Wednesday, 25 May 2011

An inbuilt bias.


If there’s one thing I learnt during the statistics lectures during the second year of university – clearly this was AFTER a group of us went to see the lecturer to explain that we hadn’t understood ANYTHING from the first year, and ask if he could start again from the beginning with extra lessons - it was the need to be rigorous whenever taking a sample.

No room for bias!

Except.

Except – I have a theory.

My theory runs like this – statistical rigorousness can be obtained from a biased sample if the biased sample is systematically adhered to for a repeated and long enough time.

In other words – the B.S. approach.

Biased sample, not bullshit.

What? – I hear you asking.

A while back, here, following encouragement I think I read into a newspaper article, I started counting the butterflies along a two km section of track that conveniently runs from my front door to the road.

The method is not statistically sound, but it is consistent with inbuilt bias.

I was at it again on Saturday.

So - May 21 2011, 11am, running down, sunny, 80, predominantly blue.

No comments:

Follow by Email