After Chapter 3.2, you will be able to:
While learning is mostly concerned with behavior, the study of memory focuses on how we gain the knowledge that we accumulate over our lifetimes. The formation of memories can be divided into three major processes: encoding, storage, and retrieval.
Encoding refers to the process of putting new information into memory. Much of the information that we gain is passively absorbed from the environment. As you walk down the street, you are constantly bombarded with information that seeps into your brain: you notice the temperature; you keep track of the route that you’re taking; you might stop at a coffee shop and realize that the same barista has been working each day this week. All of this information is gained without effort and is thus said to be the result of automatic processing.
There are, however, times when we must actively work to gain information. In studying for the MCAT, for example, you may create flashcards to memorize the enzymes of digestion or the functions of endocrine hormones. This active memorization is known as controlled (effortful) processing.
Do not allow yourself to study for the MCAT using automatic processing! Just reading the text “to get through it” won’t cut it for the MCAT. Engage with the text: fill out the MCAT Concept Checks, write notes in the margins, ask yourself questions. Scientific studies of learning have demonstrated, time and time again, that controlled processing improves comprehension, retention, and speed and accuracy on Test Day.
With practice, controlled processing can become automatic. Think back to a time when you were learning a foreign language. At first, each word required a great deal of processing to decipher: you had to hear the word and consciously translate it into your native language in order to understand what was being said. This took an amount of time and effort that was probably difficult to maintain for prolonged periods. However, as you gained more experience with the language, this process became easier until you may have been able to understand those same words intuitively, without having to think very hard about them at all. At that point, this skill that once required controlled processing became automatic.
There are a few different ways that we encode the meaning of information that requires controlled processing. We can visualize it (visual encoding), store the way it sounds (acoustic encoding), or put it into a meaningful context (semantic encoding). Of these three, semantic encoding is the strongest and visual encoding is the weakest. When using semantic encoding, the more vivid the context, the better. In fact, we tend to recall information best when we can put it into the context of our own lives, a phenomenon called the self-reference effect.
The purpose of the Real World sidebars in your MCAT Review books is semantic encoding: by putting content into a meaningful context, retention of the information is improved. Most of our Real World sidebars are related to medicine because of the self-reference effect.
Of course, grouping information into a meaningful context is only one trick that we can use to aid in encoding. Another such aid is maintenance rehearsal, which is the repetition of a piece of information to either keep it within working memory (to prevent forgetting) or to store it in short-term and eventually long-term memory—topics discussed in the next section.
Mnemonics are another common way to memorize information, particularly lists. As you’ve seen in your Kaplan study materials, mnemonics are often acronyms or rhyming phrases that provide a vivid organization of the information we are trying to remember. Two other mnemonic techniques are commonly employed by memory experts. The method of loci involves associating each item in a list with a location along a route through a building that has already been memorized. For example, in memorizing a grocery list, someone might picture a carton of eggs sitting on their doorstep, a person spilling milk in the front hallway, a giant stick of butter in the living room, and so on. Later, when the person wishes to recall the list, they simply take a mental walk through the locations and recall the images they formed earlier. Similarly, the peg-word system associates numbers with items that rhyme with or resemble the numbers. For example, one might be associated with the sun, two with a shoe, three with a tree, and so on. As groundwork, the individual memorizes their personal peg-list. When another list needs to be memorized, the individual can simply pair each item in the list with their peg-list. In this example, the individual may visualize eggs being fried by the sun (1), a pair of shoes (2) filled with milk, and a tree (3) with leaves made of butter. Because of the serial nature of both the method-of-loci and peg-word systems, they are very useful for memorizing large lists of objects in order.
Finally, chunking (sometimes referred to as clustering) is a memory trick that involves taking individual elements of a large list and grouping them together into groups of elements with related meaning. For example, consider the following list of 16 letters: E-N-A-L-P-K-C-U-R-T-R-A-C-S-U-B. Memorizing the list in order by rote might prove difficult until we realize that we can reverse the items and group them into meaningful chunks: BUS, CAR, TRUCK, PLANE.
Following encoding, information must be stored if it is to be remembered. There are several types of memory storage.
The first and most fleeting kind of memory storage is sensory memory, which consists of both iconic (visual) and echoic (auditory) memory. Sensory memory lasts only a very short time (generally under one second), but within that time our eyes and ears take in an incredibly detailed representation of our surroundings that we can recall with amazing precision. Sensory memories are maintained by the major projection areas of each sensory system, such as the occipital lobe (vision) and temporal lobe (hearing). Of course, sensory memory fades very quickly, and unless the information is attended to, it will be lost.
The nature of sensory memory can be demonstrated experimentally. Consider the following procedure: a research participant is presented with a three-by-three array of letters, such as that presented in Figure 3.5, that is flashed onto a screen for a mere fraction of a second. When asked to list all of the letters she saw, the participant is able to correctly identify three or four (a procedure known as whole-report). However, when asked to list the letters of a particular row immediately after the presentation of the stimulus (known as partial-report), she can do so with 100 percent accuracy, no matter which row is chosen. This is iconic memory in action: in the time it takes to list out a few of the items, the entire list fades, yet it is clear that all of the letters do make their way into iconic memory because any small subset can be recalled at will.
Of course, we do pay attention to some of the information that we are exposed to, and that information enters our short-term memory. Similar to sensory memory, short-term memory fades quickly, over the course of approximately 30 seconds without rehearsal. In addition to having a limited duration, short-term memories are also limited in capacity to approximately seven items, usually stated as the 7 ± 2 rule. As discussed in the previous section, the capacity of short-term memory can be increased by clustering information, and the duration can be extended using maintenance rehearsal. Short-term memory is housed primarily in the hippocampus, which is also responsible for the consolidation of short-term memory into long-term memory.
Working memory is closely related to short-term memory and is similarly supported by the hippocampus. It enables us to keep a few pieces of information in our consciousness simultaneously and to manipulate that information. To do this, one must integrate short-term memory, attention, and executive function; accordingly, the frontal and parietal lobes are also involved. This is the form of memory that allows us to do simple math in our heads.
With enough rehearsal, information moves from short-term to long-term memory, an essentially limitless warehouse for the knowledge that we are then able to recall on demand, sometimes for the rest of our lives. One of the ways that information is consolidated into long-term memory is elaborative rehearsal. Unlike maintenance rehearsal, which is simply a way of keeping the information at the forefront of consciousness, elaborative rehearsal is the association of the information to knowledge already stored in long-term memory. Elaborative rehearsal is closely tied to the self-reference effect noted earlier; those ideas that we are able to relate to our own lives are more likely to find their way into our long-term memory. While long-term memory is primarily controlled by the hippocampus, it should be noted that memories are moved, over time, back to the cerebral cortex. Thus, very long-term memories—our names and birthdates, the faces of our parents—are generally not affected by damage to the hippocampus.
There are two types of long-term memory. Implicit (nondeclarative or procedural) memory consists of our skills and conditioned responses. Explicit (declarative) memory consists of those memories that require conscious recall. Explicit memory can be further divided into semantic memory (the facts that we know) and episodic memory (our experiences). Interestingly, memory disorders can affect one type of memory but leave others alone. For example, a patient with amnesia might not remember the time he learned to ride a bicycle or the names of the parts of a bicycle (episodic and semantic memories, respectively) but may, to his surprise, retain the skill of riding a bicycle when given one. The various major categories of memory are summarized in Figure 3.6.
Of course, memories that are stored are of no use unless we can pull them back out to use them. Retrieval is the name given to the process of demonstrating that something that has been learned has been retained. Most people think about retrieval in terms of recall, or the retrieval and statement of previously learned information, but learning can be additionally demonstrated by recognizing or quickly relearning information.
Recognition, the process of merely identifying a piece of information that was previously learned, is far easier than recall. This difference is something you can take advantage of because the MCAT, as a multiple-choice test, is largely based on recognizing information. If the MCAT were a fill-in-the-blank style exam, your approach to studying would have to be vastly different and far more in-depth.
Think back to elementary school. How many of your classmates do you think you could list? Chances are, not many. On the other hand, glancing through your class photo, you would probably recognize the vast majority of your former classmates. This gap is the difference between recall and recognition.
Relearning is another way of demonstrating that information has been stored in long-term memory. In studying the memorization of lists, Hermann Ebbinghaus found that his recall of a list of short words he had learned the previous day was often quite poor. However, he was able to rememorize the list much more quickly the second time through. Ebbinghaus interpreted this to mean that the information had been stored, even though it wasn’t readily available for recall. Through additional research, he discovered that the longer the amount of time between sessions of relearning, the greater the retention of the information later on. Ebbinghaus dubbed this phenomenon the spacing effect, and it helps to explain why cramming is not nearly as effective as spacing out studying over an extended period of time.
Recalling a fact at a moment’s notice can be difficult. Fortunately, the brain has ways of organizing information so that it can take advantage of environmental cues to tell it where to find a given memory. Psychologists think of memory not as simply a stockpile of unrelated facts, but rather as a network of interconnected ideas. The brain organizes ideas into a semantic network, as shown in Figure 3.7, in which concepts are linked together based on similar meaning, not unlike an Internet encyclopedia wherein each page includes links for similar topics. For example, the concept of red might be closely linked to other colors, like orange and green, as well as objects, like fire engine and roses. When one node of our semantic network is activated, such as seeing the word red on a sign, the other linked concepts around it are also unconsciously activated, a process known as spreading activation. Spreading activation is at the heart of a retrieval cue known as priming, in which recall is aided by first being presented with a word or phrase that is close to the desired semantic memory.
Context effects are another common retrieval cue. Memory is aided by being in the physical location where encoding took place. Psychologists have shown a person will score better when they take an exam in the same room in which they learned the information. Context effects can go even further than this; facts learned underwater are better recalled when underwater than when on land.
Similarly, a person’s mental state can also affect recall. This retrieval cue is called state-dependent memory, alternately referred to as a state-dependent effect. People who learn facts or skills while intoxicated, for example, will show better recall or proficiency when performing those same tasks while intoxicated as compared to performing them while sober. Emotions work in a similar way: being in a foul mood primes negative memories, which in turn work to sustain the foul mood. So not only will memory be better for information learned when in a similar mood, but recall of negative or positive memories will lead to the persistence of the mood.
Finally, the serial position effect is a retrieval cue that appears while learning lists. When researchers give participants a list of items to memorize, the participants have much higher recall for both the first few and last few items on the list. The tendency to remember early and late items is known as the primacy and recency effect, respectively. However, when asked to remember the list later, people show strong recall for the first few items while recall of the last few items fades. Psychologists interpret this to mean that the recency effect is a result of the last items still being in short-term memory on initial recall.
Unfortunately, even long-term memory is not always permanent. Several phenomena can result in the loss of memorized information.
There are several disorders that can lead to decline in memory. The most common is Alzheimer’s disease, which is a degenerative brain disorder thought to be linked to a loss of acetylcholine in neurons that link to the hippocampus, although its exact causes are not well understood. Alzheimer’s is marked by progressive dementia (a loss of cognitive function) and memory loss, with atrophy of the brain, as shown in Figure 3.8. While not perfectly linear, memory loss in Alzheimer’s disease tends to proceed in a retrograde fashion, with loss of recent memories before distant memories. Microscopic findings of Alzheimer’s include neurofibrillary tangles and β-amyloid plaques. One common phenomenon that occurs in individuals with middle- to late-stage Alzheimer’s is sundowning, an increase in dysfunction in the late afternoon and evening.
The β-amyloid plaques of Alzheimer’s disease are incorrectly folded copies of the amyloid precursor protein, in which insoluble β-pleated sheets form and then deposit in the brain. Protein folding is discussed in detail in Chapter 1 of MCAT Biochemistry Review.
Korsakoff’s syndrome is another form of memory loss caused by thiamine deficiency in the brain. The disorder is marked by both retrograde amnesia (the loss of previously formed memories) and anterograde amnesia (the inability to form new memories). Another common symptom is confabulation, or the process of creating vivid but fabricated memories, typically thought to be an attempt made by the brain to fill in the gaps of missing memories.
Agnosia is the loss of the ability to recognize objects, people, or sounds, though usually only one of the three. Agnosia is usually caused by physical damage to the brain, such as that caused by a stroke or a neurological disorder such as multiple sclerosis.
Of course, not all memory loss is due to a disorder. Often, memories are simply lost naturally over time as the neurochemical trace of a short-term memory fades. In his word memorization experiment, Ebbinghaus noted what he called a “curve of forgetting,” as shown in Figure 3.9. For a day or two after learning the list, recall fell sharply but then leveled off.
Another common reason for memory loss is interference (also referred to as an interference effect), a retrieval error caused by the existence of other, usually similar, information. Interference can be classified by its direction. When we experience proactive interference, old information is interfering with new learning. For example, think back to a time when you moved to a new address. For a short time, you may have had trouble recalling individual pieces of the new address because you were so used to the old one. Similarly, Ebbinghaus found that with each successive list he learned, his recall for new lists decreased over time, an effect he attributed to interference caused by older lists.
Retroactive interference is when new information causes forgetting of old information. For example, at the beginning of a school year, teachers learning a new set of students’ names often find that they can no longer remember the names of the previous year’s students. One way of preventing retroactive interference is to reduce the number of interfering events, which is why it is often best to study in the evening about an hour before falling asleep (although this also depends on your personal style!).
Contrary to popular belief, aging does not necessarily lead to significant memory loss; while there are many individuals whose memory fades in old age, this is not always the case. In fact, studies show that there is a larger range of memory ability for 70-year-olds than there is for 20-year-olds. There are, however, some trends that can be demonstrated when evaluating the memories of older individuals. When asked about the most pivotal events in their lives, people in their 70s and 80s tend to say that their most vivid memories are of events that occurred in their teens and 20s, a fact that psychologists interpret to mean that this time is a peak period for encoding in a person’s life.
Even for the elderly, certain types of memory remain quite strong. People tend not to demonstrate much degeneration in recognition or skill-based memory as they age. Even certain types of recall will remain strong for most people; semantically meaningful material can be easily learned and recalled, most likely due to older individuals having a larger semantic network than their younger counterparts. Prospective memory (remembering to perform a task at some point in the future) remains mostly intact when it is event-based—that is, primed by a trigger event, such as remembering to buy milk when walking past the grocery store. On the other hand, time-based prospective memory, such as remembering to take a medication every day at 7:00 a.m., does tend to decline with age.
We often think of memory as a record of our experiences or a kind of video recording that is stored to be accessed later. Nothing could be further from the truth: memory can be faulty to the point where two people can recall the same event as occurring in completely different ways. In fact, memories are influenced heavily by our thoughts and feelings both while the event is occurring and later during recall. We’ve already discussed confabulation, a phenomenon in which we fill in gaps in our memories such that, over time and with enough rehearsal, our memories of the event can change drastically. Confabulation is one example of the creation of false memories, but our memories can also be affected by outside sources as well.
One such phenomenon is the misinformation effect. In a famous experiment, participants were shown several pictures including one of a car stopped at a yield sign. Later, they were presented with written descriptions of the pictures, some of which contained misinformation, such as describing a car stopped at a stop sign. When asked to recall the details of the pictures, many participants insisted on having seen a stop sign in the picture.
The misinformation effect can also be seen at the point of recall. In another experiment, participants were shown a video of an automobile accident. Some participants were then asked, How fast were the cars moving when they collided?, while others were asked about the accident using more descriptive language such as How fast were the cars moving when they crashed? Those participants who were asked the question with leading language were much more likely to overstate the severity of the accident than those who had been asked the question with less descriptive language.
Source-monitoring error involves confusion between semantic and episodic memory: a person remembers the details of an event, but confuses the context under which those details were gained. Source-monitoring error often manifests when a person hears a story of something that happened to someone else, and later recalls the story as having happened to him- or herself.
During a Congressional Medal of Honor ceremony in 1983, Ronald Reagan relayed a vivid story about a heroic World War II pilot who received a posthumous medal. Skeptical reporters, unaware of any incident matching the details of Reagan’s story, checked into the story and found that the pilot had existed—in the 1944 movie A Wing and a Prayer. Reagan had remembered the details of the pilot’s heroic actions but had forgotten their source.