Study

When we are presented with new information, we try and connect it to information we already hold. This is automatic. Sometimes the information fits in easily; other times the fit is more difficult — perhaps because some of our old information is wrong, or perhaps because we lack some of the knowledge we need to fit them together.

When we're confronted by contradictory information, our first reaction is usually surprise. But if the surprise continues, with the contradictions perhaps increasing, or at any rate becoming no closer to being resolved, then our emotional reaction turns to confusion.

Confusion is very common in the learning process, despite most educators thinking that effective teaching is all about minimizing, if not eliminating, confusion.

But recent research has suggested that confusion is not necessarily a bad thing. Indeed, in some circumstances, it may be desirable.

I see this as an example of the broader notion of ‘desirable difficulty’, which is the subject of my current post. But let’s look first at this recent study on confusion for learning.

In the study, students engaged in ‘trialogues’ involving themselves and two animated agents. The trialogues discussed possible flaws in a scientific study, and the animated agents took the roles of a tutor and a student peer. To get the student thinking about what makes a good scientific study, the agents disagreed with each other on certain points, and the student had to decide who was right. On some occasions, the agents made incorrect or contradictory statements about the study.

In the first experiment, involving 64 students, there were four opportunities for contradictions during the discussion of each research study. Because the overall levels of student confusion were quite low, a second experiment, involving 76 students, used a delayed manipulation, where the animated agents initially agreed with each other but eventually started to express divergent views. In this condition, students were sometimes then given a text to read to help them resolve their confusion. It was thought that, given their confusion, students would read the text with particular attention, and so improve their learning.

In both experiments, on those trials which genuinely confused the students, those students who were initially confused by the contradiction between the two agents did significantly better on the test at the end.

A side-note: self-reports of confusion were not very sensitive, and students’ responses to forced-choice questions following the contradictions were more sensitive at inferring confusion. This is a reminder that students are not necessarily good judges of their own confusion!

The idea behind all this is that, when there’s a mismatch between new information and prior knowledge, we have to explore the contradictions more deeply — make an effort to explain the contradictions. Such deeper processing should result in more durable and accessible memory codes.

Such a mismatch can occur in many, quite diverse contexts — not simply in the study situation. For example, unexpected feedback, anomalous events, obstacles to goals, or interruptions of familiar action sequences, all create some sort of mismatch between incoming information and prior knowledge.

However, all instances of confusion aren’t necessarily useful for learning and memory. They need to be relevant to the activity, and of course the individual needs to have the means to resolve the confusion.

As I said, I see a relationship between this idea of the right level and type of confusion enhancing learning, and the idea of desirable difficulty. I’ve talked before about the ‘desirable difficulty’ effect (see, for example, Using 'hard to read' fonts may help you remember more). Both of these ideas, of course, connect to a much older and more fundamental idea: that of levels of processing. The idea that we can process information at varying levels, and that deeper levels of processing improve memory and learning, dates back to a paper written in 1972 by Craik and Lockhart (although it has been developed and modified over the years), and underpins (usually implicitly) much educational thinking.

But it’s not so much this fundamental notion that deeper processing helps memory and learning, and certain desirable difficulties encourage deeper processing, that interests me as much as idea of getting the level right.

Too much confusion is usually counter-productive; too much difficulty the same.

Getting the difficulty level right is something I have talked about in connection with flow. On the face of it, confusion would seem to be counterproductive for achieving flow, and yet ... it rather depends on the level of confusion, don't you think? If the student has clear paths to follow to resolve the confusion, the information flow doesn't need to stop.

This idea also, perhaps, has connections to effective practice principles — specifically, what I call the ‘Just-in-time rule’. This is the principle that the optimal spacing for your retrieval practice depends on you retrieving the information just before you would have forgotten it. (That’s not as occult as it sounds! But I’m not here to discuss that today.)

It seems to me that another way of thinking about this is that you want to find that moment when retrieval of that information is at the ‘right’ level of difficulty — neither too easy, nor too hard.

Successful teaching is about shaping the information flow so that the student experiences it — moment by moment — at the right level of difficulty. This is, of course, impossible in a factory-model classroom, but the mechanics of tailoring the information flow to the individual are now made possible by technology.

But technology isn't the answer on its own. To achieve optimal results, it helps if the individual student is aware that the success of their learning depends on (or will at least be more effective — for some will be successful regardless of the inadequacy of the instruction) managing the information flow. Which means they need to provide honest feedback, they need to be able to monitor their learning and recognize when they have ‘got’ something and when they haven’t, and they need to understand that if one approach to a subject isn’t working for them, then they need to try a different one.

Perhaps this provides a different perspective for some of you. I'd love to hear of any thoughts or experiences teachers and students have had that bear on these issues.

One of the points I mention in my book on notetaking is that the very act of taking notes helps us remember — it’s not simply about providing yourself with a record. There are a number of reasons for this, but a recent study bears on one of them. The researchers were interested in whether physically writing by hand has a different effect than typing on a keyboard.

In a fascinating experiment, adults were asked to learn to write in an unknown alphabet, with around twenty letters. One group was taught to write by hand, while another group used a keyboard. Participants were tested on their fluency and recall after three and six weeks. Those who had learned the letters by handwriting were significantly better on all tests. Moreover, Broca's area, a brain region involved in language, was active when this group were recognizing the letters, but not among those who had learned by typing on a keyboard.

The findings point to the importance of sensorimotor processes in processes we have typically regarded as primarily intellectual.

I recently reported on another finding concerning handwriting — that the memory-blocking effect of exam anxiety could be overcome by the simple strategy of writing out your anxieties just before the exam. It’s also interesting in this context to remember the research into the benefits of gesturing for reducing the load on your working memory, with consequent assistance for memory, learning and comprehension. The writing effect on exam anxiety is also thought to be related to reducing the load on working memory.

In the case of this latest study, it seems likely that the benefits have more to do with the increased focus on the shape of the letters that occurs when writing by hand, and with the intimate connection between reading and writing.

But the message of these different studies is the same: that we ignore the physical at our peril; that cognition is “embodied cognition”, rooted in our bodies in ways we are only beginning to understand.

The limitations of working memory have implications for all of us. The challenges that come from having a low working memory capacity are not only relevant for particular individuals, but also for almost all of us at some points of our lives. Because working memory capacity has a natural cycle — in childhood it grows with age; in old age it begins to shrink. So the problems that come with a low working memory capacity, and strategies for dealing with it, are ones that all of us need to be aware of.

Today, I want to talk a little about the effect of low working memory capacity on reading comprehension.

A recent study involving 400 University of Alberta students found that 5% of them had reading comprehension difficulties. Now the interesting thing about this is that these were not conventionally poor readers. They could read perfectly well. Their problem lay in making sense of what they were reading. Not because they didn’t understand the words or the meaning of the text. Because they had trouble remembering what they had read earlier.

Now these were good students — they had at least managed to get through high school sufficiently well to go to university — and many of them had developed useful strategies for helping them with this task: highlighting, making annotations in the margins of the text, and so on. But it was still very difficult for them to get hold of the big picture — seeing and understanding the text as a whole.

This is more precisely demonstrated in a very recent study that required 62 undergraduates to read a website on the taxonomy of plants. Now this represents a situation that is much more like a real-world study scenario, and one that has, as far as I know, been little studied: namely, drawing together information from multiple documents.

In this experiment, the multiple documents were represented by 24 web pages. Each page discussed a different part of the plant taxonomy. The website as a whole was organized according to a four-level hierarchical tree structure, where the highest level covered the broadest classes of plants (“Plants”), and the lowest, individual species. However — and this is the important point — there was no explicit mention of this organization, and you could navigate only one link up or down the tree, not sideways. Participants entered the site at the top level.

After pretesting, to assess WMC and prior plant knowledge, the students were given 18 search questions. Participants were asked both to read the site and answer the questions. They were given 25 minutes to do so, after which they completed a post-test similar to their pre-test of prior knowledge: (1) placing the eight terms found in the first three levels on the hierarchical tree (tree construction task); (2) selecting the correct two items from a list of five, that were subordinates to a given item (matching task).

Neither WMC nor prior knowledge affected performance on the search task. Neither WMC nor prior knowledge (nor indeed performance on the search task) directly affected performance on the post-test matching task, indicating that learning simple factual knowledge is not affected by your working memory capacity or how much relevant knowledge you have (remember though, that this was a very simple and limited amount of new knowledge).

But, WMC did significantly affect understanding of the hierarchical structure (assessed by the tree construction task). Prior knowledge did not.

These findings don’t only tell us about the importance of WMC for seeing the big picture, they also provide some evidence of what underlies that, or at least what doesn’t. The findings that WMC didn’t affect the other tasks argues against the idea that high WMC individuals may be benefiting from a faster reading speed, or that they are better at making local connections, or that they can cope better at doing multiple tasks. WMC didn’t affect performance on the search questions, and it didn’t affect performance on the matching task, which tested understanding of local connections. No, the only benefit of a high WMC was in seeing global connections that had not been made explicitly.

Let’s go back to the first study for a moment. Many of the students having difficulties apparently did use strategies to help them deal with their problem, but their strategy use obviously wasn’t enough. I suspect part of the problem here, is that they didn’t really realize what their problem was (and you can’t employ the best strategies if you don’t properly understand the situation you’re dealing with!).

This isn’t just an issue for people who lack the cognitive knowledge and the self-knowledge (“metacognition”) to understand their intrinsic problem. It’s also an issue for adults whose working memory capacity has been reduced, either through age or potentially temporary causes such as sleep deprivation or poor health. In these cases, it’s easy to keep on believing that ways of doing things that used to work will continue to be effective, not realizing that something fundamental (WMC) has changed, necessitating new strategies.

So, let’s get to the burning question: how do you read / study effectively when your WMC is low?

The first thing is to be aware of how little you can hold in your mind at one time. This is where paragraphs are so useful, and why readability is affected by length of paragraphs. Theoretically (according to ‘best practice’), there should be no more than one idea per paragraph. The trick to successfully negotiating the hurdle of lengthy texts lies in encapsulation, and like most effective strategies, it becomes easier with practice.

Rule 1: Reduce each paragraph to as concise a label as you can.

Remember: “concise” means not simply brief, but rather, as brief as it can be while still reminding you of all the relevant information that is encompassed in the text. This is about capturing the essence.

Yes, it’s an art, and to do it well takes a lot of practice. But you don’t have to be a master of it to benefit from the strategy.

The next step is to connect your labels. This, of course, is a situation where a mind map-type strategy is very useful.

Rule 2: Connect your labels.

If you are one of those who are intimidated by mind maps, don’t be alarmed. I said, “mind map-type”. All you have to do is write your labels (I call them labels to emphasize the need for brevity, but of course they may be as long as a shortish sentence) on a sheet of paper, preferably in a loose circle so that you can easily draw lines between them. You should also try to write something by these lines, to express your idea of the connection. These labels will also provide a more condensed label for the ideas being connected. You can now make connections between these labels and the others.

The trick is to move in small steps, but not to stay small. Think of the process as a snowball, gathering ideas and facts as it goes, getting (slowly) bigger and bigger. Basically, it’s about condensing and connecting, until you have everything densely connected, and the information getting more and more condensed, until you see the whole picture, and understand the essence of it.

Another advantage of this method is that you will have greatly increased your chances of remembering it in the long-term!

In a situation similar to that of the second study — assorted web pages — you want to end up with a tight cluster of labels for each page, the whole of which is summed up by one single label.

What all this means for teachers, writers of text books, and designers of instructional environments, is that they should put greater effort into making explicit global connections — the ‘big picture’.

A final comment about background knowledge. Notwithstanding the finding of the second study that there was no particular benefit to prior knowledge, the other part of this process is to make connections with knowledge you already have. I’d remind you again that that study was only testing an extremely limited knowledge set, and this greatly limits its implications for real-world learning.

I have spoken before of how long-term memory can effectively increase our limited WMC (regardless of whether your WMC is low or high). Because long-term memory is essentially limitless. But information in it varies in its accessibility. It is only the readily accessible information that can bolster working memory.

So, there are two aspects to this when it comes to reading comprehension. The first is that you want any relevant information you have in LTM to be ‘primed’, i.e. reading and waiting. The second is that you are obviously going to do better if you actually have some relevant information, and the more the better!

This is where the educational movement to ‘dig deep not broad’ falls down. Now, I am certainly not arguing against this approach; I think it has a lot of positive aspects. But let’s not throw out the baby with the bathwater. A certain amount of breadth is necessary, and this of course is where reading truly comes into its own. Reading widely garners the wide background knowledge that we need — and those with WMC problems need in particular — to comprehend text and counteract the limitations of working memory. Because reading widely — if you choose wisely — builds a rich database in LTM.

We say: you are what you eat. Another statement is at least as true: we are what we read.

A brief round-up of a few of the latest findings reinforcing the fact that academic achievement is not all about academic ability or skills.Most of these relate to the importance of social factors.

Improving social belonging improves GPA, well-being & health, in African-American students

From Stanford, we have a reminder of the effects of stereotype threat, and an interesting intervention that ameliorated it. The study involved 92 freshmen, of whom 49 were African-American, and the rest white. Half the participants (none of whom were told the true purpose of the exercise) read surveys and essays written by upperclassmen of different ethnicities describing the difficulties they had fitting in during their first year at school. The other subjects read about experiences unrelated to a sense of belonging. The treatment subjects were then asked to write essays about why they thought the older college students' experiences changed, with illustrations from their own lives, and then to rewrite their essays into speeches that would be videotaped and could be shown to future students.

The idea of this intervention was to get the students to realize that everyone, regardless of race, has difficulty adjusting to college, and has times when they feel alienated or rejected.

While this exercise had no apparent effect on the white students, it had a significant impact on the grades and health of the black students. Grade point averages went up by almost a third of a grade between their sophomore and senior years, and 22% of them landed in the top 25% of their graduating class, compared to about 5% of black students who didn't participate in the exercise.

Moreover, the black students in the treatment group reported a greater sense of belonging compared to their peers in the control group; they were happier, less likely to spontaneously think about negative racial stereotypes, and apparently healthier (3 years after the intervention, 28% had visited a doctor recently, vs 60% in the control group).

Source: http://news.stanford.edu/news/2011/march/improve-minority-grades-031711…

Protecting against gender stereotype threat Stereotype threat is a potential factor for gender as well as ethnicity.

I’ve reported on a number of studies showing that reminding women or girls of gender stereotypes in math results in poorer performance on subsequent math tests. A new study suggests that women could be “inoculated” against such effects if their math / science class is taught by a woman. Although in these experiments, women’s academic performance didn’t suffer, their engagement and commitment to their STEM major was significantly affected.

In the first study, 72 women majoring in STEM subjects were given several tests measuring their implicit and explicit attitudes towards math vs English, plus a short but difficult math test. Half the students were (individually) tested by a female peer expert, supposedly double majoring in math and psychology, and half by a male peer. Those with a male showed negative implicit attitudes towards math, while those tested by a female showed equal liking for math and English on an implicit attitudes test. Similarly, women implicitly identified more with math in the presence of the female expert. On the math test, women who met the female attempted more problems (an average of 7.73 out of 10 compared to 6.39). There was no effect on performance — but because of the difficulty of the test, there was a floor effect.

In the second study, 101 women majoring in engineering were given short biographies of 5 engineers, who were either male or female, or descriptions of engineering innovations (control condition). Again, women presented with female engineers showed equal preference for math and English in the subsequent implicit attitudes test, while those presented with male engineers or innovations showed a significant implicit negative attitude to math. However, implicit identification with math wasn’t any stronger after reading about female engineers. However, those who read about female engineers did report greater intentions to pursue an engineering career, and this was mediated by greater self-efficacy in engineering. Again, there was no effect on explicit attitudes toward math.

In the third study, the performance of 42 female and 49 male students in introductory calculus course sections taught by male (8 sections) and female instructors (7 sections) were compared. Professors were yoked to same-sex teaching assistants.

As with the earlier studies, female students implicitly liked math and English equally when the teacher was a women, but had a decidedly more negative attitude toward math when their instructor was a man. Male students were unaffected by teacher gender. Similarly, female showed greater implicit identification with math when their teacher was a woman; male students were unaffected. Female students also expected better grades when their teacher was a woman; male students didn’t differ as a function of teacher gender (it should be noted that this wasn’t because they thought the women would be more generous markers; marking was pooled across all the instructors, and the students knew this). There was no effect of teacher gender on final grade (but there was a main effect of student gender: women outperformed men).

In other words, the findings of the 3rd study confirmed the effects on implicit attitudes towards STEM subjects, and demonstrated that male students were unaffected by the interventions that affected female students.

Now we come to engagement. At the beginning of the semester, female students were much less likely than male students (9% vs. 23%) to respond to questions put to the class, but later on, female students in sections led by women were much more likely to respond to such questions than were women in courses taught by men (46% vs 7%). Interestingly, more male students also responded to questions posed by female instructors (42% vs 26%). That would seem to suggest that male instructors are much more likely to engage in strategies that discourage many students from engaging in the class. But undeniably, women are more affected by this.

Additionally, at the beginning of the courses, around the same number of female students approached their instructors, regardless of their gender (12-13%). But later, while this percentage of female students approaching female instructors stayed constant, none of them approached male instructors. This could be taken to mean male instructors consistently discouraged such behavior, but male students did not change (an average of 7% both at Time 1 and Time 2).

The number of students who asked questions in class did not vary over time, or by student gender. However it did vary by teacher gender: 22% of both male and female students asked questions in class when they were taught by women, while only 15% did so in courses taught by men.

Some of these effects then seem to indicate that male college instructors are more inclined to discourage student engagement. What the effects of that are, remains to be seen.

Source: http://www.insidehighered.com/news/2011/03/03/study_suggests_role_of_ro…

Social and emotional learning programs found to boost student improvement

A review of 213 school programs that enhance students' social and emotional development, has found that such programs not only significantly improved social and emotional skills, caring attitudes, and positive social behaviors, but also resulted in significant improvement on achievement tests (although only a small subset of these programs actually looked at this aspect, the numbers of students involved were very large).

The average improvement in grades and standardized-test scores was 11 percentile points —an improvement that falls within the range of effectiveness of academic interventions.

Source: http://www.physorg.com/news/2011-02-social-emotional-boost-students-ski…

http://www.edweek.org/ew/articles/2011/02/04/20sel.h30.html

Boys need close friendships

Related to this perhaps (I looked but couldn’t find any gender numbers for the SEL programs), from the Celebration of Teaching and Learning Conference in New York, developmental psychologist Niobe Way argues that one reason why boys are struggling in school is that they are experiencing a "crisis of connection." Stereotypical notions of masculinity, that emphasize separation and independence, challenge their need for close friendships. She's found that many boys have close friendships that are being discouraged by anxiety about being seen as gay or effeminate.

Way says that having close friendships is linked to better physical and mental health, lower rates of drug use and gang membership, and higher levels of academic achievement and engagement. When asked, she encouraged teachers to allow boys to sit next to their best friends in class.

Source: http://blogs.edweek.org/teachers/teaching_now/2011/03/psychologist_boys…

High rate of college students with unrecognized hearing loss

On a completely different note, a study involving 56 college students has found that fully a quarter of them showed 15 decibels or more of hearing loss at one or more test frequencies — an amount that is not severe enough to require a hearing aid, but could disrupt learning. The highest levels of high frequency hearing loss were in male students who reported using personal music players.

Source: http://www.physorg.com/news/2011-03-college-students.html

Our society gives a lot of weight to intelligence. Academics may have been arguing for a hundred years over what, exactly, intelligence is, but ‘everyone knows’ what it means to be smart, and who is smart and who is not — right?

Of course, it’s not that simple, and the ins and outs of academic research have much to teach us about the nature of intelligence and its importance, even if they still haven’t got it all totally sorted yet. Today I want to talk about one particular aspect: how important intelligence is in academic success.

First of all, to simplify the discussion, let’s start by pretending that intelligence equals “g” and is measured by IQ testing. (“g” stands for “general factor”, and reflects the shared element between multiple cognitive tests. It is a product of a statistical technique known as factor analysis, which measures the inter-correlation between scores on various cognitive tasks. It is no surprise to any of us that cognitive tasks should be correlated — that people who do well on one task are likely to do well on others, while people who do poorly on one are likely to perform poorly on others. No surprise, either, that some cognitive tasks will be more highly correlated than others. But here’s the thing: the g factor, while it explains a lot of the individual differences in performance on an IQ test, accounts for performance on some of the component sub-tests better than others. In other words, g is more important for some cognitive tasks than others. Again, not terribly unexpected. Some tasks are going to require more ‘intelligence’ than others. One way of describing these tasks is to say that they are cognitively more complex. In the context of the IQ test, the sub-tests each have a different “g-loading”.)  

Now there is no doubting that IQ is a good predictor of academic performance, but what does that mean exactly? How good is ‘good’? Well, according to Flynn, IQ tests that are heavily-loaded on g reliably predict about 25% of the variance in academic achievement (note that this is about variance, that is the differences between people; this is not the same as saying that IQ accounts for a quarter of academic performance). But this does vary significantly depending on age and population — for example, in a group of graduate students, the relative importance of other factors will be greater than it is in a cross-section of ten-year-olds. In the study I will discuss later, the figure cited is closer to 17%.

Regardless of whether it’s as much as 25% or as little as 17%, I would have thought that these figures are much smaller than most people would imagine, given the weight that we give to intelligence.

So what are the other factors behind doing well at school (and, later, at work)?

The most obvious one is effort. One way to measure how hard people work is through the personality dimension of Conscientiousness.

One study involving 247 British university students compared the predictive power of the “Big Five” personality traits (Neuroticism, Extraversion, Openness to Experience, Agreeableness, Conscientiousness) on later exam performance, and found that Conscientiousness had a significant effect, and was the only trait to have a significantly positive effect. Illuminatingly, of Conscientiousness’s components (Competence, Order, Dutifulness, Achievement striving, Self-discipline, Deliberation), only Dutifulness, Achievement striving, and (to a lesser extent), Self-discipline, had significant effects.

There were also, smaller and less reliable, negative effects of Neuroticism and Extraversion. The problems here came mainly from Anxiety and Impulsiveness, and Gregariousness and Activity.

Overall, Dutifulness, Achievement striving, and Activity, accounted for 28% of the variance in overall exam grades (over the three years of their undergraduate degrees).

But note that these students were highly selected — undergraduates were (at this point in time) accepted to the University College London at an application: acceptance ratio of 12:1 — so IQ is going to be less important as a source of individual difference.

In another study by some of the same researchers, 80 sixth-formers (equivalent to grade 10) were given both personality and intelligence tests. Conscientiousness and Openness to Experience were found to account for 13% of unique variance in academic performance, and intelligence for 10%. Interestingly, there were subject differences. Intelligence was more important than personality for science subjects (including math), while the reverse was true for English language (literature, language) subjects.

The so-called Big Five personality dimensions are well-established, but recently a new model has introduced a sixth dimension: Honesty-Humility. Unexpectedly (to me at least), a recent study showed this dimension also has some implications for academic performance.

The first experiment in this study involved 226 undergraduate students from a School of Higher Education in the Netherlands. Both Conscientiousness and Honesty-Humility were significantly and positively correlated to grade point average (with Conscientiousness having the greater effect). All the components of Conscientiousness (in this model, Organization, Diligence, Perfectionism, Prudence) were significantly related to GPA. Three of the four components of Honesty-Humility (Greed Avoidance, Modesty, Fairness) were significantly related to GPA (in that order of magnitude).

In the second experiment, a wider data-set was used. 1262 students from the same school were given the Multicultural Personality Test—Big Six, which measures Emotional Stability, Conscientiousness, Extraversion, Agreeableness, Openness, and Integrity (a similar construct to Honesty-Humility, involving the facets Honesty, Sincerity, Greed Avoidance). Again, Conscientiousness and Integrity showed significant and positive correlations to GPA. In this case, Conscientiousness was divided into Need for Rules and Certainty, Orderliness, Perseverance, and Achievement Motivation — all of which were separately significant predictors of GPA. For Integrity, Greed Avoidance produced the largest effect, with Honesty being a smaller effect but still highly significant, while Sincerity was of more marginal significance.

In summary, personality traits such as Diligence, Achievement Motivation, Need for Rules and Certainty, Greed Avoidance, and Modesty, were the traits most strongly associated with academic performance.

Of course, one flaw in personality tests is that they rely on self-reports. A much-discussed longitudinal study of eighth-graders found that self-discipline accounted for more than twice as much variance as IQ in final grades. Moreover, self-discipline also predicted which students would improve their grades over the course of the year, which IQ didn’t.

Again, however, it should be noted that this is a selected group — the students came from a magnet public school in which students were admitted on the basis of their grades and test scores.

This study measured self-discipline not only by self-report, but also by parent report, teacher report, monetary choice questionnaires (in an initial experiment involving 140 students), a behavioral delay-of-gratification task, a questionnaire on study habits, (in a replication involving 164 students).

One personality trait that many have thought should be a factor in academic achievement is Openness to Experience, and indeed, in some experiments it has been so. It may be that Openness to Experience, which includes Fantasy (vivid imagination), Aesthetic Sensitivity, Attentiveness to Inner Feelings, Actions (engagement in novel activities), Ideas, and Values (readiness to reexamine traditional values), is associated with higher intelligence but not necessarily academic success (depending perhaps on subject?).

It may also be that, as with Neuroticism, Extraversion, and Conscientiousness, only some (or even one) of the component traits is relevant to academic performance. The obvious candidate is Ideas, described as the tendency to be intellectually curious and open to new ideas. Supporting this notion, recent research provides evidence that Openness incorporates two related but distinct factors: Intellect (Ideas) and Openness (artistic and contemplative qualities, embodied in Fantasy, Aesthetics, Feelings, and Actions), with Values a distinct marker belonging to neither camp.

A recent meta-analysis, gathering data from studies that have employed the Typical Intellectual Engagement (TIE) scale (as a widely-used proxy for intellectual curiosity), has found that curiosity had as large an effect on academic performance as conscientiousness, and together, conscientiousness and curiosity had as big an effect on performance as intelligence.

Of course, while research has shown (not unexpectedly) that Conscientiousness and Intelligence are quite independent, the correlation between Intelligence and Curiosity is surely significant. In fact, this study found a significant correlation between both TIE and Intelligence, and TIE and Conscientiousness. Nevertheless, the best-fit model indicated that all three factors were direct predictors of academic performance.

More to the point, these three important attributes all together still accounted for only a quarter of the variance in academic performance.

Regardless of the precise numbers (this area of study depends on complex statistical techniques, and I wouldn’t want to rest any case on any specific figure!), it is clear from the wealth of research (which I have barely touched on), that although intelligence is an important attribute in determining success in the classroom and in employment, it is only one among a number of important attributes. And so is Diligence. Perhaps we should spend less time praising intelligence and hard work, and more time encouraging engagement and curiosity, and a disinterest in luxury goods or a high social status.

Read more about the curiosity study at https://medicalxpress.com/news/2011-10-curiosity-doesnt-student.html

Michael Gove is reported as saying that ‘Learning facts by rote should be a central part of the school experience’, a philosophy which apparently underpins his shakeup of school exams. Arguing that "memorisation is a necessary precondition of understanding", he believes that exams that require students to memorize quantities of material ‘promote motivation, solidify knowledge, and guarantee standards’.

Let’s start with one sturdy argument: "Only when facts and concepts are committed securely to the working memory, so that it is no effort to recall them and no effort is required to work things out from first principles, do we really have a secure hold on knowledge.”

This is a great point, and I think all those in the ‘it’s all about learning how to learn’ camp should take due notice. On the other hand, the idea that memorizing quantities of material by rote is motivating is a very shaky argument indeed. Perhaps Gove himself enjoyed doing this at school, but I’d suggest it’s only motivating for those who can do it easily, and find that it puts them ‘above’ many other students.

But let’s not get into critiquing Gove’s stance on education. My purpose here is to discuss two aspects of it. The first is the idea that rote memorization is central to education. The second is more implicit: the idea that knowledge is central to education.

This is the nub of the issue: to what extent should students be acquiring ‘knowledge’ vs expertise in acquiring, managing, and connecting knowledge?

This is the central issue of today’s shifting world. As Ronald Bailey recently discussed in Reason magazine, Half of the Facts You Know Are Probably Wrong.

So, if knowledge itself is constantly shifting, is there any point in acquiring it?

If there were simple answers to this question, we wouldn’t keep on debating the issue, but I think part of the answer lies in the nature of concepts.

Now, concepts / categories are the building blocks of knowledge. But they are themselves surprisingly difficult to pin down. Once upon a time, we had the simple view that there were ‘rules’ that defined them. A dog has four legs; is a mammal; barks; wags its tail … When we tried to work out the rules that defined categories, we realized that, with the exception of a few mathematical concepts, it couldn’t be done.

There are two approaches to understanding categories that have been more successful than this ‘definitional’ approach, and both of them are probably involved in the development of concepts. These approaches are known as the ‘prototypical’ and the ‘exemplar’ models. The key ideas are that concepts are ‘fuzzy’, hovering around a central (‘most typical’) prototype, and are built up from examples.

A child builds up a concept of ‘dog’ from the different dogs she sees. We build up our concept of ‘far-right politician’ from the various politicians presented in the media.

Some concepts are going to be ‘fuzzier’ (broader, more diverse) than others. ‘Dog’, if you think about St Bernards and Chihuahuas and greyhounds and corgis, has an astonishingly diverse membership; ‘banana’ is, for most of us, based on a very limited sample of banana types.

Would you recognize this bright pink fruit as a banana? Or this wild one? What about this dog? Or this?

I’m guessing the bananas surprised you, and without being told they were bananas, you would have guessed they were some tropical fruit you didn’t know. On the other hand, I’m sure you had no trouble at all recognizing those rather strange animals as dogs (adored the puli, I have to say!).

To the extent that you’ve experienced diversity in your category members, the concept you’ve built will be a strong one, capable of allowing you to categorize members quickly and accurately.

In my article on expertise, I list four important differences between experts and novices:

• experts have categories

• experts have richer categories

• experts’ categories are based on deeper principles

• novices’ categories emphasize surface similarities.

How did experts develop these deeper, richer categories? Saying, “10,000 hours of practice”, may be a practical answer, but it doesn’t tell us why number of hours is important.

One vital reason the practice is important is because it grants the opportunity to acquire a greater diversity of examples.

Diverse examples, diverse contexts, this is what is really important.

What does all this have to do with knowledge and education?

Expertise (a word I use to cover the spectrum of expertise, not necessarily denoting an ‘expert’) is rooted in good categories. Good categories are rooted in their exemplars. Exemplars may change — you may realize you’ve misclassified an exemplar; scientists may decree that an exemplar really belongs in a different category (a ‘fact’ is wrong) — but the categories themselves are more durable than their individual members.

I say it again: expertise is rooted in the breadth and usefulness of your categories. Individual exemplars may turn out to be wrong, but a good category can cope with that — bringing exemplars in and out is how a category develops. So it doesn’t matter if some exemplars need to be discarded; what matters is developing the category.

You can’t build a good category without experiencing lots of exemplars.

Although, admittedly, some of them are more important than others.

Indeed, every category may be thought of as having ‘anchors’ — exemplars that, through their typicality or atypicality, define the category in crucial ways. This is not to say that they are necessarily ‘set’ exemplars, required of the category. No, your anchors may well be different from mine. But the important thing is that your categories have such members, and that these members are well-rooted, making them quickly and reliably accessible.

Let’s take language learning as an example (although language learning is to some extent a special case, and I don’t want to take the analogy too far). There are words you need to know, basic words such as prepositions and conjunctions, high-frequency words such as common nouns and verbs. But despite lists of “Top 1000 words” and the like, these are fewer than you might think. Because language is very much a creature of context. If you want to read scientific texts, you’ll want a different set of words than if your interest lies in reading celebrity magazines, to take an extreme comparison.

What you need to learn is the words you need, and that is specific to your interests. Moreover, the best way of learning them is also an individual matter — and by ‘way’, I’m not (for a change) talking about strategies, which is a different issue. I’m talking about the contexts in which you experience the words you are learning.

For example, say you are studying genetics. There are crucial concepts you will need to learn — concepts such as ‘DNA’, ‘chromosomes’, ‘RNA’, epigenetics, etc — but there is no such requirement concerning the precise examples (exemplars) you use to acquire those concepts. More importantly, it is much better to cover a number of different examples that illuminate a concept, rather than focus on a single one (Mendel’s peas, I’m looking at you!).

Genetics is changing all the time, as we learn more and more. But that’s an argument for learning how to replace outdated information (an area of study skills sadly neglected!), not an argument for not learning anything in case it turns out to be wrong.

To understand a subject, you need to grasp its basic concepts. This is the knowledge part. To deal with the mutability of specific knowledge, you need to understand how to discard outdated knowledge. To deal with the amount of knowledge relevant to your studies and interests, you need skills in seeing what information is important and relevant for your studies and interests and in managing the information so that it is accessible when needed.

Accessibility is key. Whether you store the information in your own head or in an external storage device, you need to be able to lay hands on it when you need it. And here’s the nub of the problem: you need to know when you need it.

This problem is the primary reason why internal storage (in your own memory) is favored by many. It’s only too easy to file something away in external storage (physical files; computer documents; whatever) and forget that it’s there.

But what all this means is that what we really need in our memory is an index. We don’t need to remember what a deoxyribose sugar is if we can instantly look it up whenever we come across it.

Or do we?

This is the point, isn’t it? If you want to study a subject, you can’t be having to look up every second word in the text, you need to understand the concepts, the language. So you do need to have those core concepts well understood, and the technical vocabulary mastered.

So is this an argument for rote memorization?

No, because rote memorization is a poor strategy, suitable only for situations where there can be no understanding, no connection.

We learn by repetition, but rote repetition is the worst kind of repetition there is.

To acquire the base knowledge you need to build expertise, you need repetition through diverse examples. This is the art and craft of good instruction: providing the right examples, in the right order.

This post is the fourth and last part in a four-part series on how education delivery is changing, and the set of literacies required in today’s world. Part 1 looked at textbooks; Part 2 at direct instruction/lecturing; Part 3 at computer learning.. This post looks at learning models and types of literacy.

Literacy. What does it mean?

Literacy is about being able to access information locked up in a code; it's also about being able to use that code. To be literate is to be able to read and write.

There's also another aspect of literacy that goes beyond mere decoding. This is about reading with understanding, with critical awareness.

Argument around the dangers of modern technology tends, in the way of arguments, to simplistically characterize the players: Internet = short, shallow; Social media = frivolous, distracting; Games = frivolous; Textbooks, Lectures = serious, deep, instructive.

But of course this is ridiculous even if we restrict ourselves to the learning context. Even social media have their uses. Even games can teach. And even textbooks and lectures can be shallow, or uninstructive, or inaccurate. (Indeed, way back in my first year of university I experienced a calculus lecturer who, I believe, reduced my understanding of calculus!)

The internet is, as we all recognize, a two-sided tool (but every tool is). Many people worry about the misinformation, the shallowness of much of the information, the superficiality of surfing, the way people might get stuck in a little corner that reinforces their vulnerabilities or prejudices, and so on.

We can say the same about infographics (data visualization, visual communication, call it what you will). It’s fostered as a way of helping us deal with the complexity and quantity of information (and I’m a big fan of it), but some people have criticized it for its potential for misinformation. Of course, text (wherever found) is far from pure in this respect!

But we don’t deal with misinformation by banning it (well, some of us don’t); we deal with it by providing the tools and the education so that people can recognize when something is being dangerously misleading or just plain wrong.

So, one of the important aspects of literacy (once you get beyond the decoding level) is being able to evaluate the information.

Why do we talk about digital literacy? Do we really need a new term (or terms)?

It comes down to skills. Because that is what literacy is: it's a skill (with all that that implies). And the new literacies do, undeniably, require new skills.

As far as the decoding aspect is concerned, well, text is still text. And textbooks have always included illustrations, so you could say that that is not new either. But that would be a mistake. The problem with visualizations is that it is not obvious that there's a skill to reading them — they're not as transparent as most believe (hence the misinformation claim). Humans have always used pictures to communicate; it is only recently that these have become sufficiently sophisticated to warrant the term 'language'.

So one of the modern literacies must be visual language, which like verbal language (and math and music), comes in different flavors. We wouldn’t use the same strategies to interpret and analyze a novel as we would a chemistry text, or a poem. We need to develop the same understanding of the taxonomy of visual language.

So I think we should include visual literacy in our new literacy set.

But of course, the new information delivery systems have requirements that go beyond content. Being able to use the code goes beyond reading text and pictures. It involves being able to navigate the delivery system. With a book, you just have to turn the pages. But with hyperspace, learning spaces, video-books, and so on, 'reading' is more complicated.

This is the important thing, the qualitative shift: the shift from linearity. Having a space, be it the whole of the internet or a confined learning space, in which you can go in many directions, in which there is no one path, may be empowering and richly layered, but it is not a place you can throw anyone into without training. Not if they are going to truly benefit from it. Like the need for visual literacy, this is another under-recognized need.

The complexity of these spaces and their navigation has, however, led to a number of useful distinctions being made — between digital literacy and computer literacy, information literacy, and media literacy (among others). Basically, these point to the need to distinguish between an ability to use technology (know the language of software — What’s a window? What’s the difference between a browser and a search engine? Do you hashtag your tweets? Do you use folders?) from the ability to find, filter, and evaluate information, and from the ability to actively participate in the information flow across media (Do you change your verbal style appropriately when you move from a tweet to a YouTube script to a written report to a comment on someone’s blog? Do you use different modes of analysis and evaluation when viewing different media?)

Given that we want students to become adept at all of these, how should we teach them?

In an interview, Will Richardson, a teacher whose experiences with interactive Web tools in the classroom led him to write Blogs, Wikis, Podcasts, and Other Powerful Web Tools for Classrooms talks about the need for teachers to have a visible presence on the Web, to participate in learning networks, and about how this openness is a huge culture shift for the closed shop of teachers. About network literacy as a key skill: “students should be able to create, navigate, and grow their own personal learning networks in safe, effective, and ethical ways. It’s really about the ability to engage with people around the world in these online networks, to take advantage of learning opportunities that are not restricted to a particular place and time, and to be conversant with the techniques and methodologies involved in doing this.” About how kids may be more technologically savvy, but they need help sorting out which information, and which people, to trust.

My favorite bit: he talks about Rethinking Education in the Age of Technology: The Digital Revolution and Schooling in America (Amazon affiliate link), which apparently discusses how historically we used to have an apprenticeship model of education, which moved to the factory model, where it’s all about training everyone the same way, and now we’re moving back to a more individualized, self-directed and flexible lifelong-learning model. Put in those terms, it seems clear that we can’t just keep tweaking; the changes are more fundamental than that.

He also asks why no one is consciously teaching kids how to read and write in linked environments — which relates back to my point about learning to traverse non-linear spaces.

But the onus shouldn't (and can't) be all on the teacher. They need a structure that supports them.

But as with learning networks and digital tools (Facebook, Twitter, blogs, RSS, Scribd, Flickr, Tumblr,  Mashable, ...), the structures too keep changing under their feet. Blackboard, Moodle, Udemy, Instructure (to pick out some old and some new).

It's perhaps easier when the structure is purely online. (In the U.S., the Keeping Pace with K-12 online learning 2010 report tells us that state virtual schools or state-led online learning initiatives now exist in 39 states, and 27 states plus Washington DC have at least one full-time online school operating statewide.) But mostly online learning occurs side-by-side with face-to-face learning. (The report estimates that about 50% of all districts are operating or planning online and blended learning programs.)

A report profiling 40 schools that have blended-learning programs has found six basic models of blending learning:

• Face-to-face-driver: face-to-face teachers deliver most of the curricula. The physical teacher deploys online learning on a case-by-case basis to supplement or remediate, often in the back of the classroom or in a technology lab.
• Rotation: within a given course, students rotate on a fixed schedule between learning online in a one-to-one, self-paced environment and sitting in a classroom with a traditional face-to-face teacher. The face-to-face teacher usually oversees the online work
• Flex: uses an online platform to deliver most of the curricula. Teachers provide on-site support on a flexible and adaptive as-needed basis through in-person tutoring sessions and small group sessions.
• Online-lab: relies on an online platform to deliver the entire course but in a lab environment. Usually these programs provide online teachers. Paraprofessionals supervise, but offer little content expertise.
• Self-blend: encompassing any time students choose to take one or more courses online to supplement their traditional school’s catalog. The online learning is always remote.

As we can see (and as was also discussed in the Keeping Pace report), online learning is not about making the teacher redundant! No surprise when you consider that a major aspect of online learning (and its attraction for many students) is that it personalizes learning.

This is also echoed at university level. A spokesman for the Pearson Foundation, discussing a survey of over 1,200 college students, said: "There seems to be this belief among students that tablets are going to fundamentally change the way they learn and the way they access what they are learning. Students see these devices as a way to personalize learning." Students don't see tablets as means of accessing digital textbooks as much as a means to access e-mail, manage assignments and schedules, and read non-textbook materials such as study aids, reports, and articles. (You might also like to read about one university's experience introducing iPad's into the classroom)).

In the same way, a study involving students in China and Hong Kong found that Facebook was being used to let them connect with faculty and other students, provide comments to peers/share knowledge, share feelings with peers, join Groups established for subjects, share course schedules and project management calendars, and (via educational applications) organize learning activities.

So one aspect of online learning is management and collaboration.

Of course online learning is not only about personalizing learning. It's also about broadening access to quality educational resources. The open course movement is perhaps more advanced at university level (exemplified by MIT's OpenCourseWare (updated: now Open Education Global), Yale's Open Courses, the Open University's Learning Space), but in Iowa, schools will soon have access to wide variety of digital materials from a central repository using Pearson Education's Equella. Certainly the internet is rife with educational materials aimed at K-12, but there are great benefits from the more formal structure of such a repository.

But this wonderful cornucopia is also the biggest problem. So many resources. And so many structures, programs, digital tools. It takes a lot of time and effort to master each one, and who wants to put in that effort unless they’re sure it’s really important and going to last?

There's no good answer to that, I'm afraid. We are living in a time of transition, and this is the price of that.

But we can try and develop our own 'rules of engagement'. Something to filter out the deluge of new tools and new systems and new resources.

When doing so, we need to consider the two principal, and different, issues involved in this revolution in information delivery systems, which should be kept quite distinct when thinking about them (however muddled together they will be in application). One concerns their use in learning — do textbooks need 'bells and whistles' to be more effective means of learning? what is the best way to frame this information so that students (at their grade level) can understand and remember it? This is the how question. For this we need to work out the different strategies that each delivery system needs to be an effective learning tool, and the different contexts in which each one is effective.

The other issue concerns the world for which the education system is supposedly training students. How is information delivered today? How do people work with information? This is the what question; the issue of content — though not in the 'core knowledge' sense.

Although, part of the issue does concern this question of core content. Because the fact is, however we may pine for the days when we all knew the same things, read the same books, could recite the same poems (no, we never really had those days; we just had smaller groups), there is too much information in the world for that to be possible. And society needs the diversity of many people knowing different things, because there's too much for us all to know the same thing. So what we need from our education system — and I know it's a truism but there you go, doesn't make it less true — is for our students to learn how to learn. Which means they need to know the best strategies for learning from the various information delivery systems they're going to be trying to learn from.

And there's something else, that stems from this point that there's too much for us all to know the same thing. We have this emphasis on doing well as an individual — individuals graduate, become famous, get Nobel Prizes, get remembered in the history books. But science and scholarship, and politics and community development, have always benefited from the stimulation of different minds. The complexity of the world today means that we need that more than ever. The complexity of science today means that most discoveries are the results of a team rather than a single person. Even in mathematics, the archetypal home of the solitary genius.

For example, the Polymath Project began with one mathematical genius who decided to take one of the complex mathematical problems he had struggled to solve to his blog. He threw it out there. And readers threw ideas back. Since then, several papers have been published in journals under the collective name DHJ Polymath.

An example of the open science movement (see the Open Knowledge Foundation and the Open Science Summit), raising the question — is the ‘traditional’ way of doing science really the best way? Let’s bear in mind that the ‘traditional’ way is not in fact all that traditional. It’s a product of its times (and rather recent times at that). We shouldn’t confuse the process of scientific thinking with the institutionalization of science. Proponents of Open Science argue that the advent of the internet can break right through the inertia of the institutions, can allow collaboration and the processing of huge data-sets in ways that are far quicker and more efficient.

This is the world we need to educate for. Educate ourselves and our children. And the heart of it is collaboration. Which is one of the reasons we shouldn't be keeping social media out of the classroom. We just have to use it in the right way.

I began this series with Denmark allowing internet access during exams. So let's finish by returning to this issue.

As with the wider question of education, we need to ask ourselves what testing is for. First of all there's the point that, like note-taking, testing has an obvious purpose and a less obvious one. The obvious one is that it provides a measurement of how well a student knows something (we’ll get to the squirrelly ‘knows’ in a minute); the less obvious is that testing helps students learn. (For note-taking, the obvious purpose is that it provides a record; the less obvious is the same as for testing: it helps you learn.) Many tests may be (or perhaps should be) primarily for learning.

Final exams, on the other hand, are usually solely about assessment. But then we must ask, assessment of what? What do we mean by 'know'? There are topics within subjects which are 'core' — crucial details and understandings without which the subject cannot be understood — cell division in biology; atomic structure in chemistry. But there are many other details that you don't need to have in your head — but you do need to 'know' them enough so that you can find them readily, and fit them into their place readily.

Anyone who can write well and develop an argument in depth on a specialist topic in a three-hour exam period from the internet deserves to pass (I'm assuming, of course, that there are adequate guards against plagiarism!). As with course-work, access to the internet simply raises the standard.

These posts have all been rather a grab-bag. This is such a wide topic, with so many issues and everything is such a state of flux. To write coherently on this would require a book. Here I have simply tried to raise some issues, and point to a random diversity of articles and tools that might be of interest. Do add any others (issues, articles, tools) in the comments.

This post is the third part in a four-part series on how education delivery is changing, and the set of literacies required in today’s world. Part 1 looked at the changing world of textbooks; Part 2 looked at direct instruction/lecturing. This post looks at computer learning.

The use of computers in schools and for children at home is another of those issues that has generated a lot of controversy. But like e-readers, they’re not going back in the box. Indeed, there’s apparently been a surge of iPads into preschool and kindergarten classrooms. There are clear dangers with this — and equally clear potential benefits. As always, it all depends how you do it.

But the types of guidance and restrictions needed are different at different ages. Kindergarten is different from elementary is different from middle grade is different from high school, although media reports (and even researchers) rarely emphasize this.

Media reports last year cited two research studies as evidence that home computers have a negative effect on student achievement, particularly for students from low-income households. One involved 5^th to 8^th students in North Carolina ; the other Romanian students aged 7 to 22.

The Romanian study concerned low-income families who won government vouchers for the purchase of a personal computer. The study found that, although there was an increase in computer skills and fluency and even an apparent increase in general cognitive ability, academic performance (in math, English, and Romanian) was negatively affected. Use of the computers was mostly focused on games, at the expense of doing homework and reading for pleasure (and watching TV).

Interestingly, children with parents who imposed rules on computer use were significantly less skilled and fluent on the computer, but no better on homework or academic achievement. On the other hand, those who had parents who imposed rules on homework retained the benefits in terms of computer skills, and the negative impact on academic achievement was significantly reduced.

Additionally, there was some evidence that younger children showed the biggest gains in general cognitive ability.

Similarly, the North Carolina study (pdf) found that students who gained access to a home computer between 5th and 8th grade tended to show a persistent decline in reading and math test scores. But these results are very specific and shouldn’t be generalized. Those who already had computers prior to the 5^th grade scored significantly above average, and showed improvement over time.

An Italian study also found positive benefits of computer ownership - PISA achievement significantly correlated with 15-year-olds' use of computers at home as an educational tool. However, there seemed to be an optimal level, with the effect becoming smaller the more often they used the computer and even becoming negative if they used school computers almost every day.

The North Carolina and Romanian studies indicate that the problem appears to be when computer use knocks out more beneficial activities such as doing homework and reading for pleasure. It's unsurprising that this might be more likely to occur among children and adolescents who gain ready access to a computer after many years of "deprivation".

In Britain the e-Learning Foundation has recently come out claiming that over a million children will perform significantly worse on exams (an average grade lower) because they don’t have internet access at home. This idea is based on research showing that students who use revision materials on the internet to help them revise have an advantage over those students who don’t have access to such materials. Surely no surprise there! And no contradiction to the previous research. There is undoubtedly a lot of very good educational material on the internet, and even if you have a good teacher, getting a different take on things can help you understand more fully. If you have a poor teacher, this is even more true!

So it all comes down to how computers are being used (and what their use is knocking out, for there is only so much time in the day). Bearing on this point, two programs in the U.S. have with some apparent success introduced computers into disadvantaged homes in such a way that they support a more effective home-learning environment and thus improve academic achievement.

There’s also an argument that laptops have shown little benefit in general because the schools in which they’re used have, by and large, good teachers and good students. But the true value of laptops is for those without access to good teachers. For ten years, computers have been placed into brick walls in public places in hundreds of villages and slums in India, Cambodia and Africa, with apparently very successful results.

An extension of the project has involved British grandparents, many of them retired teachers, volunteering their time to talk, using Skype, to children in the slums and villages of India. From this has developed the model of a Self-organized learning environment (Sole), where children work in self-organized groups of four or five, exploring ideas using computers, the exploration triggered (but not constrained) by questions set by teachers.

I must admit, while I applaud this sort of thing, I have to shake my head at the surprise that this sort of activity is effective, and the comment that the students “maintain their own order”. My children had a Montessori education in their early years — in Montessori schools children habitually “self-organize” and teach themselves (with of course the teachers’ guidance, and the use of the resources provided).

But of course, it helps to have the right resources. Five years gathering data from math-tutoring programs has revealed how 10th and 11th grade students use a help button, which offers progressively more in-depth hints and eventually gives the answer to the question. Basically, most students (70-75%) strenuously resist seeking help, even after several errors. When they do eventually give in and ask for a hint, they do so only because they have given up trying to solve the problem and are aiming to cheat — 82% of those using the hint tool didn’t stop to read it, just clicked through all the hints to get to the answer.

Most recently, then, the researchers changed a geometry tutoring program so that the help tool would encourage students to reflect on their problem-solving strategies — for example, by opening a help window if a student seems to be guessing, or doesn’t seem to reading the hints. In pilot studies, the new help tutor significantly improved students’ help-seeking behavior.

But perhaps these children wouldn’t so misunderstand the use of the help button if they’d been taught in a learning environment that encouraged peer-tutoring. As any teacher knows, the best way to learn something is to teach it!

Teachable Agents software allows students to customize a virtual agent and teach it mathematics or science concepts. The agent questions, misunderstands, and otherwise learns realistically. Pilot studies of these programs have included kindergarten through to college.

Additionally, the virtual agent always explains how it came to an answer, and this seems to transfer to the student-teachers, helping them learn how to reason.

But I'd like to note (because it sounds a wonderful program) that you don’t need fancy software to harness the power of peer-tutoring. The Learning Community Project (English translation) operates in nearly 600 rural schools in Mexico and is planned to go into nearly 7000 rural and urban schools. In this model, students choose a learning project and explore it, guided by adult tutors. They then formally present the results of their inquiry to fellow students, tutors, and parents. When they have developed mastery in an area, they tutor other students who are exploring that area. The learning of students and the training of tutors builds a fund of common knowledge that is available in the community of neighboring schools.

But anyway, the message seems clear, if rather obvious: computers and the internet can be a very positive tool for learning, but, as with books and lectures, there are right ways and wrong ways of implementing these delivery systems.

In the next and lash post in this series, I'll discuss what literacy means in today's world, and the new  learning models that are being developed.

[Update: Note that some links have been removed as the linked article is no longer available]

This post is the second part in a four-part series on how education delivery is changing, and the set of literacies required in today’s world. Part 1 looked at the changing world of textbooks. This post looks at the oral equivalent of textbooks: direct instruction or lecturing.

There’s been some recent agitation in education circles about an article by Paul E. Peterson claiming that direct instruction is more effective than the ‘hands-on’ instruction that's so popular nowadays. His claim is based on a recent study that found that increased time on lecture-style teaching versus problem-solving activities improved student test scores results (for math and science, for 8^th grade students). Above-average students appeared to benefit more than below-average, although the difference was not statistically significant.

On the other hand, a college study found that a large first-year physics class taught in a traditional lecture style by an experienced and highly rated professor performed more poorly on several measures than another class taught only by engaging in small-group problem-solving tasks. Attendance improved by 20% in the experimental class, and engagement (measured by observers and "clicker" responses) nearly doubled. Though the experimental class didn’t cover as much material as the traditional class, dramatically more students showed up for the unit test, and they scored significantly better (average score of 74% vs 41%).

It must be noted, however, that this experiment only ran for a week (3 hours instruction).

But the researchers of the middle-grade study did not conclude that lecturing was superior, or that their results applied to college students. Their very reasonable conclusion was that “Newer teaching methods might be beneficial for student achievement if implemented in the proper way, but our findings imply that simply inducing teachers to shift time in class from lecture-style presentations to problem solving without ensuring effective implementation is unlikely to raise overall student achievement in math and science. On the contrary, our results indicate that there might even be an adverse impact on student learning.”

The whole issue reminds me of the phonics debate. I don’t know what it is about education that gets people so polarized, when it seems so obvious that there are no simple answers. What makes an effective strategy is not simply the strategy itself, but how it is carried out, who is using it, and when they are using it.

In this case, the quality and timing of these ‘problem-solving activities’ is perhaps central. The rule of thumb that twice as much time should be allocated to problem-solving activities as to direct instruction is perhaps being applied with too little understanding about the role and usefulness of specific activities.

But it’s obvious that there are going to be strong developmental differences. The ‘best’ means of teaching 18-year-olds is not going to suit 5-year-olds, and vice versa. So we can’t conclude anything about middle school by looking at college studies, or college by looking at middle school studies.

So, bearing in mind that a discussion of college lecturing has little to do with direct instruction in schools, let’s look a little further into college lectures, given that this is the predominant method of instruction at this level.

First of all, we must ask what students are doing during lectures. Given many teachers’ distress at their students’ activity on phones and laptops during class, it’s worth noting the findings of two recent studies that spied on college students in class rather than relying on self-reporting.

The first study involved 45 students who allowed monitoring software to be installed. Distinguishing between “productive” applications (Microsoft Office and course-related websites) from “distractive” ones (e-mail, instant messaging, and non-course-related websites), the researchers found that non-course-related software was active about 42% of the time. However only one type of these distractive applications was significantly correlated with poorer academic performance: instant messaging. This despite the fact that IM windows had the shortest average duration. (It’s also worth noting that instant-messaging use was massively under-estimated by students (by 40% vs, for example, 7% for email use)).

It seems likely that this has to do with switching costs. For those who read my recent blog post on working memory, you might recall that switching focus from one item to another has high costs. Moreover, it seems that the more frequently (and thus briefly) you switch focus, the higher the cost.

The other study used human observers rather than spyware, with obvious drawbacks. But the finding I found interesting was the dramatic jump between first-year and second- and third-year law students: more than half of the latter who came to class with laptops used them for non-class purposes more than half the time, compared to 4% of first-year students. While the teacher took this as a signal to ban laptops in his upper-year courses, perhaps he should have rather taken it as evidence that his students had become more discerning about what was relevant. We need to know how this laptop use mapped against performance before drawing conclusions.

But not all teachers are reflexively against distractive technology. The banning of cellphones from classrooms, and general distress about social media, is starting to be offset by teachers setting up “backchannels” in their classes. These digital channels are said to encourage shy and overwhelmed students to ask questions and make comments during class.

Of course, most teachers are still anti, and a lot of that may be driven by a fear of losing control of the class, or being unable to keep up with the extra stream of information (particularly in the face of the students’ facility in multitasking).

And maybe some teachers are so antagonistic toward distractive technology because they feel it’s insulting. It implies they’re boring.

Well, unfortunately, many students do find a lot of their lectures boring. A 2009 study of student boredom suggested that almost 60% of students find at least half their lectures boring, of which half found most or all of their lectures boring.

But I don’t think the answer to this is to remove their toys. Do you think they’ll listen if they don’t have anything else to do? The study found bored students daydream (75% of students), doodle (66%), chat to friends (50%), send texts (45%), and pass notes to friends (38%).

It’s not that teachers have to entertain them! Granted it’s easier to hold students’ attention if you’re doing explosive chemistry experiments, but students really aren’t so shallow that you have to provide spectacles. They are there (at college level at least) because they want to learn. But you do have to present the information in a way that facilitates learning.

One of the main contributors to student boredom is apparently the (bad) use of PowerPoint.

But even practical sessions, supposedly more engaging than lectures, appear to bore students. Lab work and computer sessions achieved the highest boredom ratings in the study.

Because boredom is not as simple a concept as it might appear. Humans are designed for learning. This is our strength. Other animals may be fast, may be strong, may have sharp claws or teeth, or venom. Humans are smart, and curious, and we know that knowledge is power. Humans like to learn. So what goes wrong during the education process?

Well, one of the problems is that there’s a cognitive “sweet spot”. If you make something too difficult, most people will be put off. If you make something too easy, they won’t bother. The sweet spot of learning is that point where the amount of cognitive effort is not too little and not too great — of course you have to find that point, and a complicating factor is that this varies with individuals.

One area where creators have had a lot of success in finding that sweet spot (because they try very hard) is video games.

How can we harness the power that video games seem to have? A book called "Reality Is Broken: Why Games Make Us Better and How They Can Change the World" points out that creating Wikipedia has so far taken about 100 million hours of work, while people spend twice that many hours playing World of Warcraft in a single week.

Some of the features of good games that researchers believe are important are: instant feedback, small rewards for small progress, occasional unexpected rewards, continual encouragement from the computer and other players, and a final sense of triumph. Most of this is no news to educationalists, but there’s a quote I really love: “One of the most profound transformations we can learn from games is how to turn the sense that someone has ‘failed’ into the sense that they ‘haven’t succeeded yet.” (Tom Chatfield, British journalist and author)

That quote is a guide to how to find that sweet spot.

Providing motivation, of course, as we all know, is crucial. Where’s the relevance? Traditionally, it may have been enough to simply tell students that they needed to know something, and they’d believe you. But it’s not just that students have become cynical and less respectful (!) — the fact is, they have good reason to question whether traditional content and traditional strategies have any relevance to what they need to know.

Here’s a lovely example of the importance of motivation and relevance. In India Bollywood musicals are madly popular. For nine years, these movies have had karaoke-style subtitles. The first state to broadcast the subtitles was Gujarat. Because viewers were so keen to sing along, they paid attention to these captions, often copying them out to learn. As a consequence, literacy has improved. Newspaper reading in one Gujarat village has gone up by more than 50% in the last decade; women, who can now read bus schedules themselves, have become more mobile, and more children are opting to stay in school. Viewers in India have shown reading improvement after watching just eight hours of subtitled programming over six months.

This has apparently worked in more literate nations as well. Finland (and we all know how well it scores in education rankings) attributes much of its educational success to captions. For several decades now, Finland has chosen to subtitle its foreign language television programs (in Finnish) instead of dubbing over them. And Finnish high school students read better than students from European countries that dub their TV programs, and are more proficient at English.

But songs, it seems, are better for this than dialog.

Of course this strategy is only useful at a certain stage — when learners have basic skills, but are having trouble moving beyond.

This is the point, isn’t it? Different situations (a term encompassing the learners, their prior knowledge, and their goals, as well as the content and its context) require different strategies. For example, I recently read a discussion on Education Week prompted by a teacher being forced by his/her institution to use PowerPoint in a class for ESL students to improve their English speaking skills.

Powerpoint slides can be very effective, but far too many aren’t. Similarly, lab sessions can be true learning experiences, or simply “paint-by-numbers” events for which the result is known. Lectures can be a complete waste of time, or true learning experiences.

Consider marathon oral readings of famous texts. A recent article on Inside Higher Ed said that such “events help convey messages, engage students, and foster community on their campuses in ways that reading alone cannot do”. And there was a nice quote from a student: "Until you hear another student read it in his or her own voice, you don't really understand the vast possibilities for interpretation."

What’s the difference between this and a lecture? Well, in one sense none. Both depend on delivery and presentation. I’ve been to some very engaging and inspirational lectures, and some readings can be flat and uninspiring. But the critical difference is that one is literature (a story) and the other is expositional. To make instructional text engaging, you have to work a lot harder. And this is true regardless of the mode of delivery — lectures and textbooks are the oral and written variants of linear exposition.

Is it fair to dismiss a strategy just because some people perform it badly? Is it smart to require a strategy because in some circumstances it is better than another?

We need a better understanding of the situations in which different strategies are effective, and the different variables that govern when they are effective. And we need more flexibility in delivery.

Which brings us to computer learning, which I’ll discuss in the next post.

[Update: Please note that some links have been removed as the articles on other sites are no longer available]

As we all know, we are living in a time of great changes in education and (in its broadest sense) information technology. In order to swim in these new seas, we and our children need to master new forms of literacy. In this and the next three posts, I want to explore some of the concepts, applications, and experiments that bear on this.

Apparently a Danish university is going to allow students access to the internet during exams. As you can imagine, this step arouses a certain amount of excitement from observers on both sides of the argument. But really it comes down, as always, to goals. What are students supposed to be demonstrating? Their knowledge of facts? Their understanding of principles? Their capacity to draw inferences, make connections, apply them to real-world problems?

I’m not second-guessing the answers here. It seems obvious to me that different topics and situations will have different answers. There shouldn’t be a single answer. But it’s a reminder that testing, like learning, needs to be flexible. And education could do with a lot more clear articulation of its goals.

For example, I came across an intriguing new web app called Topicmarks, that enables you to upload a text and receive an automated précis in return. On the one hand, this appalls me. How will students learn how to gather the information they need from a text if they use such tools? How can a summary constructed automatically possibly elicit the specific information you’re interested in? (Updated: this no longer appears to exist, but you can see an example at the end of this post, where I’ve appended the summary produced of a Scientific American article.)

Even if we assume it actually does a good job, it is worrying. And yet … There is too much information in the world for anyone to keep up with — even in their own discipline. There’s a reason for the spate in recent years of articles and books on how the invention of printing brought about a technological revolution — a need for new tools, such as indices, the idea of using the alphabet to order them, meaningful titles and headings, tables of contents. Because the flood of information, as we all know, requires new tools. This one (which will assuredly get better, as translation software apparently has) may have its place. Before we get all excited about the terrible consequences of automated summaries, and internet-access during exams, we should think about the world as it is today, and not the world for which the education system was designed.

The world for which the education system was designed was a simpler one, in terms of information. You gained information from people you knew, or from a book. Literacy was about being able to access the information in books.

But that’s no longer the case. Now we have the internet. We have hyperlinked texts and powerpoint slides, multimedia and social media. Literacy is no longer simply about reading words in a linear, unchanging text. Literacy is about being able to access information from all these new sources (and the ones that will be here tomorrow!).

Even our books are changing.

The simplest ‘modernized’ variant of the traditional textbook is the traditional textbook on a digital device. But e-readers are not well designed for textbook reading, which is quite different from novel reading.

A study involving 39 first-year graduate students in Computer Science & Engineering (7 women and 32 men; aged 21-53) who participated in a pilot study of the Kindle DX, found that, seven months into the study, less than 40% of the students were regularly doing their academic reading on the e-reader. Apart from the obvious – students wanted better support for taking notes, checking references and viewing figures – the really interesting thing was the insight it gave into how students use academic texts.

In particular, students constantly switch between reading techniques, such as skimming illustrations or references before reading the text. They also use physical cues to help them remember where certain information was, or even to remember the information itself (this is something classic and medieval scholars relied on heavily; I have spoken of this in the context of the art of memory). Both of these are problematic with the Kindle.

Consequently, in a survey of 655 college students, 75% said that, if the choice was entirely theirs, they would select a print textbook. (The article also lists some of the digital textbook providers, and some open-access educational resources, if you’re interested).

But e-readers are the future. (Don’t panic! This is not an either/or situation. There will still be a place for physical books — but that place is likely to become more selective.) The survey found a surge in the number of students who have a dedicated e-reader (39% vs 19% just five months earlier). Another, more general survey of over 1,500 end users in the US, the UK, Japan, India, Italy, and China, found that the amount of time spent reading digital texts now nearly equals time spent reading printed materials.

Nearly everyone (94%) who used tablets (such as iPads) either preferred reading digital texts (52%) or found them as readable as print (42%). In contrast, 47% of laptop users found digital text harder to read than print. While 40% of respondents had no experience of e-readers, this varied markedly by country. Surprisingly, the country with the highest use of e-readers was China. Rates in the US and the UK were comparable (57% and 56% had no experience of e-readers).

The age-group unhappiest about reading on screen were 40- to 54-year-olds. Falling into that age-group myself, I speculate that this has something to do with the way our eyesight is beginning to fail! We’re not at the point of needing large font (or at least of accepting that we need it), but we find increasing difficulty in comfortably reading in conditions that are less than optimal.

So, we have a mismatch between e-readers and the way textbooks are read. There’s also the issue of the ‘textbook model’. Many think it’s broken. Because of their cost, because some subjects move so fast (and publishing moves so slowly) that they’re out of date before they come out, even because of their weight. And then there’s the question of whether students actually learn from textbooks, and how relevant they are to student learning today.

This is reflected in various attempts to revolutionize the textbook, from providing interactive animations (see, for example, a new intro biology textbook) to the ‘learning space’ being developed (again in biology — is this just happenstance, or is biology leading the way in this?). Here information is organized into interconnected learning nodes that contain all of the baseline information a textbook would include, plus supplemental material and self-assessments. So there are videos, embedded quizzes, information flow between students and the teacher.

One aspect of this I find particularly interesting: both students and teachers can write new nodes. So for example, in a pilot of this biology program, 19 students wrote 130 new nodes in one semester — clearly demonstrating their engagement in the course, and hopefully their much greater learning.

Another attempt at providing more user-control is that of “flexbooks”. Flexbooks for K-12 classes enables teachers to easily select specific chapters from the content on the website, and put them together into a digital textbook in three formats (pdf, openreader, and html — this format is interactive, with animations and videos). You can also change the content itself.

Multimedia is of course all the rage. But, as I discuss in my book on effective note-taking, it’s not enough to simply provide illustrations or animations — it has to be done in the right way. Not only that, but the reader needs to know how to use them. Navigating a ‘learning space’ or multimedia environment is not the same as reading a book, and it’s not something our book-literacy skills directly transfer to.

And it’s not only a matter of textbooks. Textbooks have their own particular rules, but any expositional text has the potential to be recreated as a multimedia experience.

Here, for example, is a “video-book”: Learning From YouTube , is "large-scale online writing that depends upon video, text, design, and architecture for its meaning making." The author, Alexandra Juhasz, talks about how “common terms of scholarly writing and publishing must be reworked, modified, or scare-quoted to most effectively describe and traverse the "limits of scholarship" of the digital sphere.”

She talks about how scholars should ask which book medium is best suited for their study (rather than simply assuming it must be a traditional book). Reading and writing practices are changing on the internet — rather than deploring or embracing the new habits, we should ask ourselves which practices are most appropriate for the specific material.

She also talks about the need to educate readers in new ways of doing things. We don’t want to simply equate internet use with surfing, with hyperactive jumping and skimming. That has a place, but the internet is also home to material (like her video-book) that requires lengthy and deep study.

And of course there’s an obligation on the net to actually provide the deeper information (at least in the form of links) that in print books we can fob off with references and recommended reading lists.

Note this point: scholars should ask which book medium is best suited for their study. It applies to textbooks too. Books are not being transformed into something different; they are blossoming. There is still room for straight texts. Nor should it — it most certainly should not — be assumed that throwing a bunch of animated videos into the mix is enough to turn a book into an exciting new learning experience. As with books, some are going to be created that are effectively presented, and some are not.

I’ve said we should think of this as a blossoming of the book concept. But are these new, blossoming variants, still books? Where are we going to draw the lines? Video-books and learning spaces are more like courses than books. Indeed, the well-known textbook publisher Pearson has recently partnered with the lecture capture provider Panopto — another sign of the movement from traditional textbooks to cloud-based “educational ecosystems”.

Perhaps it’s premature to try and draw any lines. Let’s consider the oral equivalent of textbooks: lecturing, or as it’s known at K-12 level, direct instruction. My post tomorrow will look at that.

Topicmarks summary (Scientific American article)

In humans, brain size correlates, albeit somewhat weakly, with intelligence, at least when researchers control for a person's sex (male brains are bigger) and age (older brains are smaller). Many modern studies have linked a larger brain, as measured by magnetic resonance imaging, to higher intellect, with total brain volume accounting for about 16 percent of the variance in IQ. But, as Einstein's brain illustrates, the size of some brain areas may matter for intelligence much more than that of others does. Studying the brains of 47 adults, Haier's team found an association between the amount of gray matter (tissue containing the cell bodies of neurons) and higher IQ in 10 discrete regions, including three in the frontal lobe and two in the parietal lobe just behind it. In its survey of 146 children ages five to 18 with a range of IQs, the Cincinnati group discovered a strong connection between IQ and gray matter volume in the cingulate but not in any other brain structure the researchers examined.

In a 2006 study child psychiatrist Philip Shaw of the National Institute of Mental Health and his colleagues scanned the brains of 307 children of varying intelligence multiple times to determine the thickness of their cerebral cortex, the brain's exterior part. Over the years brain scientists have garnered evidence supporting the idea that high intelligence stems from faster information processing in the brain. Underlying such speed, some psychologists argue, is unusually efficient neural circuitry in the brains of gifted individuals. The researchers used electroencephalography (EEG), a technique that detects electrical brain activity at precise time points using an array of electrodes affixed to the scalp, to monitor the brains of 27 individuals while they took two reasoning tests, one of them given before test-related training and the other after it. The results suggest that gifted kids' brains use relatively little energy while idle and in this respect resemble more developmentally advanced human brains.

Some researchers speculate that greater energy efficiency in the brains of gifted individuals could arise from increased gray matter, which might provide more resources for data processing, lessening the strain on the brain. In a 2003 trial psychologist Jeremy Gray, then at Washington University in St. Louis, and his colleagues scanned the brains of 48 individuals using functional MRI, which detects neural activity by tracking the flow of oxygenated blood in brain tissue, while the subjects completed hard tasks that taxed working memory. The researchers saw higher levels of activity in prefrontal and parietal brain regions in the participants who had received high scores on an intelligence test, as compared with low scorers. Lee and his co-workers measured brain activity in 18 gifted adolescents and 18 less intelligent young people while they performed difficult reasoning tasks. These tasks, once again, excited activity in areas of the frontal and parietal lobes, including the anterior cingulate, and this neural commotion was significantly more intense in the gifted individuals' brains.