Chapter 5 Working Memory and Executive Control 5

Chapter 5 Working Memory and Executive Control

5. 1 Behavioral Processes

5. 1 Behavioral Processes • Transient Memories • Working Memory • The Central Executive • Unsolved Mysteries—Is Working Memory the Key to Intelligence? 3

Transient Memories • Temporary representations of information either just sensed or just retrieved from long-term memory. Need active attention or decay occurs. Two types: Sensory Memory Short-Term (Working) Memory 4

Sensory Memory • Temporary storage of sense information in its physical form as it reaches the brain. The pool of sensory information to which we pay selective attention. There may be a sensory memory for each sense. e. g. , visual sensory memory—temporary storage for information perceived by your visual system. 5

Visual Sensory Memory • Sperling (1960) studied visual sensory memory. Whole (Full) Report: Display 3 rows of items; report any items recalled. Participants recall about 35% of the items. Partial Report: Display 3 rows of items; use sound cues to report items recalled from only one row. Participants recall about 75% of the items. 6

The Sperling Task 7

The Sperling Task 8

The Sperling Task 9

The Sperling Task 10

The Sperling Task 11

Iconic Memory • Iconic memory—the rapidly decaying visual sensory memory. Critical for recognizing and processing briefly presented information. Temporary buffer for information to undergo further processing. 12

Short Term Memory (STM) • Maintains information registered by sensory memory. • Limited by attention; if distracted, will likely forget information. • STM can hold 5 to 9 chunks of information. 7 items, plus or minus 2 13

Atkinson-Shiffrin Memory Model • Interaction of sensory memory, STM, and long-term memory (LTM) 14

Transferring Information from STM to LTM • Depth of processing—deep, meaningful processing that increases the likelihood of transfer from STM to LTM. Shallow, superficial processing (e. g. , passive rehearsal) has little effect on transfer to LTM. 15

Depth of Processing: Craik and Tulving 1975 Study • Shows direct relationship between depth of processing and amount of recall. 16

Depth of Processing: Craik and Tulving 1975 Study • Shows direct relationship between depth of processing and amount of recall. 17

Depth of Processing: Craik and Tulving 1975 Study In 1975, the researchers conducted an experiment in which participants were shown a list of 60 words. • They were then asked to recall certain words by being shown one of three questions, each testing a different level of processing, similar to: Was the word in capital letters or lower case? (Tests structural processing SHALLOW PROCESSING) Does the word rhyme with (another word)? (Tests phonemic/auditory processing, as the participant has to listen to the word judge whether it rhymes with another word) Does the word fit in the following sentence. . . ? (Tests semantic processing; UNDERSTANDING THE MEANING OF THE WORD DEEP PROCESSING/ ELABORATE REHERSAL) • Out of another larger list, the participants were asked to pick out the appropriate word, as the original words had been mixed into this list. 18

Working Memory • Working memory When STM is used as a buffer to briefly maintain information, before it is manipulated to affect behavior. Executive control is the maintenance and manipulation of working memory. 19

Baddeley’s Working Memory Model • Phonological loop Maintains acoustic (verbal) information. • Visuo-Spatial sketchpad Maintains visual and spatial information. • Central executive Material-specific; manipulates information from these visuo-spatial and audio buffers. 20

Baddeley’s Working Memory Model 21

The Phonological Loop • Contains a passive, short-term buffer. Stores about 2 seconds of acoustic or verbal information. • Can be refreshed by rehearsal. Usually silent, internal rehearsal • Word-length effect—the longer the words, the harder they are to remember. Longer words take longer to rehearse. 22

The Visuo-Spatial Sketchpad • Mental workspace for storing and manipulating visual and spatial information • Has a limited capacity; is independent of the phonological loop. Dual-task experiments test ability to process information in visual and verbal memory buffers simultaneously. Filling one buffer has minimal effect on the capacity of the other. 23

The Visuo-Spatial Sketchpad • Delayed nonmatch-to-sample task—a visual memory test Monkey learns to remember the image of an original object to identify a new object. Courtesy of David Yu, Mortimer Mishkm, and Janita Turchi, Laboratory of Neuropsychology, NIMH/NIH/DHHS 24

The Central Executive • Directs attention and rehearsal. • The cognitive decision-maker • Often works automatically. • Decides what you pay attention to and what you ignore. 25

The Central Executive • Typical measures: N-back task: Participant is read a random list of items. Must remember what was read N (e. g. , 2) items back. Petrides’ self-ordered memory tasks: Participant is shown a stack of cards with a set of target items. Must point to a different item on each card without repeating any items. 26

*Example* N-Back Task • Task: When target number 7 appears, tell what item appeared 2 items ago. • For example: 4 8 3 7 8 2 5 6 7 8 0 2 4 6 7 8 5 4 • Try it now… 27

2 28

6 29

3 30

9 31

5 32

1 33

9 34

6 35

4 36

2 37

5 38

1 39

7 40

3 41

4 42

6 43

9 44

5 45

1 46

8 47

4 48

7 49

9 50

3 51

1 52

Petrides’ Self-Ordered Tasks • Human task: Eight cards; each card has the same 8 items in a different spatial order. Participant sees one card at a time. Must choose a new item on each consecutive card; must keep track of previous choices to avoid repetitions. • In monkey version, monkey chooses a container with food after containers are covered and shuffled. 53

Petrides’ Self-Ordered Tasks for Humans Adapted from Petrides, 2000. 54

Petrides’ Self -Ordered Tasks for Monkeys 55

Setting Goals and Planning • Setting goals and priorities. • Setting steps to accomplish goals. • Monitoring progress, making adaptations. 56

Setting Goals and Planning: Tower of Hanoi • Tower of Hanoi puzzle requires manipulation of working memory. What subgoals have been accomplished? What subgoals remain? What is the next subgoal to be addressed? 57

Task Switching • Flexibility in changing rules. How well can you switch gears? • Example: Wisconsin Card Sort Test Participant is shown cards with three different characteristics (color, shape, number). First, cards are sorted by one characteristic. Later, task is changed and cards must be sorted by another characteristic. 58

Wisconsin Card Sort Test 59

Stimulus Selection and Response Inhibition • How well can you inhibit an automatic response and replace it with a new response? • Example: the Stroop task Need to inhibit reading the word (response inhibition). Report the color of the letters of the word (stimulus selection). 60

*Example* The Stroop Task Non-Reflexive Response Control • Task: Read next list as fast as you can. • Try it now… 61

BLACK YELLOW GREEN ORANGE RED PINK BLUE WHITE 62

*Example* The Stroop Task Much Harder Variant • Task: Say the color of the LETTERS of the color-name, not the word. e. g. , RED => “Blue” • Try it now… 63

BLACK YELLOW GREEN ORANGE RED PINK BLUE WHITE 64

BLACK YELLOW GREEN ORANGE RED PINK BLUE WHITE 65

Unsolved Mysteries: Is Working Memory the Key to Intelligence? • Intelligence—capacity for learning, reasoning, and understanding • May be associated with a strong working memory, including: Executive control. Manipulation of a large number of multiple rules, concepts, goals, and ideas. 66

5. 1 Interim Summary • Sensory memories—brief, transient sensations produced when you see, hear, feel, or taste something. • Short-term memory can be used as a buffer for maintaining information temporarily over short delays so that it can be manipulated to guide and control behavior. 67

5. 1 Interim Summary • Baddeley's model of working memory includes two independent buffers: Visuo-spatial sketchpad: holds visual and spatial images. Phonological loop: temporary storage for auditory memory; uses internal speech rehearsal. • Model also includes central executive system: Manipulates memory buffers by adding/deleting items, selecting items to guide behavior, storing/retrieving information in long-term memory. 68

5. 1 Interim Summary • Baddeley's model of working memory suggests: Process-specific dissociation between manipulation of information by central executive and maintenance of information by rehearsal buffers. Material-specific dissociation between maintenance of verbal-phonological information and visuo-spatial information. 69

5. 2 Brain Substrates

5. 2 Brain Substrates • Behavioral Consequences of Frontal-Lobe Damage • Frontal Brain Activity during Working. Memory Tasks • Mapping Baddeley’s Model onto PFC Anatomy • Prefrontal Control of Long-Term Declarative Memory 71

Prefrontal Cortex Prefrontal cortex (PFC)—front of frontal lobes Critical for working memory and executive control. Adapted from Fuster, 1995. • 72

Behavioral Consequences of Frontal-Lobe Damage • Damage from tumors, surgery, strokes, frontal head trauma can lead to dysexecutive syndrome. Deficits in working memory and executive function Act impulsively! Can impair STM for series of numbers, colors, shapes, object locations. 73

Frontal Lobe Damage • Impaired performance on planning tasks, such as the Tower of Hanoi puzzle. Move disks randomly. • Impaired ability to switch rules in tasks such as the Wisconsin Card Sort Test. e. g. , first sort cards by color, then by shape. Continues to sort by color, even though reminded to sort by shape. Perseveration on the initial rule 74

Primate Frontal Lobes 75

Functional Neuroanatomy of the Prefrontal Cortex • PFC is made up of orbital, medial, and lateral prefrontal cortex. • Orbital and medial PFC involved in many memory functions. • Lateral prefrontal cortex (LPFC) is the key area for working memory. Lateral = sides LPFC includes the dorsolateral PFC and the ventrolateral PFC. 76

Frontal Brain Activity during Working Memory Tasks • Goldman-Rakic recorded dorsolateral PFC neural activity in monkeys during spatial and object working memory tasks (delayed response tasks). Used eye-tracking technology. After several second delays, monkeys learned to shift their gaze to the former location (or opposite location) of an object or visual pattern for a reward. 77

Golman-Rakic’s Eye-Gaze Response Test 78

The Spatial Delayed-Response Eye-Gaze Task Data from Funahashi, Bruce, & Goldman-Rakic, 1989. 79

From Funahashi et al. , 1989. Response of One PFC Neuron in Eye-Gaze Delayed-Response Task 80

Frontal Brain Activity during Working Memory Tasks • Electrophysiological records of their PFC suggested: Sensory neurons fired during cue presentation. Movement-response neurons fired during eyeglaze response. A neural combination fired during the interim delay period. 81

Frontal Brain Activity during Working Memory Tasks • During delay, corresponding activity in primary and secondary sensory and motor regions of the temporal and parietal lobes. These areas interact with PFC. • Dorsolateral PFC provides focused control over working memory, despite distractions. • Damaged PFC = High degree of distractibility 82

Mapping Baddeley’s Model onto PFC Anatomy • Does Baddeley’s functional distinction between manipulation and rehearsal correspond to anatomical distinctions in brain regions? • Are there anatomical distinctions between the visual-spatial sketchpad and the phonological loop? 83

Maintenance (Rehearsal) versus Manipulation (Executive Control) • Monkeys with dorsolateral PFC lesions chose one of three containers to hold a treat. After 10 second delay, they could not remember which container held the treat—even if they could see the containers during the delay. 84

Maintenance (Rehearsal) versus Manipulation (Executive Control) • In basic delayed recognition tasks, monkeys maintained memories of objects for varying delay. • Increasing number of items to be ordered (not delay time) associated with decreasing memory accuracy. • Dorsolateral PFC lesions are associated with severe temporal ordering deficits. 85

Functional Brain Imaging of Humans on Self-Ordered Tasks • Abstract design items—significantly active Dorsolateral PFC, especially right side. • Verbal items—significantly active Dorsolateral PFC on both sides. Right side may be important for all monitoring processes. Left side may be specialized for verbal material. 86

Functional Brain Imaging of Humans on Self-Ordered Tasks • Premotor cortex is active during rehearsal of visuo-spatial information. • Ventrolateral PFC is active during simple rehearsal, especially internal rehearsal. • Posterior parietal and occipital areas are active in temporary maintenance (not rehearsal) of spatial working memory. Posterior = storage mechanisms Anterior = rehearsal mechanisms 87

Brain Substrates of Working Memory 88

Visuo-Spatial and Phonological-Verbal Buffers • PET images of humans during spatial working-memory tasks and simple perceptual tasks revealed dorsal activation in right hemisphere. Humans have more dorsal and ventral activation than monkeys. 89

Visuo-Spatial and Phonological-Verbal Buffers • Participants completed spatial and object instructions with same stimuli: Spatial “Is the location of the target face identical to any previous face? ” Object “Is the target face identical to the identity of any previous face? ” In humans only: Spatial location = activation within right premotor cortex Object identity = activation of right dorsolateral PFC 90

Prefrontal Control of Long-Term Declarative Memory • Searching LTM involves: Strategic control and manipulation of memory processes. Maintaining awareness of the purpose for the search. • Neuroimaging shows activation of the dorsolateral PFC. 91

Dorsolateral PFC Activity during Recollection of Source Data from Dobbins et al. , 2002. ; Images from Dobbins, I. G. , Foley, H. , Schacter, D. L. , & Wagner, A. D. (2002). Exectutive control during episodic retrieval: Multiple prefrontal processes subserve source memory. Neuron, 35, 989 – 996, with permission from Elsevier. 92

Prefrontal Control of Long-Term Declarative Memory • Functional imaging shows active ventrolateral PFC during intentional encoding of new memories. • Encoding semantic stimuli (e. g. images of nameable objects) is associated with left ventrolateral PFC activation. 93

Prefrontal Control of Long-Term Declarative Memory • Anterior regions are activated during tasks processed semantically. • Shallow phonological processing is associated with posterior ventrolateral PFC activation. 94

5. 2 Interim Summary • Animal and human studies indicate the brain’s frontal lobes are critical for workingmemory and executive-control processes. Especially prefrontal cortex (PFC; the most anterior section of the frontal lobes). • Severe deficits found in task-switching tests (associated with certain lesions) suggest: Purposeful shifts in processing may be especially demanding on executive-control processes mediated by the frontal lobes. 95

5. 2 Interim Summary • Electrophysiological studies in animals by Fuster and Goldman-Rakic suggested PFC is critical for maintaining an internal representation in working memory over a delay, prior to making some response. 96

5. 2 Interim Summary • Primate PFC divided into three main regions: Orbital PFC Medial PFC Lateral PFC, • Lateral PFC is further subdivided into: Ventrolateral PFC (lower region) Dorsolateral PFC (upper region) 97

5. 2 Interim Summary • Miller argued that a key to understanding the PFC’s “cognitive” contribution to working memory is its ability to sustain activity despite distractions. 98

5. 2 Interim Summary • Petrides et. al. suggested that Baddeley’s process-specific functional dichotomy is to be found in the organization of the PFC. Ventrolateral PFC supports encoding and retrieval of information (including rehearsal for maintenance), akin to Baddeley’s visuo-spatial sketchpad and phonological-rehearsal loops. Dorsolateral PFC supports higher order executivecontrol functions such as monitoring and manipulating stored information, akin to Baddeley’s central executive. 99

5. 2 Interim Summary • Neuroimaging studies confirm distinction between storage and rehearsal. Storage mechanisms located in posterior brain regions. Rehearsal mechanisms located in anterior brain regions (including PFC). • Language tends to be left-lateralized and frontal-lobe patients with left-side damage are likely to show verbal (as opposed to visuospatial) deficits in working memory. 100

5. 2 Interim Summary • Working memory interacts with LTM (especially with LTM for episodes or facts). Studies show dorsolateral PFC is activated when people attempt to remember past events. • Functional role of left ventrolateral PFC during encoding of new information: Anterior regions activated during tasks that involve semantic processing. Posterior regions activated during phonological processing. 101

5. 3 Clinical Perspectives

5. 3 Clinical Perspectives • Schizophrenia • Attention-Deficit/Hyperactivity Disorder • Learning and Memory in Everyday Life: Improving Your Working Memory 103

Schizophrenia • A psychiatric disorder Hallucinations (hearing voices) Delusions (that others can read their thoughts) • Disturbances in working memory and executive control • Impaired dorsolateral PFC Trouble keeping large numbers of items in working memory during delay. 104

Schizophrenia • Close to average performance on: Phonological or visuo-spatial memory tasks Memory tasks involving minimal delays or few items • No increased activity in dorsolateral PFC during the Wisconsin Card Sort Test. Positive relationship between sorting performance and dorsolateral activation. 105

Schizophrenia • Weak activity in the dorsolateral PFC during poor N-back task performance. Average ventral and posterior PFC activation • Memory deficits may be related to defective cortical dopamine processing. • PET imaging using chemicals sensitive to dopamine found more D 1 dopamine receptors in dorsolateral PFC. 106

Schizophrenia • Patients with highest number of D 1 receptors in PFC performed worst on the Nback test. Suggests association between dopamine regulation in the dorsolateral PFC and working memory. • Mutation in the COMT gene may affect dopamine metabolism in frontal lobes. 107

Schizophrenia • The number (0, 1, 2) of the bad COMT gene corresponded with: Number of perseverative errors on the Wisconsin Card Sort Test. Depressed PFC activity during the N-back task. • Mutations may contribute to individual’s vulnerability to schizophrenia. 108

Schizophrenia : Gerald • http: //www. youtube. com/watch? v=g. Gnl 8 dq Eo. PQ • http: //www. youtube. com/watch? v=i 6 h 8 Ic. I 7 R 0&feature=related 109

Attention-Deficit/ Hyperactivity Disorder (ADHD) • Difficulties with: Planning, organizing time Keeping attention focused on a task Inhibiting response to distracting stimuli • May involve dysfunction in PFC cortical and subcortical connections, including cerebellum and basal ganglia. • http: //www. youtube. com/watch? v=YBFH 81 zb. Ci. Y 110

ADHD • Children with ADHD may have smaller right PFC than other children. Area is associated with spatial attention and working memory. Ritalin and other medications are stimulants that increase dopamine release or block its synaptic reuptake for 3– 4 hours. • Genes associated with ADHD may be involved in dopamine regulation in the brain. 111

ADHD • Children with ADHD may have smaller right PFC than other children. Area is associated with spatial attention and working memory. Ritalin and other medications are stimulants that increase dopamine release or block its synaptic reuptake for 3– 4 hours. • Genes associated with ADHD may be involved in dopamine regulation in the brain. 112

ADHD • http: //www. youtube. com/watch? v=MKZXH 7 MOwj. I • http: //www. youtube. com/watch? v=Rkzytidh P 1 M&feature=related 113

Learning and Memory in Everyday Life: Improving Your Working Memory • Use both verbal and visuo-spatial stores. e. g. , translate verbal information into images • Minimize multitasking, especially with a dangerous task, such as driving a car. Focusing on one task at a time increases the effectiveness of working memory. • Avoid stress. May increase dopamine in the PFC, impairing ability to efficiently monitor/update information. 114

5. 3 Interim Summary • Working-memory and schizophrenia: Impairments apparent during attempts to maintain large number of items over a temporal delay. Require functions associated with the dorsolateral PFC. Functions attributed to the ventrolateral PFC seem relatively unimpaired. Performance on phonological or visuo-spatial memory tasks, and on memory tasks involving only minimal delays or few items, appears normal. 115

5. 3 Interim Summary • Schizophrenia patients with highest number of D 1 dopamine receptors in PFC (relative to controls) exhibit worst performance on the Nback assessment of working memory. Evidence for link between dopamine regulation of dorsolateral PFC function and working memory. 116

5. 3 Interim Summary • Adults with ADHD show deficits in mental calculations that require use of working memory. • Neuroimaging of children with ADHD indicates they have smaller right PFC region. Region is associated with spatial attention and working memory. 117