The Seven Sins of Memory: An Update. Daniel L. Schacter Harvard University

The Seven Sins of Memory: An Update Daniel L. Schacter Harvard University

Henry Molaison (1926-2008)

New Zealand Herald, June 18, 2009 Two men have had a night in the cells to reflect on the importance of Two men have had a night in the cells to reflect on the importance of planning, after one of them forgot his mask and gun while allegedly attempting an armed hold-up.n while allegedly attempting an armed hold-up.

Donald Thomson: A Disturbing Accusation

The Seven Sins of Memory (Schacter, 1999, 2001) Transience: decreasing accessibility of memory over time Absentmindedness: lapses of attention; forgetting to do things Blocking: temporary inaccessibility of stored information Misattribution: attributing memories to incorrect source; false recognition Suggestibility: implanted memories Bias: retrospective distortions produced by current knowledge and beliefs Persistence: unwanted recollections that people cannot forget

Wall Street Journal, April 15, 2014 Why We Keep Losing Our Keys Everyday Memory Lapses Hit at Any Age Losing things is irritating, though you think we d adjust: The average person misplaces up to nine items a day, and one-third of respondents said they spend an average of 15 minutes day searching. Cellphones, keys, and paperwork top the list,according to an online survey of 3,000 people published in 2012 by a British insurance company.

Absentminded Memory Errors Often Result from Mind Wandering Mind wandering occurs when people engage in thoughts that are unrelated to the primary cognitive task at hand (see Smallwood, Psychological Bulletin, 2013) Working memory performance is severely impaired when people mind wander during the task (Kane & McVay, Current Directions in Psychological Science, 2012) Reduced comprehension when mind wandering during reading and poor subsequent memory (Smallwood, McSpadden, & Schooler, Memory & Cognition, 2008)

Mind Wandering Occurs Frequently During Classroom Lectures (for review, see Szpunar, Moulton, & Schacter, Frontiers in Psychology, 2013) Mind wandering during psychology lectures reported in response to ~33% of direct auditory probes sounded during a lecture; more frequent toward end of lecture (44%) than beginning (25%), and negatively correlated with learning of lecture content (Lindquist & McLean, Learning & Individual Differences, 2011).

Is there anything than can be done to reduce mind wandering and increase attention during lectures?

The Testing Effect Testing of memory not only assesses what we know, but changes it (Roediger & Karpicke, 2006). Many studies show that testing can enhance later performance, even more than additional study. Some evidence that giving tests after each of several word lists can enhance attention to studied information and increase performance on a final test. When people expect to be tested, and are tested, attention to target material is increased - fewer memory errors on a final test.

Can Interpolated Testing Reduce Mind Wandering and Improve Retention of Information from a Lecture? (Szpunar, Khan, & Schacter, PNAS, 2013) *Used a videotaped, 21 minute statistics lecture. *Divided lecture into four segments; separated by either brief tests, math problems, or re-study. N=16/group; for tested and re-study groups, 6 items/segment. *Participants in each condition instructed at the beginning that the lecture would be divided into four segments, that it would be randomly determined whether they were tested after each segment, and that there would be a final test. *Also told that they could take notes on the lecture and that they would be probed occasionally about whether they were mind wandering during the lecture.

Figure 1 B Experiment 2 Segment 1 Segment 2 Segment 3 Segment 4 Tested Lecture Math Test Lecture Math Test Lecture Math Test Lecture Math Test Re-Study Lecture Math Re-study Lecture Math Re-study Lecture Math Re-study Lecture Math Test Non-Tested Lecture Math Math Lecture Math Math Lecture Math Math Lecture Math Test Mind Wandering Probes Are you mind wandering? YES or NO Are you mind wandering? YES or NO Are you mind wandering? YES or NO Are you mind wandering? YES or NO

Prop. of Probes Figure 2 Prop. of Slides w/ Notes Proportion Correct B Experiment 2 0,5 0,4 0,3 0,2 0,1 0 Mind Wandering T RS NT 0,3 0,2 0,1 0 Notes T RS NT 1 0,8 0,6 0,4 0,2 0 Correct Responses - 4th Segment T RS NT

Additional findings: *Significantly higher levels of retention on final test of all four segments (5 min delay) for tested group (86%) than restudy group (66%) or nontested group (72%). *Participants in tested group reported feeling less anxious about final test than participants in the other two groups. *Participants in tested group rated their experience of learning the lecture as less cognitively demanding than participants in the other two groups.

Judgments of learning People overestimate learning from videorecorded modules (Choi & Johnson, 2005; Salomon, 1984).

Judgments of learning People overestimate learning from videorecorded modules (Choi & Johnson, 2005; Salomon, 1984). Interpolated tests should provide feedback necessary to calibrate predictions.

Judgments of learning Tested Segment 1 Segment 2 Segment 3 Segment 4 Lecture Math Test Lecture Math Test Lecture Math Test Lecture Math Test Non-tested Lecture Math Math Lecture Math Math Lecture Math Math Lecture Math Math

Judgments of learning 5 Tested Segment 1 Segment 2 Segment 3 Segment 4 Lecture Math Test Lecture Math Test Lecture Math Test Lecture Math Non-tested Lecture Math Math Lecture Math Math Lecture Math Math Lecture Math Test Math m i n u t e F I N A L b r e a k T E S T

Judgments of learning How well will you do? 0-100% 5 Tested Segment 1 Segment 2 Segment 3 Segment 4 Lecture Math Test Lecture Math Test Lecture Math Test Lecture Math Non-tested Lecture Math Math Lecture Math Math Lecture Math Math Lecture Math Test Math m i n u t e F I N A L b r e a k T E S T

Final Test Performance Szpunar, Jing, & Schacter (2014) JARMAC

Szpunar, Jing, & Schacter (2014) JARMAC Final Test Performance 75% 48%

Szpunar, Jing, & Schacter (2014) JARMAC Final Test Performance 77% 75% 78% 48%

Mind Wandering, Interpolated Testing, and Lecture Learning: Summary Study 1: Interpolating brief quizzes in a videorecorded lecture reduced task-irrelevant mind wandering, increased task-relevant behaviors such as note taking, and improved retention. Study 2: Interpolating brief quizzes improved calibration between judgments of learning and actual learning.

Eyewitness Misidentification and Wrongful Conviction: DNA Evidence www.innocenceproject.org Results extend to over 250 cases (B.L. Garrett, Convicting the Innocent, 2011)

False Recognition of Semantic Associates (Deese, 1959; Roediger & McDermott, 1995) Participants study semantic associates that all converge on a non-presented theme word: candy, sour, sugar, bitter, good, taste, tooth, nice, honey, soda, chocolate, heart, cake, eat, pie Tested with words from the study list (taste), unrelated words that were not presented (point), and associative related theme word or critical lure (sweet). Extremely high levels of false alarms to theme word, accompanied by high confidence

True and False Recognition Proportion Old Responses 1 0.8 0.6 0.4 0.2 0 True Target False Target Young Old Norman & Schacter, 1997

Neuroimaging of True vs. False Memories

Schacter et al. (Neuron, 1996; see also Cabeza et al., PNAS, 2001 )

True Recognition > False Recognition Abe N et al. Cereb. Cortex 2008;18:2811-2819

True and False Recognition of Visual Shapes (Slotnick & Schacter, Nature Neuroscience, 2004) Subjects study 16 sets of 9 exemplars of a prototype. Old/new recognition test including old, related new, and unrelated new Nonstudied Prototype Exemplar Exemplar

% Signal change True and False Recognition 0.2 0.1 0-0.1 BA 39/40 0 4 8 12 16 Time (sec) Old-hits Related-FAs New-CRs Prefrontal (BA45/46/47, BA8/9) Parietal (BA7/39/40) Hippocampus True recognition False recognition Slotnick & Schacter (2004, Nature Neuroscience)

% Signal change True recognition > False recognition Ventral View Left fusiform gyrus (BA18) Early visual regions (BA17, BA18) 0.3 0.2 0.1 0-0.1-0.2-0.3 0 4 8 12 16 Time (sec) LH Old-hits Related-false alarms X Old-hits > Related-false alarms Related-false alarms > Old-hits

Nature Neuroscience, February 2013

Neuroimaging of True/False Memory: Issues Related to Legal Application -Relation of lab results to real-world memories: Short delay between encoding & retrieval in the lab; typically much longer delays in legal cases. -Target populations: Lab studies of true vs.false memories have typically used undergraduate participants; more diverse populations encountered in the courtroom. -Averaging: Groups vs. individuals; collections of items/events vs. particular items/events. -Countermeasures: Can strategies be used to beat the test?

Memory Reactivation and Misattribution

Memory Reactivation and Misattribution *Reactivating a memory for a past experience can affect subsequent memory for that experience *Reactivation can increase subsequent memory, e.g., the testing effect *But reactivation can also bring a memory into a labile state where it is prone to disruption and possible distortion: reconsolidation (for review, see Hardt et al., 2010, Annual Review of Psychology) *Little known about how varying properties of reactivation affects true or false memories. St. Jacques & Schacter (2013, Psychological Science) investigated the issue using a novel real-world memory paradigm

Participants went on guided tours of exhibits in adjoining Harvard Museum of Natural History and Peabody Museum. Tour guided them to view some items in each exhibit (total=32), and not others. During the tour, participants wore a Vicon revue camera, which takes photos every 15 seconds.

*48 hours after tour, participants came to Memory Lab for reactivation session, in which they viewed an event movie consisting of photos of exhibit items they had seen on the tour, and made a judgment about the relatedness of an item from the same exhibit that they had not seen. *Properties of reactivation were manipulated. *48 hours after reactivation, participants were given a recognition test that included items from the tour, reactivation session, and novel baseline items.

Manipulating Reactivation

Memory Reactivation: Results

NEURAL CORRELATES OF AM RETRIEVAL Lateral Prefrontal Cortex (PFC) Posterior Parietal Cortex Retrosplenial Cortex Medial Temporal Lobe (hippocampus & parahippocampus) Medial Prefrontal Cortex (PFC) St. Jacques, Kragel & Rubin (2011). Neuroimage; Cabeza & St. Jacques (2007). Trends in Cognitive Sciences.

Lateral Prefrontal Cortex (PFC) NEURAL CORRELATES OF AM RETRIEVAL Posterior Parietal Cortex Retrosplenial Cortex Medial Temporal Lobe (hippocampus & parahippocampus) Medial Prefrontal Cortex (PFC) Parahippocampal, Retrosplenial & Posterior Parietal Cortices:Recollection/ Reactivation Quality (e.g., Ranganath & Ritchey, 2012, Nature Reviews Neuroscience) St. Jacques, Kragel & Rubin (2011). Neuroimage; Cabeza & St. Jacques (2007). Trends in Cognitive Sciences.

FMRI PREDICTIONS 1. Reactivation-related effects on subsequent True Memory and False Memory supported by brain regions associated with recollection: Parahippocampal, Retrosplenial Cortex & Posterior Parietal Cortex

Brain Activity in AM Regions During Reactivation Session Predicts Subsequent True and False Memories St. Jacques, Olm & Schacter (Proceedings of the National Academy of Sciences, 2013)

Conclusions 1. Quality of reactivation modulates subsequent memory effects. 2. Reactivation-related enhancement and distortion is associated with neural recruitment in regions sensitive to the quality of reactivation.

Are the Seven Sins Really Sins? (Schacter, 1999, 2001) Transience: decreasing accessibility of memory over time Absentmindedness: lapses of attention; forgetting to do things Blocking: temporary inaccessibility of stored information Misattribution: attributing memories to incorrect source; false recognition Suggestibility: implanted memories Bias: retrospective distortions produced by current knowledge and beliefs Persistence: unwanted recollections that people cannot forget

Misattribution: An Adaptive Sin? 1999/2001 view: Many misattributions occur because we typically don t need to remember every detail of every experience. Sometimes we re better off remembering the gist, which can promote retention of meaning and general themes of experience, but sets the stage for misattribution errors. More recent developments: New evidence for an adaptive view of misattribution - e.g., in the museum paradigm, false recognition may reflect updating of memory (see also Loftus on the misinformation effect).

Schacter, Guerin, & St. Jacques(2011). Trends in Cognitive Sciences

A new hypothesis: Episodic memory - traditionally defined as the ability to recollect our past experiences - is important for the future as well as the past. We use episodic memory to imagine possible future scenarios, and this constructive activity requires flexible access to elements of past experience that may sometimes produce misattributions and related memory distortions (Schacter & Addis, 2007, Philosophical Transactions of the Royal Society; Schacter, 2012, American Psychologist)

Acknowledgements Memory Lab & Collaborators Helen Jing Roland Benoit Kevin Madore Haley Dodds David Maillet Peggy St. Jacques Karl Szpunar Alexis Carpenter Dahua Wang Taiga Abe Anne Crosby Zoe Galindo Alexis Gomez Stella Tu Scott Guerin Scott Slotnick Sam Moulton Supported by NIH and Harvard Initiative in Learning & Teaching (HILT)