We've moved

Sorry people... we moved again. We're now situated at wordpress; we have more options there, statistics, the whole shebang.
Update your bookmarks, and sorry for the inconvenience.

Science and Law - Part 0

In the Journal of Law, Medicine & Ethics, Volume 34, Number 2, there’s an article titled “Just evidence: The Limits of Science in the Legal Process”, written by Sheila Jasanoff. It talks about the reliance on science in the legal process. What follows is a summary, with some short notes added.

It starts with an introduction about the Massachusetts governor, who tried to reintroduce the death penalty, mainly because science would produce failsafe/infallible results. Science produces a lot of facts, and with the help of the self-corrective nature of science, the reliance on those facts is high. Peer reviews are important in this respect. Organized skepticism, communalism, universalism and disinterestedness are important notions in the science community, as noted by the sociologist of science Robert K. Merton.

The law and science have different frameworks, different contexts, for producing facts. Therefore, the law shouldn’t always defer to “science’s overriding commitment to self-correction”. Trial judges should act as “surrogates for the scientific community in determining admissibility”. This isn’t a perfect solution, however. For example, in the post about the Monty Hall problem, there were a lot of math professors who didn’t agree with the solution. When science ultimately decides about someone’s life or death, these failures can’t be tolerated.

Science can’t proceed the same way in the courtroom, as it would outside it. It simply can’t remove the uncertainty that the law itself would have when convicting/judging a suspect.

DNA and Truth-Telling

“Science may be a social activity, but when executed correctly, the results are viewed as no longer bearing traces of human subjectivity.” Because of the removal of human elements, the facts that are produced by scientists, are very reliable kinds of evidence. The transition from the subjective legal definitions to more objective scientific notions, through the removal of fallible human interpretations in criminal law through diagnostic instruments, is a process we already see happening with the advance of DNA technology. “The hope is that technology, through its mechanical reproducibility, will be impervious to context and will provide unbiased and reliable evidence about the facts of the matter.” Again, DNA technology is a good example. The enormous discrimination possible with DNA (random match probabilities of 1 in a billion for a complete profile) are negligible with respect to the chance a mistake is made by a human factor: problems with taking samples, mixed up profiles, contamination, holes in the chain of custody, etc. Those factors are far more likely to occur, and illustrate that an overreliance of DNA profiles is dangerous. Also, the human element in the law, and the urge of the public prosecutor to convict somebody (e.g. bias), are noteworthy components that shouldn’t be forgotten. The ability of DNA to establish identity is not questioned; it’s the interpretation of the results that should be questioned. Alternative explanations, no matter how unlikely, should all be removed before there’s a certainty for a rightful conviction.

Three more propositions will be investigated:

• “Truth-seeking in science is equivalent to truth-seeking in the law” (See part 1)
  
• “Law enforcement (or forensic) science establishes the truth as reliably as science in pure research contexts” (See part 2)
  
• “Genetic science is a particularly dependable source of truth, especially in disputes concerning human identity” (See part 3)
  

Vegetables

In the May edition of Scientific American, there's an article about consciousness of coma / vegetable patients. It's titled "Eyes Open, Brain Shut", and written by Steven Laureys.

When people slip into a coma, they don't open their eyes, but some of them may show some reflex movements of the limbs. If people come out of their coma, they can enter a vegetative state, in which they remain unconscious; they are awake, but not aware. For instance, they have sleep/wake cycles, and some form of movement which is not purposeful but only reflexive.


How do you measure the awareness of a patient? How do you diagnose a vegetative state? This may be helpful to distinguish between patients who may recover or not. A MRI or CT scan can show how damaged the brain is, but it's impossible to see if the patient has some level of consciousness. An EEG (ElectroEncephaloGram) measures the brain's electrical activity which is able to show the state of wakefulness, but not a reliable change in awareness.

With the use of a PET (Positron Emission Tomography) scanner, the metabolic activity can be viewed (measured by its consumption of glucose). In the vegetative state this metabolism is lower. When the patient is at rest, it can't successfully distinguish between the vegetative and minimally conscious state. This changes when external stimuli like pain and spoken words are added to the equation. This makes sense, because the awareness in a vegetative state is lower then in a minimally conscious state.
Persons in a vegetative state may very well understand commands: there is a "conscious linguistic processing in the vegetative patient". Furthermore, in some "mental imagery tasks", the patient understood the tasks (tasks like: imagine walking through the rooms of your house). These responses were indistinguishable from that seen in the healthy subjects. However, there may be a possibility that the patient was transitioning to a minimally conscious state, in which the awareness is raised.

It's a well written article; so if you have access to it, I'd recommend reading it.

Received LPs / CDs

A few days ago (saturday), just 5 days after it was shipped, I received the LPs/CDs I ordered a few days before that from constellation records. A few albums from Godspeed You Black Emperor!, Do Make Say Think, Exhaust, Fly Pan Am, and a compilation CD.

They're amazing; especially the LPs look great, and the F#A#∞ album includes a penny crushed by a train, hand-glued picture on the front, de-bossed titles ... Simply amazing.
Some cd's look silkscreened; it has a very nice touch... I'm not sure if the word is correct, but suffice it to say it looks good.

These people really deserved my money; they're really in it for the music, and you can hear it.

Monty Hall problem

I was thinking about the Monty Hall problem, a problem my statistics teacher once told me about. In short it's about the following:

You appear on a game show. There are three doors, behind one there is a prize. The other two are empty.
Assuming you want to win the prize, you guess which door holds the prize: A,B or C. The gamehost subsequently shows one door behind which the prize doesn't reside. The question is, should you alter your choice or not?

The counterintuitive but correct answer is: Yes.
It's sufficient to look at the two situations: either you alter your choice, or you don't.

If you don't swap:
Two things can happen:

• You choose a door. You have a chance of 1/3 to choose the right door immediately.
• Hence you have a chance of 2/3 to choose the wrong door.

Chance of failing: 2/3
Chance of winning: 1/3


If you do swap:
Two things can happen:

• You choose the door behind which the prize is (chance is 1/3). The gamehost then shows you a door behind which the prize doesn't reside. Two closed doors remain. You change your choice to the other door, so that your final choice will be the wrong one (you fail).
• You choose a door behind which the prize doesn't reside (chance is 2/3). The gamehost then shows you a door behind which the prize doesn't reside. Again two closed doors remain. You change your choice to the other door, and hence your final choice will be the right one (you win).

Chance of failing: 1/3
Chance of winning: 2/3


You see it's better to change your choice in such a situation.

Small matlab code:
(Note: ceil(3.*rand(1,1)) means that a random number is generated, not higher than 3, rand(1,1) is a random 1x1 matrix).
Sorry for the fucked up code; blogger doesn't like tabs :\.

function findprize=findprize(repeat, swap)

WIN=0;
FAIL=0;
if swap==1;
for i=1:1:repeat
prizenumber=ceil(3.*rand(1,1));
choosenumber=ceil(3.*rand(1,1));

if prizenumber==choosenumber;
FAIL=FAIL+1;
else
WIN=WIN+1;
end
end
end

if swap==0;
for i=1:1:repeat
prizenumber=ceil(3.*rand(1,1));
choosenumber=ceil(3.*rand(1,1));

if prizenumber==choosenumber;
WIN=WIN+1;
else
FAIL=FAIL+1;
end
end
end

FAIL, WIN

This gives:
no swap:
findprize(100000,0) gives: 66625 failures, 33375 wins

swap:
findprize(100000,1) gives: 33438 failures, 66562 wins

There ya go.

Sisters wallpapers

I was a little bored this afternoon, and while browsing the heartland forum, I thought about making some wallpapers myself. Here they are. Credits go to the person who vectorized the logos, and the persons who made the background layer.
Hope you like them; most of them are 1600x1200, in PNG.
The first 10 or so are almost the same, only a different layer style.
(please leave a comment if you like them)

PER3 Polymorphism Predicts Sleep Structure and Waking Performance

The latest edition of current biology holds an article about the connection of VNTRs on genes and sleep behaviour. (Current Biology 17, 1–6, April 3, 2007)
The genetic background about sleep and waking patterns is largely unknown. This paper writes something about it; not everything is understandable for me, as I have a limited background in biology (actually, no background in biology), but with the use of wikipedia I could draw some conclusions.
Individuals were monitored in their sleep-wake cycles, after which some intensive physiological tests were done. This was done in normal conditions, and in conditions of sleep loss. The persons, selected on basis of their genotype and homozygosity for the PER3 -gene, showed no significant differences in bed time, wake time or sleep duration.

Note: PER3[4/4] means that the person is a homozygote, with 4 repeats of the characteristic amino acid.

In their normal patterns, there was no significant difference in the different stages of sleep (REM sleep, stage 1 sleep, stage 2 sleep, total sleep time), but

"PER3[5/5] subjects fell asleep more readily than PER3[4/4] subjects"

When the PER3[5/5] were kept awake for a long time, the subjects performed worse than PER3[4/4] persons on spatial, reaction-time, and logic tests, especially in the late night and early morning hours:

"Most strikingly, PER3[5/5] homozygotes performed very poorly during the hours after the melatonin midpoint. The decrement in waking performance in the PER3[4/4] homozygotes was far less. These major differences in performance between the two genotypes occurred during the late-night and early-morning hours, a time known from both laboratory and field studies as the nadir of the circadian timing system and during which performance is poorest and sleep propensity at its peak. "

"The PER3 5-repeat allele, which is the less frequent one in most ethnic groups, has been associated with extreme morning preference, while the 4-repeat allele has been linked with DSPS in our previous study."

DSPS is a delayed sleep phase syndrome; people with this syndrome tend to fall asleep late at night, and have difficulty waking up in the morning. Furthermore, for a lot of these persons it doesn't matter at what time they go to bed, because they fall asleep at approximately the same time anyway. DSPS is a syndrome from the bigger family of Circadian rhytm sleep disorders; a well known member from this Circadian rhytm sleep syndrome is the jet lag. This may mean that there's a problem with a part of the brain that produces melatonin, which receives information from the eyes about light and dark.

These results, among others,

" (...) led us to consider it as a candidate for mediating some of the marked individual differences in sleep-wake regulation. These individual differences include the preferred timing of sleep-wake cycles, the structure of sleep, EEG patterns during sleep and wakefulness, and their response to sleep loss and circadian-phase misalignment."
"Our results indicate that the PER3 polymorphism may contribute to the marked individual differences in performance decrement during sleep loss."

All in all, this may signify that there's a relation between day- and nightpeople.

"Conclusions
The effects of the PER3 polymorphism on SWS (slow wave sleep), SWA (slow wave activity), and the decrements of waking performance during the biological night, as observed in this study, are significant and substantial. This implies that this polymorphism may be an important marker for individual differences in sleep and susceptibility to sleep loss and circadianphase misalignment, which are major causes of health problems and accidents in our society."

Of course, there are still a lot of open questions: what happens for example with heterozygotes? Or with a smaller number of VNTRs on the PER3 gene? Is there a connection, or is it just a correlation? I'm not in the position to answer these questions, as I could barely understand the article ;).

From: Viola et al., PER3 Polymorphism Predicts Sleep Structure and Waking Performance, Current Biology (2007), doi:10.1016/j.cub.2007.01.073

Abstract pictures

It has been a long week, without much time to post. But I made a few pictures, and to be honest I like them. Click on them to get the large version.

New name

We changed the address of this blag; now it's zerofication.blogspot.com, as you're probably aware of. The new name is more original, and stands for everything in life that is diminished into nothing. Nullified. Zerofied. Zerofication: the process of zerofying. Or something like that. Like television news reporting on 'human interest'-stories, instead of the really important news.
By the way, a good news source is The New Standard, in their own words "independent. non-profit. commercial-free. uncompromised".

Lou Reed in Amsterdam

I was happy when I saw that Lou Reed is coming to town. Lou Reed. To Amsterdam. Performing Berlin. Live. Until I discovered the ticketprices: approximately between 60 and 73 euros. Fucking hell. I expected it to be around 45-50. Not sure if I still want to go; it's a lot of money. I can (or I have to) eat for a month with that money. But still haven't made up my mind.

New pictures

Here is some new experimental art I created. Maybe you'll like it.

(No, it's not some kind of love-toy ... what women think of these days ...)

Omega-3 Fatty Acids and Mood Disorders

Another interesting article from the American Journal of Psychiatry, titled

"Omega-3 Fatty Acids and Mood Disorders"

Here are the main points; below that there's a summary of the article.


Objective: This article is an overview of epidemiological and treatment studies suggesting that deficits in dietary-based omega-3 polyunsaturated fatty acids may make an etiological contribution to mood disorders and that supplementation with omega-3 fatty acids may provide a therapeutic strategy.

Method: Relevant published studies are detailed and considered.

Results: Several epidemiological studies suggest covariation between seafood consumption and rates of mood disorders. Biological marker studies indicate deficits in omega-3 fatty acids in people with depressive disorders, while several treatment studies indicate therapeutic benefits from omega-3 supplementation. A similar contribution of omega-3 fatty acids to coronary artery disease may explain the well-described links between coronary artery disease and depression.

Conclusions: Deficits in omega-3 fatty acids have been identified as a contributing factor to mood disorders and offer a potential rational treatment approach. This review identifies a number of hypotheses and studies for consideration. In particular, the authors argue for studies clarifying the efficacy of omega-3 supplementation for unipolar and bipolar depressive disorders, both as individual and augmentation treatment strategies, and for studies pursuing which omega-3 fatty acid, eicosapentaenoic acid (EPA) or docosahexaenoic acid (DHA), is likely to provide the greatest benefit.
(Am J Psychiatry 163:6, June 2006)

I learned a few things from this paper: first, the omega-3 fatty acids can be either marine-based or from plants.
The rapid growth in population has been associated with a change in diet, resulting in a decrease in omega-3 while the uptake of omega-6 (e.g. from vegetable oils) has increased. Some authors suggest that there's a connection between the depression-rate and other neurological disorders and the increase of the omega-6 fatty acids; there's a possible link between fish consumption and mood disorders. Both Iceland and Japan, which both have high consumption rates of fish, have low rates of "seasonal affective disorders". As the consumption rate of fish declines (and the consumption of 'Western' food rises), the rates of seasonal affective disorders increases, even when they receive more winter sunlight. Furthermore, the "likelihood of having depressive symptoms was significantly higher among infrequent fish consumers than among frequent consumers." Some research also opposes these findings, but there's a general tendency to support the previously mentioned claims. Other research state that this is especially so with women. But in short, 5 out of 6 researches find that there's a correlation between fish consumption (omega-3 intake) and mood disorders.
During pregnancy the fetus accumulates more DHA (primary component of omega-3) than the intake of the mother, and after the birth the depletion of the omega-3 acids continues by breast-feeding. This contributes to the risk for depression around the time of birth. So feed your wife some fish during pregnancy (finally a reason to be happy someone eats dead animals);

• it's safe
  
• it may have "additional benefits for the infant’s neurodevelopment"
  

Then there's a lot of biological mumbo-jumbo, which I don't fully understand, but which still support the previous claims. Some possible mechanisms are presented, to continue to the treatment studies. Bipolar/depressed/borderline patients which took omega-3 had in general greater symptom reduction. On the other hand, there are some experiments which don't reproduce this effect, but again, in general, the results support the previous idea.

Interesting stuff.

Spatial memory and depression

In my first post I already mentioned that this blag is mostly for myself, and that I'd add notes for myself about things I like or want to remember. This is such a piece.
In the American Journal of Psychiatry I found an interesting article titled:

Performance on a Virtual Reality Spatial Memory Navigation Task in Depressed Patients

The title doesn't leave much for the imagination on what it's about, but the findings are interesting nonetheless:

Method: Performance on a novel virtual reality navigation task and a traditional measure of spatial memory was assessed in 30 depressed patients (unipolar and bipolar) and 19 normal comparison subjects.

Results: Depressed patients performed significantly worse than comparison subjects on the virtual reality task, as assessed by the number of locations found in the virtual town. Betweengroup differences were not detected on the traditional measure. The navigation task showed high test-retest reliability.

Conclusions: Depressed patients performed worse than healthy subjects on a novel spatial memory task. Virtual reality navigation may provide a consistent, sensitive measure of cognitive deficits in patients with affective disorders, representing a mechanism to study a putative endophenotype for hippocampal function.
(Am J Psychiatry 2007; 164:516–519, link )

Well, there you go. I think this was interesting, and I want to remember it.
(The only question I have that the article didn't answer is, if there existed a correlation between the familiarity with the game and the people who were depressed. I.e. whether the depressed people could be more familiar with the game)

Common sense?

Scientific American: Fact or Fiction?: Living People Outnumber the Dead
Booming population growth among the living, according to one rumor, outpaces the dead.

I read in the above article about the question whether the living people outnumber the dead. From the article:
"The human population has swelled so much that people alive today outnumber all those who have ever lived, says a factoid whose roots stretch back to the 1970s. Some versions of this widely circulating rumor claim that 75 percent of all people ever born are currently alive."

Unless you believe the world started populating around the time Jesus supposedly died on the cross, this can't be true.
A highly simplified model could be the following:
If we start with 1 couple (2 people), which produce slightly more than 2 children who (when their time comes) produces more than 2 children ... and so on, we can ask the following:

When is the sum of all the people produced equal to 6 billion (current population)?
If we take as an exponent 1.025 (so that 1000 couples produce 1025 new couples), it takes less than 800 generations. Of course, we didn't take anything in account (plagues, food shortage, wars, infertile men/women, gay couples, abortions, women who are never in the mood, ugly/stupid people who couldn't get a date, etc), but the exponent is very low.
Even if we take as an exponent 1.002 (1000 couples produce 1002 new couples), we need less than 8200 generations to accomplish the total amount of 6 billion people who are dead. The first homo sapiens originated from about 200.000 years ago from Africa (Wiki).

Estimations for the year 2050 range between 7.3 and 10.7 billion people, giving a much larger exponent than we took.
I think we can safely say it's bullshit to say that more people are alive now than that ever lived.
(The article comes to the same conclusion; it just surprises me that people could ever believe such a statement)

Experimental art

Experimental art using two forks.