 |
 |
You probably saw the above "shock, horror" headline. It set me thinking as follows ..... If true, then the last three years must have been --- (where - denotes a decrease i.e. none of the last three years was a +) Now, given that three years in a row have been ---, the 3rd value must be pretty low. So the chance of getting another - is surely far less than the chance of getting a +, even if all values are i.i.d.? I tried to work this out analytically, using iid uniform (0,1) variables but eventually did some simulation (3000 values). I got a conditional ratio of - to + in year 4 as 185:50 i.e. almost 4:1. Moreover, the --- pattern only appears about once every 25 years. (Ditto with +++.) Any advice please on how best to write up this general phenomenon for a popular audience? Does it have a name? Other examples? How best to do the maths? Any useful extensions for non-iid time-series? Thanks for any help you can give. JOHN BIBBY ****************************************************** Please note that if you press the 'Reply' button your message will go only to the sender of this message. If you want to reply to the whole list, use your mailer's 'Reply-to-All' button to send your message automatically to [log in to unmask] Disclaimer: The messages sent to this list are the views of the sender and cannot be assumed to be representative of the range of views held by subscribers to the Radical Statistics Group. To find out more about Radical Statistics and its aims and activities and read current and past issues of our newsletter you are invited to visit our web site www.radstats.org.uk. *******************************************************