PSY 368 Human Memory Implicit memory Outline Theories

PSY 368 Human Memory Implicit memory

Outline • Theories accounting for Implicit vs. Explicit memory • Experiment 2 Signal detection analysis • Process-dissociation procedure, working through our example (probably not time, so after break)

Memory Tasks Test Instructions incidental Study Instructions intentional indirect implicit memory expts. Levels of Processing expts. ? explicit memory expts. Implicit Memory: Often defined as "memory without awareness” • Also “Non-declarative” & “procedural” (Squire, Knowlton, & Mesen, 1993)

Implicit/Explicit Dissociations • Many demonstrations of different effects depending on whether implicit or explicit tasks are used • Amnesic patients • Levels of processing manipulations • Pleasantness vs. vowel comparisons • Generation effect • Divided attention • Picture-word superiority • Note. Most of we’ve talked about concern repetition priming effect (study “horse” and respond “horse”)

Accounting for Implicit/Explicit Dissociations • Four major approaches have been proposed • • The Activation view Multiple Memory systems view Transfer appropriate processing view Bias View

Accounting for Implicit/Explicit Dissociations • Four major approaches have been proposed • • The Activation view Multiple Memory systems view Transfer appropriate processing view Bias View

The Activation View • Priming on indirect tests is attributable to the temporary automatic activation of preexisting representations. • Because it is automatic, it occurs without elaborative processing and thus has little to no contextual information • Weak Point • Can not explain priming over long time periods • Some implicit priming over days or even weeks (e. g. , Sloman, et al, 1988) • Can not explain priming without pre-existing representations • The least popular of the four views

Accounting for Implicit/Explicit Dissociations • Four major approaches have been proposed • • The Activation view Multiple Memory systems view Transfer appropriate processing view Bias View

Multiple Memory Systems • Many dissociations between direct and indirect tests of memory arise because the tests tap different underlying memory systems. Squire (1987)

Multiple Memory Systems • Many dissociations between direct and indirect tests of memory arise because the tests tap different underlying memory systems. Tulving (1984)

Multiple Memory Systems • What is a system? Schacter and Tulving (1994) • It is NOT a process • It is NOT a task • Some different ways that systems have been defined

Multiple Memory Systems • What is a system? Schacter and Tulving (1994) • Functional Dissociations • Task that taps into system A that has no effect (or a different effect) in System B • Different neural substrates • System A involves different brain areas than System B (brain damage cases and neural imaging studies) • Stochastic independence • Performance on System A task uncorrelated with performance on a System B task • Functional incompatibility • Could involve different rates of forgetting • Function carried out by System A can not be done by System B

Multiple Memory Systems • What is a system? If you “know how to do something” Schacter and Tulving (1994) System Other Name Subsystems Characteristics Procedural Nondeclarative Motor skills Non-conscious operation (indirect) Cognitive skills Simple conditioning Simple associative learning Allows you to automatically recognize things See earlier in the semester Factual information (chpt 10) Memory of events Perceptual representation Nondeclarative Visual word form Auditroy word form Structural description Primary memory Working memory Semantic Generic Spatial Factual Relational Auditory Knowledge Episodic Visual Personal Autobiographical Event memory Conscious operation (direct)

Multiple Memory Systems • Brain areas • Brain imaging studies found that different areas of the brain are used when completing implicit and explicit tasks • Note: more than one structure involved in each type of memory Buckner et al (1995) PET study

Multiple Memory Systems • Brain areas • Brain imaging studies found that different areas of the brain are used when completing implicit and explicit tasks Gabrieli et al (1995) Case study of MS • Studied lists of words • Perceptual identification and recognition task • Intact performance on explicit tests of recognition and cued recall • Intact performance on implicit test of conceptual memory MRI of MS’s brain • Impaired performance on implicit tests of visual perceptual memory • Suggests a specific deficit in visual implicit memory

Multiple Memory Systems • Brain areas • Different kinds of implicit tasks seem to involve different areas • Perceptual vs. conceptual tasks appear to use different brain areas • Conclusion: brain area involvement may be a function of type of processing and type of memory Buckner & Koutstall (1998) f. MRI study

Multiple Memory Systems • Stochastic Independence • Hayman and Tulving (1989) • Measure correlation between explicit and implicit task performance • If not correlated (independent), then tasks measure different processes • Forgetting • Tulving et al. (1989) showed a difference in forgetting rate for recognition and fragment completion • Confirmed with other tasks (stem completion)

Multiple Memory Systems • Strengths • Fits well with dissociations found • In patients • In experiments • Problem • Hard to find consensus on what the systems are • May be “too easy” to posit a new system

Accounting for Implicit/Explicit Dissociations • Four major approaches have been proposed • • The Activation view Multiple Memory systems view Transfer appropriate processing view Bias View

Transfer Appropriate Process • The key to good performance is similarity of processes involved in encoding vs. retrieval, be it implicit or explicit, perceptual or conceptual test • Implicit and explicit may refer to different processes, yet the key to performance is matching processes. A consequence: conceptual processing is the common Overlap core in free recall and determines implicit conceptual tasks, Processes retrieval hence performance on at encoding at test success these two types of task should be equal.

Transfer Appropriate Process Assumes: • Performance depends of match between processing at study and processing at test. Analogous to encoding specificity. • Two-types of Processes (Jacoby, 1990) • Data-driven (perceptual) – processing of physical features. • Conceptually-driven (semantic) – processing for meaning Typically confounded, however, it is possible to unconfound test-type from process-type

Mixing Implicit and Explicit Effects • Jacoby (1990) proposed that implicit vs. explicit memory is confounded with two different kinds of memory processes (associated with two kinds of information) Memory system Declarative (Episodic) Mode of Processing Perceptual identification Word Fragment Completion Perceptual (Data-driven) Meaning based (conceptually-driven) Procedural (Priming) Free Recall Recognition

Transfer Appropriate Process Data-driven Conceptually-driven (Perceptual): (Semantic): • fragment completion • word association • stem completion • anagram completion • lexical decision • perceptual identification doctor ? ? • category-instance generation “name a mammal” • general knowledge “The capital of the US is …? ”

Transfer Appropriate Process Blaxton (1989) • Goal to demonstrate • data-driven processing can affect direct tests • data-driven processing do not necessarily affect indirect tests Data-driven Conceptually-driven Direct Graphic-cued Recall Free Recall Indirect Fragment Completion General Knowledge

Transfer Appropriate Process Blaxton (1989) S’s saw or heard lists of words (key IV here) Target word: bashful • • graphic-cued recall: looks like “bushful” free recall frag completion: b_sh_u_ General knowledge: “Name one of the 7 dwarfs” Data-driven Conceptually-driven Direct Graphic-cued Recall Free Recall Indirect Fragment Completion General Knowledge

Transfer Appropriate Process Blaxton (1989) Predictions • Systems view: modality match should affect only indirect tests (if indirect tap separate system, then modality should affect them in the same way) • for both implicit tests: visual > auditory • for both explicit test: visual = auditory Data-driven Conceptually-driven Direct Graphic-cued Recall Free Recall Same pattern of results regardless of modality Indirect Fragment Completion General Visual better than auditory Knowledge for both

Transfer Appropriate Process Blaxton (1989) Predictions • TAP View: modality match should affect data-driven tasks only. (priming depends on match between study/test processing match & not on indirect vs direct): • for both data-driven tests: visual > auditory • for both conceptually-driven tests: visual = auditory Data-driven Direct Indirect Conceptually-driven Graphic-cued Free Recall Visual should General Fragment Completion be Knowledge better than auditory Visual and auditory should be about the same

Transfer Appropriate Process Blaxton (1989) Results Priming Effect (V > A) for datadriven tasks only: • indirect: frag completion • direct: graphemic-cued recall Not all indirect tests display priming effect. • Gen Know (indirect, conceptual): V = A Conclusions Support view that processing rather than system is what is important

Accounting for Implicit/Explicit Dissociations • Four major approaches have been proposed • • The Activation view Multiple Memory systems view Transfer appropriate processing view Bias View

The Bias View • Proposed to account for repetition priming effects. • Prior presentation of an item can bias subsequent processing of the item on later presentations (if you see it once, you are more likely to interpret in the same way later) • Multiple systems attributes this to 3 separate systems, but doesn’t really offer an explanation • TAP’s answer is considered circular (you respond faster the second time because of transfer appropriate processing, which was developed to account for priming effects)

The Bias View • Bias View’s account for repetition priming effects. • Bias entails both cost and benefits • Cost : There will be an advantage if prior processing is appropriate for the current task • Benefits : There will be a disadvantage if prior processing is inappropriate for the current task 2. Second See ambiguous woman 1. First See one of old woman and young woman 3. People are more likely to interpret the ambiguous picture as the same person as the unambiguous picture 1. First See Old Woman Young Woman 2. Second See Ambiguous -> Old Woman Ambiguous -> Young Woman

Comparing theories TAP • Strengths • Processing perspective • No “need” for separate systems (true of Bias view too) • Bias View may be seen as a complement to the TAP view • Weaknesses • Doesn’t explain impact of conscious awareness • Trouble with finer grain distinctions between tasks Multiple Systmes • Strengths • Good fit for deficit data (but may be too easy to propose “new systems”) • Weaknesses • Has troubles with some data showing differential decline in memory performance with aging • Sometimes difficult to make specific predictions in advance

Implicit Memory Summary • Implicit memory is memory without awareness. • Implicit and explicit tasks are not “process pure” • PDP offers a measurement method for processes • Implicit memory is different memory from explicit memory by experimental dissociations. • There is 4 main accounts for implicit memory • Probably a complex relationship between systems and processes

Experiment 2 • Recall that for experiment 2 you each collected data from three participants. • IV – levels • Prediction: our instructions would lead participants to shift their criterion for what counts as old vs. new. • Signal detection analysis

Signal Detection Theory • Recognition accuracy depends on: • Whether a signal (noise/target memory) was actually presented • The participant’s response INCORRECT • Thus, there are four possible outcomes: • Hits • Correctly reporting the presence of the signal • Correct Rejections • Correctly reporting the absence of the signal • False Alarms • Incorrectly reporting presence of the signal when it did not occur • Misses • Failing to report the presence of the signal when it occurred

Signal Detection Theory • Calculating d’ and C (or β) • Discriminability (d’): • Step 1) Look up the z-score for the average Hit and False Alarm rates. • Step 2) Apply the formula d’ = z. HIT – z. FA, where z. FA is the z-score for FAs and z. HIT is the z-score for Hits. • Criteria C (or β): • Take the negative of the average of z. HIT and z. FA. This is the criterion value C. • Remember that positive C values indicate a conservative response bias, while negative C values indicate a liberal response bias. http: //memory. psych. mun. ca/models/dprime/

Experiment 2 Neutral Conservative Liberal “Old” Target Lure Hit 15. 05 0. 75 False Alarm 2. 48 0. 12 Target Lure Hit 12. 05 0. 60 False Alarm 1. 14 0. 06 Target Lure Hit 16. 95 0. 85 False Alarm 4. 38 0. 22 N=21 per condition Total possible hits or false alarms = 20 Averages Proportions (avg/20)

Experiment 2 http: //memory. psych. mun. ca/models/dprime/ Neutral Conservative Liberal “Old” Target Lure Hit 15. 05 0. 75 False Alarm 2. 48 0. 12 Target Lure Hit 12. 05 0. 60 False Alarm 1. 14 0. 06 Target Lure Hit 16. 95 0. 85 False Alarm 4. 38 0. 22 d’ = 1. 85 C = 0. 25 d’ = 1. 81 C = 0. 65 d’ = 1. 81 C = -0. 13

Experiment 2 probability d’ Neutral d’ = 1. 85 C = 0. 25 Signal (remember) Noise stimulus intensity Conservative d’ = 1. 81 C = 0. 65 Liberal d’ = 1. 81 C = -0. 13

Experiment 2 - Criterion + probability New Neutral d’ = 1. 85 C = 0. 25 Old Signal (remember) Noise stimulus intensity Conservative d’ = 1. 81 C = 0. 65 Liberal d’ = 1. 81 C = -0. 13

Experiment 2 - Criterion + probability New Neutral d’ = 1. 85 C = 0. 25 Old Signal (remember) Noise stimulus intensity

Experiment 2 - Criterion + probability New Old Signal (remember) Noise stimulus intensity Conservative d’ = 1. 81 C = 0. 65

probability Experiment 2 - Criterion + New Old Signal (remember) Noise stimulus intensity Liberal d’ = 1. 81 C = -0. 13