The Hand LabMeNeuroBiographyWorkTeachingStatisticsMiscellaneous
A random quotation (don't like it? refresh!)

"Absurdity, n. a statement or belief manifestly inconsistent with one's own opinion"

Ambrose Bierce

My latest work:

Holmes NP (unpublished) Multisensory integration: inverse effectiveness and noisy signals. Current Biology, 0:0
  [NBArticle #45652] [Cites 1]

Reader A, Tamè L, Holmes NP (unpublished) Two instances of presyncope during magnetic stimulation of the median nerve, and evaluation of resting motor threshold with transcranial magnetic stimulation (tms). , 0:0
  [NBArticle #45584]

Blanchard CCV, McGlashan HL, French B, Sperring RJ, Petrocochino B, Holmes NP (2017) Online control of prehension predicts performance on a standardised motor assessment test in 8-12 year old children. Frontiers in Psychology, 8:374
    [NBArticle #41777] [Cites 43] [CitedBy 3]

Paper of the Week

Every now and again, I choose one paper that I've read to highlight and briefly review*

3810th Jun, 2011Silvia PJ (2006) How to Write A Lot: A Practical Guide to Productive Academic Writing. APA Books
  [NBArticle #32411]
I am the target of this book, and I have made all its targeted mistakes. I complain. I make excuses. I blame my computer being old, I say that I will allocate blocks of time - weekends, summer holidays - to make up for it. I will do it tomorrow. What have I done wrong? I've written nothing in two years, that's what. If you are an academic or researcher, then you are likely the same. Why aren't you writing? Hard, you say? Can't do it? Waiting for inspiration? Myths all. If you aren't writing but should be, read this short book today. Then start writing tomorrow. Make a schedule, sit down, move your fingers up and down over a keyboard. Write.
3730th Jan, 2011Perfiliev S, Isa T, Johnels B, Steg G, Wessberg J (2010) Reflexive limb selection and control of reach direction to moving targets in cats, monkeys, and humans. Journal of Neurophysiology, 104(5):2423-2432
    [NBArticle #31760] [Cites 37]
If you've ever watched a cat swiping a ball on the end of a string, leftwards and rightwards, rightwards and leftwards. And if you've ever wondered how they do it or why they do it. And if you're a scientist who thinks it would be cute to put pictures of such a cat in a manuscript and send it to the Journal of Neurophsyoiology, then you are probably Sergei Perfiliev. Not content, as many researchers are, with studying only humans, or only monkeys, or only cats, Sergei and co have taken a very simple task (reaching very quickly to grasp a suddenly moving target ball) and studied how these three animal species perform it. Simple, but effective. This article is worth downloading just to see the cats, but the scientific message is appealing: Whether human, cat, or monkey, we reach and grasp red balls in a very similar way!
3615th Aug, 2010Maunsell JHR (2010b) Announcement regarding supplemental material. Journal of Neuroscience, 30(32):10599-10560
  [NBArticle #31101]
Last week, I watched a 'video abstract' for an article published in 'Current Biology'. The idea is simple: Rather than forcing people to read about your work, you can simply record a video abstract of it, and your audience can watch it during the TV commercials, over a cup of coffee, or while waiting for their hair to dry. Yesterday I found myself wondering what's next in this race to the bottom, and a friend suggested that, every day at 12:00 local time, the authors of an article should be available online for a Skype abstract, where interested IMers can chat about the article. Tweets would obviously come a little later in the afternoon, when the short of attention span are awake.

But it doesn't have to be like this. We can change. We can fight this progressive neophilia and the infantilisation of science. We can go Back. Backwards, not Forwards. One great step in this direction is to ban all Supplementary Materials, as the editor of The Journal of Neuroscience has just done. Reviewers can no longer request experiments that will only see the Supplementary lime light, authors can no longer hide critical experiments in the Supplementary shadows. Now we have to take the article, all of it, and no more or less, at face value. With one step, John Maunsell has taken us into the new century of scientific publishing: The Twentieth.
358th Aug, 2010Harris JA, Clifford CWG, Miniussi C (2008) The functional effect of transcranial magnetic stimulation: Signal suppression or neural noise generation? Journal of Cognitive Neuroscience, 20(4):734-740
      [NBArticle #21086] [Cites 16]
Transcranial magnetic stimulation (TMS) is often described as producing a temporary 'virtual lesion' in the brain area over which it is applied. Some also describe it as introducing random neural 'noise' into the signals carried by the brain area. Others say that it simply disrupts ongoing processing in that area. Which is the correct description? Harris and colleagues applied TMS over the occipital pole, then presented sine-wave grating stimuli at the location in the visual field where the TMS produced a 'phosphene' image. By modulating the contrast of the sine wave, and adding a variable amount of random noise to the stimulus image, the effects of TMS could be compared with the effects of adding noise, on whether participants could perceive the orientation of the grating. Three hypotheses were entertained: 1) TMS adds noise, leading to a leftwards shift in the noise-performance curve; 2) TMS interferes with the signal, increasing the slope of the noise-performance curve; 3) TMS adds noise to a signal that has already been gain-modulated, both shifting and increasing the slope of the curve. Hypothesis 2 was supported: TMS disrupts or weakens the signal, and does not simply add noise to the signal.
341st Aug, 2010Eickhoff SB, Schleicher A, Zilles K, Amunts K (2006c) The human parietal operculum. I. Cytoarchitectonic mapping of subdivisions. Cerebral Cortex, 16(2):254-267
      [NBArticle #30848] [Cites 40] [CitedBy 1]
How many areas are there in the human brain? Most modern answers will probably start by taking a look at Korbinian Brodmann's (1909) maps and counting them. The areas are numbered up to 52, but some are missing or rarely mentioned (16, 50, 51). Other brain maps are available and parcellate the brain in different ways on different criteria (e.g., Vogt & Vogt, 1919; Von Economo & Koskinas, 1925; Sarsikov et al., 1949), but in this age of brain imaging, high-performance computing and image analysis, it is ridiculous that we predominantly still use a century-old brain classification scheme based on the visual inspections of one scientist. Enter Eickhoff and colleagues who have, over the years, automated and developed an observer-independent classification scheme for cytoarchitectural anatomy. This paper, which is just one of many, picks apart Brodmann's areas 40 and 43 to discover not two but four reliably different brain areas. The anatomical territory covered corresponds approximately to the functional territory of the second somatosensory cortex. Probabilistic cytoarchitectural maps are now available in brain imaging software packages. Anatomy is not dead. Long live anatomy.
3325th Jul, 2010Smith CUM (2010) Does history repeat itself? Cortical columns: 4. Déja vu? Cortex, 46(8):947-948
    [NBArticle #30773]
A short and beautiful essay on the misuse of analogy, and the loss of detail with time in neuroscience. Examines the case for 'columns' as functional modules of the visual cortex and other cortices, and notes how the same mistakes are made, unmade, then remade with the passing decades:

"So, does history repeat itself? It seems that scientific investigators, like everyone else, are prone to make the same mistakes. George Santanyana may have been right here as elsewhere. Those who know no history are condemned to repeat it. Will columns go the same way as the phrenologist's bumps and protruberances, of von Economo and Koskinas' 109 architectonic areas? We all tend to see what we want to see. The lessons of Kanizsa's triangle and similar illusions need to be learnt. Historians might thus urge caution. They might point out that the same mistake has been made by analytical minds seeking to connect function and structure in the cortex several times before. Few doubt that cerebral cortices have a vertical as well as a horizontal organisation. There is overwhelming evidence that this is so. But a regular repeating modular structure? We might, indeed, take the issue as a useful justification for a concern with the history of neuroscience. For history, as they say, tends to repeat itself when no one listens the first time." p948
3211th Jul, 2010Kranjec A, Lehet M, Bromberger B, Chatterjee A (2010) A sinister bias for calling fouls in soccer. Public Library of Science ONE, 5(7):e11667
    [NBArticle #30638]
A World Cup Special Paper Of The Week
Psychologists have long known that humans are terrible at making precise judgements about events. We can't tell the relative timing or locations of events very well, we are biased by our beliefs, intentions, our current focus of attention, and by other events than have just happened. Our spatial judgements are even biased by the position of our eyes, head, and other body parts. So, who better to put in charge of adjudicating a multi-billion-dollar enterprise like professional football (soccer), than three humans? Kranjec et al. show how judgements of whether one player fouls another are influenced by the relative direction of movement. Players moving from right to left are more likely to be judged as having been fouled, by about 2.5%, than players moving left to right. With the standard system of referees moving diagonally across a pitch, this means the referee's foul criterion is likely to be lower than the linesman's, thus benefiting the attacking team. The explanation is in terms of habitual reading directions - events are usually seen as moving from left to right, so events moving in the opposite direction are debased in some manner. This explains England's disallowed goal against Germany (see also Ernst 2010)
3113th Jun, 2010Lyon DC, Nassi JJ, Callaway EM (2010) A disynaptic relay from superior colliculus to dorsal stream visual cortex in macaque monkey. Neuron, 65(2):270-279After a substantial temporary intermission, PotW is back!

Every undergraduate psychology student knows that visual information is received by the retina, relayed via the thalamus, then visual perception begins in V1. But this is not the only visual pathway, and in some senses it is not the main one. Apparently, there are more neurons sending information back to the thalamus than there are sending information from the thalamus to the visual cortex. This week's papers provide key anatomical and physiological evidence for a very important second pathway, leading from the retina to the superior colliculus, the inferior pulvinar of the thalamus, then to the middle temporal visual area MT and V3. The physiology of this pathway suggests a functional role in the processing of rapid visual information for the control of eye movements.
Berman RA, Wurtz RH (2010) Functional identification of a pulvinar path from superior colliculus to cortical area MT. Journal of Neuroscience, 30(18):6342-6354
3015th Feb, 200958 Humanities Journals (2009) Journals under threat: A joint response from history of science, technology and medicine editors. History of the Human Sciences, 22(1):1-4
    [NBArticle #26265]
Which is the 'best' journal? Ever since Eugene Garfield invented the Impact Factor in 1955, students, researchers, administrators, and funders have been debating this question furiously. To provide a final answer, at least within the field of humanities, Europe is coming together to build a ranking of journals, with Premier League, Second, and Third divisions. Soon, we can all avoid publishing in or even reading from 'bad' journals, and consider only the 'best' ones. Phew! That will save us all a lot of time and effort.
Sadly, not everyone feels this is such a good thing. For example, the editors of 58 European humanties journals. And the British Academy. And PotW.
History of the Human Sciences and the 57 other journals state their disapproval and boycott of this new ranking system in a recent joint editorial. PotW wishes them all good luck in fighting the tide of scientometrics!
298th Feb, 2009Mazziotta JC, Woods RP, Iacoboni M, Sicotte N, Yaden K, Tran M, Bean C, Kaplan JT, Toga AW (2009) The myth of the normal, average human brain: The ICBM experience: (1) Subject screening and eligibility. NeuroImage, 44(3):914-922
      [NBArticle #26323]
Do you have a normal brain? If you answer 'yes', and are willing to phone up some medical researchers, then the answer is probably 'no' (only 32% of such people may have normal brains). If you then go on through several more hours of interviews and medical examinations, then you're only 11% likely to have a normal brain overall. Mazziotta and colleagues spoke to 1685 people on the telephone and in person. At the end of the study, only 11% were declared normally brained and were accepted into the brain scanning session (whether they did, in fact, have a normal brain will be the subject of a study to be published in the future...).
If only 1 in every 9 volunteers has a medically normal brain, it makes any study based on normal participants hard to interpret. It also questions the usefulness of defining 'normal' at all. A survey of recent neuroimaging articles revealed that 75% of studies do not report screening their normal, healthy participants for medical problems. Mazziotta and colleagues argue that ignoring the large variability in medical conditions, medication use, and medical history contaminates any study of supposed normal brain function. It's hard to disagree, but rather than excluding 90% of potential participants from our studies, perhaps we can model this variability explicitly?
Top Tip: measure participants' blood pressure, since this was the most common basis for exclusion.
2831st Dec, 2008As organisations and institutions around the world, both scientific and 'other', gear up to give and receive their annual awards ('person of the year', 'woman of the year', 'chef of the year', 'molecular entomologist of the year', etc), the great responsibility falls on me to choose 'paper of the year'. And what a year it has been. PotW started after I read the massive paper by Nickerson (1998, Issue #1). Since then, I've delved into a number of subjects, journals, authors, and years. To be fair, the PotY should probably have been published in 2008, but since this is the first, I shall include any recent papers. Two papers vie for this prestigious award: Yuval-Greenberg et al. (2008, #10), and Koch et al. (2007, #22).

The first paper showed that 'transient induced gamma-band responses' in the human electro-encephalogram signal may be (almost entirely) explained by microsaccades. Neurophysiologists and electrophysiologists may have been able to predict such a result (the 'saccade potential' has been known about for decades, and skeptical electrophysiologists may scoff at such broad-band coherent EEG signals), but to get their result, Yuval-Greenberg and collleagues took some harsh neurophysiological criticisms into account, re-ran, and re-analysed their cherished and already-published experiments, this time with high-speed optical eye-tracking. In the process, they may have overturned nearly two decades of (woolly) thinking on consciousness and the binding problem, based on these transient, induced gamma-band responses (which just happen to occur at the same time as micro-saccades...).

The second outstanding paper, by Koch and colleagues, showed that transcranial magnetic stimulation over the human posterior parietal cortex changes the excitability of the primary motor cortex. This is not an amazing new result in iteself, and certainly not unexpected, but  the fact that they replicated their effect (at least) five times in the same paper, using very different, but convergent techniques, leaves very little open room for interpretation, and perhaps no need for replication. This is proper science.

So, if you want to be PotY in 2009, either publish something that overwhelmingly contradicts your previous work (and the work of many others), or else show some newish result in an overwhelming fashion, at least 5 times in the same paper. Happy new year!

2730th Nov, 2008Schiermeier Q (2008) Self-publishing editor set to retire. Nature, 456(7221):432
      [NBArticle #25504]
Rarely does anything scientific make me lol. This week, Schiermeier tells the heart-rending tale of the editor of the physics journal "Chaos, Solitons and Fractals". The editor is a prolific writer himself. In 2008 alone, no less than 58 of his best works were published in Chaos, Solitons, and Fractals (with an impressive 9 articles in Volume 35, Issue 2!). Some would criticise an editor who abuses their position to publish their own work, boosting the journal's impact factor, by selective publishing and high self-citation rates (in his last paper, 21 of the 30 references were to his own work), or who fabricates honorary degrees and memberships of scientific societies (e.g., from Johann Wolfgang Goethe University, Frankfurt). Not me: Here at PotW we like a good laugh, and trail-blazing editors such as the above are most welcome to provide us with some seasonal cheer. Best of luck in your next job, Ed. Apparently, this means 'laugh out loud'
2623rd Nov, 2008Galton F (1886) Regression towards mediocrity in hereditary stature. Journal of the Anthropological Institute, 15:246-263The method of 'regression' in statistical analysis is one of the most commonly used techniques for assessing whether and how two or more variables are related to each other. It is so commonly used that it is hard to imagine a time when it didn't exist. Francis Galton, who made his debut PotW appearance in August, invented regression over 130 years ago. The first demonstrations were based on sweet-pea seeds, but he later moved to humans. The laws of regression remained applicable to many situations. Apart from its historical importance, delving into the ancient scientific literature makes you realise just how different scientific communication was in Victorian times: "I may be permitted to say that I never felt such a glow of loyalty and respect towards the sovereignty and magnificent sway of mathematical analysis as when his [Mr J Hamilton Dickson of St Peters, Cambridge] answer reached me, confirming, by purely mathematical reasoning, my various and laborious statistical conclusions, with far more minuteness than I had dared to hope, for the original data ran somewhat roughly, and I had to smooth them with tender caution." p 255 We should all take a leaf from Galton's book, and smooth our data with tender caution.
2516th Nov, 2008Duchesne S, Jannin P (2008) Proposing a manuscript peer-review checklist. NeuroImage, 39(4):1783-1787
      [NBArticle #20353]
How should you review a manuscript? Despite the primary importance of the reviews that a new manuscript receives for determining whether and where it is published, very little has been published in the way of an answer to this question. The majority of reviewers learn by trial and error, and few are specifically instructed or taught. This is inefficient and unhelpful, and with increasing demands on reviewer's time, a gear-change in productivity is required. Towards this end, journals and journal families now frequently ask reviewers to complete checklists, to give numerical or categorical ratings, and to respond to specific questions, but there still remains a wide scope for interpreting the role of the reviewer. By collating the recommendations of a wide range of publishing, writing, and scientific organisations, Duchesne & Jannin have produced a 71-item checklist as an aide memoire for assessing neuroimaging and other closely related manuscripts. The checklist (Version 1.0) is available at the author's website, as well as in the published article. Try it out, then send the authors some feedback.
249th Nov, 2008Maslov S, Redner S (2008) Promise and pitfalls of extending Google's PageRank algorithm to citation networks. Journal of Neuroscience, 28(44):11103-11105
      [NBArticle #24384]
Several weeks ago, the Journal of Neuroscience began a special series of short articles on 'Impact Factors'. Impact factors (IFs) are a currency of gossip, ego, evaluation, and promotion in science, and reflect how popular the articles in a journal are within the first two years after they are published. (For more information, see my page on the dreaded IFs). The first two articles in this series have both been about an alternative method for evaluating journals (and individual articles), based on the PageRank algorithm that Google uses to rank webpages. The basic limitation of IF, in these authors' views, is that it rates all citations equally - a citation in a 'lowly' journal counts the same as a citation in a 'high-ranking' journal such as Nature or Science. So, instead of just counting all citations, the PageRank/CiteRank/EingenfactorTM methods create a matrix that shows which journals cite which other journals. Further, the influence of a citation is weighted according to both the ranking of the journal that the article was published in and the number of citations in the article - a commentary in an important journal that cites just a few articles would therefore give a lot of weight to those citations, while a lengthy review in a technical journal listing hundreds of articles would not. Importantly, the absolute number of citations to an article is not necessarily as important as the nature of the citation network that the article belongs to - some moderately-cited articles from many years ago will be highlighted by these methods.

Editor's opinion: This is an improvement on the IF, but I fear two things: 1) There seems to be an inherent circularity in rating the importance of a citation based on the importance of the citing journal - presumably this leads to a positive feedback in which, e.g., mirror-neurons citations are rated as more important than other citations about the brain; 2) Counting citations will only ever give you a measure of popularity and ease-of-citation in the short-term. However, it is very interesting to note that the 'best' physics papers identified by Maslov & Redner's algorithm were all from 1929-1979 - as if nothing important has been published in physics for 30 years!
Bergstrom CT, West JD, Wiseman MA (2008) The eigenfactorTM metrics. Journal of Neuroscience, 28(45):11433-11434
      [NBArticle #25020]
232nd Nov, 2008Rothwell JC, Day BL, Thompson PD, Kujirai T (2009) Short latency intracortical inhibition: One of the most popular tools in human motor neurophysiology. Journal of Physiology, 587(1):11-122
      [NBArticle #24338]
Not a busy week of reading for me, but TMS fans will be happy with this choice: A historical and methodological perspective on a widely-used set of techniques in TMS studies of the human motor system. The basic finding is that applying one, low-intensity TMS pulse 1-5ms before a second, high-intensity TMS pulse, reduces the effect (measured by muscular activity) of the second TMS pulse. The first pulse activates intracortical inhibitory circuits, but is not sufficient to activate the corticospinal outputs (which would result in muscle twitches). Increasing the delay between these two pulses to 10-15ms leads to intracortical facilitation. Rothwell and co provide a brief, useful summary of the history and current thinking about the mechanisms and effects of motor cortical TMS.
2226th Oct, 2008Koch G, Del Olmo MF, Cheeran BJ, Ruge D, Schippling S, Caltagirone C, Rothwell JC (2007) Focal stimulation of the posterior parietal cortex increases the excitability of the ipsilateral motor cortex. Journal of Neuroscience, 27(25):6815-6822
      [NBArticle #21261] [CitedBy 1]
Is this the best transcranial magnetic stimulation (TMS) paper ever published? It could well be. I'm an extremely critical reader - of the hundreds of articles that I read every year, I am completely convinced by about, err, a few. I am left feeling a little unsatisfied by most papers. At the end of this one, though, I was totally blown away. Koch and colleagues took two TMS coils, put one at the posterior end of the intraparietal sulcus (IPS) over the angular gyrus in the right hemisphere, and the other over the right primary motor cortex (M1). They stimulated both coils with different delays between the parietal and the M1 TMS. And with different intensities. And with different coil discharge directions. And at different positions over the posterior part of the IPS. And at the posterior, middle, and anterior parts of the IPS. And recorded from single motor units with needle electrodes to show changes in the timing of muscle firing. And tested the effects of parietal stimulation on the H-reflex. And tested the effects in the left hemisphere. All 6 of their experiments produced consistent results: Stimulating the posterior IPS at 90% of the resting motor threshold and 4-6ms before M1 enhances the amplitude of the motor evoked potentials (MEPs) caused by M1 stimulation. The authors replicated their effect five times within the same paper. Researchers often find a small effect in their first experiment, then refuse to try to replicate it in later 'control' experiments. Not Koch and colleagues. Is this the 'Paper of the Year'?
2119th Oct, 2008Pruszynski JA, Kurtzer IL, Scott SH (2008) Rapid motor responses are appropriately tuned to the metrics of a visuospatial task. Journal of Neurophysiology, 100(1):224-238
    [NBArticle #21867] [CitedBy 4]
What is a 'reflex'? Doctors with hammers will tell you, while bashing your knees, that it is a sensory-motor reaction that cannot be altered by voluntary intervention. But studies have shown that even spinal 'reflexes' can be modified by the voluntary state of the subject. Pruszynski and colleagues took a detailed look at the 'rapid motor responses' (a better term than 'reflexes') that follow brief perturbations of hand position when people are trying to move their hand towards a target. They recorded the electrical activity of 6 arm muscles, and examined differences in activity in four time periods following the perturbation: 20-45, 45-75, 75-105, and 120-180ms. The first three periods were the short-, medium-, and long-latency rapid responses, the latest period was the 'voluntary' period - reflecting the earliest activity for a simple reaction time task. In three careful experiments, the authors showed that the first period was always insensitive to target position, but that, from around 60ms after the perturbation, the muscles began to distinguish between different target directions and distances. Within the blink of an eye (literally), you can control your own reflexes. That seems important to know, especially if you're carrying drinks in a crowded bar.
2012th Oct, 2008Fein D (1987) Systematic misidentification of toes in normal adults. Neuropsychologia, 25(1, 2):293-294
      [NBArticle #23975]
Perhaps the simplest and shortest paper that is likely to appear on PotW, this paper was well-hidden in the depths of Neuropsychologia from 20 years ago. Deborah Fein touched the toes of 18 healthy adults, and asked them, with their eyes closed, to move the corresponding finger - if the big toe was touched, move the thumb, if the fourth toe, move the ring finger. Easy, you might think? Well, the participants were surprisingly bad at this task, making at least 16% errors for every toe. People systematically mis-localised touches towards the little toe by 1-3 toes. Most surprisingly, performance was best for the 'ring' toe (89% correct), and worst for the 'index' toe (16% correct) - a pattern opposite that of the tactile abilities of our fingers. Fein was at a loss to explain these data. Never mind, it's a light, interesting read anyway.
195th Oct, 2008Zampini M, Spence C (2004) The role of auditory cues in modulating the perceived crispness and staleness of potato chips. Journal of Sensory Studies, 19(5):347-363
    [NBArticle #16054]
(I think I lost a week somewhere in the UK...) Honoured this week with the Ig Nobel Prize in Nutrition Science " for electronically modifying the sound of a potato chip to make the person chewing the chip believe it to be crisper and fresher than it really is", Zampini and Spence asked 20 participants (myself included) to bite into a Pringles crisp*, while holding their mouth near to a microphone. The sounds of the biting, breaking, and crunching were simultaneously re-played to the participants over headphones, but with either the overall intensity, low-, or high-frequency components of the sounds amplified or attenuated. For overall intensities of -20dB and above, amplifying the high-frequencies made the crisps feel crunchier and fresher, while attenuating high-frequencies made the crisps feel soft and stale. Eating is a multisensory experience. *Legally speaking, Pringles are not crisps [read more]. Personal anecdote: I was doing my Ph.D. in Oxford when these studies were being performed in the lab. At the end of each day's experiments, there was an excess of crisps lying around, which proved very handy in sustaining late-evening working. To discourage any brand-recognition effects during the experiments, the crisps were placed into a bowl. Unfortunately, for several days until I learnt better, I ate from the waste bowl - the bowl that contained the floor-sweepings and post-crunch spittings.
1821st Sep, 2008Cantello RM, Civardi C, Cavalli A, Varrasi C, Vicentini R (2000) Effects of a photic input on the human cortico-motoneuron connection. Clinical Neurophysiology, 111(11):1981-1989In a week dominated by reading about mirror-neurons, this paper was really the only one to stand out from the crowd. How soon can a visual stimulus affect the human motor cortex? That is the basic question being addressed here. The answer: 55ms. The clinical background of the study concerns patients with 'photic cortical reflex myoclonus' - people whose muscles over-react to sudden visual stimuli. But the implications of this study go further than that, for example, by placing a maximum speed limit on visual cortical processing in healthy people. Subjects relaxed on a couch with their eyes closed while bright lights, magnetic pulses, and electric shocks were presented to different parts of their body at different time intervals. The authors tested each subject twice, up to 2 years apart, with several different stimulation techniques. The results were replicable and informative. Well, certainly not the sexiest paper to grace PotW, but this is good, solid science that probably does not require any mopping-up or replication or qualification by other researchers (unlike all the mirror-neuron TMS studies that I read this week, hmmphh).
1714th Sep, 2008Gardner JL, Merriam EP, Movshon JA, Heeger DJ (2008) Maps of visual space in human occipital cortex are retinotopic, not spatiotopic. Journal of Neuroscience, 28(15):3988-3999I am never quite sure whether papers which tell us things that we already know, but in an overwhelming and beautiful way, are important or not. This week's paper comes from a leading, well-established visual neuroimaging laboratory, and tells us that 12 areas in human visual cortex are retinotopic. But we knew that, so why is this paper interesting? At the most superficial level, just take a look at the pictures, they're really pretty, more than justifying the article's 12MB. At a somewhat less superficial level, the approach is simple (measuring the visuo-topic organization of visual areas while the eyes are in three different positions), the analysis is robust, the results are clear and powerful (all 12 areas are retinotopic, not spatiotopic), and the paper as a whole clears up a problem introduced by a previous paper (d'Avossa et al., 2007 in Nature Neuroscience, which suggested that human MT was spatiotopic, not retinotopic). So, if you ever need to say that the human visual cortex represents the visual world in a retinotopic reference frame, you can cite this article.
167th Sep, 2008Pruessmann KP (2008) Medical imaging: Less is more. Nature, 455(7209):43-44
  [NBArticle #23354]
Most of what I read this week was related to a revision of an article that I'm writing, hence not really worth mentioning. However... This little 'News and Views' gem was published in Nature the other day. It struck me because I've occasionally overhead friends, colleagues, and myself talking about the latest super-high strength brain scanning magnets. Not so long ago, magnets of 0.5-1.5 Tesla (T) in strength were standard. 1T is about 10,000 times the strength of the Earth's magnetic field. Nowadays, healthy people are routinely being put inside magnets of 4T, and even up to 7T. There are reasons for and against using these different strengths of magnets, but one thing about this area of research is the often nave presumption that stronger is always better, and the sharp intake of breath when someone describes their latest high-power magnet is a little odd, scientifically. So, hail the studies that Preussmann describes: Structural brain imaging at 0.03T. The benefits of such ultra-low-field imaging are the portability of the system and the reduced invasiveness or awkwardness that brain scanning typically entails. Let's hope that cheap, safe brain imaging will soon come in a briefcase format so that we can take it home and scan our pets.
Charlton BG (2008) Figureheads, ghost-writers and pseudonymous quant bloggers: The recent evolution of authorship in science publishing. Medical Hypotheses, 71(4):475-480
  [NBArticle #23357]
1531st Aug, 2008Ingram JN, Körding KP, Howard IS, Wolpert DM (2008) The statistics of natural hand movements. Experimental Brain Research, 188(2):223-236
    [NBArticle #23013] [CitedBy 2]
It's nice to get out of the lab and walk around, once in a while. James Ingram, two of his colleagues, and three others in Cambridge did just that, strapping on a Cyberglove and a laptop-backpack, and got on with their daily manipulative lives. 17 cumulative hours later, these special data-gloves had recorded the angles of 19 joints in the user's hands, resulting in a dataset of 5 million individual hand postures. Then they opened up Matlab and did some fancy statistics. One piece of good news for lab-based scientists is that the results of this naturalistic study of hand movements are largely in agreement with what had come from previous more standard, limited, laboratory studies. Despite the wide range of hand movements recorded, 60% of the variability in the 19 joint angles of the six different participants could be explained by just two independent statistical components, and over 90% with just 7 more. This is good news for understanding motor control: The ~20 degrees of freedom present anatomically in the hand may easily be reduced to single-digits without much loss of complexity. The results of this study were largely known already, but the methods they used have propelled it to PotW status. In a year or two, I hope to be following in their handprints with a similar set-up - watch this space...
1417th Aug, 2008Cumming G (2008) Replication and p intervals: p values predict the future only vaguely, but confidence intervals do much better. Perspectives on Psychological Science, 3(4):286-300Geoff Cumming is a major force in the 'Statistical Reform' movement in psychology, behavioural, social, and other sciences. One major argument of the movement is that the p-value, the arbitrary measurement of differences between experimental conditions, is so often abused, misinterpreted, worshipped and revered, that it is long past time to replace it. p tells us the likelihood that the data we collected could have arisen by chance, if the null hypothesis was true. Because the reference-point for p is the null hypothesis, any interpretation of p, strictly speaking, should also be in terms of the null hypothesis. Of course, very few researchers actually do this, and when they do, scientific prose becomes dull, tortuous, and full of double or triple negatives: "The evidence was insufficient to reject the null hypothesis that God doesn't exist" or "These data failed to show that the Earth isn't flat". (This is the equivalent of the British Football Pools lottery - where punters wager money on the games they think will result in 0-0 final scores.)p-value of 0.05 from a first experiment means in terms of replicability for the second experiment, given that the effect is real, and they will likely say that it means you have a good chance of repeating the finding. Sadly not. Cumming demonstrates how, in fact, given p=0.05 in Experiment 1, we can only be ~50% sure (or 'hopeful') of replicating this significant p-value in Experiment 2. This is true even if the effect is real, if nothing in the experiment or the study population has changed, and regardless of the sample size. To be just ~80% certain that our effect would be replicable in the next experiment, we should instead adopt p-value thresholds of 0.005 or lower. This article is scary. Read it before you next open SPSS... Instead, Cumming and the other Knights of the Statistical Reform encourage us all to talk about effect sizes, confidence intervals, variance explained, and, the subject of this week's paper: Replicability. Ask a load of scientists what a Following this, I have decided to start creating a page on: Statistical Reform. Don't expect me to abandon p-values just yet, however...
1310th Aug, 2008Vul E, Pashler H (2008) Measuring the crowd within: Probabilistic representations within individuals. Psychological Science, 19(7):645-647
  [NBArticle #22751]
After a long tour of Europe, PotW returns! Without counting, how many characters are there on this page? Your guess will almost certainly be wrong. Very wrong. But, intriguingly, ask a few other people, or a whole crowd of people to guess, and the error of the average of the guesses will be much closer to the correct answer than the average error of an individual's guess. This has been called 'the wisdom of crowds' effect (for balance, see also: Mackay C (1841). Extraordinary popular delusions and the madness of crowds). As with most things, Francis Galton noted this a hundred years ago (1907), but the new finding of Vul & Pashler is that simply asking the same person to guess a second time, whether immediately or three weeks after the first guess, results in a more accurate average guess for that person. Subjects weren't simply looking up the correct answers in the delay between questionnaires, since their second guesses were on average less accurate than the first. The finding is explained by considering that every judgement, every decision we make, comes with a bit of internal noise. So long as that internal noise is at least a little bit uncorrelated between repeated judgements, then averaging across judgements should lead to an improved response. It also implies that, when guessing, we may know a little more than we admit to. Does this mean that in dictatorships, where some individuals may cast many thousands of 'votes', the result is better than in a democratic election? No: repeating your guess improves your accuracy by only 10-30%, relative to the improvement you would get by asking another, independent, person to guess. So, it seems that democracy is here to stay, just as Galton intended.
1225th May, 2008St Amant R, Horton TE (2008) Revisiting the definition of animal tool use. Animal Behaviour, 75(4):1199-1208An entirely self-interested choice this week: I spent my PhD in Oxford studying the multisensory spatial attentional consequences of tool use in humans. 3 years, 23 experiments, and 4 papers later, I still have to argue with my reviewers about the basic questions: 'what is a tool' and 'what is tool use'. Since my data and conclusions run counter to the dominant hypothesis about tool use (i.e., that tool use changes, specifically extends, 'peripersonal space'), I frequently encounter long, and somewhat pointless, arguments with reviewers about how to interpret my data. The most frequent, and in my view circular, argument is that, since the tasks I studied are not really tool use, the results I found are therefore tangential to any particular hypotheses about tool use. I've devoted several thousand published words in the last two years to justify why the tasks were, in fact, tool use, but, alas, it seems to no avail. In the stead of my own articles, I can now cite Amant & Horton (2008) as independent witnesses for why, in fact, the tasks I studied constituted Tool Use in it proper, ethologically meaninful form. Ahhhhh. Robert St Amant is now studying tool use in robots: I wish him all the best.
1118th May, 2008Bones AK, Johnson NR (2007) Measuring the immeasurable: Or "Could Abraham Lincoln take the implicit association test?". Perspectives on Psychological Science, 2(4):406-411
  [NBArticle #21996]
An important tool in social and cognitive psychology is the Implicit Association Test (IAT). In this test, you have to press two buttons, for each of two categories of information. For example, you press the left button for 'pink' and the right for 'blue' in one test, then you press the left for 'boy' and the right for 'girl' in another. In further tests, these two categories are mixed up so that you have to respond to both categories simultaneously. The IAT measures how easy it is for you to do this last part - if the left button is used for both 'boy' and 'blue', you may find this easier than if the button was for both 'boy' and 'pink'. If it is easier, then we have an implicit association between 'boy' and 'blue' and/or between 'girl' and 'pink'. Bones and Johnson lament the unstoppable rise in IAT test research in the USA, but also suggest ingenious ways to overcome the shortage in IAT participants. One is to test participants in utero, the other to test the dead. With admirable clarity and expertise, Bones and Johnson show just how measurable the immeasurable can be.
1011th May, 2008Yuval-Greenberg S, Tomer O, Keren AS, Nelken I, Deouell LY (2008) Transient induced gamma-band response in EEG as a manifestation of miniature saccades. Neuron, 58(3):429-441When Saturday night comes, my heart rate increases, my palms start sweating. 'What, oh what', I ask myself, 'will be the Paper of the Week'? What responsibility, that choice. This week, it's only Thursday evening, and the choice was not really mine to make. There are several reasons why this paper made it to the Top 1 this week: 1) I know most of the authors and work in their lab; 2) The methods and analysis are as thorough and rigorous as you could ask for from wishy-washy psychologists; 3) The colours are really pretty; 4) There's nothing like a massive artefact to keep us all on our toes. When a visual stimulus appears where we are looking, our eyes at first stop moving, then about 200-300ms later, we make some very small eye movements. Get participants in an experiment to fixate while flashing visual stimuli at them, repeat this hundreds of times, and record the electrical potentials across their brain. At the same time as people make those small eye movements, there is a substantial, broad-frequency electrical response across the scalp. Mostly this is gamma-band activity (20-100Hz). The trouble with all this is that eye-movements tend to correlate with attention, perception, consciousness, and all sorts of other fascinating behavioural and psychological things. A certain flavour of gamma-band activity also correlates with these phenomena. Unless you record gamma-band and mini-eye-movements at the same time, you'll not notice just how closely they are related to each other (actually, they're practically identical). These authors did, and that's why this paradigm-busting paper exists. Go forth, and buy an eye-tracker.
94th May, 2008Veerman MM, Brenner E, Smeets JBJ (2008) The latency for correcting a movement depends on the visual attribute that defines the target. Experimental Brain Research, 187(2):219-228One of the modern cognitive neuroscientific dogmas is to divide the brain into any of a number of dissociated processing streams, and to find any single experiment that can support this distinction. Anything will do: 'what' vs. 'where', 'how' vs. 'who', 'if' vs. 'but'. Rather less common are well-controlled, systematic, parametric attempts to uncover the stimulus characteristics that lead to such dissociations. This paper is a good example of the latter, investigating how quickly the 'dorsal' or 'where' or 'how' stream can react, in visually-guided pointing movements, to a change in the target, defined by its position and some other visual aspect. In support of the rapid processing of magnocellular-specific stimulus attributes in the control of action, changes in orientation, luminance, and size resulted in faster movement corrections (by 50ms) than changes in colour, shape, and texture.
827th Apr, 2008Schalk G, Kubánek J, Miller KJ, Anderson NR, Leuthardt EC, Ojemann JG, Limbrick D, Moran DW, Gerhardt LA, Wolpaw JR (2007) Decoding two-dimensional movement trajectories using electrocorticographic signals in humans. Journal of Neural Engineering, 4(3):264-275
  [NBArticle #20122]
For several reasons, I've read about as much this month as I usually do in a week (data collection, visitors, programming, lab retreat). So this is the paper of the month. There has been a lot of research on Brain-Computer-Interfaces (BCIs) in the last few years, however I was recently surprised to find that the field is at least 20 years old (e.g., Farwell & Donchin, 1988). The advance presented by Schalk and colleagues is that a different, less invasive signal recorded from the surface of the brain (electrocorticography) may be just as useful as previous deep/penetrating-electrode techniques. Such experiments must typically be performed on epileptic patients undergoing pre-surgical preparation and exploration: Stimulating the brain with electrodes is a reasonably good way to find out if a particular part is in working order (i.e., if the stimulus makes you feel sensations, move a muscle, or stop talking) before you cut it or other bits out. Recording from the brain surface provides much information about how the brain controls behaviour. Correlate the signals from the brain with the behaviour, use a portion of these signals to generate a model to predict the behaviour in another portion, and you have yourself a BCI. Happy experimenting! [but don't try it at home]
730th Mar, 2008Beauchamp MS (2005) Statistical criteria in fMRI studies of multisensory integration. Neuroinformatics, 3(2):93-113The use of statistics in fMRI is extremely important, yet relatively few researchers have an in-depth knowledge of just how those statistics work. Similarly, statistics can be used in the field of multisensory integration to force your data to follow your prejudices. The impressive thing about this week's paper is the use of multiple criteria and equations to analyse the same data. By systematically varying the analytic technique, while keeping the data constant, Beauchamp provides important insights about both fMRI in general, and the phenomenon of 'superadditivity' in multisensory integration in particular.
616th Mar, 2008Gjedde A (2008) Functional brain imaging celebrates 30th anniversary. Acta Neurologica Scandinavica, 117(4):219-223
  [NBArticle #20362]
It's hard to imaging a time in which brain imaging did not exist. But this was only ~30 years ago. A good reminder of this is to read a PET article from the early 80s - the ultra-low spatial resolution colour figures of the brain now seem shockingly imprecise and almost useless. The physical, computational, and statistical progress in functional neuroimaging in the last few decades outstrips that in any other cognitive neuroscience technique (that I can think of). Albert Gjedde recounts the social and academic dynamics of the Brain77 conference, and is proud to point out that functional imaging is a Swedish invention from the 1970s. He also points to the long history of work on cerebral blood flow that made neuroimaging possible. In such a view, neuroimaging is well over 50 years old. Gjedde doesn't go further back in time that this. But he should have - William James, that over-quoted review writer from the late 19th century, described in his Principles of Psychology, experiments on 'thermometry' - placing numerous thermometers across the scalp, and recording changes in temperature due to sensory, motor, and cognitive tasks. Functional brain imaging based on blood flow is at least 120 years old!
59th Mar, 2008Weisman RG (2008) Advice to young behavioral and cognitive scientists. Behavioural Processes, 77(2):142-148
  [NBArticle #20522]
Newton apparently said: "If I have seen further, it is by standing on the shoulder of giants." I can't imagine anyone that I know at present who in a few hundred years will still be quoted, but the last two weeks have nevertheless seen a rash of advice-giving (relative) giants of science. To reflect this, I re-write Newton's quip in a modern air: 'If I have seen further, it is by sitting on the knee of giants.' Good advice is surprisingly hard to come by in science. Those arrogant enough to give it are probably wrong, and those learned enough to know better have little to give. So, blessed are the occasions when (usually at retirement) scientists dare to offer a few hints. Robert G. Weisman's advice is: 1) Don't try to explain your experiments with nature, but experiment to explain nature; 2) Test several alternative hypotheses with every experiment; 3) Include positive controls to prove your methods; 4) Write well; 5) Anticipate your reviewers at every stage, but be polite to the end; 6) Leave two days before drafting your reply to an editorial decision. There is nothing special or deep about these insights, but to read them in a scientific journal is all the more striking because of it.
42nd Mar, 2008Cowey A (2008) Q & A: Alan Cowey. Current Biology, 18(4):R141-143
    [NBArticle #20740]
I was too busy to read much this week, but I was glad to take a few minutes over these two closely-related biographical sketches which appeared by what seems coincidence in the same week. Alan Cowey was Lawrence Weiskrantz's Ph.D. student. Both have worked on visual perception for a combined total of over 100 years. Cowey recollects influences from Darwin, Crick & Watson, Virgil, & Huxley, and regrets his decision not to discover the visual area "MT", while bemoaning "the regrettable and increasing modern tendency to use phrases such as 'here we show for the first time' ". Weiskrantz focuses on the things that surprise him, some 'historical accidents': That visual aftereffects can occur after just imagining a stimulus, that brain-damaged patients with 'no' memory could nevertheless recognise stimuli better than chance, and, perhaps most famously, that 'blind' patients may nevertheless react and respond correctly to visual stimuli. Weiskrantz quotes Niels bohr: 'To progress in science, what one really needs is enemies. Competent, zealous enemies'. If you can, make enemies of these two eminent scientists.
Weiskrantz L (2008) Surprises. Neuropsychologia, 46(3):771-773
  [NBArticle #20751]
324th Feb, 2008Silvanto J, Muggleton NG (2008) A novel approach for enhancing the functional specificity of TMS: Revealing the properties of distinct neural populations within the stimulated region. Clinical Neurophysiology, 119(3):724-726
  [NBArticle #20489]
Of all the methods in modern cognitive neuroscience, transcranial magnetic stimulation (TMS) can seem, at first glance, the most brutal: A high-powered electromagnet is placed directly on the head of a volunteer, and discharged very rapidly. This rapid electromagnetic discharge stimulates neurons in the chunk of brain directly below the magnet, interrupting whatever they were doing just before the stimulation. Despite the apparent coarseness of this technique, in practice, the effects of TMS are extremely sensitive to the precise position, orientation, and intensity of the magnetic stimulus, and changes in just a few millimetres, degrees, or percentage points can make all the difference between actual and sham stimulation. Silvanto & Muggleton discuss a relatively new approach which pushes the specificity of TMS even further: By first behaviourally adapting a certain sub-population of neurons (for example, those involved in seeing a particular direction of motion, or moving a certain muscle), TMS can be used selectively to stimulate just these neurons, gaining valuable new insights into their functional properties. This new 'state-dependent-TMS' methodology may be extremely important in the next few years, just as fMRI-Adaptation has been in the last decade of functional neuroimaging.
217th Feb, 2008Stetson C, Fiesta MP, Eagleman DM (2007) Does time really slow down during a frightening event? Public Library of Science ONE, 2(12):e1295
    [NBArticle #19497]
A few years ago I attended a one-day workshop that celebrated 100 years since Henry Head cut the nerves in his own wrist for the sake of science (he wanted to study how they regenerated over months and years). This week's paper reminded me a little of this sort of gung-ho approach to science. If you want to know whether time slows down when people are scared, drop them 31m from a crane and ask them to read flashing numbers on a wristwatch. If they can read the numbers, then time has slowed down. If they can't, it hasn't. Simple enough. While subjects estimated that their own plummet to earth felt 36% longer than the falls of others, their ability to read the flashing numbers was the same whether on the ground or accelerating downwards at 9.8m.s-2. Time doesn't slow down when we're scared, but we remember it as being exactly 36% longer than it was: It's the memory, stupid. [see also Frankenhaeuser M (1960). Subjective time as affected by gravitational stress. Scandinavian Journal of Psychology, 1(1):1-6]
110th Feb, 2008Nickerson RS (1998) Confirmation bias: A ubiquitous phenomenon in many guises. Review of General Psychology, 2(2):175-220Clear two days in your diary, grab your slippers and sink into this enlightening, amusing, scholarly, and, in my unbiased opinion, extremely important review. 'Confirmation bias' is the bias that people commonly display in weighing new evidence for or against one's beliefs. In short, we are more likely to seek to confirm hypotheses we entertain, rather than to disconfirm them. All the more intriguing, this occurs even if one states no vested interest in the hypothesis. Along the way, Nickerson provides countless examples from social and experimental psychology, economics, medicine, the history of science, and more. Scientists beware: Despite the classical Popperian notion that scientists seek disconfirmatory evidence in order to test their hypotheses, Nickerson points out time and again just how unnatural this is and how unwilling we are to do it. Further, those who seek to disconfirm the hypotheses of others may just as likely be prone to the same bias, but in the opposite direction. No one is safe: Read this article immediately. Ed: Anu Garg, from Wordsmith's 'A-Word-A-Day' (AWAD) service, sent this quotation on 03/09/2008: "Like a lawyer, the human brain wants victory, not truth; and, like a lawyer, it is sometimes more admirable for skill than virtue. -Robert Wright, author and journalist (b. 1957)"
* The main motivation for doing this is that I've occasionally been asked: "so is there any research that you actually like?" (my work tends to be on the critical side...). This page is the answer to that question. Note: It's really not a comprehensive in-depth review of the topic, just the most interesting &/or novel thing I read each week.