Cognitive Psychology Chapter 5
Semantic Memory
Another type of long-term memory is semantic memory—memories of facts such as an address or a birthday or the names of different objects ("that's a bicycle").
Short-Term Memory/ Working Memory
Information that stays in our memory for brief periods, about 10 to 15 seconds if we don't repeat it over and over as Christine did, is short-term memory or working memory.
Episodic Memory
Long-term memories of experiences from the past, like the picnic, are episodic memories.
Long-Term Memory
Long-term memory is responsible for storing information for long periods of time—which can extend from minutes to a lifetime.
Sensory Memory
Sensory memory is the retention, for brief periods of time, of the effects of sensory stimulation.
Short-Term Memory
Short-term memory (STM) is the system involved in storing small amounts of information for a brief period of time. Because of the brief duration of STM, it is easy to downplay its importance com-pared to LTM, but, as we will see, STM is responsible for a great deal of our mental life.
Duration of Short-Term Memory
Short-term memory, as conceived by cognitive psychologists, lasts 15 to 20 seconds or less.
Sensory Memory
This brief persistence of the image, which is one of the things that makes it possible to perceive movies, is called sensory memory.
Modal Model of Memory (Atkinson and Shiffrin, 1968)
1. Sensory memory is an initial stage that holds all incoming information for seconds or fractions of a second. 2. Short-term memory (STM) holds five to seven items for about 15 to 20 seconds. 3. Long-term memory (LTM) can hold a large amount of information for years or even decades.
Chunk
A chunk has been defined as a collection of elements that are strongly associated with one another but are weakly associated with elements in other chunks.
Working Memory and Distributed Representation
Another current idea about working memory is that it involves physiological processes that extend beyond the PFC. It isn't hard to see why working memory would involve brain areas in addition to the frontal lobes. Working memory involves an interplay between a number of areas of the brain. This interplay is symbolized by the interaction between brain areas which depicts a network based on the research on a large number of experiments (Curtis & Espisoto, 2003; Ericsson et al., 2015; Lee & Baker, 2016; Riley & Constantinidis, 2016). This idea that a number of areas of the brain are involved in working memory is an ex-ample of distributed representation.
Evidence for Phonological Loop: Articulatory Suppression
Another way that the operation of the phonological loop has been studied is by determining what happens when its operation is disrupted. This occurs when a person is prevented from rehearsing items to be remembered by repeating an irrelevant sound, such as "the, the, the . . .". This repetition of an irrelevant sound results in a phenomenon called articulatory suppression, which reduces memory because speaking interferes with rehearsal.
Problems with the Modal Model of Memory
As research on STM progressed, it be-came apparent that the concept of STM as presented in the modal model was too narrow to explain many research findings. The problem was that STM was described mainly as a short-term storage mechanism. As we will see next, more goes on in short-term memory than storage. Information doesn't just sit in STM; it can be manipulated in the service of mental processes such as computation, learning, and reasoning.
Interference in the Visuospatial Sketchpad
Brooks (1968) did some experiments in which he demonstrated how interference can affect the operation of the visuospatial sketch pad. This demonstration involves visualizing a large "F" which has two types of corners, "outside corners" and "inside corners," two of which are labeled. Task 1: Cover Figure 5.16, and while visualizing F in your mind, start at the upper-left corner (the one marked with the o), and, moving around the outline of the F in a clockwise direction in your mind (no looking at the figure!), point to "Out" for an outside corner and "In" for an inside corner. Move your response down one level in Table 5.2 for each new corner. Task 2: Visualize the F again, but this time, as you move around the outline of the F in a clockwise direction in your mind, say "Out" if the corner is an outside corner or "In" if it is an inside corner. Most people find that the pointing task is more difficult. The reason is that holding the image of the letter and pointing are both visuospatial tasks, so the visuospatial sketch pad becomes overloaded. In contrast, saying "Out" or "In" is an articulatory task that is handled by the phonological loop, so speaking doesn't interfere with visualizing the F.
Luck and Vogel (1997)
Change detection has also been used with simpler stimuli to determine how much information a person can retain from a briefly flashed stimulus. An example of change detection which shows stimuli like the ones used in Luck and Vogel's experiment. The display on the left was flashed for 100 ms, followed by 900 ms of darkness and then the new display on the right. The participant's task was to indicate whether the second display was the same as or different from the first. This task is easy if the number of items is within the capacity of STM but becomes harder when the number of items becomes greater than the capacity of STM.
Reading Span Test
Daneman and Carpenter (1980) carried out one of the early experiments on individual differences in working memory capacity by developing a test for working memory capacity and then determining how individual differences were related to reading comprehension. The test they developed, the reading span test, required participants to read a series of 13-to 16-word sentences. Each sentence was seen briefly as it was being read, then the next sentence was presented. Immediately after reading the last sentence, the participant was asked to remember the last word in each sentence in the order that they occurred. The participant's reading span was the number of sentences they could read, and then correctly remember all of the last words. Participants' reading spans ranged from 2 to 5, and the size of the reading span was highly correlated with their performance on a number of reading comprehension tasks and their verbal SAT score. Daneman and Carpenter concluded that working memory capacity is a crucial source of individual differences in reading comprehension. Other research has shown that higher working memory capacity is related to better academic performance, better chance of graduating from high school, the ability to control emotions, and greater creativity.
Chunking
Describes the fact that small units (like words) can be combined into larger meaningful units, like phrases, or even larger units, like sentences, paragraphs, or stories. Chunking in terms of meaning increases our ability to hold information in STM. We can recall a sequence of 5 to 8 unrelated words, but arranging the words to form a meaningful sentence so that the words become more strongly associated with one another increases the memory span to 20 words or more.
Alvarez and Cavanaugh (2004)
Did an experiment using Luck and Vogel's change detection procedure. But in addition to colored squares, they also used more complex objects. For example, for the shaded cubes, which were the most complex stimuli, a participant would see a display containing a number of different cubes, followed by a blank interval, followed by a display that was either the same as the first one or in which one of the cubes was different. The participant's task was to indicate whether the two displays were the same or different.
Control Process
Dynamic processes associated with the structural features that can be controlled by the person and may differ from one task to another. An example of a control process that operates on short-term memory is rehearsal—repeating a stimulus over and over, as you might repeat a telephone number in order to hold it in your mind after looking it up on the Internet. Other examples of control processes are (1) strategies you might use to help make a stimulus more memorable, such as relating the digits in a phone number to a familiar date in history, and (2) strategies of attention that help you focus on information that is particularly important or interesting.
Prefrontal Neurons that Hold Information
Funahashi and coworkers (1989) conducted an experiment in which they re-corded from neurons in a monkey's PFC while the monkey carried out a delayed-response task. The monkey first looked steadily at a fixation point, X, while a square was flashed at one position on the screen. In this example, the square was flashed in the up-per-left corner (on other trials, the square was flashed at different positions on the screen). This caused a small response in the neuron. After the square went off, there was a delay of a few seconds. The nerve firing records show that the neuron was firing during this delay. This firing is the neural record of the monkey's working memory for the position of the square. After the delay, the fixation X went off. This was a signal for the monkey to move its eyes to where the square had been flashed. The monkey's ability to do this provides behavioral evidence that it had, in fact, remembered the location of the square. The key result of this experiment was that Funahashi found neurons that responded only when the square was flashed in a particular location and that these neurons continued responding during the delay.
Anders Ericsson et al. (1980)
K. Anders Ericsson and coworkers (1980) demonstrated an effect of chunking by showing how a college student with average memory ability was able to achieve amazing feats of memory. Their participant, S.F., was asked to repeat strings of random digits that were read to him. Although S.F. had a typical memory span of 7 digits, after extensive training (230 one-hour sessions), he was able to repeat sequences of up to 79 digits without error. S.F. used chunking to recode the digits into larger units that formed meaningful sequences. S.F. was a runner, so some of the sequences were running times. For example, 3,492 became "3 minutes and 49 point 2 seconds, near world-record mile time." He also used other ways to create meaning, so 893 became "89 point 3, very old man." This example illustrates an interaction between STM and LTM, because S.F created some of his chunks based on his knowledge of running times that were stored in LTM.
Memory
Memory is the process involved in retaining, retrieving, and using information about stimuli, images, events, ideas, and skills after the original information is no longer present. Memory is active any time some past experience has an effect on the way you think or behave now or in the future.
Change Detection
More recent measures of STM capacity have set the limit at about four items. Experiments in which two pictures of a scene were flashed one after the other and the participants' task was to determine what had changed between the first and second pictures. The conclusion from these experiments was that people often miss changes in a scene.
Items in Short-Term Memory
Not only is information lost rapidly from STM, but there is a limit to how much information can be held there. Estimates for how many items can be held in STM range from four to nine.
Neural Dynamics of Working Memory
One idea, proposed by Mark Stokes (2015), is that information can be stored by short-term changes in neural networks. Figure 5.23a shows the activity state, in which information to be remembered causes a number of neurons, indicated by the dark circles, to briefly fire. This firing doesn't continue, but causes the synaptic state in which a number of connections between neurons, indicated by the darker lines, are strengthened. These changes in connectivity, which Stokes calls activity-silent working memory, last only a few seconds, but that is long enough for working memory. Finally, when the memory is being retrieved, the memory is indicated by the pattern of firing in the network, shown by the dark circles. Thus, in Stokes's model, information is held in memory not by continuous nerve fir-ing but by a brief change in the connectivity of neurons in a network.
Digit Span
One measure of the capacity of STM is provided by the digit span—the number of digits a person can remember. According to measurements of digit span, the average capacity of STM is about five to nine items—about the length of a phone number.
Echoic Memory
Other research using auditory stimuli has shown that sounds also persist in the mind. This persistence of sound, called echoic memory, lasts for a few seconds after presentation of the original stimulus. An example of echoic memory is when you hear someone say something, but you don't understand at first and say "What?" But even before the person can repeat what was said, you "hear" it in your mind. If that has happened to you, you've experienced echoic memory.
Recall
Participants are presented with stimuli and then, after a delay, are asked to report back as many of the stimuli as possible. Memory performance can be measured as a percentage of the stimuli that are remembered. Participants' responses can also be analyzed to determine whether there is a pattern to the way items are recalled. Recall is also involved when a person is asked to recollect life events, such as graduating from high school, or to recall facts they have learned, such as the capital of Nebraska.
Persistence of Vision
Persistence of vision is the continued perception of a visual stimulus even after it is no longer present. This persistence lasts for only a fraction of a second, so it isn't obvious in everyday experience when objects are present for long periods. However, the persistence of vision effect is noticeable for brief stimuli, like the moving sparkler or rapidly flashed pictures in a movie theater.
Episodic Buffer
Research has shown that there are some things the model can't explain. One of those things is that working memory can hold more than would be expected based on just the phonological loop or visuospatial sketch pad. For example, people can remember long sentences consisting of as many as 15 to 20 words. The ability to do this is related to chunking, in which meaningful units are grouped together, and it is also related to long-term memory, which is involved in knowing the meanings of words in the sentence and in relating parts of the sentence to each other based on the rules of grammar. Baddeley decided it was necessary to propose an additional component of working memory to address these abilities. This new component, which he called the episodic buffer, is shown in Baddeley's new model of working memory. The episodic buffer can store in-formation (thereby providing extra capacity) and is connected to LTM (thereby making interchange between working memory and LTM possible).
Working Memory and Damage to the PFC
Researchers have noted that damage to the frontal lobe causes problems in controlling attention, which is an important function of the central executive. An example of animal research that explored the effect of frontal lobe damage on memory tested monkeys using the delayed-response task, which required a monkey to hold information in working memory during a delay period. The monkey sees a food reward in one of two food wells. Both wells are then covered, a screen is lowered, and then there is a delay before the screen is raised again. When the screen is raised, the monkey must remember which well had the food and uncover the correct food well to obtain a reward. Monkeys can be trained to accomplish this task. However, if their PFC is removed, their performance drops to chance level, so they pick the correct food well only about half of the time. This result supports the idea that the PFC is important for holding information for brief periods of time.
Purpose of Sensory and Short-Term Memory
Sensory memory is important when we go to the movies (more on that soon), but the main reason for discussing sensory memory is to demonstrate an ingenious procedure for measuring how much information we can take in immediately, and how much of that information remains half a second later. The purpose of short-term memory will become clearer as we describe its characteristics, but stop for a moment and answer this question: What are you aware of right now? Some material you are reading about memory? Your surroundings? Noise in the background? Whatever your answer, you are describing what is in short-term memory. Everything you know or think about at each moment in time is in short-term memory.
Sperling's Partial Report Method
Sperling reasoned that if participants couldn't report the 12-letter display because of fading, perhaps they would do better if they were told to just report the letters in a single 4-letter row. Sperling devised the partial report method to test this idea. Participants saw the 12-letter display for 50 ms, as before, but immediately after it was flashed, they heard a tone that told them which row of the matrix to report. Because the tones were presented immediately after the letters were turned off, the participant's attention was directed not to the actual letters, which were no longer present, but to whatever trace remained in the participant's mind after the letters were turned off. When the participants focused their attention on one of the rows, they correctly reported an average of about 3.3 of the 4 letters (82 percent) in that row. Because this occurred no matter which row they were reporting, Sperling concluded that immediately after the 12-letter display was presented, participants saw an average of 82 percent of all of the letters but were not able to report all of these letters because they rapidly faded as the initial letters were being reported.
Sperling's Delayed Partial Report Method
Sperling then did an additional experiment to determine the time course of this fading. For this experiment, Sperling devised a delayed partial report method in which the letters were flashed on and off and then the cue tone was presented after a short delay. The result of the delayed partial report experiments was that when the cue tones were delayed for 1 second after the flash, participants were able to report only slightly more than 1 letter in a row. This graph indicates that immediately after a stimulus is presented, all or most of the stimulus is available for perception. This is sensory memory. Then, over the next second, sensory memory fades. Sperling concluded from these results that a short-lived sensory memory registers all or most of the information that hits our visual receptors, but that this information decays within less than a second.
Procedural Memory
The ability to ride a bicycle, or do any of the other things that involve muscle coordination, is a type of long-term memory called procedural memory.
Central Executive
The central executive is the component that makes working memory "work," because it is the control center of the working memory system. Its mission is not to store information but to coordinate how information is used by the phonological loop and visuospatial sketch pad. Baddeley describes the central executive as being an attention controller. It determines how attention is focused on a specific task, how it is divided between two tasks, and how it is switched between tasks.
Information Held in Short-Term Memory
The idea that the capacity of short-term memory can be specified as a number of items, as described in the previous section, has generated a great deal of research. But some researchers have suggested that rather than describing memory capacity in terms of "number of items," it should be described in terms of "amount of information." When referring to visual objects, information has been defined as visual features or details of the object that are stored in memory.
Central Executive
The phonological loop and the visuospatial sketch pad are attached to the central executive. The central executive is where the major work of working memory occurs. The central executive pulls information from long-term memory and coordinates the activity of the phonological loop and visuospatial sketch pad by focusing on specific parts of a task and deciding how to divide attention between different tasks. The central executive is therefore the "traffic cop" of the working memory system.
Phonological Loop
The phonological loop consists of two components: the phonological store, which has a limited capacity and holds information for only a few seconds, and the articulatory rehearsal process, which is responsible for rehearsal that can keep items in the phonological store from decaying. The phonological loop holds verbal and auditory information.
Evidence for Phonological Loop: Phonological Similarity Effect
The phonological similarity effect is the confusion of letters or words that sound similar. In an early demonstration of this effect, R. Conrad (1964) flashed a series of target letters on a screen and instructed his participants to write down the letters in the order they were presented. He found that when participants made errors, they were most likely to misidentify the target letter as another letter that sounded like the target.
Results of Luck and Vogel (1997)
The result of Luck and Vogel's experiment indicates that performance was almost perfect when there were one to three squares in the arrays, but that performance began decreasing when there were four or more squares. Luck and Vogel concluded from this result that participants were able to retain about four items in their short-term memory.
Results of Alvarez and Cavanaugh (2004)
The result was that participants' ability to make the same/different judgment depended on the complexity of the stimuli. Memory capacity for the colored squares was 4.4, but capacity for the cubes was only 1.6. Based on this result, Alvarez and Cavanaugh concluded that the greater the amount of information in an image, the fewer items that can be held in visual short-term memory.
Structural Features
The types of memory listed above, each of which is indicated by a box in the model, are called the structural features of the model.
Visuospatial Sketchpad
The visuospatial sketch pad handles visual and spatial information and is therefore involved in the process of visual imagery—the creation of visual images in the mind in the absence of a physical visual stimulus. Shepard and Metzler's (1971) mental rotation task demonstrates visual imagery. Another demonstration of the use of visual representation is an experiment by Sergio Della Sala and coworkers (1999) in which participants were asked to recall visual patterns. In this demonstration, the patterns are difficult to code verbally, so completing the pattern depends on visual memory. Della Sala presented his participants with patterns ranging from small (a 2 3 2 matrix with 2 shaded squares) to large (a 5 3 6 matrix with 15 shaded squares), with half of the squares being shaded in each pattern. He found that participants were able to complete patterns consisting of an average of 9 shaded squares before making mistakes. The fact that it is possible to remember the patterns in Della Sala's matrix illustrates the operation of visual imagery.
Visuospatial Sketchpad
The visuospatial sketch pad holds visual and spatial information. When you form a picture in your mind or do tasks like solving a puzzle or finding your way around campus, you are using your visuospatial sketch pad.
Evidence for Phonological Loop: Word Length Effect
The word length effect occurs when memory for lists of words is better for short words than for long words. It takes more time to pronounce and rehearse longer words and to produce them during recall. In another study of memory for verbal material, Baddeley and coworkers (1975) found that people are able to remember the number of items that they can pronounce in about 1.5-2.0 seconds.
Iconic Memory/ Visual Icon
This brief sensory memory for visual stimuli, called iconic memory or the visual icon (icon means "image"), corresponds to the sensory memory stage of Atkinson and Shiffrin's modal model.
Peterson and Peterson (1959)
Used the method of recall to determine the duration of STM. Peterson and Peterson presented participants with three letters, such as FZL or BHM, followed by a number, such as 403. Participants were instructed to begin counting backwards by threes from that number. This was done to keep participants from rehearsing the letters. After intervals ranging from 3 to 18 seconds, participants were asked to recall the three letters. Participants correctly recalled about 80 percent of the three letter groups when they had counted for only 3 seconds, but recalled only about 12 percent of the groups after counting for 18 seconds. Results such as this have led to the conclusion that the effective duration of STM (when rehearsal is prevented, as occurred when counting backwards) is about 15 to 20 seconds or less.
Vogel et al. (2005)
What is it about differences in working memory capacity that results in these outcomes? Vogel and coworkers (2005) focused on one component of working memory: the control of attention by the central executive. They first separated participants into two groups based on their performance on a test of working memory. Participants in the high-capacity group were able to hold a number of items in working memory; participants in the low-capacity group were able to hold fewer items in work-ing memory. Participants were tested using the change detection procedure. Sequence of stimuli: (1) they first saw a cue indicating whether to direct their attention to the red rectangles on the left side or the red rectangles on the right side of the displays that followed. (2) They then saw a memory display for one-tenth of a second followed by ... (3) a brief blank screen and then ... (4) a test display. Their task was to indicate whether the cued red rectangles in the test display had the same or different orientations than the ones in the memory display. While they were making this judgment, a brain response called the event-related potential was measured, which indicated how much space was used in working memory as they carried out the task. The fact that adding the blue bars had only a small effect on the response of the high-capacity group means that these participants were very efficient at ignoring the distractors, so the irrelevant blue stimuli did not take up much space in working memory. Because allocating attention is a function of the central executive, this means that the central executive was functioning well for these participants. The fact that adding the two blue bars caused a large increase in the response of the low-capacity group means that these participants were not able to ignore the irrelevant blue stimuli, so the blue bars were taking up space in working memory. The central executive of these participants is not operating as efficiently as the central executives of the high-capacity participants. Vogel and coworkers concluded from these results that some people's central executives are better at allocating attention than others'.
Baddley's Working Memory Model
What kind of model can take into account both (1) the dynamic processes involved in cognitions such as understanding language and doing math problems and (2) the fact that people can carry out two tasks simultaneously? Baddeley concluded that working memory must be dynamic and must also consist of a number of components that can function separately. He proposed three components: the phonological loop, the visuospatial sketch pad, and the central executive.
Sperling (1960)
Wondered how much information people can take in from briefly presented stimuli. He determined this in a famous experiment in which he flashed an array of letters on the screen for 50 milliseconds (50/1000 second) and asked his participants to report as many of the letters as possible. This part of the experiment used the whole report method; that is, participants were asked to report as many letters as possible from the entire 12-letter display. Given this task, they were able to report an average of 4.5 out of the 12 letters.
Working Memory
Working memory is de-fined as "a limited-capacity system for temporary storage and manipulation of information for complex tasks such as comprehension, learning, and reasoning." Short-term memory is concerned mainly with storing information for a brief period of time (for example, remembering a phone number), whereas working memory is concerned with the manipulation of information that occurs during complex cognition (for example, remembering numbers while reading a paragraph).
