Standard Deviation...ok, try to think about it this way....its all to do with statistics. Lets say we were doing a study into how many drinks the average Brit drinks on a saturday night....easy enough to associate with? Ok, so we go out and ask people how much they drank on saturday and plot it on a graph...this is were you have to use your imagination....on the x axis we put number of pints, and on the y axis we put the number of people who had that number of pints. With me? Ok, what you'll find is that the graph peaks around an average like this....
If that peak is at say "5 pints", then we can say that the average person drank 5 pints on that saturday night. Make sense? Ok, now, standard deviation. We could re-examine our stats with the "standard deviation formula", and what we will acheive is we will find out what "most people" drank, rather than what "the average person drank". For example, we know that the average person drank 5 pints....but what standard deviation will tell us is that MOST people drank between 3 and 8 pints. Get it? So if we were to plot the graph again....
the red area would show the value of standard deviation....lets say for arguments sake, 70% of the population....the red together with the green area would show us two standard deviations, lets say totalling 85% of the population....and the red, green AND blue areas would show us three standard deviations, effectively accounting for 98% of the population. The other 2% would be people who drank nothing and people who drank ridiculous amounts (2%=TLR members methinks!
)
And its as simple as that! So why do we want to find standard deviation? Well its all to do with companies etc that rely on statistics. If this drinking survey was done in every town in the country, you would be able to calculate how much each town "deviates" from the national average....to find out things like "the most drunk town in britain" or scarily "the most sober town in britain"...you'd even be able to work out which town is the most "average" by looking at the town, wait for it, with the lowest standard deviation.
Get it? No? Well I'm not explaining it again!!