Long time, no blog...
An AMAZING Horizon program on the nature of scientific debate and the representation of science in the media. An absolute must watch!!!
Tuesday, 1 February 2011
Wednesday, 8 December 2010
The Joy of Stats...
There was a great program last night on BBC 4 on The Joy of Stats. It was a great overview of the history of statistics and how they are relevant and used in current research.
Continuing with the stats theme, the new series of More or Less starts on Radio 4 this Friday (10th), and Lady Tasting Tea is a great book on the history of statistics and the key people involved in developing this field.
Continuing with the stats theme, the new series of More or Less starts on Radio 4 this Friday (10th), and Lady Tasting Tea is a great book on the history of statistics and the key people involved in developing this field.
Friday, 3 December 2010
Finger length is associated with EVERYTHING...
There seems to regularly be reports in the media of a new studying showing something to do with finger length and illness, psychological disorders, personality, sexuality, the list goes on... The Daily Mail have even attempted to provide a summary of what this is all about!!! And obviously the responses are full of "rubbish" and "why is our money wasted on this research"...
So what is this all about?
First, it is not actually finger length, it is the relative lengths of the index and ring fingers. If the index finger is shorter than the ring finger, this indicates a more "masculine" 2D:4D.
Second, digit ratio (as it is properly called) is certainly not going to "cause" any kind of illness, personality or any other psychological trait. Instead, digit ratio and all of these "outcomes" are associated with a third variable - prenatal testosterone exposure.
Levels of testosterone prenatally are associated with determining the relative lengths of the fingers, and potentially in a great number of different psychological and medical outcomes. So if someone has high levels of prenatal testosterone exposure, they are likely to have more masculine traits and may be at risk of some illnesses associated with higher levels of testosterone.
Before dismissing the relationship between prenatal testosterone and digit ratio, there is very strong evidence to support this relationship. One study measured levels of testosterone from amniotic fluid and found that, in later life, that children with higher levels of prenatal testosterone had more masculine digit ratios (Lutchmaya et al., 2004). Two other studies artificially manipulated prenatal hormonal levels in rats, and found that digit ratio changed accordingly (Talarovicova et al., 2009; McMechan et al., 2004).
So there is not a relationship between finger length and cancer/personality/sexuality/etc. They have the same causal factor.
It is also really important to remember that having a particular digit ratio will not 100% predict a particular illness or personality, it might just indicate an increased risk factor. Just like obesity increases the risk of various illnesses, but it does not totally determine or predict the illness, and people who are not obese might also develop the illness.
Similarly, the levels of prenatal testosterone exposure are not the only determiner of digit ratio. Many other factors are likely to be involved, but there is a clear and reliable relationship between levels of prenatal testosterone and digit ratio. Other factors under examination includes the functioning of hormone receptors (meaning some people may be more sensitive to, or absorb more, testosterone) and the relative exposure to testosterone in relation to other hormones, particularly oestrogen.
Rant over - thanks!!!
So what is this all about?
First, it is not actually finger length, it is the relative lengths of the index and ring fingers. If the index finger is shorter than the ring finger, this indicates a more "masculine" 2D:4D.
Second, digit ratio (as it is properly called) is certainly not going to "cause" any kind of illness, personality or any other psychological trait. Instead, digit ratio and all of these "outcomes" are associated with a third variable - prenatal testosterone exposure.
Levels of testosterone prenatally are associated with determining the relative lengths of the fingers, and potentially in a great number of different psychological and medical outcomes. So if someone has high levels of prenatal testosterone exposure, they are likely to have more masculine traits and may be at risk of some illnesses associated with higher levels of testosterone.
Before dismissing the relationship between prenatal testosterone and digit ratio, there is very strong evidence to support this relationship. One study measured levels of testosterone from amniotic fluid and found that, in later life, that children with higher levels of prenatal testosterone had more masculine digit ratios (Lutchmaya et al., 2004). Two other studies artificially manipulated prenatal hormonal levels in rats, and found that digit ratio changed accordingly (Talarovicova et al., 2009; McMechan et al., 2004).
So there is not a relationship between finger length and cancer/personality/sexuality/etc. They have the same causal factor.
It is also really important to remember that having a particular digit ratio will not 100% predict a particular illness or personality, it might just indicate an increased risk factor. Just like obesity increases the risk of various illnesses, but it does not totally determine or predict the illness, and people who are not obese might also develop the illness.
Similarly, the levels of prenatal testosterone exposure are not the only determiner of digit ratio. Many other factors are likely to be involved, but there is a clear and reliable relationship between levels of prenatal testosterone and digit ratio. Other factors under examination includes the functioning of hormone receptors (meaning some people may be more sensitive to, or absorb more, testosterone) and the relative exposure to testosterone in relation to other hormones, particularly oestrogen.
Rant over - thanks!!!
Thursday, 2 December 2010
Science journalism...
Usually I'm somewhat critical of science journalism - but this article is 100% perfection!!! A great parody from The Lay Scientist.
Tuesday, 23 November 2010
Gender and development...
The BBC have examined a really interesting case study of a boy who was brought up as a girl.
It is often difficult to distinguish between genetic/biological and environmental influences, but the child had a twin brother - possibly the best control participant ever!
It is often difficult to distinguish between genetic/biological and environmental influences, but the child had a twin brother - possibly the best control participant ever!
Tuesday, 16 November 2010
Statistical happiness???
With the government announcing that it will start to measure the nation's happiness and wellbeing, the latest Go Figure article by Michael Blastland looks at some of the problems with this analysis. Obviously the "correlation does not imply causation" mantra returns...
Monday, 15 November 2010
Real time world statistics...
Worldometers is a great site showing real time changes in global statistics to do with world population, society and media, food, health, energy and a lot more. Very interesting, quite mesmerising, but slightly scary in places.
Subscribe to:
Posts (Atom)