Welcome to the statsART blog!

Welcome to the statsART blog where we highlight recent research, mainly work that has been discussed in the media.
Please note that the views expressed here only represent the views of the statsART blogger.
To vist or return to our website: www.statsART.com

Monday, 31 May 2010

Statistics in the media...

Generally there are two problems with the way that science is portrayed in the media. First, it is over simplified and this can lead to serious misinterpretation of the research done and the "real world" implications it has. Second, the presentation of the statistics often is wildly inaccurate and leads to scare stories.

Occasionally the BBC publishes fantastic articles about how to interpret statistics and science in the media. This is a great article about the recent scare story about the relationship between eating sausages and increased risk of heart disease.

Monday, 24 May 2010

Autism and migration...

A very interesting piece of research has suggested that the risk of a child developing autism if the mother has immigrated into the UK.

This was one of those studies that I initially was quite sceptical about - but toward the end of the report, they say the following...

"One theory is that the stress of migrating could act as a "trigger" for the disability, a factor discovered in similar studies looking at the causes of schizophrenia."

This is a really interesting idea, and it could explain the relationship that the researchers report. Going even further than they do - stress is associated with hormonal fluctuations, and there is research showing that there is a higher risk of autism if the child is exposed to higher levels of testosterone prenatally.

So, the news article reports the relationship between autism and migration as a relatively direct one. However, there may actually be four, or even more, stages...

Migration > Higer levels of stress > Changes in hormonal levels > Increased risk of autism.

A lovely example of the "correlation does not imply causation" mantra!!!

Wednesday, 19 May 2010


Apparently men lie more than women.

It's probably best for me not to comment on that...

Wednesday, 12 May 2010

But can you study that???

This newly published research examined how people judge research that opposes their own views.

Interestingly they found that people are likely to entirely dismiss the research and claim that "the topic could not be studied scientifically"!!!

Nice - if I like it we can study it scientifically, if I don't then you just need to trust me...

Tuesday, 4 May 2010

TV junkies...

A study has apparently shown that young children who watch too much TV are likely to suffer from a whole load of negative consequences later in life.

Interesting, and I don't doubt the studies findings, but I do think this could be an interesting example of "correlation does not cause causation". Just because there is a relationship, it doesn't mean that toddlers watching lots of TV causes them to have higher BMI's, poorer performance at school, etc.

It could be that watching lots of television may also be one of the outcomes of some other causal factor. What might cause a child to watch lots of TV, be overweight and do poorly at school? Now that would be an interesting study...