by how often we forget the past. What’s the point of having this memory system if I have wander aimlessly around the parking lot looking for my car?? Memory researchers actually argue that memory is not for remembering the past; memory is for helping us predict the future (Szpunar & McDermott, 2008). Our memory system does this by taking into account complex patterns based on environment, context, and frequency. This information helps us generate the best possible information in response to a given cue (Karpicke, Lehman, & Aue, 2014). This theory of memory is supported by how we forget information. Information that doesn’t get used very often is forgotten – unless it is associated with an emotionally charged and therefore important event. Information that is used frequently is more easily remembered. Our memory systems have recognized that this comes up a lot in this context and make it easier to remember. The best predictor of the future is the past. Something that we encounter everyday is more likely to happen tomorrow than something that happens every few years.
This may feel a bit disappointing at first. After all, it’s frustrating when we can’t find our keys (or our car). Wouldn’t it be nice if we could always remember everything? Not so much. Consider for a minute what it would be like to remember everything all the time. Every time you saw or heard the word “key” you thought of every key you have ever encountered. You would need to consciously consider each option before deciding which key was most relevant to your situation. Instead of remembering where your current car keys are, you instead reminisce about the keys to your grandfather’s truck that were last seen on the refrigerator of the house you lived in 10 years ago. By conveniently forgetting most of the keys you have encountered, our memory system saves us time and energy. Your memory system helps remember the keys to your car and not your grandfather’s truck.
A simpler way of describing memory is that a memory, or an association between a cue and information, becomes better the more we use it. Practice makes perfect. Learning and memory researchers who study this process of cues activating information call it retrieval. Every time something is recalled and we go through this process of retrieval we get better at rebuilding the memory and make it easier to remember in the future (Roediger, McDermott, & McDaniel, 2011).
There are 3 important aspects of retrieval to keep in mind when trying to improve memory and learning. 1) Retrieval is cue-dependent, 2) Retrieval is a process that works best when spaced, and 3) Retrieval (and therefore testing) is NOT a neutral learning event.
Retrieval is cue-dependent.
Cue-dependent refers to the association between a cue in the environment and the information it is associated with. For example, if you were learning Japanese you would try to associate the cue “Ohayou” with “morning”. So the more times you are greeted with “ohayou!” and you recall that it means “morning!” the easier it will be for you remember that “ohayou” means “morning”. Because memory is cue-dependent that means it can be influenced by the context we are in when we encounter the cue. Often we are good at remembering something in one context, but struggle when we switch to a different context. For example, if art students are going to be tested on their knowledge of color theory and linear perspective by creating a piece of artwork, then they might struggle if they only memorized the definitions of color theory and linear perspective because the terms are being used in a different context. The cue-dependent nature of memory means that the best way for us to study and prepare for tests is to study in the way we expect to be tested. If our knowledge of art is going to be tested based on our ability to create certain types of art, then we should practice creating pieces with strong colors and the feeling of distance created by linear perspective. Not by memorizing the definitions of terms (though knowing the definitions will certainly be helpful).
Retrieval is a process that works best when spaced.
Take a moment to consider the following math problem:
3 x 2 + 7 = ?
If you are a bit rusty on your algebra you might take a second to remember what the order of operations is. You might then feel a small sense of accomplishment when you pull PEMDAS out of some corner of your mind. Then, armed with confidence that you can solve this math problem, you multiply 3 x 2 and get 6. Then you add 6 and 7 to get 13. Now that you have brushed up on your algebra skills, I invite you to attempt a second problem:
3 x 2 + 7 = ?
Was this second problem easier to solve than first? I’m willing to bet you were much faster at solving the second problem than the first problem. But did you apply the same processes that you did while solving the first problem? Probably not. You didn’t have to go through the same steps because you just solved the exact same problem. The same thing happens when we practice retrieval. When we retrieve information from memory we go through a number of steps similar to solving the math problem. If we immediately retrieve the same information from memory we don’t have to go through the same process again. But the more times we retrieve that information, the better we will be at remembering it in the future (Roediger, McDermott, & McDaniel, 2011). So the best thing we can do to improve our memory is to space out our studying, or retrieval practice (Karpicke & Bauernschmidt, 2011; Benjabmin & Tullis, 2010). This is a complicated way of saying what teachers already know: cramming doesn’t work for the long term. Cramming the night before, or minutes before might give you a temporary boost in memory, but spacing out your studying and preparing in advance gives you a much stronger grasp of the material.
Retrieval (and therefore testing) is NOT a neutral learning event.
Whether we’re learning Japanese, practicing for an art test, or solving math problems, we are using the active process of retrieval to make gains in learning. Anytime we retrieve a piece of knowledge we change our ability to recall that piece of knowledge in the future. Tests, quizzes, practicums, etc. can asses learning by asking students to retrieve knowledge to recognize a correct answer, produce a correct answer, or demonstrate a skill. However, this also means that tests, quizzes, practicums, etc. also changes their ability to retrieve that knowledge to recognize a correct answer, produce a correct answer, or demonstrate a skill. Every retrieval attempt (especially when they are spaced out) improves the association between a cue and the information in memory. This has some really straightforward and practical implications. If we want to get better at a certain skill, then we need to practice that skill. If we want students to learn basic skills because we feel they are useful in the real world, then we should have them practice those skills in real-world scenarios. For example, basic math is useful for keeping track of personal finances. One way to have students practice math in finance is to set up a small token economy in the classroom where they may earn tokens for good behavior and grades and then spend them to receive prizes and rewards. Students can keep track of their tokens in a checkbook. Activities like this are beneficial because they are more engaging than completing worksheets, but also because they take away some of the challenges that students have when they need to transfer skills they learned in one context (math problems in class) to another context (balancing a checkbook/tracking finances).
Before I finish talking about retrieval, I want to stress that testing is just one way that you can use or practice retrieval to help learning. Tests are often thought of as high-stakes, state mandated, and very limited ways of assessing students. By advocating for retrieval practice I am not necessarily advocating the use of more of these types of tests*. I am advocating for more retrieval practice. Retrieval practice could come in the form of an engaging classroom discussion where students have to use their knowledge of a text to argue for or against some stance. Retrieval practice could come in the form of review quiz games where students play on teams and compete for bonus points. Retrieval practice could come in the form of practicing skills, like in the example of creating artwork above. Retrieval practice can take many forms in and out of the classroom. As long we are mindful of what and how often we are practicing retrieval we’ll see substantial gains in learning and memory in and out of the classroom.
*In a second post later this month I will explore some the issues surrounding these types of tests and explain why, somewhat paradoxically, that means we need MORE tests.