Secrets to a happy marriage

These articles are everywhere today exclaiming loudly that "scientists" have found the secret to a happy marriage. The articles tend to follow the same idiotic script, with a beautifully seamless dismissal of women and any importance they may have to the relationship process, choose a younger smarter women and you will stay married longer. It immediately set off my bullshit detector so I started looking into it.

Firstly the article is published in the European Journal of Operational research, if this doesn't sound like a psychology journal or a journal studying social interactions it's because it isn't, it's a journal about the application of mathematics to engineering and computer programming. More typical of the articles it produces are:

A hybrid immune multiobjective optimization algorithm
and
An adaptive memory methodology for the vehicle routing problem with simultaneous pick-ups and deliveries

Both of these sound more interesting to me than "the secret to a happy marriage" but I'm weird that way. The article is titled: Optimizing the marriage market: An application of the linear assignment mode, which is far better than the BBC's title for the article:

'Younger wife' for marital bliss

Anyway I think we can all agree that most journalists are idiots when it comes to any application of science so lets ignore the content of most articles and focus on the study itself.

Firstly they assume that homogamy is the best predictor of marriage success. What's homogamy? homo means same, homogamous couples are couples that are very similar (in social origin, education, race, ethnicity, religion, and age, at least they are according to the makers of this study).

They start out with 1534 cohabiting couples from Switzerland then 5 years later did a follow up on 1074 of those couples. This immediately raises some red flags: firstly losing 1/3 of your couples does not bode well for the statistical power of your study, there could be a reason applicable to the questions you are asking which could have influences the drop out rate. Secondly no differentiation is made between married and non married couples, nor for how long the relationship has been going on, in particular young people tend to have less homogamous relationships but also due to their age these relationships tend to have been forged more recently, you might expect relationships which have already stood the tests of say 20 years of togetherness to be more likely to withstand another 5. Also Switzerland has it's own culture, history and social issues and these may not cross over into other societies. Lastly and most importantly little is said about how the couples were selected, selection bias can destroy your data, there is little here to ward off the problems this could cause. Still it may be able to give us a general idea of some things provided we do not take the results too seriously.

Of the couples they were able to get a hold of they compared those who had stayed together for the five years to those who had broken up and tested for homogamy in both groups in different areas. They then took the factors they believed would influence the viability of a relationship and ranked them according to how likely they were to be present in a broken up couple but not in a still together couple. The highest scoring factors were: mixed cultures, woman older than man, husband previously divorced, non western couples. They then reassigned partners (theoretically) according to minimize these factors they moved 98% of couples and found that in 68% of couples they could reduce the factors that occurred most often in break ups. The researches then go into great detail about how the different combinations and the most efficient way to match up potential partners, it is the most interesting part of the study and in fairness to the authors I believe this was the point of the study and the above analysis seems to be just their way of getting some parameters for their algorithms.

Again I just want to stress that the large drop out rate and the lack of proper randomisation (and controls) mean that the data here cannot be taken to have a high degree of certainty. However I do have some other critiques:

One thing most articles discussed was that if the woman had a higher level of education than the man the marriage was more stable, they failed to mention that the least stable marriages were between men and women who both had a high level of education and that all of the educational links to break ups were extremely tenuous. Also and I cannot stress this enough, correlation does not imply causation, higher levels of education generally imply lower levels of religiousness and lower levels of stigma associated with divorce. It may be that those with lower levels of education are not happy in their marriage but feel pressured against divorce, or there could be a whole other mess of confusing factors that influence the final outcome. Also again the study did not make any differentiation between cohabiting and married couples.

Also I have a problem with the way the culture groups were divided, the three groups, Western, non-Western and Swiss, that is a woefully inadequate way to divide cultures, it puts for example an orthodox Jew and a fundamentalist Muslim in the same category, I think it is safe to completely dismiss therefore this line of inquiry, I find it of little surprise that the only properly identified cultural group (the Swiss) had the best link to a stable relationship.

This leaves us with the last statement we can extract from the report, that relationships where the woman is younger are more stable. This seems to be a solid proposition and is supported by other, much better research. There are many reasons why this should be so and it is important not to confuse correlation with causation; for example, this is a more traditional view of marriage and so will put less strain on a relationship, it is favoured by religious groups who put pressure on couples not to divorce, and I'm sure there are other reasons which I am unable to think of on the spot.

So what does this study and science in general say on the subject of marriage? Sadly both this study and science in general tell us that some people just aren't suited to long term relationships, for those of us that are suited to long term relationships homogamy is a useful concept, far from the be all and end all of relationships but certainly it is useful to have something in common with your partner. Unfortunately there just doesn't seem to be a magic formula for the perfect relationship, Richard Wiseman in his book 59 seconds goes through the studies that have been done and looks at what scientifically we can say about relationships and attraction and I would recommend his book for anyone who wants to know more about this.

OCA - The did a study!

Sorry, seriously I'll stop soon. The Organic Consumers association published a study they conducted, not in a peer reviewed journal, trust me, none would take it; check it out, seriously.

It's two pages long, has no references and contains 1 two page long table. Introduction, method, discussion, results and conclusions are crammed into half a paragraph at the top of the page.

Here is another one:
http://www.organicconsumers.org/bodycare/DioxaneResults08.cfm

I love it; they are measuring an "aliquot" of substance (wow a whole aliquot!), at an unnamed "reputable third-party laboratory".

Now I'm not saying they fudged the results but when you don't include anywhere near enough information to make the tests reproducible it makes it very hard to tell whether the results were fudged. Also I've seen higher standards of reporting in a primary school classroom.

Oganic Consumers Association, more from the worst website on the internet... sorry

I couldn't resist I was bored, (I'm sick) so I flicked over to their site and saw this:

Homeopathy Successfully Treated Flu Epidemic of 1918

Bam, best article ever!

For those of you unfamiliar with homoeopathy it works like this:

You get 10 people and you give them a weird chemical hoping that they get sick, then you look at their symptoms and you decide that the chemical they took will cure any diseases that cause those symptoms if it is diluted (why? because some guy said so 200 years ago). So you get that chemical and then dilute it to the point where there is no chemical left and you have only water (take one drop put it in a small container of water, then take one drop of that put it in another container of water, repeat 30 times), then you dissolve some sugar in the water and evaporate off the water so all you have left is a sugar tablet. Sorry I missed out the most important bit those containers of water need to be hit against a horsehair and leather board as you are diluting (literally otherwise it's not homoeopathy and it loses it's magic powers).

Now if you go back far enough in time, to when doctors sucked at doing anything and tended to kill more people then they helped homoeopathy was a better treatment, after all sugar is only dangerous if you are dying of tooth decay. Anyone who actually believes homoeopathy does anything at all is an idiot, and not just a regular idiot but the kind of super-strength idiot they normally save for reality TV shows.

Organic consumers association: Are people really this stupid?

Never attribute to malice what is adequately explained by incompetence.

Words I try to keep in mind when I'm reading something sciency (philosophical thought for the day; if it's not a real word can it be spelt incorrectly?). I was pointed in the direction of this site http://www.organicconsumers.org/ by an organic food defender after a brief discussion about the rather robust report produced by the Food Standards Agency (an independent body set up by the UK government to look at food standards) which rubbished claims that organic food contains extra nutrients. This site is possibly the worst website on the internet in terms of information content, every story I have checked so far has been wrong, clearly obviously wrong. They never link directly to the source of their information but to news services mentioning reports, sorry that statement seems to give the impression that they actually read the reports and not just the newspaper coverage. Anyway here is a quick summary of some of the stories on their site:

Rural Well Water Linked to Parkinson's; California Study Implicates Farm Pesticides

The title should read: Rural well water not linked to parkinsons, weird result obtained; one common farm pesticide may prevent parkinsons.
Here is the study they do not reference (but the article they reference does):
http://www.ehponline.org/members/2009/0900852/0900852.pdf
The methods seem to be quite solid (except for the fact that they don't actually measure the amount of pesticides in the wells but that would not be feasible in this kind of study) but the conclusion completely ignores the results. For all pesticides tested besides diazinon the 95% confidence level included the null. In layman's English; when you are looking at statistics the answer is never one value but a spread of possible values, the general rule is to take a spread of values from the middle large enough that there is a 95% chance of containing the real answer. As long as the spread of values does not include the null (or zero value) then there is a 95% chance that you have found a link. It needs to be this high because the system is rather complex, there are a lot of other factors that could influence the result despite the researchers best efforts to keep them out. Even with a 95% confidence level 1 in 20 tests will return a result outside the range.

Anyway enough stats back to the goof balls. Diazinon had a 95% link to parkinsons, that's some proof isn't it? Diazinon had a link to parkinsons all right and at 95% confidence, the only problem here is that higher doses of diazinon REDUCED the likelyhood of parkinsons. This is likely one of those wacky results you get when you measure a whole lot of crap and I would not advise those at risk of parkinsons to take diazinon at any dosage level unless a lot more testing is done.

Study Finds Pesticide-Free Diet May be Beneficial for Children

No it didn't it didn't even test that. If you click on the "straight to the source" link you are taken to that hallowed hall of scientific data MSNBC. MSNBC do not link to the study because they didn't want to bias there report by actually reading it, when the report was written the study was not published, MSNBC were giving readers information straight off the press release. The study is out now, here it is:
http://www.ehponline.org/members/2003/5754/5754.html

What it does test is whether pesticides from non-organic food were actually getting into the bodies of the people eating them, it found that they weren't with one exception dimethyl phosphate. Dimethyl phosphate was found at low levels and is used commonly as a pesticide because it is extremely harmful to insects and almost harmless to humans (compared to other pesticides, everything carries some harm, even water will kill you if you drink too much).

I could go on and I will because I think it's funny.

Modified Corn Seeds Sow Doubts

There's been a study on genetically modified corn? Actually there's been like 20 (this is a guess they mention multiple studies but don't say how many), all pointing towards GM corn being safe. In fact pretty much all good studies on GM food have shown that it is probably safe (by good I mean following proper protocols to remove error). What are they whining about? This particular combination of genes has not been tested before, all the individual genes have but somehow combining them may produce a harmful effect. I'm never one to rubbish calls for more research, testing this is a great idea but it has to be balanced with the fact that research budgets are limited and there is no scientific basis for believing that stacking these genes together could have any negative health results. Maybe it's a quantum effect?

Cat-Food Irradiation Banned After Cats Die

This one actually carries a small kudos to the Organic Consumers Association, they did better with this headline than their "source" the Sydney Morning Herald who had:

Cat-food irradiation banned as pet theory proved

Let's start with what actually happened, a bunch of cats died, a Sydney vet thinks this may be linked to the high irradiation pet food has to undergo in case it contains some life-forms that may be harmful to the Australian environment. Human food receives much lower irradiation. The government says hey this could be the problem it's not much effort to change it and won't increase the risk greatly, makes the change.

SCIENCE: irradiation has to be direct to cause any damage, eating something that has been irradiated in the past is not an issue in any way shape or form. The issue is that when you irradiate something not all the radiation goes into causing damage, some of it just excites particles into a higher energy state. At some random point in the future these particles will return to their base state and in doing so emit radiation. The radiation coming from irradiated objects drops off very fast, the higher energy state would have to be very stable for the some radiation to be present when the cats were eating the food. Irradiation does lower nutrition in food, this could plausibly cause health problems.

Most likely though this was a random statistical fluctuation and a proper study would need to be carried out to figure out what was going on. The fact that this has only affected one brand, in one species probably means it's garbage. Kudos to Champion Petfoods the maker of the food under attack for agreeing to pay compensation to pet owners regardless.

This post is too long already so just quickly here are some more garbage headlines:

African Chickens Refuse to Eat Genetically Modified Corn
my dog won't eat dry dog food, who cares? Also it's bullshit.

Free Trade Makes You Fat
No, just no.

Swine Flu Vaccine Will Contain Mercury Linked to Autism and Neurological Disorders
Mercury isn't linked to shit

A Deadly Ingredient in a Chicken Dinner
Your chicken contains arsenic! We're all gonna die. The source isn't even a news story it's an editorial opinion piece with no reference to any data what so ever.

Nobody Knows What Nanoparticles Do - Yet They Are in Your Food, Cosmetics, and Toys
Full of scary imagery, Carole Bass, should be a horror writer. What are nanoparticles, it literally means anything small. OH NO, THERE ARE SMALL THINGS! RUN!

Scientists Warn of Hazards of GMOs
This time they actually linked to something they called a study, it's not a study, it's a call for a study, there is a slight difference. I guess the headline, scientists want to study things didn't cut it.

That's enough for today, I may come back to this if I need a laugh.

Climate Change: the basics on the evidence and the problems

Lets start off with a saying my brother is fond of, emotions high intelligence low, as soon as you mention climate change a lot of people seem to lose their ability to think rationally. Please try to leave your emotions at the door at look at the evidence rationally which I have attempted to summarize. A lot of people out there are quite vocal about climate change without having had a careful look at the evidence so here it is:

Evidence for climate change:

The temperature is rising, the polar ice caps are melting and it is impossible to deny the evidence for climate change. The main question now is how much of it we are causing and what can we do about it. I’m going to take up the logical position here and say that though there almost certainly is anthropogenic (man made) climate change the evidence isn’t all that compelling.

There is a strong correlation between warming and human carbon production (they happen at the same time) but it’s a very complicated system there are other factors that could influence climate change, some of those factors also correlate strongly with the temperature increase. Then there are big problems with the data the oceans in particular tend to throw everything completely out of whack. Water has a high specific heat and that means small changes in water temperature require large amounts of heat energy. If the oceans cool slightly and release that energy onto the surrounding environment it will heat up significantly. That’s why we don’t see a clear trend in energy, day to day, year to year; the energy stored or released by the water around us controls the temperature.

There is a recognized method for carbon to cause temperature increase, the greenhouse effect. Most of us are familiar with it so I’ll just throw in the bits most people miss:

Radiation penetration and wavelength:

Below is a lovely image I have stolen from I forget where (if I could remember I would ask for permission or at least credit them). It has the electromagnetic spectrum and the amount that gets through the big cloud of gas surrounding the earth we call the atmosphere. Now we are going to talk a bit of physics (I’ll make it as painless as possible) the wavelength of light most strongly emitted by a body depends on its temperature. The sun has a surface temperature (that’s the one that counts) of around 6000 degrees and so emits mostly in the yellow part of the visible spectrum. This will come as no great surprise to anyone with vision. The earth is a little bit colder (about 14 degrees Celsius average surface temperature) and emits radiation in the infra-red at about 10 micrometers (the symbol for micro is that funny u like latin character μ pronounced mew, one micrometer is 1/1000 of a millimeter) I’ve gone ahead and labeled these on the diagram.

You would be forgiven for thinking wow they’re both in little windows which allow the radiation to pass through, isn’t that lucky? Luck however has nothing to do with it, the earth should be radiating further to the right on this little graph because it should be a lot colder however thanks to the greenhouse effect the Earth has warmed up enough to radiate happily and harbor life. Anyone want to have a guess what carbon dioxide and methane do to light at 10 microns? You win if you guessed that the light is not going to make it out of the atmosphere, and unsurprisingly that’s bad.

Carbon Production

We’re putting up all that carbon right? Wrong! Most of it comes from volcanoes, decaying plant matter and forest fires. Looking at the actual numbers there is a bit of disagreement but it’s fair to say that approximately 90% of the carbon output is natural and about 10% anthropogenic (again the numbers are approximate). This is what I mean when I say that the evidence isn’t very compelling.

The major factor this leaves out though is what absorbs the carbon back out of the atmosphere. Forests are great for sucking carbon out of the atmosphere. Deforestation is removing some of our best carbon sinks. Pollution is not helping the other carbon reducers, bacteria and algae, mostly in the oceans. Interestingly deforestation was taken out of the Kyoto protocol, apparently what is possibly the major human contribution to climate change wasn’t important enough or something. Honestly who keeps letting politicians make decisions?

Carbon Reducers

Before man started interfering there was a kind of equilibrium in the carbon cycle, carbon went out through the digestive and pulmonary (breathing) systems of animals and through decay and was absorbed by autotrophs (plants, algae, and some bacteria). Then we (humans) cut down a lot of forests, and destroyed a lot of natural vegetation, this stuff literally eats carbon; it uses energy from sunlight to break up carbon dioxide into carbon which is the basic building block of life and releases the left over oxygen. Also we have increased the amount of carbon going out (even if it's not a lot). This explains at least some of amount of carbon we measure in the atmosphere and why it is increasing; we are responsible for at least some of that increase, the exact amount can not be determined with a great degree of accuracy.

The Future

The issue with climate change is not what is happening now, we can deal with current carbon output and absorption, the issue is what we predict will happen in the future. China, India, the rest of Asia and Africa have not been big carbon producers in the past. They are going to have to start churning out massive amounts of carbon if they ever want to leave poverty behind and give their people the basic amenities we take for granted, things like electricity, transport and food (food is kind of important to these people). To give a “back of the envelope” figure the 10% carbon output we currently have only really services around 1/10 of the world’s population. If we increase the carbon output to the point where we can give everyone basic access to power, transport and food there will be a problem, not to mention the fact that the developed countries aren’t reducing the amount of power, transport and food they consume, we’re increasing it and increasing it exponentially, and we’ve reduced the Earths ability to soak up carbon to the point where we are already noticing a problem.

What do we do?

The main cause carbon output is the energy industry, the main cause of deforestation is the food industry these are the two things we need to work on.

Energy

Renewable energy is the future, particularly solar power, however there is a major problem with all renewable sources; they aren’t reliable. Some days are cloudy, some days aren’t windy, some times it doesn’t rain, and sometimes there are no waves. The obvious solution is to store the excess energy but here we run into an issue. At the moment our best shot at energy storage is chemical energy, like we have in batteries. Simplifying the science there are two chemicals, they release energy when they are combined and it takes energy to break them back up. For large cities we are going to need a lot of chemicals, and the more efficient they are at storing energy the more dangerous they are (they release more energy and are less stable and more likely to react with the surrounding environment). There is a lot of very interesting and very promising research going on in this field but this problem may be very difficult to solve.

The way forward is however clear, we can have a large proportion of our power created by renewable energy as long as we have something to back it up and quickly get extra power into the system when our renewable sources fall short. Nuclear power is safe, cheap and environmentally friendly and is the obvious choice. The alternative, sticking with fossil fuel and hoping the research bears fruit quickly is not as bad as it seems. Though we may not solve this problem for quite a while nuclear will require a lot of resources to implement and if we are lucky we will not need it. I favour nuclear power and in the future I intend to blog about why.

Food

Meat is the major culprit here, not only is methane a particularly active greenhouse gas but meat uses up a lot more land in order to feed people than the equivalent calories in vegetables. When you eat a plant you only take in about 10% of the energy it took to create that plant and the number is the same (roughly) for energy conversion when you eat meat. However the meat we eat not only needs room to move around when it is growing (unlike plants) but it needs to eat a significant amount of plants in order to grow thus taking up even more land in order to grow feed. I’m not saying we should give up meat, mainly because it is delicious but a reduction in the amount of meat we eat in the developed world would be a big help to climate change, and we haven’t even taken into account the carbon dioxide that animals breathe out as opposed to the fact that plants will suck carbon dioxide out of the air in order to grow (it’s roughly equivalent to 90% of the carbon the plants that were the animals food sucked out of the atmosphere minus the plant parts that aren’t digestible, the other 10% goes into the animal).

That’s pretty much it for climate change, if you see a lot of fancy graphs which “prove” that climate change does or does not occur or is or is not caused by humans then they are either graphs of the vague effects I have discussed or you’re either viewing speculative data or being bullshitted. As much as I like Al Gore the system is just too complicated to draw definitive data and results from. Much of the speculation is rather interesting and I encourage you to have a look but just remember the basic principles, if something seems too good to be true it probably is.

Quantum Physics it's really not that complicated

Richard Feynman once said science is like chess, you can know all the basic moves but that doesn’t mean you can play like Alekhine. Like chess you can spend a lifetime looking into quantum physics and never know all there is to know here I’m going to give you a run down of the basic moves, enough that you should understand the basics without any difficulty.

There is no such thing as particles

Most of us, when we went through high school learnt that there were such things as particles (matter) and waves (like waves on the beach, primarily light). We learnt that they were two separate things with separate behaviors, waves were black particles were white. What we actually see is that all behaviors are random and the probability is determined by a wave function (a squiggly line) and that rather than a black and white distinction between waves and particles we see instead only shades of grey. Here is why:

Here are some sketches of what we like to call wave functions, all that means is that where the graph is really high or really low then that’s probably where the thing is. (I’ll explain this in more detail later)

Waves are wavy:

Particles or what we think of as matter (i.e. atoms, molecules, stuff in general) are simply waves that are spiky:

Now we need to use a little bit of physics that Einstein used to win the nobel prize, namely the momentum of a wave is related to it’s wavelength. (You’re going to have to trust me on this it’s easy enough to explain but it would take as long as this whole quantum physics explanation).

So now looking at matter we can tell about where it is:

but we have no idea what it’s wavelength is.

Looking at waves we can work out it’s wavelength:

But we have no idea where it is.

If you make the wave more spiky then you have a better idea of where it is and vice versa. This is called the uncertainty principle. The better you know the location the worse you know the momentum and vice versa.

We can take anything, say a light wave, and make it more spiky, but where will the main spike be?

It’s a bit harder to see with the spike that the wavelength of the wave could be almost anything, we aren’t interested in the best move to make in the chess match though just what moves are allowed. Just to show you why this can be almost anything I have a graph here that is just a bunch of waves added together:

I hope you can see now that when we take a spike and try to make it wavy that we can’t be sure exactly how the wave is going to look and most importantly what the wavelength will be.

That’s pretty much it for quantum mechanics, now some quick answers to some obvious questions:

1. If matter is going in different directions all the time why doesn’t stuff just explode?

It’s like having 1 thousand people jammed in a department store so tight they can barely move, when the stuff inside a solid moves it interacts with all the other stuff. Also since the movement is completely random it averages out over the whole crowd. I suppose it’s more like a mosh pit then a department store.

2. If everything is so random how do we get laws of nature like gravity?

The random movement averages out over the billions of atoms and molecules the force gravity exerts on each atom is random but since there are so many atoms in stuff it averages out to a pretty certain number, it moves the whole mosh pit. All the other laws of nature are the same, they are just the most likely thing to happen in a certain circumstance, since there are so many atoms/light rays in the stuff we observe it almost always averages out to the most likely outcome.

Any questions or corrections leave a comment, I’m going to try and update this blog semi regularly. I’ve got quick tutorials about nuclear power and alternative medicine coming up next

Update - Something Positive for a Change

I'm a dummass (considering dummass isn't a real word can it be spelt incorrectly? – philosophical thought for the day, why not find a dummass and ask them). Gravity does cause energy loss (probably) it is just extremely small though it has been measured in the periods of binary stars. Regardless of this we still know next to nothing about gravity. More updates will follow shortly (about cows).