Tag Archive for dark chocolate

Delusional idiots, choice blindness & jam

I love how stupid the brain can be. The way it’s wired can often result in some pretty stupid behaviour in, say, the human possessing the brain. And I love it. I find it entertaining, in a bizarre and somewhat despairing way, and I find it informative, because it tells us about the elaborate, multi-coloured, lumpy-textured blanket of self-serving sham and deception that our brains crochet for us to protect our precious egos and make us think we’re, like, totally awesome ‘n’ stuff and so much better than, I don’t know, that random guy over there – just look at that idiot. He thinks he’s so good, but he’s not. But I am. Seriously. You probably don’t even know how good I am, because you’re probably an idiot too. Poor, stupid idiot.

My favourite example is the better-than-average effect, a convenient self-serving cognitive bias that means that a lot of people think they’re, hmm, better than average. A lot of people, as in 80% of the population… which is interesting when, by definition, only 50% of people can be better than average. So for things like sense of humour, or driving ability, 80% of the population thinks they’re better than at least 50% of the population. Awesome, well done on the mathematics there.

An extension of the better-than-average effect is the Dunning-Kruger effect, which makes me want to laugh until I vomit in disgust or something, because it is bitterly funny and depressingly amazing. The Dunning-Kruger effect describes the tendency for incompetent, unskilled people (i.e. the worse-than-average people on a given task or skill) to completely overestimate their ability. So now it’s not just people thinking they’re kind of more awesome than the majority – it’s the certified idiots thinking they’re great. The effect is perfectly summed up in the title of the original paper by Dunning and Kruger, “Unskilled and unaware of it: how difficulties in recognizing one’s own incompetence lead to inflated self-assessments”. There was also another chap who put it quite nicely, more than a century before:

“Ignorance more frequently begets confidence than does knowledge.”
-Charles Darwin

(And following on from that, another twist in the story is that competent, skilled people are also unaware of their competence or skill – but they tend to underestimate it. So, basically, we’re mostly kind of useless at judging the difficulty of tasks and at assessing our own competence when it comes to said tasks. There are people who are good at those judgement calls though – people with depression. Sigh.)

So our brains like to protect us from realising how incompetent or inaccurate we are, which isn’t surprising, because if you spent your life not being confident that your judgement in general was at least “adequate”, if not “amazingly blindingly fantastic and infallible”, you’d be crippled by self-doubt and unable to make decisions. Consequently, our brains have become pretty good at telling us “Oh hey, you know that decision you made? Brilliant. Well done, you. No doubt you made the right choice there. Help yourself to some ice-cream or something.”

For instance, you give me two photos to look at briefly:

I have to choose which one I think is more attractive:

Then, in a sleight of hand, you swap the photos without me knowing, and give me the one I didn’t choose:

And you ask me to explain why I chose that one. And I’ll do a fantastic job of explaining, and I won’t have noticed it wasn’t the one I originally chose. This is called choice blindness, and the theory is that I don’t realise there’s been a deception and a swap but I think that I must’ve chosen that particular photo for a reason, so my brain does a fantastic job of retrospectively justifying that “choice” because, hey now, I wouldn’t make a bad choice. Not me. Oh no.

Kind of unsurprisingly, this extends into the realm of taste, flavour and food. In an experiment by Hall and colleagues, participants had to make a choice between two types of jam or two types of flavoured tea. The bit of the experiment that thrills me no end is the way the sleight-of-hand was conducted – jars with secret compartments. Ah, brilliant!

So in one instance of the experiment, there would be two jars and the participant would take a spoonful of jam from an opaque white plastic jar (and maybe it was cinnamon-apple jam) and taste it, and the experimenter would ask the participant to rate how they liked the jam on a scale of 1 to 10. While the experimenter was distracting the participant by asking for the rating, another experimenter surreptitiously turned the jar upside-down. Then the participant tasted a sample from an opaque blue jar, this time containing a different jam (maybe grapefruit). Again the participant was asked to rate how they liked the jam, and again the other experimenter turned the jar upside-down without the participant noticing. And the marvellous thing is that the jars contain two compartments, so when you turn them upside-down and take a sample of jam, you’re getting the other jam – turn the cinnamon-apple jam jar upside down and you get a jam that looks identical, but is actually the grapefruit jam. Turn the grapefruit jam jar upside down and you’ll get the cinnamon-apple jam. Gah! So cunning!

So the participant was then asked to re-sample the jar that they had given the higher rating to out of the pair. Say they rated the cinnamon-apple jam higher. The cinnamon-apple jam jar has now been turned over, and the participant takes a sample of jam (which is now grapefruit jam) from the same jar and tastes it, and now… they’re asked to explain why they preferred that jam to the other one.

And they do explain.

And they generally haven’t noticed that the jam has changed.

In fact, in two-thirds of these jam-tastings, the participants didn’t realise the jam had been changed.

It was the same case when the participants sniffed the flavoured teas. They’re less likely to detect the change when the two samples are somewhat similar (e.g. blackcurrant jam vs blueberry jam) but even when the difference seems like it would be obvious (e.g. cinnamon-apple jam vs grapefruit jam, or aniseed-flavoured tea vs mango-flavoured tea), more than half the time the change was not detected.

In the case of the cinnamon-apple jam vs grapefruit jam, just over 20% of people immediately noticed that the jam had been changed and that they had been given the jam that they hadn’t originally preferred. Another 20% reported, when asked, that some quality of the jam they had preferred had tasted different the second time, so maybe it was a bit stronger or a bit sweeter, but they still didn’t notice the change. And less than 10% said “Hey… umm… was that a different jam?” or something along those lines towards the end of the experiment, well after the change had occurred. For apple pie-flavoured tea vs honey-flavoured tea, the change was only detected immediately about 7% of the time, it was detected retrospectively about 7% of the time, and no other change in the qualities of the tea was reported. So the change went completely undetected 86% of the time. The change between ginger jam and lime jam went completely undetected about 63% of the time. The change between blackcurrant jam and blueberry jam went completely undetected about 80% of the time. Seems kind of unbelievable, doesn’t it?

It’s not that the flavours were difficult to distinguish: another part of the experiment showed that participants had no trouble telling apart the different jam or tea flavours. Participants were often very surprised or voiced disbelief when they were told that the samples had been switched – they genuinely (unless they are very good actors with strange motivations) did not notice the change.

And in addition to this experiment being dastardly and clever (everything you want in an experimental psychology study), the authors are people I want to hang out with, because they made this paper fun to read with its rhetorical questions and random embellishments. I’ll leave it to them to outline the limits of the insight with which this experiment provides us:

Obviously, an experimental finding like choice blindness is bound at the limits by choices we know to be of great importance in everyday life. While it lies close at hand to speculate about couples at the altar solemnly affirming their choice of partner, and then (after the minister pulls some unearthly sleight-of-hand!) bringing home a complete stranger, no one would fail to notice such a change (and this, we hope, includes even those involved in the most hasty of Las Vegas marriages).

So true, Hall et al. (2010), so true.

In my own clunky attempt at sleight-of-hand, I’ve created what I consider a nose-bleedingly amazing triumph: Vegemite in a dessert. These Vegemite caramel pots with dark chocolate ganache are are nose-bleedingly amazing because Vegemite isn’t one of my favourite things (I don’t mind it scraped over water crackers, but that’s about it) yet I have made a dessert that, to be frank, impressed me.

It impressed me because the flavour works amazingly well (Vegemite and caramel – the new flavour pairing that will take the world by storm, just you wait and see) but also… it may be the world’s richest dessert. It’s so rich it could probably buy several hotels in Dubai. The Vegemite jars I served the stuff in that you can see in the photos – that much of this stuff would feed about 4 people. I ate about three tablespoons before I started to feel like my stomach was going to start convulsing in protest. Which means: my work here is done.

Read on for the recipe for Vegemite caramel pots with dark chocolate ganache.


Ok after all the serious discussion and meta-analysis and breaking out the graphs of the last post, it’s time for some fun! (Warning: my definition of fun almost guaranteed not to match your definition of fun.) Time for some mythbusting tangentially related to food and braaaaaaains!

Myth: Sugar makes you hyperactive (especially if you’re a kid).

Perhaps my favourite medical myth ever! It’s more like a pandemic than a myth, because it is just everywhere. So let me say it clearly: sugar does not cause hyperactivity. No ifs or buts. It doesn’t. End of story. Twelve well-conducted trials have found no relationship between sugar and hyperactivity. It doesn’t depend on the amount of sugar. It doesn’t depend on whether the kid has ADHD. It doesn’t matter if the sugar is processed and refined or natural. It doesn’t matter if it’s sweets or chocolate or fruit. There is no link.

“But but but,” you say, clearly ignoring my stipulation of no buts. “But my kid goes crazy after having something sugary. Every time! It’s a clear-cut case of cause-and-effect!” Ahh, my anecdote-spewing friend. You must remember the sheer power of the brain to see the things it wants to see or expects to see.

Scientists conducted a particularly cute little study to look at this. Parents and their kids participated in this one. In one condition, kids were given a placebo pill with no active ingredient and their parents were told this. In another condition, kids were given a pill containing a big dose of sugar and their parents were told this. Each kid would then have a bit of a play and his or her parent was asked to rate their kid’s behaviour. Parents whose child had received the big dose of sugar reported that their child was off the walls with hyperactivity, definitely more energetic than usual, more frenetic, way more active, more crazy. Too bad that in a charming little twist of the kind we have come to expect of psychology experiments like this, all the children had been given placebo. The only thing that varied was whether parents thought their kid had had sugar or not.

And before you say “Oh no, that doesn’t apply to me! My kid really does go hyper on sugar!”, please just stop and don’t say it. Because it does apply to you. Unless you’ve conducted your own strictly controlled trial where your kid was randomly given either a pill full of sugar or an inactive placebo on multiple occasions (in a way that hid the identity of the pill from the kid as well) and a panel of independent raters who had no knowledge of whether your kid had had sugar or placebo evaluated the kid’s behaviour and when the codes were broken, your kid consistently scored higher for hyperactivity symptoms in the sugar condition to a statistically significant extent. Did you do that? Huh? Huh? No, I didn’t think so.

I rest my case.

(But if you did do all that, well done you on your double-blind randomised placebo-controlled study and your weird kid.)

Vreeman, R.C. & Carroll, A.E. (2008) Festive medical myths. British Medical Journal, 337:a2760. (Seriously, I love this paper. Read the whole thing here.)
Hoover, D.W. & Milich, R. (1994). Effects of sugar ingestion expectancies on mother-child interactions. Journal of Abnormal Child Psychology, 22(4), 501-515.

Recipe for chocolate chai madeleines (which contain 3/4 of a cup of sugar) after the cut.


(Epic post ahoy!)

Hello, come in, come in, please, come in and make yourself comfortable. How are you? Good? Good. Are those new mittens? They’re lovely. Can I offer you a drink, maybe a Pimms No. 1 with over-sized cucumber garnish? No? Would you like a delicious biscuit? Ok, here’s a delicious biscuit.

Now that you’re comfortable and have a delicious biscuit…

… we need to talk.

Please assume the brace position, as this is heavy news.

Chocolate is probably not that good for you. The antioxidants and all that good-for-you-heart stuff, well… it’s not all it’s made out to be.

There’s been a gargantuan 18-wheeler truckload of stuff in the media about how eating chocolate is good for you. Why is it good for you? It’s got antioxidants! Antioxidants that, well, they stop your body from ageing… and stuff. They stop your cells from getting attacked by, um… nasty things! They prevent cardiovascular problems. Chocolate is full of antioxidants so chocolate prevents cardiovascular problems! Right? Right?

Well, maybe, maybe not. I’m sorry, but it’s just not black and white like that. (It so rarely is with science. Sigh.) There is plenty of research being conducted into the cardiovascular-protective properties of chocolate consumption, and it’s been going on for a fair few years now. Unfortunately, the evidence is not conclusive, and we’re starting to realise that the research that’s gone on so far isn’t anywhere near detailed enough to tell us about whether we should eat chocolate for its cardio-protective qualities, and if so, how much.

It’s true that a fair few studies have found results like “consumption of dark chocolate is correlated with a reduction in blood pressure”. But! (There’s always a but in science, damn it.) When you look at the studies overall, there are a lot of problems.

The number of people in a lot of the studies is quite small, so that increases the chances that we might be seeing false positives in the results, i.e. we think we see a relationship between chocolate consumption and cardiovascular health but actually there isn’t one. The results we have so far just don’t allow us to reliably conclude that chocolate consumption definitely does (or does not) improve cardiovascular health in any specific way.

Some of the studies are conducted over a pretty short time period. These are acute intervention studies that only look at the effects of dark chocolate consumption over a few days (and for those few days, maybe people in the study are paying better attention to their health and stress levels purely by virtue of being in a study that draws attention to those aspects of their lives). These short studies don’t tell us much about the long-term effects of chocolate consumption on either cardiovascular health or other important aspects of health. Need more info.

The amounts of chocolate that should be consumed to achieve any reported improvement in blood pressure or whatever are all over the place. Some studies say you need to eat 2 squares a day to improve blood pressure, but don’t have more than 2 because that doesn’t improve blood pressure. Some studies say you need to eat a few squares daily, whereas others say you only need 10 square per month. It’s a big confusing mess, really. In fact, we’re not even sure if there are enough antioxidants in chocolate to have any impact on overall antioxidant levels in the body — studies that show antioxidants improve cell health in cell cultures in a dish in a lab have used antioxidant concentrations many many times greater than what could be achieved in the human body by eating chocolate. Basically your stomach would be exploding in chocolate pyrotechnics if you ate enough chocolate to get the same concentrations as the cells in the lab were being dosed with. I’d prefer not to explode, quite frankly.

And something that’s emerging as a major flaw in this research is that we don’t know the exact levels of antioxidants in different chocolates. We know dark chocolate is probably highest in antioxidants, but there’s a huge range of variation in the amounts depending not just on percentage cacao, but also how the cacao was treated and processed and where it originally came from. For example, some processing methods can radically reduce the key antioxidant compounds such as catechin and epicatechin. So essentially, when a study is done to see whether chocolate consumption results in improved health measures, we don’t actually know what sort of dose of antioxidants these people in the study are getting!

The upshot of this is that we can’t possibly recommend a particular amount of chocolate to consume in order to get health benefits — we don’t know what amount of antioxidants is in any given chocolate, and we don’t know what the ideal dose of antioxidants is anyway.

Alright, so antioxidants play an undeniably important role in the health of the human body. But here’s the thing: just because something is necessary for health, doesn’t mean that big amounts of it are good for you, and that even bigger amount are better. In fact, it’s usually quite the opposite.

If there’s one thing you learn this year about science and health, make it this: a lot of things are on what’s called an inverted-U curve. It’s not a linear trend of more = better. It’s more like not enough is bad, enough is good, too much is bad again. However, we have to do an enormous amount of research to establish what “enough” actually is. At the moment, we’re trying to work out what “enough” is for antioxidants, but it’s safe to assume that too much is probably a bad thing.

In fact, an enormous review was done of the antioxidant research. Bjelakovic and colleagues did a systematic review and meta-analysis of studies that looked at what happened when people were given antioxidants supplements. This means that they looked at a whole heap of studies, kicked out the ones that weren’t rigorous enough, then tried to figure out the bigger picture that the well-conducted, rigorous studies were painting. They looked at 68 of these good-quality studies, which altogether tested the effects of antioxidant supplementation (in the form of beta carotene, vitamin A, vitamin C, vitamin E or selenium) in 232,606 people. With that many people, surely you’ve got the ability to detect even tiny little relationships between the antioxidant supplementation and health outcomes.

And the results of this enormously powerful meta-analysis? Some antioxidant supplementation might cause bad health outcomes, whereas others still have an unclear effect. Still! After all those studies and all those people! There was no clear benefit to taking extra antioxidants. This could be because: (a) there’s no relationship between extra antioxidants and the health variables being measured, (b) the relationship is so small and subtle it’s difficult to detect or (c) the effects are dependent on other variables that we aren’t taking account. My money’s on (c).

So it’s kind of scary when you think about how frequently the media and various companies (with a financial interest in getting us to buy their products) tell us we should be indiscriminately cramming antioxidants every which-way into our bodies. (Goji berries! You need goji berries! Or was it acai berries that were the latest superfood du jour? Oh god just get some pomegranate before it’s too late! Don’t you want to live forever?) It’s entirely unsurprising that most of us have absorbed the inaccurate message that antioxidants are always good for us and we need as big a dose as we can get.

But even though we don’t know what the right amount of antioxidants is for optimal health benefits (yet), and even though we don’t know what precise levels of antioxidants are in various products, and even though we’re surrounded by products with labels zealously screaming “FULL OF ANTIOXIDANTS!”, and even though we don’t know if we’re getting too little or too much, don’t despair! What should you do? It’s old and tired and not that interesting, but it’s the truth: eat a balanced diet and eat things in moderation. If you do that (along with some exercise) you’ve got a very good chance of getting pretty much everything you need to achieve very good health. If you’re eating tomatoes and carrots and that sort of thing and maybe a bit of dark chocolate here and there, you’re probably getting the antioxidants you need. Don’t fret!

So maybe, eventually, research will be able to pin-point exactly what daily dose of antioxidants we should aim to get, and whether we can eat chocolate to get some of that dose. Until then, don’t use “it’s good for me!” as an excuse to eat chocolate. If you’re leading a generally healthy lifestyle, you can eat it chocolate just because you enjoy it — you don’t need much more justification than that, right?

Thank you for your time. And now: a recipe.

Recipe for not-actually-superfood cookie sandwiches under the cut.

Rosewater brownies with roasted pistachios

These ridiculously dense brownies are a variation of the brownie recipe from the Max Brenner Chocolate: A Love Story cookbook (the brownie recipe is properly titled, in true Max style, “A philosophical highly concentrated fudge brownie made of 70 percent dark chocolate thoughts”). Even I was disturbed by the fact that they had almost a kilogram of chocolate in them, so much so that I didn’t believe it at the time and only bought 500g of chocolate because no recipe could possibly need more than that. I ended up rummaging around in the cupboard to find enough chocolate to make up the full 900g. Amazing.

The idea with the rosewater is to just add enough that the flavour isn’t obvious (or even necessarily identifiable) but it’s still there and changes the overall perception of the brownie flavour subtly. And I added pistachios because they pair well with rosewater as kind of Middle Eastern ingredients. Plus I just like pistachios. Although these brownies are retrospectively dedicated to Tash because she likes both rosewater and pistachios (impeccable taste). Recipe after the cut!

Chocolate brandy mousse with vanilla pear purée and pear tart milk pudding

One day I’ll go back to coming up with recipes that don’t have a short essay as a title. Today is not that day.

Again I’ve gone with the mousse combined with a pudding made from infused milk, but I’ve added fruit in as well this time. The mousse is made with dark chocolate and brandy; I would have used Poire William pear brandy, but guess what? I don’t have any. The pear purée is something I’d eat by the bowlful on its own – a tiny bit of brandy, a tiny bit of vanilla paste, a quick stab with the stick blender and it’s kind of amazing. Totally gestalt. And the pudding is a milk pudding made from milk that I infused with a pear tart, using the same method I tried out here. It worked reasonably well, with the flavour of the pastry coming through subtly in the final pudding. Recipes after the cut.