Statistical analysis for beginners

Following my startling revelation in an earlier post (see “Just one word”) that the average IQ of regular participants in the Lounge was 235, I have received a deluge of e-mails asking how this figure was arrived at.

Well, over the past few months I have been running a small but sophisticated AI program which analyses linguistic and logical content, identifies markers for mathematical, scientific, cultural, artistic and social awareness and creativity, and measures the speed, relevance and coherence of reponses to Lounge posts. It can also make heuristic estimates of the numbers of people who regularly log in and read the posts but don’t reply because (a) they think the posts are too stupid or (b) they are too stupid to understand them or (c) they can type their own name and password  - possibly with some assistance and patient encouragement - but have never mastered basic grammar and syntax. And can never quite find the “Mmmphhh” or “Doh!” keys on their home computer.

The raw figures were then subjected to a series of iterative stochastic procedures to minimise statistical bias. This is a quite complex process and I have to admit that sometimes, at the end of a well-lubricated evening of socio-cultural analysis, a little carelessness can creep in. So when I re-ran the numbers this morning after correcting for some minor discrepancies I found the true average IQ figure was closer to 58. Which make a lot of sense. (Or, much more to the point, makes very little sense at all.)

By the way, in order to protect the fragile self-esteem of those who are differently blessed, intellectually speaking, the results have been fully anonymized, so individual scores cannot be extracted. This will be frustrating for those of you with a thirst for truth and transparency. On the other hand, it could just save you a lot of embarrassment.

Sorry, this is just me having a bit of idle fun and I know it will leave many of you totally bewildered.

QED.