What your choice of best ever footballer says about human memory

Cruijff - the best ever player?

Ask a friend to name the best ever footballer and they're likely to pick someone who was mid-career when they (your friend) was aged around 17. That's according to a new investigation into the "reminiscence bump". This term describes the fact that when you ask people to name the most memorable events in their lives, they tend to refer to things that happened to them in their teens and early twenties. Recently it's been shown that a similar effect occurs when you ask people to name their favourite music, books and films, with them tending to pick out content from their youth. Now David Rubin and his colleagues have extended this line of research to people's judgement of the best footballers of all time.

Six hundred and nineteen people (aged 16 to 80) took part in the study online, conducted in Dutch and hosted on the website of the University of Amsterdam. Participants were presented with the names of 190 all-time leading football players and asked to name their judgement of the five best players of all time. They could either select from the list or choose their own.

The researchers calculated the mid-career point of the 172 players named by the participants and compared this against the participants' age at that time. Participants overwhelming tended to name players whose career mid-point coincided with participants' teens and early twenties. The modal age (i.e. the most common) of the participants at their chosen players' mid-career was 17 years. The researchers said this was the most appropriate statistic to use because the average (22 years) and median (20 years) stats are more susceptible to the bias to name currently active players.

Another way of reporting the results is to say that participants recalled more players who were mid-career in the second decade of the participants' lives than ones who were mid-career in the participants' third decade. And they named more players from the period in which they were aged 11 to 30 than from the period in which they were aged 1 to 10 or aged 31 to 40.

Focusing on the most frequently chosen players, Johan Cruijff was most often selected by participants who were aged 9 to 18 when he was at his career midpoint; Pelé was most often selected by participants who were aged between 12 and 21 years when he was mid-career. Incidentally, currently active players who made the list of twenty most frequently chosen players were: Lionel Messi, Cristiano Ronaldo, Ronaldinho, Ronaldo and David Beckham (go Becks!). Only the youngest cohort (born between 1986 to 1995) chose more players who were mid-career in 2000s than players who were mid-career in the 90s.

"The results of this study are another example of the robustness of the reminiscence bump phenomenon," the researchers said.

Several theories have been put forward to explain the reminiscence bump, including that our memories are more efficient in our teens and twenties. Others think it's because more novel things happen to us at that time of life, such as our first kiss or first job, causing them to get lodged in memory. Rubin and his team say their findings are inconsistent with this "cognitive account", as it's known, because children typically start to play and follow football between the ages of 5 and 15, so if the cognitive account were true you'd think they'd pick players who were mid-career at that time.
_________________________________

Janssen, S., Rubin, D., and Conway, M. (2011). The reminiscence bump in the temporal distribution of the best football players of all time: Pelé, Cruijff or Maradona? The Quarterly Journal of Experimental Psychology, 1-14 DOI: 10.1080/17470218.2011.606372

Post written by Christian Jarrett for the BPS Research Digest.

Did I do that? The psychology of alcohol-induced blackouts

The morning after. Photo credit: Sophie Scott

When our autobiographical memory lets us down, how do we reconstruct the lost chapters? Two psychologists Robert Nash and Melanie Takarangi have identified the perfect population for investigating this very question. "Colleges and universities teem with amnesiacs of a sort," they write, referring to the large number of students who drink quantities of alcohol sufficient to wash away whole episodes from memory.

Nash and Takarangi surveyed 280 students about their alcohol-induced blackouts and found the students were highly motivated to reconstruct what happened. In fact, their desire to fill in the blanks often led them to rely on unreliable sources, such as drunk friends, and to therefore form false memories of the blacked-out period. "Such errors could have enormous impact," the researchers said, "not least because during blackouts people engage in ... risky behaviours such as drug use, fighting and sexual intercourse."

Of the surveyed students, 85 per cent described themselves as drinkers and 61 per cent reported having experienced a total or partial memory blackout whilst drunk. Men were more likely to have had a blackout than women (75.4 vs. 56.7 per cent).

The researchers presented the students with a hypothetical party scenario in which they'd experienced a blackout and asked them to say how motivated they'd be to try each of eight strategies for filling in the blanks. Unsurprisingly, the students tended to say they were motivated most strongly to seek the help of a sober friend who'd been there. Other favoured strategies included: checking photos or videos, consulting a drunk friend who'd been present, and thinking hard about what had happened. Less favoured were: returning to the scene of the party, asking a sober or drunk party guest other than a friend.

Comparing students who'd experienced blackouts in real life with those who hadn't, an intriguing difference emerged - the blackout sufferers were more motivated to rely on drunk friends and there was a slight trend for them to judge drunk friends as more reliable. Blackout sufferers also judged drunk non-friends as more reliable than did non-sufferers.

Turning to the students' reports of how they'd actually attempted to reconstruct boozy blanks in real life, consulting a drunk friend was more common than consulting sober people (77 per cent vs. 69.6 per cent). Forty-three per cent said they'd seen a photo or video of what had happened on at least one occasion; 20.9 per cent had found other physical evidence.

The blackout sufferers said that their reconstructions of boozy blanks sometimes turned out later to have been inaccurate - 16.9 per cent admitted to this having happened, and they said the most frequent reason was relying on drunk friends. Some of the students (11.5 per cent of blackout sufferers) said they'd previously had confidence in the incorrect account of what had happened; 3.4 per cent said they'd actually formed (false) memories for events that hadn't happened.

A curious paradox to emerge in the results was that students who said they'd relied on drunk friends in the past were more likely to admit having been exposed to misinformation, but at the same time were more confident in the future reliability of drunk friends and non-friends. The researchers speculated that perhaps drunk friends had been the only source of information in the past and "because people are highly motivated to reconstruct forgotten experiences, it is possible that such circumstances might encourage individuals to believe that the available sources of evidence are more reliable" - a kind of self-serving bias.

Finally, Nash and Takarangi asked the students if they'd ever knowingly given blackout sufferers false information about blanks in their memories. Seventy-six per cent of the sample said they might have unintentionally done so; 13.7 per cent said they'd deliberately made up details; 7.1 per cent had fabricated an entire event.

The researchers end their study on a sombre note. "We can only speculate about the consequences that blackout sufferers' false beliefs and memories could have in some cases," they said. "For instance, archival studies suggest that numerous innocent people have confessed to crimes after being led to believe they committed acts while drunk, and flawed reconstructions might also lead blackout sufferers to make false accusations against others."
_________________________________

Nash, R., and Takarangi, M. (2011). Reconstructing alcohol-induced memory blackouts. Memory, 19 (6), 566-573 DOI: 10.1080/09658211.2011.590508

Post written by Christian Jarrett for the BPS Research Digest.

What triggers an Earworm - the song that's stuck in your head?

PYT was triggered by the letters EYC 

The brain has its own jukebox. A personal sound system for your private listening pleasure. The downside is that it has a mind of its own. It often chooses the songs and it frequently gets stuck, playing a particular tune over and over until you're sick of it. Psychologists have nicknamed these mental tunes "earworms" (from the German Ohrwurm). A study from 2009 found that they can last anywhere between minutes to hours, but that they're only unpleasant in a minority of cases. Now a team led by Victoria Williamson, in partnership with BBC 6 Music and other international radio stations, has surveyed thousands of people to try to find out the various triggers that cause earworms to start playing. Radio listeners and web visitors were invited to fill in an online form or email the station about their latest earworm experience and the circumstances that preceded it.

Just over 600 participants provided all the information that was needed for a detailed analysis. Predictably, the most frequently cited circumstance was recent exposure to a particular song. "My bloody earworm is that bloody George Harrison song you played yesterday," one 6 Music listener wrote in. "Woke at 4.30 this morning with it going round me head. PLEASE DON'T EVER PLAY IT AGAIN." In relation to this kind of earworm-inducing exposure, the survey revealed the manifold ways that we come into contact with music in modern life, including: music in public places, in gyms, restaurants and shops; radio music; live music; ring tones; another person's humming or singing; and music played in visual media on TV and on the Internet.

However, a song doesn't have to be heard to worm its way inside your head. Many listeners described how earworms had been triggered by association - contact with certain people, rhythms, situations, sounds or words - sometimes with quite obscure links. "On my journey, I read a number plate on a car that ended in the letters 'EYC' which is NOTHING LIKE 'PYT' (by Michael Jackson)," said another listener, "but for some unknown reason, there it was - the song was in my head."

Memories also triggered earworms - for example, driving along the same stretch of road that a song was first heard. And also anticipation. Another listener had "Alive" by Pearl Jam stuck in their head in the days before attending a Pearl Jam concert.

Mood and stress were other triggers. "Prokofiev 'Montagues and Capulets' opening theme. I was writing an email about a distressing subject. I suspect the mood of the piece matched my mood at the time," said an amateur musician. Another listener had Michael Jackson's Man in the Mirror playing in her mind ever since she'd been thinking about the star non-stop and feeling sad (the survey coincided with his death in 2009).

A final theme to emerge from the survey was the way that earworms start playing when we're in a "low attention state", bored or even asleep. "My earworm is 'Mulder and Scully' by Catatonia. In fact I dreamt about running through woods and this was the sound track in my head," said a 6 Music listener. Another survey respondent experienced K'naan "Waving Flag" when mind wandering through a monotonous lab task.

Theoretically, Williamson and her colleagues said earworms can be understood as another manifestation of what Ebbinghaus in the nineteenth century identified as "involuntary memory retrieval". They could even provide a new window through which to study that phenomenon.

"While musical imagery is a skill that many (especially musicians) can utilise to their advantage, involuntary musical imagery (INMI) is an involuntary, spontaneous, cognitive intrusion that, while not necessarily unpleasant or worrying, can prove hard to control," the researchers concluded. "The present study has classified the breadth of circumstances associated with the onset of an INMI episode in everyday life and provided insights into the origins of the pervasive phenomenon, as well as an illustration of how these different contexts might interact."

What about you? What earworms have you experienced lately and what was the context? Please use comments to share your earworm experiences.
_________________________________

Williamson, V., Jilka, S., Fry, J., Finkel, S., Mullensiefen, D., and Stewart, L. (2011). How do "earworms" start? Classifying the everyday circumstances of Involuntary Musical Imagery Psychology of Music DOI: 10.1177/0305735611418553

Link to Earwormery, the website used by the authors of this study to survey participants' experiences.
Link to previous Digest item on earworms, "A natural history of the Earworm - the song that won't get out of your head."
Link to previous Digest item: "Hearing music that isn't there."

Post written by Christian Jarrett for the BPS Research Digest.

How walking through a doorway increases forgetting

Like information in a book, unfolding events are stored in human memory in successive chapters or episodes. One consequence is that information in the current episode is easier to recall than information in a previous episode. An obvious question then is how the mind divides experience up into these discrete episodes? A new study led by Gabriel Radvansky shows that the simple act of walking through a doorway creates a new memory episode, thereby making it more difficult to recall information pertaining to an experience in the room that's just been left behind.

Dozens of participants used computer keys to navigate through a virtual reality environment presented on a TV screen. The virtual world contained 55 rooms, some large, some small. Small rooms contained one table; large rooms contained two: one at each end. When participants first encountered a table, there was an object on it that they picked up (once carried, objects could no longer be seen). At the next table, they deposited the object they were carrying at one end and picked up a new object at the other. And on the participants went. Frequent tests of memory came either on entering a new room through an open doorway, or after crossing halfway through a large room. An object was named on-screen and the participants had to recall if it was either the object they were currently carrying or the one they'd just set down.

The key finding is that memory performance was poorer after travelling through an open doorway, compared with covering the same distance within the same room. "Walking through doorways serves as an event boundary, thereby initiating the updating of one's event model [i.e. the creation of a new episode in memory]" the researchers said.

But what if this result was only found because of the simplistic virtual reality environment? In a second study, Radvansky and his collaborators created a real-life network of rooms with tables and objects. Participants passed through this real environment picking up and depositing objects as they went, and again their memory was tested occasionally for what they were carrying (hidden from view in a box) or had most recently deposited. The effect of doorways was replicated. Participants were more likely to make memory errors after they'd passed through a doorway than after they'd travelled the same distance in a single room.

Another interpretation of the findings is that they have nothing to do with the boundary effect of a doorway, but more to do with the memory enhancing effect of context (the basic idea being that we find it easier to recall memories in the context that we first stored them). By this account, memory is superior when participants remain in the same room because that room is the same place that their memory for the objects was first encoded.

Radvansky and his team tested this possibility with a virtual reality study in which memory was probed after passing through a doorway into a second room, passing through two doorways into a third unfamiliar room, or through two doorways back to the original room - the one where they'd first encountered the relevant objects. Performance was no better when back in the original room compared with being tested in the second room, thus undermining the idea that this is all about context effects on memory. Performance was worst of all when in the third, unfamiliar room, supporting the account based on new memory episodes being created on entering each new area.

These findings show how a physical feature of the environment can trigger a new memory episode. They concur with a study published earlier this year which focused on episode markers in memories for stories. Presented with a passage of narrative text, participants later found it more difficult to remember which sentence followed a target sentence, if the two were separated by an implied temporal boundary, such as "a while later ...". It's as if information within a temporal episode was somehow bound together, whereas a memory divide was placed between information spanning two episodes.
_________________________________

Radvansky, G., Krawietz, S., and Tamplin, A. (2011). Walking through doorways causes forgetting: Further explorations. The Quarterly Journal of Experimental Psychology, 64 (8), 1632-1645 DOI: 10.1080/17470218.2011.571267

Post written by Christian Jarrett for the BPS Research Digest.

Exploring people's beliefs about their memory problems

We think of memory complaints as being more common among older people. A recent colloquialism has even emerged for older folk to refer good humouredly to their "senior moments". Performance on lab-based memory tests also tends to deteriorate with age. So how come researchers have found that subjective complaints about memory don't correlate reliably with lab-based memory performance? And why are the links between age and subjective memory complaints not as robust as we'd expect?

Part of the answer may have to do with the complicating influence of factors like personality and depression. But to probe deeper, Peter Vestergren and Lars-goran Nilsson have surveyed hundreds of people of various ages about their memory concerns and their perceived reasons for their memory problems.

Three hundred and sixty-one participants (aged 39 to 99) answered a simple question about their memory: "Do you experience problems with your memory?", by choosing between "no problems at all", "small problems", "moderate problems", "big problems", and "very big problems". Anyone answering "moderate problems" or above was categorised as feeling that they had a memory problem, and they were further asked to say what they felt the causes were for their memory problems.

Thirty per cent of the sample said they had memory problems, and the proportion increased with rising age (although age only accounted for 4% of the variance in subjective memory problems). Cited reasons for memory problems fell into three main categories: ageing (26.6%), stress (20.2%), and multi-tasking (12.8%), with the reasons given varying with age. Older participants (aged 69-99) tended to say that ageing was the cause of their problems more often than did middle-aged participants (aged 39-64) - the proportions being 61 vs. 18 %. By contrast, stress and multi-tasking were more often given as reasons by the middle-aged group than the older group (50% vs. 8.3%).

Vestergren and Nilsson think these results could help explain past inconsistencies in the literature. Perhaps, they reasoned, subjective memory complaints are more frequent in middle age, versus older age, than we might expect, because of the stress and work demands experienced by people in mid life. The results "may also to some extent explain a lack of relations between subjective and objective measures of memory," the researchers said. "Assuming that many subjective measures of memory are sensitive to transient effects of varying degrees of stress and cognitive load on memory performance, events influenced by these variables will not be replicated by laboratory tests under constant conditions." Based on this, the researchers called on their fellow memory researchers to gauge subjective and objective stress levels and multi-tasking demands alongside their tests of objective memory - to do so will help illuminate instances when subjective memory scores diverge from objective memory.

The current study complements an earlier Digest item that covered diary research into people's memory lapses or "d'oh moments". Healthy participants were found to experience an average of 6.4 such lapses per week - with younger participants actually reporting more than older participants.
_________________________________

Vestergren, P., and Nilsson, L. (2011). Perceived causes of everyday memory problems in a population-based sample aged 39-99. Applied Cognitive Psychology, 25 (4), 641-646 DOI: 10.1002/acp.1734

Post written by Christian Jarrett for the BPS Research Digest.

How psychology helped locate HMAS Sydney II, lost for over 60 years

The next time an ignoramus asks you what psychology has ever achieved, here's a new answer for you: it only helped in the 2008 discovery of the Australian light cruiser HMAS Sydney II, lost in deep water off the west coast of Australia since its sinking in November 1941.

John Dunn and Kim Kirsner have documented in a new paper how they used insights from research into memory transmission to analyse the testimony from the German survivors of the ship, HSK Kormoran, that battled with Sydney not long before both vessels were lost. Whereas, tragically, all the crew of Sydney perished, 317 of the German crew survived and many were interrogated by Australian authorities about what happened. Finding Kormoran was the key that would unlock the location of Sydney, as the ships were proximate at the time of their sinking.

Dunn and Kirsner applied many principles from cognitive psychology to the testimony provided by the German survivors, which included 72 references to the last known location of Kormoran, many of them contradictory. One of these principles is that as memory becomes degraded, either over time in an individual, or through transmission from one person to another - it becomes progressively influenced by a person's top-down expectations and expertise. Consider a study in which participants were asked to recall pictures of fruit and veg, some portrayed larger, some smaller, than their real-life sizes. People's memories for the pictures were distorted in the direction of prior knowledge, so that large vegetables were recalled as having been portrayed as larger.

Based on this idea, and with reference to the status and opportunity of the various witnesses, Dunn and Kirsner identified seven "source statements" about the location of Kormoran which had informed the testimony of the other witnesses and been (further) distorted by them. For example, one of the statements, now known to be inaccurate, was from the Kormoran captain Theodor Detmers.

To confirm this assessment of the available data, the researchers exploited techniques used in the analysis of species evolution, to identify clusters of statements, with each cluster containing statements of various levels of degradation or "mutation" from the key source statements. Once the source statements were confirmed, the researchers tested candidate locations for Kormoran and worked out the potential of each one in relation to its distance from the seven source statements.

A key facet of Dunn and Kirsner's approach was to use all the available testimony to arrive at a prediction of where Kormoran would be found. By contrast, other non-psychological experts involved in the search had tended to rely on just one or two key witnesses, such as Detmers.

By combining the best fit approach from the seven source statements with two further physical landmarks - drift objects lost from Kormoran and an emergency signal sent by Kormoran just prior to battle - Dunn and Kirsner identified a recommended search area. On 16 March 2008, the Finding Sydney Foundation located Kormoran just 5km from Dunn and Kirsner's best prediction of where she lay. Five days later, Sydney was found 21km away. The discovery helped heal a scar in Australia's history.

"The method we developed in response to the problem that was placed before us was necessarily tailored to the specific details of that problem," the researchers said. "Nevertheless, it may provide a blueprint for potential solutions to other similar problems. Such problems may include, but would not necessarily be restricted to, search problems for missing objects. In our view, the critical feature of a problem that would make it suitable for our methodology would be a set of statements or similar data that can be regarded as a set of constraints on a state of affairs that can be evaluated quantitatively. For example, and to move away from the present spatial domain, a relevant problem may involve the evaluation of eyewitness descriptions of a particular person, e.g. a criminal."
_________________________________

Dunn, J., and Kirsner, K. (2011). The search for HMAS Sydney II: Analysis and integration of survivor reports. Applied Cognitive Psychology, 25 (4), 513-527 DOI: 10.1002/acp.1735

Post written by Christian Jarrett for the BPS Research Digest.

Why you should go for a brisk walk before revising

The exam season may be over, but here's a simple piece of advice for next semester. Go for a brisk walk before studying and your memory of the material is likely to benefit.

Carlos Salas and his colleagues had dozens of students study 30 nouns, each displayed for 6 seconds. Some of the students went for a ten-minute walk before being presented with the words. They were told to adopt "the walking speed one would use when late to an appointment, but without the anxiety caused by such a scenario". Other students spent the same time sitting quietly looking at pictures of natural landscapes. After the study phase, some of the students went for another ten-minute walk before attempting to recall as many of the words as they could; other students sat quietly for ten minutes before their recall attempt. This meant there were four experimental groups (walk-walk, walk-sit, sit-sit, and sit-walk, depending on how the participants behaved before the study and recall phases).

The key finding is that those students who went for a walk before the study period recalled 25 per cent more words correctly compared with students who sat still before the study period. By contrast, walking versus sitting before the attempt at recall made no difference to the students' performance.

Past research has shown context-dependent effects on memory. For example, if you chew gum while learning, your recall performance will benefit if you also chew gum when attempting to retrieve memories. No evidence for this was found in this study in the sense that the students' performance was no better when their pre-recall activity (walk vs. sit) matched their pre-learning activity, perhaps because the recall test followed too soon after the learning phase, so that the effects of the earlier walk or sitting period were still ongoing.

Another detail of this study is that the researchers asked the students to report their levels of arousal and tension after the periods of sitting or walking. Arousal was higher after walking than sitting, but tension was no different. So increased arousal is a possible physiological mechanism underlying the benefits of a pre-study walk (see earlier Digest item: "Memory performance boosted while walking").

Salas and his team also looked at meta-memory: this is people's insight into their own memory processes. During the study phase, after each word appeared, the participants were asked to indicate their likelihood of recalling it correctly. Students who sat for ten minutes before studying tended to significantly overestimate their later performance. By contrast, the walkers were much more accurate. However, there was no absolute difference in the predictions made by the two groups. In other words, it seems the walkers only had superior meta-memory because walking boosted their performance to match their confidence.

"Overall, these results suggest that individuals can gain a memory advantage from a ten-minute walk before studying," the researchers said. "Given [these] positive results ... and [their] potentially important practical applications, we hope that researchers will continue to explore the relationship between walking, memory, and meta-memory."
_________________________________

Salas, C., Minakata, K., and Kelemen, W. (2011). Walking before study enhances free recall but not judgement-of-learning magnitude. Journal of Cognitive Psychology, 23 (4), 507-513 DOI: 10.1080/20445911.2011.532207

If you liked this post, you might also like our round-up of 9 evidence based study tips.

[This post was written by Christian Jarrett for the BPS Research Digest.]

Improving people's memory by punishing their correct answers

A well-established finding in psychology is that successfully retrieving information from memory serves to consolidate the storage of that information. Each time your brain's filing clerk tracks down the right information, the more likely he is to find it another time. Psychologists call this the testing effect - practising retrieval of information is far more effective than simply re-studying that same material.

Can this benefit of testing be enhanced? Yes it can. A new study has provided the first ever demonstration of how to enhance the memory consolidation that occurs after correctly answering a test question. Bridgid Finn and Henry Roediger's important and somewhat surprising new finding is that following a correct answer with an aversive stimulus serves to enhance the consolidation of that memory. It's like punishing the filing clerk after each correct retrieval makes him even more accurate in the future.

Forty undergrads studied multiple lists of ten word-pairs, each featuring a Swahili word and its English translation. After each list of ten, they were tested. Presented with the Swahili, they had to answer with the English. Here's the important bit. If they answered correctly, one of three things happened immediately: a blank screen appeared, a neutral picture appeared (e.g. a fork) or a negative, aversive picture appeared (e.g. a dead cat).

After this pattern of study period and test had been followed for ten lists of ten word-pairs, the participants were then given a jumbo test of all 100 Swahili words. Here's the key result: for those items answered correctly in the earlier mini-tests, it was those that were followed by a nasty picture that were most likely to be accurately recalled in the final jumbo test. Earlier correct answers that had been followed by a neutral pic or blank screen were not so well remembered (and performance was equivalent across the blank/neutral conditions).

"These data are the first to show that arousal following successful retrieval of information enhances later recall of that information," the researchers said.

A follow-up study was similar to the first but this time correct answers in the initial mini-tests were followed by neutral or aversive pictures that appeared two seconds later, as opposed to appearing immediately as they did in the first study. This was to see if there was a narrow window beyond which a negative stimulus wouldn't any longer enhance the consolidating effect of correct retrieval. The results were just the same as for the first study, so even two seconds later, a nasty picture is still able to enhance the memory consolidating effect of a correct retrieval. Future studies are needed to test just how long after a correct retrieval this process is still effective, and to see if positive images exert a similar benefit.

Finally, the researchers looked to see if the presentation of a negative pic has its memory enhancing effect after items are merely re-studied, as opposed to recalled. A similar protocol with Swahili-English word pairs was followed as before, but this time, instead of mini-tests after each set of ten word pairs, the participants were simply given the pairs to study again, with each pair proceeded either by a blank screen, neutral picture or nasty picture. This time, there was no benefit of the negative pics. In fact, there was a trend for pairs to be recalled less often if they'd been followed by a nasty pic in the earlier study phase.

Why should negative images boost the consolidating effects of answering a test item correctly? Finn and Roediger aren't sure but think it has to do with links between the amygdala, which is involved in fear learning, and the hippocampus - a brain area involved in long-term memory storage. This is a rather vague account and doesn't explain why aversive stimuli only enhance memory after correct retrieval, not further study. By way of further context, a 2006 study showed the presentation of aversive images after to-be-learned stimuli was beneficial during the initial study of that material.

Post-script

I couldn't help wondering what Milgram would have made of this study. Recall that participants in his classic obedience research thought they were taking part in an investigation of the effects of punishment on learning. In Milgram's mock set-up, the "learner" was subjected to an electric shock each time they answered incorrectly. Of course, Milgram wasn't really studying memory, but this new article suggests that he could have been onto something. Somewhat paradoxically, though, it seems it's correctly answered items that ought to be followed by an aversive stimulus, not incorrect answers.
_________________________________

Finn, B., and Roediger, H. (2011). Enhancing Retention Through Reconsolidation: Negative Emotional Arousal Following Retrieval Enhances Later Recall. Psychological Science, 22 (6), 781-786 DOI: 10.1177/0956797611407932

This post was written by Christian Jarrett for the BPS Research Digest.

Has the Internet become an external hard drive for the brain?

Last year's annual question posed by Edge was "How is the Internet changing the way you think?" Several psychologists answered that it was becoming an extension of their minds. "The Internet is a kind of collective memory,’ wrote Stephen Kosslyn (Harvard University). "When I write with a browser open in the background, it feels like the browser is an extension of myself."

A research team led by Betsy Sparrow has now tested the idea that the Internet really has become a kind of memory prosthesis. First they showed that difficult questions prompted dozens of undergrad participants to think automatically of computers and search engines. Participants tackled either easy or difficult trivia questions and then completed a version of the classic Stroop task: they had to look at a series of words and say what colour ink they were written in. After difficult questions, participants were extra slow at naming the colour of words like "Google". This is a sign that the search engine concept was salient in their minds and therefore interfered more with the process of colour naming.

Next, a group of dozens more undergrad participants read 40 trivia statements and then typed them into a computer. Half the participants were told that the computer would save their entry, the others were told the entries would be deleted. Participants in the "saved" condition performed worse at a subsequent recall test of the statements, as if they'd relied on the computer as an external memory store. Half the participants in both conditions had been instructed explicitly to try to remember the statements, but this made no difference to their memory performance. "Participants were more impacted by the cue that information would or would not be available to them, regardless of whether they thought they would be tested on it," the researchers said.

In another task, a group of participants read trivia statements and then typed them out, with a message telling them which folder the statement had been saved in. Ten minutes later they wrote out as many of the statements as they could, and then they attempted to recall which folder each statement, identified by a single prompt, had been saved to (e.g. "What folder was the statement about the ostrich saved in?"). The striking finding here is that participants were better at remembering the location of the statements than the statements themselves. What's more, they were more likely to remember the location of statements which they'd failed to recall. It's as if we've become adept at using computers to store knowledge for us, and we're better at remembering where information is stored than the information itself.

"This is preliminary evidence that when people expect information to remain continuously available (as we expect with Internet access), we are more likely to remember where to find it than we are to remember the details of the item," the researchers said. "One could argue that this is an adaptive use of memory - to include the computer and online searches as an external memory system that can be accessed at will."

Post-script

The issue of whether and how the Internet is changing our brains and the way we think tends to generate a lot of hyperbole and hot air. There is in fact a long history of technology exciting such reactions. Against that context, it's refreshing to have some new, relevant data (also see here) as opposed to yet more excitable conjecture. However, it's important to keep these new findings in perspective: they hint at how the Internet could be altering our memory habits, but they haven't demonstrated that this is any different from other forms of memory support. For example, similar results might have been obtained if trivia statements had been written in notebooks or told to friends, as opposed to typed into a computer. Of course it pays to note that the present study didn't actually involve the Internet at all. And there's also no evidence here of any irreversible effects  - our minds are likely adapting to technology all the time, as they do to everything else, but there's no reason they couldn't adapt back again if necessary.
_________________________________

B Sparrow, J Liu, and M Wegner (2011). Google effects on memory: Cognitive consequences of having information at our fingertips. Science : 10.1126/science.1207745

[This post was written by Christian Jarrett for the BPS Research Digest.]

Test how much you know about the reliability of memory

In the latest in a series of investigations into how much people know about eye witness memory, Svein Magnussen and Annika Melinder have compiled 12 questions about memory and put them to 857 licensed members of the Norwegian Psychological Association. The correct answers were based on the latest consensus findings in the field of memory research. The Norwegian psychologists scored an average of just 63 per cent correct, no better than achieved by Norwegian judges (63 per cent) in prior research, and only slightly ahead of the general public (they scored 56 per cent on average).

This blindspot for understanding memory isn't a uniquely Norwegian problem. Past research has established that US and Chinese judges, US law students and undergrads all have limited knowledge about the factors that affect eye witness testimony.

The findings have serious implications for the understanding of memory processes in court, especially the limitations of eye witness accounts. Magnussen and Melinder said their findings support the official guidance of the British Psychological Society's Research Board that being a fully credentialed psychologist does not by itself make someone a memory expert. "A memory expert is someone whose expertise is recognised," states the 2010 version of the report. "Recognition of relevant expertise should usually be in the form of outputs that are publicly verifiable, for example, peer-reviewed publications, other publications, and presentations at professional meetings. Of these, peer-reviewed publications are the most important."

So how would you have fared at the memory quiz? Here are the test items in shortened form:

1) Is a person's confidence in their memories a good predictor of the accuracy of those memories?
2) Is it true that eye witness testimony reflects not just what a witness originally saw and heard, but also other information obtained later on from the police, other witnesses etc?
3) Is a witness's ability to recall minor details about a crime an indication of the accuracy of their identification of the perpetrator?
4) Does intense stress at the time of an event impair the accuracy of the memory of that event?
5) Can their attitudes and expectations affect a person's memory of an event?
6) Does the presence of a weapon tend to impair a witness's memory for a perpetrator's face?
7) Does most forgetting tend to occur soon after an event?
8) Do children have better memories for events than adults?
9) How far back into their childhood can most people remember?
10) Are traumatic memories from childhood that are "recovered" in therapy (having never before been recalled) likely to be false?
11) Are dramatic events more or less likely to be forgotten?
12) Is it possible for a perpetrator to have forgotten their criminal act because they've suppressed that specific memory?

Here are the answers: 1) No, 2) Yes, 3) No, 4) Yes, 5) Yes, 6) Yes, 7) Yes, 8) No, worse, 9) to the age of three to four years 10) Yes, 11) Less, 12) No

So how did you do? As well as their overall relatively poor performance, Magnussen and Melinder found that psychologists performed better than the public on question (1) but actually performed worse on questions (6) and (7). "It is particularly surprising that so few psychologists were familiar with the normal course of forgetting, the classic Ebbinghaus function," they said.

Some of the answers may strike you as more controversial than others. On the question of recovered memories, Magnussen and Melinder wrote: "Repression is not among the mechanisms of forgetting acknowledged by current memory science, and the available evidence does not support the idea of repression." They go on to say that well-controlled prospective studies of childhood sexual abuse victims suggest strongly that memories of abuse are not forgotten.

What about the idea that criminals can't selectively forget a criminal act? Although psychogenic amnesia is a real phenomenon (that is, amnesia in the absence of any detectable brain damage or disease), Magnussen and Melinder argue that these "mnestic blocks" typically cover periods of weeks or even years, not the specific instances in time that are often claimed by offenders.

Another point of clarification is that whilst trauma interferes with the details of a memory, it actually makes that memory more persistent and vivid, hence the apparent contradiction of questions (4) and (11).

The researchers call for better scrutiny of memory expertise by the courts and they lament that "psychlore appears to be a stronger determinant of the theoretical ideas [about memory] than are the results of empirical research", even among the majority of qualified Norwegian psychologists.
_________________________________

Magnussen, S., and Melinder, A. (2011). What Psychologists Know and Believe about Memory: A Survey of Practitioners. Applied Cognitive Psychology DOI: 10.1002/acp.1795

This post was written by Christian Jarrett for the BPS Research Digest.

How reliable is our memory for our own previous intentions

Why did I buy this?

The fallibility of eye-witness memory is well documented. But what about people's memories of their own past intentions? This is an unexplored issue in memory research with real-life implications.

Consider the copyright infringement case in 2002, in which French composer Jacques Loussier sued Eminem, claiming that the track Kill You sampled beats from Loussier's work. Loussier further claimed that the success of the album was due in large part to the popularity of that specific track. Eminem's team responded by conducting a survey of people who'd bought the album in the last three years, only one per cent of whom stated they'd bought the album for the specific song Kill You.

The survey appeared to undermine Loussier's claim, but the trouble is that without any research on the topic, we don't know whether those survey responses can be trusted. Now a team led by Suzanne Kaasa and including Elizabeth Loftus has made a start on plugging this gap in the literature.

Nearly six hundred undergrads answered open-ended questions about why they'd purchased, downloaded or copied their most recently acquired album (the vast majority had acquired one within the last two weeks), and then they provided the same information again six months to a year later. The participants' answers fell into five main categories: because they liked the artist, liked the music, liked a specific song or songs, someone had recommended the album, or they needed the album for a specific purpose.

The key finding was that only one in five participants gave a consistent reason or reasons at both time points. The researchers had anticipated that memory for some reasons might prove more durable over time than others, but this wasn't the case. Overall, the most common form of change was simply to invent new reasons at the later time point. Sometimes participants also forgot reasons they'd mentioned earlier. Unsurprisingly perhaps, participants who recalled more reasons at the first time point tended to be more prone to forgetting reasons when quizzed again later. This was also true of participants who reported liking their CD more, perhaps because they'd felt less need to dwell on their motives at the time they acquired the album.

A subset of 82 of the participants also gave their reasons at a third time point, approximately six months to a year after the second time of questioning. Although still evident, changes in memory between the second and third time points were far reduced compared with between the first and second time points. This is important for real-life legal situations because consistency of answers across later interviews could be interpreted as a sign of memory reliability. 'It appears critical to have an accurate and complete record of the very first interview given by a witness,' the researchers said.

The study had some limitations, including the fact that the precise time between album acquisition and the first questioning session was unknown. However, the researchers observed that 'although individuals may not be able to accurately recall the reasons for their behaviours ... the real world continues to rely on self-reported motivations in a variety of circumstances, including police investigations and court proceedings.'
_________________________________

Kaasa, S., Morris, E., and Loftus, E. (2011). Remembering why: Can people consistently recall reasons for their behaviour? Applied Cognitive Psychology, 25 (1), 35-42 DOI: 10.1002/acp.1639

How is autobiographical memory divided into chapters?

How does the mind file life's episodes?

Autobiographical or 'episodic' memory describes our ability to recall past experiences and is distinct from semantic memory, which is our factual knowledge about the world. So far so good, but according to Youssef Ezzyat and Lila Davachi, psychology until now has largely neglected to investigate exactly how the brain organises the continuity of lived experience into a filing system of discrete episodes.

Ezzyat and Davachi have made a start. They had 23 participants read six narratives containing dozens of sentences about a protagonist performing everyday activities. Each sentence was displayed one at a time on a screen. Crucially, a minority of sentences began: 'A while later ...', thereby conveying a temporal boundary in the narrative; the end of one episode and start of another. For comparison, a small number of control sentences began: 'A moment later ...', indicating that the ensuing sentence was part of the same episode, not a new one.

After a ten minute break, the participants were given a surprise memory test. Presented with one sentence from the earlier narratives, their task was to recall the sentence that had followed. The key finding here was that the participants were poorer at recalling a sentence that came after a temporal boundary. It's as if information within an episode was somehow bound together, whereas a memory divide was placed between information spanning two episodes.

A second study was similar to the first except that nineteen participants had their brains scanned during the initial read-through of the sentences. Ezzyat and Davachi identified patterns of neural activity in distinct regions of the prefrontal cortex and the middle-temporal gyrus that either correlated with within-event processing or with forming boundaries between events. These neural activity patterns were more distinct in those participants who showed larger behavioural effects of episode boundaries in their memory performance.

'Our experiments are an important step toward understanding how event perception and segmentation influence the structure of long-term memory,' the researchers concluded. 'The behavioural results support the hypothesis that event segmentation shapes the organisation of long-term memory; the fMRI [brain scanning] results link these memory effects to brain activity consistent with information maintenance and integration within events.'
_________________________________

Ezzyat, Y., and Davachi, L. (2010). What Constitutes an Episode in Episodic Memory? Psychological Science DOI: 10.1177/0956797610393742

The woman whose new memories are erased each night

Psychologists have documented what they believe to be a clinical first - the case of an amnesic woman whose memory for new material is erased each night that she goes to sleep (movie fans will recognise this as a plot device in the 2004 film 50 First Dates). Referred to as case FL, the woman developed these symptoms after she hit her head in a car accident in 2005, aged 48. Brain scans and neurological exams revealed no signs of brain damage, thus suggesting the woman is exhibiting what's known as psychogenic or functional amnesia - that is, symptoms in the absence of any detectable organic cause.

FL claims that on any given day her memory for newly acquired material is fine until she has a night's sleep, during which the new memories are erased (unlike standard cases of psychogenic amnesia, she says her memories from before her accident are preserved). FL's performance on lab-based memory tests was largely in keeping with her claims, with one key exception. Christine Smith and her team deployed some trickery, intermingling test items (scenes) from earlier in the day with items from previous days. FL's memory for items that she thought were from earlier in the day, but were actually seen on earlier days, was intact and comparable to the memory performance of healthy controls.

So was FL faking it, perhaps in pursuit of a compensation claim? Smith's team don't think so. Although healthy controls who were asked to fake FL's symptoms performed similarly on the memory tests, there were also differences. For example, unlike the healthy fakers, FL showed deficits in motor learning, and her confidence for test items dropped with repeated testing whereas theirs increased.

The researchers' theory is that FL truly believes she has the memory deficit that she describes and that unconscious processes may be involved in its manifestation. FL denied having seen the film 50 First Dates, which was released a year before her accident. However, she admitted that the film's female lead, Drew Barrymore, was her favourite actress, so she may have been aware of its plot. The film 'may have influenced FL's concept of how memory could fail after a car accident', the researchers said. 'The brain uses preexisting concepts of memory and through altered brain function creates a particular constellation of symptoms.'

What about treatment? Reassuring FL that evidence had been found for the intact functioning of her overnight memory proved unsuccessful. What did work was testing the limits of FL's memory-washing system. Thirty-six hours without sleep and her memories were okay. An hour's nap during the day and they were okay. In the end, it was established that FL can sleep at night for up to four to six hours at a time without experiencing the sense that she's lost the day's memories. By setting an alarm each night to wake her after bouts of three and a half hours sleep, FL has managed to overcome her strange condition. 'At our most recent contact (March 2010), she and her husband reported that she continues to use this regimen successfully,' the researchers said.
_________________________________

Smith, C., Frascino, J., Kripke, D., McHugh, P., Treisman, G., & Squire, L. (2010). Losing memories overnight: A unique form of human amnesia. Neuropsychologia, 48 (10), 2833-2840 DOI: 10.1016/j.neuropsychologia.2010.05.025

Further reading: Amnesia at the movies.

Flynn effect for memory could invalidate neuropsychologists' tests

In Western countries, scores on IQ tests have been rising for several decades - the Flynn effect, named after the political scientist James Flynn. Now Sallie Baxendale at the Institute of Neurology has provided evidence that a similar effect has occurred for the standardised memory tests that are used by clinical neuropsychologists, a finding with implications for the diagnosis of memory problems in contemporary patients.

Baxendale focused on the Adult Memory and Information Processing Battery (AMIPB) - 'the most commonly used memory battery amongst clinical neuropsychologists in the UK' - published in 1985, and its successor, the Brain Injury Rehabilitation Trust Memory and Information Processing Battery (BMIOB), published in 2007. The two tests feature different wording and design but they both make equivalent demands: learning and recalling lists of words, and learning and recalling abstract line drawings.

Baxendale compared the performance of the two participant samples that provided the original normative data (the 'norms') for the two tests. These are the healthy participants, spanning four age ranges, whose average performance provides the benchmark for assessing patients. The normative data for the AMIPB was provided in 1985, or thereabouts, by 184 British people aged 18 to 75; the normative data for the BMIPB was collected in 2007 or thereabouts from 300 British people aged 16 to 89.

On one hand, there was little evidence of any difference in average performance on verbal learning and recall between the 1985 and 2007 samples (the exceptions were verbal learning in the 31-45 years age range and verbal recall in the oldest age range, both of which were superior in the 2007 sample). By contrast, visual learning and recall were both superior in the 2007 sample compared with the 1985 sample, at all four age ranges: 16-30; 31-45; 46-60; and 61-75. This is consistent with the traditional Flynn effect, which is most pronounced for non-verbal intelligence tests.

Baxendale said her findings have implications for diagnosis because present-day patients may, pre-trauma or pre-illness, have had elevated non-verbal learning and recall scores in comparison to the old normative data. Therefore, such patients could be impaired relative to their own healthy baseline, and yet appear unaffected compared with the out-of-date normative data. 'This may present a confound for neuropsychologists concerned with the lateralising and localising significance of memory test profiles,' Baxendale said.
_________________________________

Baxendale, S. (2010). The Flynn effect and memory function. Journal of Clinical and Experimental Neuropsychology, 32 (7), 699-703 DOI: 10.1080/13803390903493515

It's never too late to memorise a 60,000 word poem

Pounding the treadmill in 1993, John Basinger, aged 58, decided to complement his physical exercise by memorising the 12 books, 10,565 lines and 60,000 words that comprise the Second Edition of John Milton's epic poem Paradise Lost. Nine years later he achieved his goal, performing the poem from memory over a three-day period, and since then he has recited the poem publicly on numerous occasions. When the psychologist John Seamon of Wesleyan University witnessed one of those performances in December 2008, he saw an irresistible research opportunity.

Seamon and his colleagues tested Basinger's memory systematically in the lab. They provided two lines as a cue and then 'JB' (as they refer to him in their report) had to reproduce the next ten. With the exception of books VII, his least favourite, and XI, JB's performance was uniformly exceptional - regardless of whether the researchers revealed which book and book section the cue lines were from or not, and regardless of whether they tested portions of the poem in sequence or picked them randomly, JB displayed an accuracy of around 88 per cent in terms of correctly recalled words. When mistakes were made, they tended to be omissions rather than altered or added words. The researchers also tested JB's everyday memory and found that in all non-Milton respects it was age-typical.

Seamon and his co-workers claim JB's feat shows that 'cognitive expertise in memorisation remains possible even in later adulthood, a time period in which cognitive researchers have typically focused on decline.'

Just how did JB manage to pull off this incredible feat? He studied for about one hour per day, reciting verses in seven-line chunks, consistent with Miller's magic number seven - the capacity of short-term, working memory. Added together, JB estimates that he devoted between 3000 to 4000 hours to learning the poem. Seamon's team interpret this commitment in terms of Ericsson's 'deliberate practice theory', in which thousands of hours of perfectionist, self-critical practice are required to achieve true expertise.

JB didn't use the mnemonic techniques favoured by memory champions, but neither, the researchers say, should we see his achievement as a 'demonstration of brute force, rote memorisation'. Rather it was clear that JB was 'deeply cognitively involved' in learning Milton's poem. JB explained:

'During the incessant repetition of Milton's words, I really began to listen to them, and every now and then as the whole poem began to take shape in my mind, an insight would come, an understanding, a delicious possibility. ... I think of the poem in various ways. As a cathedral I carry around in my mind, a place that I can enter and walk around at will. ... Whenever I finish a "Paradise Lost" performance I raise the poem and have it take a bow.'

_________________________________

Seamon, J., Punjabi, P., & Busch, E. (2010). Memorising Milton's Paradise Lost: A study of a septuagenarian exceptional memoriser. Memory, 18 (5), 498-503 DOI: 10.1080/09658211003781522

Repression debunked

Psychologists in Denmark have hammered another nail into the coffin containing 'repression' - the idea, made popular by psychoanalysis, that negative, emotional memories are particularly prone to be being locked up out of conscious reach.

Simon Nørby and his colleagues at the University of Copenhagen presented dozens of undergrad participants with word pairs, each made up of a cue word and an unrelated target word. Past research has suggested that people are able to deliberately forget some target words while remembering others. But this has been over very short time periods. Nørby's team wanted to test the effects of deliberate forgetting over a longer time period - a week - and they also wanted to revisit the question of whether emotional words can be deliberately forgotten as easily, or more easily, than neutral words. Past research has suggested they can, but these studies have tended to block emotional word pairs altogether in series of themed trials, thus raising the possibility that their impact may have been diminished by habituation. Nørby's team avoided this problem by jumbling up neutral and emotional words altogether.

The participants spent time learning 70 word pairs, then they were informed which target words were to be deliberately forgotten and which to be retained. An ensuing training process helped them with this. Participants repeatedly gave the target words when presented with cues for to-be-remembered pairs (if they couldn't remember it, they were told the target word), whereas they repeatedly withheld and attempted to suppress target words when presented with the cues for to-be-forgotten pairs. After all this, the participants were tested once again on all the word pairs, with their task to recall even those they had deliberately forgotten.

The results of this immediate test suggested that the participants had succeeded, to some extent, in deliberately forgetting those neutral words that they were supposed to forget. Recall for to-be-forgotten neutral words dropped from a baseline of about 80 per cent to about 70 per cent, whereas accurate recall for to-be-remembered words had increased to 95 per cent (unsurprisingly, the final training phase had acted as memory aid for these words). By contrast, suppressed, to-be-forgotten negative emotional words like 'massacre' and 'incest' remained unforgotten and were recalled just as accurately as to-be-remembered emotional words.

On retesting a week later, to-be-forgotten emotional and neutral words were recalled just as often as to-be-remembered words. In fact, over the course of a week, there was evidence that memory for to-be-forgotten words had deteriorated less than memory for to-be-remembered words. This could be another manifestation of the ironic 'suppression rebound effect' which is the finding that deliberately suppressing certain thoughts can make them come back stronger.

Taken altogether, the results suggest that neutral material can be deliberately suppressed over short time periods, but not for as long as a week. Negatively emotional material, by contrast, appears to be stubbornly resistant to deliberate suppression. This flies in the face of the psychoanalytic idea of repression, but is consistent with trauma research suggesting that emotionally salient memories are more persistent than normal, not less.
_________________________________

Nørby S, Lange M, & Larsen A (2010). Forgetting to forget: on the duration of voluntary suppression of neutral and emotional memories. Acta psychologica, 133 (1), 73-80 PMID: 19906363

Image credit: fancy

Previously on the Digest:

Can we deliberately forget specific parts of what we've read?
How remembering can lead to forgetting.

Intrusive images and intrusive verbal thoughts are different phenomena

The vivid, intrusive visual images that are a hallmark of post-traumatic disorder (PTSD) are based on a separate memory system from intrusive verbal thoughts. That's according to a new study that claims to provide empirical support for psychologist Chris Brewin's dual-representation theory of PTSD.

Brewin's theory posits two memory systems, one that's largely sensation-based, inflexible and automatically accessed and another that's more deliberately accessible, containing material that is contextualised and can be easily put into words. By this account, a traumatic event can end up lodged in the sensation-based memory system, leading to sensory intrusions - 'flashbacks' - of the event being easily triggered by sights, sounds and smells that are reminiscent of the original experience.

The new study involved 79 participants watching traumatic video footage of car crash scenes, including commentary on the accidents and people involved. Crucially, some of the participants were told to keep still while they watched the footage and others were hypnotised so that they couldn't move. Past research has shown that keeping still cranks up the trauma simulation, perhaps because it is reminiscent of being frozen in terror or trapped. A final group were free to move. For a week after watching the car-crash videos the participants kept a diary of intrusive verbal thoughts and visual images associated with the videos. The key finding was that participants who had to keep still while watching the videos had significantly more intrusive visual images than the participants who were allowed to move. By contrast, the number of intrusive verbal thoughts did not differ between the groups.

A second study largely replicated the first, except rather than some participants being allowed to move while others kept still, this time some participants watched a neutral film while others watched the traumatic car-crash film. The traumatic film led to more intrusive visual images, but not more intrusive verbal thoughts, than the neutral film.

In both studies, participants who reported feeling more anxious and horrified after the traumatic videos tended to also experience more intrusive visual imagery. In contrast, intrusive verbal thoughts were not connected to mood effects in this way.

Taken altogether Hagenaars and her team said their findings suggest that intrusive visual imagery is a separate phenomenon from intrusive verbal thoughts and can be manipulated independently. 'Understanding these basic processes is likely to be valuable in developing more effective treatments for PTSD that focus on maximising change in verbal thoughts and intrusive images separately,' they concluded.
_________________________________

Hagenaars, M.A., Brewin, C.R., van Minnen, A., Holmes, E.A., & Hoogduin, K.A.L. (2010). Intrusive images and intrusive thoughts as different phenomena: Two experimental studies. Memory, 18 (1), 76-84 DOI: 10.1080/09658210903476522

Image credit: Symphonie