I have been doing some reading for a while now on the ethics of research involving model organisms, particularly the potential for studies on lower species to offer insights into human disease (and thereby contribute to the 3Rs). Some of my musings on the topic can be found here.

Aware of this interest, a colleague recommended that I read a 2004 paper published in the journal Cell. I am very grateful that he did, since the study really has the “wow” factor – demonstrating beautifully the potential of comparative genomics, experiments on model organisms and knowledge of human disease to work together to produce new insights that would have been much harder if any one component was missing. The paper is Comparative genomics identifies a flagellar and basal body proteome that includes the BBS5 human disease gene by Li JB et al. The following notes are my attempt to summarise the best bits.

The importance of cilia and basal bodies in disease
The role of cilia in respiration (and the detrimental effects of smoking on their function) were features of the school biology curriculum when I was a child. However, research over the last ten years or so has demonstrated that cilia have surprisingly diverse roles in development, from determination of left-right symmetry in the body, through to formation and function of specific organs such as the kidneys (for more detail see the Wikipedia entry on Ciliopathy or, if you have access permissions,  Badano et al (2006), The ciliopathies: an emerging class of human genetic disorders Annual Review of Genomics and Human Genetics 7:125-148). Bardet-Biedl syndrome (BBS) is one disorder associated with non-functional or malfunctional cilia. The clinical features can be varied, but include obesity, mental retardation, progressive-onset blindness and polydactylism (i.e. possession of extra digits).

Cilia, and flagella in organisms that possess these related structures, develop from, and remain attached to, modified centrioles called basal bodies. These complex structures involve upwards of 150 proteins.

Comparative genomics
The complete genome sequence for a growing collection of organisms is now known. The authors of the Cell paper – headed up by Susan Dutcher of Washington University, St Louis -  reasoned that they could identify novel genes involved in basal body development by judicious use of such genomic information.

The logic goes like this: basal bodies are needed for both cilia and flagella, therefore genes necessary for proteins in the basal body should be present in the genomes of an organism with cilia and an organism with flagella – even if otherwise morphologically dissimilar – but not in the genome of a species with neither cilia nor flagella. A Venn diagram can be helpful in visualising this subset of genes.

researchers were interested in finding genes expressed in humans (an organism with cilia) and Chlamydomonas (an organism with a flagellum) but not Arabidopsis (which has neither cilia nor flagella) - see text for details

When they looked at the genomes of humans (ciliated, approx 32,ooo genes) and Chlamydomonas (a unicellular flagellate with approximately 20,000 genes) they homed in on a set of about 4,350 genes that were common to both species despite their obvious diversity.

Next, they subtracted from that list all of the genes that were also present in the genome of the thale cress Arabidopsis (approximately 29,000 genes in total).  This plant species has neither cilia nor flagella and therefore ought not to have the genes necessary for their development.

It turned out that a total of 686 genes fit into the subset “in human and in Chlamydomonas but not in Arabidopsis“, termed the flagellar apparatus-basal body (FABB) group by the researchers, and corresponding to the yellow sector in the Venn Diagram above. This may still seem a large number, but when compared with the total of at least 20,000 genes it represents elimination of over 96% of the genome as not being involved in basal bodies.  In this way they hoped that they had shortened the odds on identifying the role(s) played by genes of previously unknown function.

.

Alzheimer’s disease (AD) remains the most common form of dementia, particularly the late-onset version which typically develops in patients aged over 65. Although there is believed to be a strong genetic basis to the disease, the only gene previously identified as a susceptibility factor in all version of the disease was APOE, coding for Apolipoprotein E. In addition, genes for Amyloid precursor protein (APP), Presenilin 1 (PSEN1) and Presenilin 2 (PSEN2) have been noted as factors in the less common early-onset form of AD, which has a strong pattern of familial inheritance. Other attempts to find genes influencing the more common late onset form of AD have been ‘under-powered’, i.e. have involved insufficient individuals (≤1,100) to reveal any further statistically-significant correlations.

In October 2009, however, two independent studies published “back-to-back” in the journal Nature Genetics identified a number of other genes in which Single Nucleotide Polymorphisms (SNPs) seem to be associated with development of late-onset AD.

Genome-wide association study identifies variants at CLU and PICALM associated with Alzheimer’s disease
Harold et al (2009) Nature Genetics 41:1088-1093

The authors of the first study (n = 86, headed by Julie Williams from Cardiff University) began by examining SNPs in about 19,000 research subjects (reduced to about 16,000 after quality control). These genotypes were drawn from seven different primary studies, hence the fairly stringent application of quality control leading to the exclusion of a significant number of genomes.

The investigation was conducted in two stages. In the first part, over 500,000 SNPs from nearly 12,000 individuals (4000 known AD cases, 7850 controls) were investigated. Analysis was limited to autosomal chromosomes (except for one SNP on the X-chromosome already known to be associated with AD). In addition to the established link to APOE, the study suggested involvement of two further genes CLU and PICALM. CLU (on chromosome 8) encodes clusterin, another significant brain apolipoprotein . The specific SNP identified lies within an intron.  PICALM (chromosome 11) encodes phosphatidylinositol-binding clathrin assembly protein (aka clathrin assembly lymphoid-myeloid leukaemia gene).  The specific SNP identified is 88.5 kb 5’ to the gene. The authors of this study had prior notice of the study of Lambert et al (see below) and note that CLU is also the novel gene demonstrating the strongest association with AD in their work.

During the second stage, the researchers looked at the two newly identified SNPs in five additional European cohorts, drawn from Belgium, Greece, Bonn, the Medical Research Council, UK and the Alzheimer’s Research Trust. These included just over 2,000 AD cases and about 2,300 age-matched, cognitively screened controls.

Armed with the data of Amouyel et al, these authors also re-examined their own results looking for evidence of associations identified in the other paper. They confirm “suggestive evidence” for association with CR1, the gene encoding complement receptor 1. They also report the genes for bridging integrator 1 (BIN1) and disabled homolog 1 (DAB1) as being “noteworthy”.

Finally, the authors consider the physiological implications of the various mutations.

• APOE acts as a chaperone for Ab, influencing the latter’s conformation, conversion to an insoluble form, aggregation and toxicity.
• Clusterin is a molecule that occurs in a fairly wide number of tissue types and with several likely functions. The gene product is a 449 AA protein which is later glycosylated.  It binds soluble Ab in a specific and reversible manner; the resulting complex has been demonstrated to cross the blood-brain barrier.
• PICALM is ubiquitously in all tissues, but particularly so in neurons. It is involved in clatharin-mediated endocytosis (as too is BIN1). The role of PICALM in trafficking of VAMP2, a protein involved in synaptic vesicle fusion to presynaptic membranes, may underlie its role in AD.

Genome-wide association study identifies variants at CLU and CR1 associated with Alzheimer’s disease
Lambert et al (2009) Nature Genetics 41:1094-1099

In the second paper, the authors (n = 50, plus unnamed members of the European AD Initiative Investigators, and headed by Philippe Amouyel in Lille) conducted an initial study of about 2000 French cases of AD and about 5,300 controls, also from France. As was also seen in the separate study by Harold et al, this analysis identified the CLU gene as having genome-wide significance in AD.

A number of further markers across several chromosomes had “suggestive evidence of association”. Therefore in a second stage, analogous to the methodology used by Harold et al, the researchers looked to confirm the observed association with CLU and probe more deeply the correlation with other markers in the major loci identified. The individuals in this stage (approx 4000 AD cases and 3,300 controls) came from Belgium, Finland, Italy and Spain. This second stage confirmed the significant association with CLU, and also identified a second gene CR1 (for complement receptor 1, the main protein interacting with complement protein C3b).

These authors review several previous experiments that support a suggestion that CR1 is involved in clearance of Ab. It is therefore their contention that whereas the gene errors implicated in early-onset (familial) AD result in Ab overproduction, those involved in standard late-onset AD may lead to inadequate Ab clearance.

None of the genes identified in the recent studies guarantee the development of AD – in this paper it is estimated the risk factors are 25.5% for APOE, 8.9% for CLU and 3.8% for CR1.

Two further points of note. One is the care taken to ensure results are genuine – when aware that a Belgian dataset had been examined in both this study and in the work by Harold et al the present authors recalculated their results excluding the shared data lest it introduced any unwarranted significance by being counted twice. Secondly, this paper labels the gene for phosphatidylinositol-binding clathrin assembly protein as PILCAM. Looking at the literature it seems that this is a typo, not an alternative name.

If you’ve read this far, you may also be interested in the assessment of this research published on the excellent NHS Choice Behind the headlines website.

From time to time examples of scientific fraud come to light and raise questions about the integrity of scientific endeavour. The most well-known example of recent years must surely be South Korean stem cell biologist Hwang Woo-Suk, whose ground-breaking discoveries in the field of therapeutic cloning were exposed as bogus (In addition to his science reputation being in tatters, Hwang was convicted in October 2009 of embezzlement and violation of bioethical laws, although he escaped a custodial sentence).

In physics, the multiple re-use of the same graphs as data for entirely different experiments led to the downfall of a leading young nanoscientist (this was the subject of a 2004 episode of the BBC’s Horizon series The dark secret of Hendrik Schön). Are Hwang and Schön rare examples bringing unwarranted criticism to a body of otherwise exemplary scientists, or are their crimes indicative of much wider malpractice within the scientific community?

University of Edinburgh researcher Daniele Fanelli has shed some light on the the extend of scientific fraud in an article How many scientists fabricate and falsify research? A systematic review and meta-analysis of survey data. Published in the open access journal PLoS ONE in May 2009, the research brought together data from a number of earlier smaller studies on scientific misconduct to generate “the first meta-analysis of these surveys” (p1).

Fanelli was interested in examining the rates of self-reporting of scientific misconduct and knowledge about the misconduct of colleagues. Recognising that “any boundary defining misconduct will be arbitrary” (p9), he limited discussion to incidents where there was clear “intention to deceive” (p1, p9) rather than generation of incorrect results as a consequence of shoddy experimental design and/or accidental misinterpretation of the data. For the purposes of this study, Fanelli also excluded plagiarism and other examples of “questionable research practices” (QRPs; such as failure to include a contributor amongst the list of authors for a paper) from his definition of scientific misconduct, which was instead limited to fabrication and falsification. The grounds for this decision seem valid; whereas fabrication (the invention of data) and falsification (the wilful distortion of results) change the actual body of scientific knowledge, these other unprofessional activities lead instead to changes in the distribution of credit for the work, the substance of which remains unaltered.

To identify the previous studies of misconduct, Fanelli conducted a search of citation databases, scientific journals, “grey literature” databases and internet search engines using the terms “research misconduct”, “research integrity”, “research malpractice”, “scientific fraud”, “fabrication, falsification” and “falsification, fabrication”. An initial search generated 3276 potentially relevant studies. The vast majority (3207) were easily excluded because they were not surveys of research misconduct.

The author then applied very strict criteria to limit the meta-analysis to genuinely appropriate studies. For example, papers were excluded if there was no quantitative data, if the data included no clear category of never/none/nobody (e.g. if only mean values were shown), if the sample had been generated in a non-random manner, or if undergraduate and/or other non-researchers were included in a manner that did not permit their removal from the dataset). Having done so, the initial pool of potential papers was whittled right down to 18 suitable studies.

Quantifying research malpractice

What were the conclusions of Fanelli’s analysis? The main issues addressed were the proportion of respondent admitting to misconduct or questionable practices of their own, or knowledge of similar behaviour committed by colleagues on at least one occasion.

In the various studies reviewed, between 0.3% and 4.9% of respondents confirmed that they had modified results to improve the outcomes. This led to an average of about 2% self-reporting of misconduct (although it was nearer 1% if the responses were limited to those that specifically mentioned ‘falsification’ or ‘fabrication’.

A rather larger number, 9.5%,  were willing to admit that they had carried out broader questionable practices. Again, however, the phrasing was important with more respondents willing to say they had “modified research results” than admitting that had reported results that they “knew to be untrue”. This may fit with an underlying assumption that it is okay to omit data that you “know” are outliers or otherwise “wrong”. As Fanelli puts it “many did not think that the data they “improved” were falsified” (p9).

When asked about the actions of others, a crude average of around 16.7% (range 5.2% to 33.3%, Fanelli elects to report this statistic as “up to 34%” (p10)) of scientists said they had personal knowledge that a colleague had fabricated or falsified data on at least one occasion. A much wider range (6.2% to 72%; crude mean 28.5%) said that they were aware of peers who had indulged in QRPs.

So, were Hwang and Schön isolated miscreants or does their identification mark the tip of an iceberg of scientific misconduct? The truth seems to lie somewhere in between. As Fanelli notes, usual rules of self-reporting bias – in which some people (typically older women) under-report criminal behaviour whereas others (typically younger males) over-report such activity – do not apply here. It is highly unlikely that anyone in a community where trust is taken seriously will over-report their own wrong actions. It is likely, therefore, that the calculated values of self-reported malpractice are underestimates.

The data regarding knowledge of other researchers’ actions are harder to validate. It is theoretically possible, for example, that more than one correspondent might be describing the wrongdoings of the same colleague. In contrast, the criteria was knowledge of malpractice on “at least one occasion” and therefore the data may not take into account serial offences.

Other approaches to measuring misconduct, as reviewed by Fanelli, have generated a range of figures which might be seen as very broadly equivalent. For example, about 0.02% of paper  are retracted from PubMed due to misconduct (Claxton, 2005). 1% of papers submitted to the Journal of Cell Biology were found to have been inappropriately manipulated (Steneck, 2006). 2% of clinical researchers were found guilty of serious scientific misconduct in routine US Food and Drug Administration audits (Glick, 1992).

Whatever the accuracy of these numbers, however, it remains true that the vast majority of science is carried out in a spirit of accuracy and integrity.

There are many reasons why I am grateful to have spent some of my summer reading Ben Goldacre’s excellent book Bad Science, including the fact that it brought to my attention a paper The Seductive Allure of Neuroscience Explanations. The article is an account of experiments conducted by Deena Weisberg and colleagues at Yale University, and was published in the Journal of Cognitive Neuroscience in 2008.

Recognising that neuroscience is an area of research that fascinates the public and where discoveries are frequently picked up by the general press, Weisberg et al generated four explanatory statements for each of 18 different psychological phenoma. In each case the four statements represented:

• a good explanation without specific mention of neuroscience
• the same good explanation with the addition of plausible, but logically irrelevant, neuroscientific details
• a bad explanation without specific mention of neuroscience
• the same bad explanation with the addition of the same plausible (but irrelevant) neuroscience as in the second example

Satisfaction (on a 7-point scale) engendered by the various statements was tested with three cohorts:

• naive adults (n=81), i.e. individuals with no formal neuroscience training
• students from an introductory cognitive neuroscience course (n=22)
• experts in neuroscience (n=48), although the definition of “expert” in this context was quite generous – 6 members of this group had completed an undergraduate course but were yet to start their advanced degrees, 29 were currently in graduate school and the other 13 were beyond grad school.

In developing their neuroscience explanations, the authors held to three important criteria: (1) the phrasing would indicate that this was a field in which knowledge was already established, (2) the same information would be added to both the good and the bad explanation, and (3) the neuroscience information should not alter the underlying logic of the explanation such as it was before addition of the extra ’science’.

With all three cohorts, the participants were informed that the study being reported was scientifically robust, but that the explanation offered may not be genuine. For both the ‘naive’ and ‘expert’ groups any individual always saw only statements that contained neuroscientific information (but might be a good or bad explanation), or they never had the additional details. This approach was adopted so that subjects were not alerted to the fact that some of the explanations were fuller than others. Unfortunately the relatively small size of the current undergraduate cohort meant that the method was different, any individual could be exposed to all 4 different types of explanation. This is a shame, since it does require some caution when interpreting side-by-side comparisons of the three cohorts.

So what were the findings? For all three groups, participants rated the good explanations as more satisfactory than the bad ones. Both the naive group and the students found explanations with added neuroscience to be better, and the effect was more striking for the bad explanations than for the good. With the experts, however, addition of the spurious scientific details to the good explanation actually led to a reduction in their satisfaction – an indication that with their fuller knowledge they saw through the vacuous additions.

The authors argue that the findings with novices and students may be manifestations of the “seductive details effect”. Previous studies have suggested interesting-but-irrelevant information can have a detrimental effect on cognitive tasks, e.g. memory tests.

They also point to a general observation that individuals respond more favourably when given a reason for a request, however obvious. They cite one example of such “placebic” information in which research subjects were more amenable to letting someone carry out some photocopying if they added the phrase “I have some copies to make” to a bland request “May I use your Xerox machine?”

In Bad Science, Ben Goldacre argues that ‘quacks’ exploit this phenomenon by dressing up their claims with sciencey-sounding explanations. This reminded me too of the BBC documentary Professor Regan’s Medicine Cabinet (first shown in April 2009) in which, amongst other things, Regan dresses up a test of pills for treating insomnia with all sorts of detailed instructions about what must, and what must not, be done in order for the therapy to be effective. The participants stuck to the rules and reported beneficial effects of the pills, despite the fact that they were – in fact – sugar cake decorations.

Weinberg D.S., Keil F.C., Goodstein J., Rawson E. and Gray J.R. (2008) The seductive allure of neuroscience explanations Journal of Cognitive Neuroscience 20:470-477

Amongst the major science research journals, Science magazine has consistently been the most prominent in flying the flag for science education. I was very interested, therefore, in an Editorial by Carl Wieman in the September 4th 2009 issue of the magazine. In his piece Galvanising Science Departments, Wieman describes some fairly radical innovations in Science Education currently underway at the University of Colorado and the University of Bristish Columbia. The aim is to adopt evidence-based teaching methodologies with emphasis on the development of scientific thinking and problem-solving skills rather than fact regurgitation.

I have no direct experience of teaching in the USA, either as provider or recipient. I know, for example, that much greater emphasis is placed on the recommended course text in the USA than in the UK, but beyond that I cannot speak with any authority. It does sound like some of the reported innovations are things that have taken place here for some while, such as the addition of specific (skill-centred) learning goals to modules. A cornerstone of the strategy has been appointment of science education specialists, individuals who not only have expertise in their subject discipline, but are also au fait with educational and cognitive psychology studies, a variety of effective teaching strategies and – I note with some mirth – possess diplomatic skills!  The programme is ongoing, the University of Colorado is in the 4th year of an initial six year project and so the full impact of the developments will not be known for some while.

What really struck me, however, was the extent of the commitment at an institutional level, including provision of serious money to fund these changes. Far too many pro-active educators, motivated by genuine desire to improve the learning experience for their students actual receive flak not gratitude. In many cases this is, I believe, because their approach to pedagogy is different to the cultural norm and, as such, moved students (and staff) outside their comfort zone.

So much of education, even at University level, is about recall of a prescribed body of information. This is relatively easy – for both the teacher and the student; it is the mindset that underlies that chirping questions “will this be in the test?” Development of thinking skills demands more from everyone. If innovations that require higher skills are associated with just one module, or even just one academic, then that individual may unfairly receive criticism in feedback from the class.

I believe this is why Wieman is absolutely right when he says “an entire department must be the unit of change”. Depending upon institutional structure, it may even require a larger body – School, Faculty, College – to move together with shared commitment to the new goals. So far at Colorado, 60% of academic staff in three Science departments have embraced the new teaching approaches, impacting 80% of their students’ credit hours. Faculty are reported to enthusiastically discussing teaching as a scholarly activity – that’s surely got to be a good thing. But – it needs time and it needs money.

Wieman C (2009) Galvanising Science Departments Science 325:1181

I was intrigued by a recent paper Cognitive control in media multitaskers in the highly-regarded journal Proceedings of the National Academy of Sciences. The study looked at the information processing styles of self-reported media multitaskers, defined as users of two or more content streams simultaneously, compared with those who do not multitask in this way. (I will skirt over the irony both that I was supposed to be doing something else at the time I spotted the BBC report of the research, and that my son has just switched on the TV as I type this review on my laptop.)

An electronic copy of the paper was posted online in August 2009, ahead of publication in the print journal (doi: 10.1073/pnas.0903620106)

In their research, Ophir et al asked 262 University students to complete an online self-assessment questionnaire regarding both the total number of hours spent using different media (they specified 12 formats including TV, online video use, music, print media, e-mail and text messaging) and how likely they might be to use some of these concurrently alongside a primary task. The authors then generated a numerical Media Multitasking Index (MMI) and ranked the students. Those with a score one or more standard deviations below the mean (light media multitaskers, LMMs) or one or more SDs above the mean (heavy media multitaskers, HMMs) were then invited to participate in a series of cognitive ability tests. In total there were between 30 and 41 students taking the various tests, evenly split between LMMs and HMMs.

Central to the whole study was the issue of working memory, a concept we have discussed previously on JLB. Some of the experiments were designed to look at the students abilities to filter out distractions and remain focused on the main task. In one test, for example, the students were asked to compare the orientation of two red rectangles in consecutive images shown to them 900 milliseconds apart. To investigate the influence of distractions, the image also included 0, 2, 4 or 6 blue rectangles. Each student was asked to do this 200 times in one sitting,  during which there were an equal number of slides in which the position changed and in which it did not (the order in which they saw each of the image pairs was randomised). This test revealed a difference between the LMMs and the HMMs. Contrary to the hopes of self-confessed media multitaskers everywhere, the HMM group seemed to be more easily distracted than the LMM, whose performance was largely unaffected by the presence of blue rectangles.

In a second test, the investigators looked into the ability of participants to hit the correct button “animal” v “nonanimal” when presented with 24 words. Each word was presented one at a time, and all 24 words were used twice in the sequence (ie 48 words total). The variation was then introduced whereby the students were presented with the same words again, but this time with the additional instruction that they should refrain from hitting the button if they heard a stop tone before they responded. The signal was transmitted 225 ms before the mean response time, and was used on 25% of occasions. In this test there was not a significant difference between the two groups.

There was also no significant difference between the HMMs and LMMs in the first part of a third test. Participants were shown a sequence of letter pairs 300 ms apart (the letters were red on a black background, the relevance of this will be reached in a moment). The students were to hit the “no” button on all occasions except when there was a cue “A” followed by a probe “X”, hence this is known as an AX Continuous Performance Task (AX-CPT).

A difference was reported, however, when distractors were introduced into the AX-CPT. Between the cue letter (red on black, remember) and the probe letter (also red on black), the participants were presented with three distractor letters which were white on black. The distractors were never A, X or indeed K or Y both of which were entirely omitted from the study to avoid shape recognition clashes of X v Y/K. The overall time between cue and probe was held constant (4900 ms) relative to original AX-CPT test.

In the AX-CPT test with distractors, the HMM students were found to have the same accuracy as the LMM cohort, but responded more slowly. Alongside the relative responses in the rectangle distraction test, it seems that HMMs are worse than LMMs at filtering out irrelevant environmental stimuli.

Given the inherent need to swap between activities implicit in multi-tasking, the findings from the task-switching test are probably the most illuminating. In this experiment, the students were presented with a cue that said either “letter” or “number” followed by a stimulus that always featured one letter and one number. If the “letter” cue had been shown, the participants were to ignore the number and press the left button for a vowel and the right button for a non-vowel (only 4 consonants (K, N, P, S)  and 4 vowels (A, E, I, U) were ever used to avoid bias to the non-vowel side; O was not included). conversely if the “number” cue was used then it was left button for “odd” and right button for “even” (again, only 4 of each sort of stimuli were shown; the numbers 1 and 0 were omitted. Each student saw 4 sets of 80 cue and stimulus pairs, 40% involved task switching and 60% non-switching. They were given “warm-up” activities either involving switching between numbers and letters or non-switching.  HMMs were slower to respond than LMMs in both non-switch (259 ms) and switch (426 ms) trials. This result is interpreted as showing that HMMs are less effective at suppressing irrelevant tasks.

In the final set of studies, the researchers looked at abilities to ignore distractions within memory. This was investigated by conducting so-called “two-back” and “three-back” tasks in which participants were asked to indicate whether a letter on the screen was a “target”, i.e. the same as the letter 2 or 3 previously (in respective study) or “non-target”, i.e. different. Each student has 3 x 30 goes, in which 10 were targets and 20 non-targets.

In the two/three-back tasks the HMMs again performed less well than the LMMs. Unsurprisingly, both groups were worse in the three-back tests than in the two-back tests and their relative declines between the two were similar. Interestingly, however, HMMs had a greater tendancy to identify false positives, i.e. to report a non-target as target, particularly in association with letters than had come earlier in the sequence and were thus “familiar” but not in the correct relative position. This effect became more apparent as the task progressed, presumably as there were more “familiar” letters that had been and gone.

So, the headline message was that self-reported multitaskers are less adept at multitasking than those who shun multiple media. As someone who will admit to being an HMM, this is a disappointment. We could argue that the tests used in the study were surrogates – the real worldquestion would be “can you get work done appropriately whilst Twitter and E-mail are open on your dashboard and you are listening to your iPod?” not “can you remember if the red rectangles are in the same orientation if there are lots of blue ones there too?” I would also defend my position by acknowledging that my Twitter network is now a major source of information with “bang on the money” relevance for my work. Nevertheless I do know for myself that when I’ve really got to get something written I hide myself away with minimum distractions, the only added media will be background music to block out other noise, and this must be lyric-free to be fit for purpose! So, I will reflect on my work practice and perhaps make accessing of e-mail and Twitter more punctuated and less continuous, but I don’t feel that cutting off these channels entirely would – on balance – be a benefit. As the authors themselves point out, it may be that HMMs are demonstrating “a bias towards exploratory, rather than exploitative, information processing”.

Psychology and cognitive science are not my normal field of study and I therefore hope that I haven’t misrepresented any of the methods or findings of the PNAS paper in the proceeding discussion. I thought this was a truly interesting investigation, and highly topical – it is no longer just adolescents that are habitual multitaskers. Finally, from an educational research point of view, it was curious to note that participants in this research received course credits for their involvement in the study, both for taking the questionnaire in the first place and then for their role in the tasks themselves. I’m not sure we’d get away with that in a UK context.

From time to time I find myself ruminating on exactly how and where I acquired a variety of study skills. I have no recollection, for example, of any formalised training in finding and selecting source materials and yet even as an undergraduate I seemed to be reasonably adept at choosing relevant information.

Back in those days, of course, finding the material was a bigger task than it is today.  The contemporary student is spared the need to leaf through the Index Medicus and then scourer the shelves for the relevant tomes before staggering to the photocopier burdened by the combined weight of several impressively-bound volumes. A multitude of online search engines and databases, coupled with institutional access to a plethora of electronic journals provides the 21st Century undergraduate with more information than they could ever hope to process in time to meet the assignment deadline (to say nothing of the less formal internet sources at their fingertips). 

The challenge today is much less focussed on the location of documents and much more heavily skewed towards the evaluation and selection of appropriate material; discriminating good sources from bad and the evaluating the relevance of the material to the specific task at hand.  Having recently marked a set of essays where several submissions were characterised by failure to achieve these goals, I am reminded that the challenge of panning through tonnes of data in order to extract pertinent nuggets of gold is considerable, and the skills necessary to achieve this are unlikely to be acquired by osmosis.

It was against this backdrop that I was interested to read a paper The annotated bibliography and citation behaviour: enhancing student scholarship in an undergraduate biology course, written by Molly Flashpohler and colleagues, and published in the Winter 2007 edition of CBE – Life Sciences Education. The paper describes an active intervention to develop the ”Information Literacy” skills of students taking an immunology and parasitology module at a small college in Minnesota, USA. 

In keeping with many institutions, our programme includes sessions on making the most of Web of Science, PubMed and other such search tools.  In addition to this, I’ve done some work myself on plagiarism-prevention.  What I don’t believe we’ve offered previously is any intervention of the type described by the Flaspohlers, in which students are encouraged to undertake a meta-critique of the reference materials they have assembled for a forthcoming assignment.

In the Minnesota model, the students are required to produce an annotated bibliography.  This is more than a citation list; each resource they are intending to use must be critiqued in a brief (approx 150 word) paragraph.  The students are advised that this annotation should demonstrate, as far as possible: the authority or background of the original author(s); their intended audience; the writing style of the author(s); how this article relates to their project; any bias or point of view apparent in the original work; and to highlight any tables, figures, etc which the student feels will be particularly relevant for their task.  Further to this, the annotations are supposed to cross-reference to one another as the student compares and contrasts the specific work with others that they have chosen to cite.

It strikes me that a task of this kind is “do-able” – it could be added as a training step on the way to submission of essays or dissertations that are already part of our courses.  The authors describe the improvements they have seen in the seven-year evolution of this intervention – including a shift towards better quality and more authoritative sources, and a concommitant fall in instances of plagiarism. Admittedly they have been working with a relatively small cohort (average 17 per annum), but I suspect that the knock-on merits of adding an exercise of this type would justify the effort of introducing something similar to my group, where n is nearer 100.

To re-quote John Porter, author of an earlier work in the same journal (albeit under its former name): “An awareness of the current literature is as important to scientific research as the careful design of adequate controls“.  This being the case, Information Literacy is too significant to be left to osmosis.

In preparation for a recent meeting of our School of Biological Sciences Pedagogic Research group, I’ve been reading a number of articles in the Assessment for Learning genre. My attention was particularly drawn to accounts of ELLI – the Evaluating Lifelong Learning Inventory – project.  ELLI has been developed by Ruth Deakin Crick and colleagues at the University of Bristol and is an instrument for the diagnosis and development of an individual’s learning power.  As Deakin Crick and colleagues point out in their 2004 paper (Assessment in Education 11:247-272), existing assessments tend to be focused on measuring either intelligence or educational achievement, whereas the measurement of a person’s learning power, that is their capacity for lifelong learning, may actually be the most valuable quality to assess.  “There is“, she notes in her 2007 paper (The Curriculum Journal 18:135-153), “an urgent need for our education system to foster flexible, creative, self-aware and dynamic learners who have the capacity to apply and adapt what is learned to their own lives, embedded in their local and global communities, and who can extend their learning and understanding into spheres of thought and action which demand intelligent behaviour in the real world” (p.137).

Having extensively piloted versions of the assessment instrument in a variety of Schools, the ELLI team have presently settled on a 72-item questionnaire that is used to probe students’ self-perception of their relative strengths and weaknesses in the seven core dimensions of learning power. These are: changing and learning; critical curiosity; meaning-making; dependence and fragility; creativity; relationships/interdependence; and strategic awareness.  Positive and negative manifestations of these dimensions are summarised as part of the slideshow below, along with an overview of the findings of ELLI so far.

.

I must admit the ability to have a handle on the learning potential of our students is an attractive proposition; it might also be illuminating to take the test myself and see what it showed about my own strengths and weaknesses in this regard! There would definitely be value, even if it was just as a one-off diagnostic.  Clearly, however, the most merit comes from being able to combine the diagnostic with a programme of activities that develop and enhance the learning power of the students, in a targeted and individualised way, combined with a reappraisal of their learning power at a later stage. 

This is a big ask and one which, with the best will in the world, is going to be tricky to fulfil in the Higher Education sector.  All sorts of problems stand in the way – the large cohort sizes; the fact that an individual student is receiving input and instruction from a broad range of colleagues rather than one staff member for a significant time; the impact of the secondary sector, where persistent summative assessment has led students to be entirely goal-driven and only engage if there are marks up for grabs. 

The balance between skill development and acquisition of factual knowledge is a recurring tension in HE, but I’d love to be able to utilise an instrument such as ELLI in order to shift the balance more towards learner enhancement and less on information regurgitation.