Lies, damned lies, and statistics?

Today was a wunderful day for all of us geeks: World Statistics Day! Hooray! Time for the boys to wear their “proud to be a geek” boxers and for the girls to rummage around for their “I love robots” handbag!

Well, I think not everyone is going to share my enthusiasm about statistics. In fact, statistics does not seem to be vey popular amongst most people. I don’t know if there is a similar thing in the UK, but in Germany there is a kind of “urban myth” which says that about two thirds of all statistics published in the media are fake. Which ironically must be the result of another statistic…

But there`s not just bad things about statistics and as I just said, this is just a myth (and I assume that it is a myth because I never found a relyable source for it – if you do find one, just let me know!). And although you might not have big trust in statsitics, let me convince you: Statistics are great!

First thing that I love about stats: They are completely based on logic. Not necessaraly on common sence, but if you dive deep into a statistics book you will find out that if you follow the rules of statistics, you can not go wrong. Ok, this does not mean that these rules are very easy to learn or sometimes even easy to understand. But if you dive deep into a statistics book and take a bit of time for reading it, you will find out that you may not unterstand why this formula looks how it looks like, but the principle that lies behind it. Which is based on – logic!

Another great thing about statistics is, they give you solutions to questions that you could elsewise only answer with very extensive and time-consuming methods. For example, if you want to know if women in the UK really watch more tv than men, you could of course go ahead and ask every singe person in the UK about their tv use habits. But that would probably take up ages. So the easier way to get the answer is to pick a representative sample of people, question them about who match they watch TV and than perform a statistical test (like the T-test) to test if there is a significant difference between these men and women. Now what`s significant? When a test is significant, it tells you that the difference between the two groups  also exists in your basic population, in this case everyone living in the UK. (Ok, to be correct, even a significant test might be wrong sometimes – this is called the probability of error. Every discipline has its own conventional probability of error – e.g. 5 per cent in social sciences. That is, if you perform the same test 100 times, five of your results will be wrong – that’s one of the reasons why studies should always be repeated to validate the results.)

A third reason to love statistics: If you know how to read them, you can not be fooled. If a press release tells you about “unambiguous” proofs of the effect of a new drug or “explicit” demonstration of a certain effect, a closer look at the original paper can tell you if this is true or if the press release overstates the findings.  So if someone says once again that statistics reported in the media are often wrong – it’s at least not due to you…

Can you now guess why I love statistics?

This entry was posted in science in the media. Bookmark the permalink.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s