This discussion highlights one of the biggest problems in using and interpreting statistics. Kobo have calculated an average. We don't know their methodology so cannot comment on what assumptions they've made and how their data was collected. However the discussion strongly suggests that they and we have assumed that reading quantity across a population will follow a normal (bell) distribution and that the mean is a meaningful statistic. Is it? Have they actually done a test to see if the normal distribution is appropriate? I suspect not. I would think (based on personal experience and no data whatsoever :P ) that the Pareto principle (20% of people do 80% of the reading) would be a closer fit. Actually I think it's more likely 10/90 or even 5/95, and even that may be optimistic. But you get the idea. There's every possibility that there are two peaks - avid readers and non-readers, with relatively few in the middle ground.
Quote:
Originally Posted by JoeD
Most people I know that read, will read a lot more than 100 books in their lifetime if they continue at their current rates. A conservative estimate of those readers may be around 10 books a year (they're not heavy readers unlike I expect most of us on mobileread). However, the number of people I know that don't read at all or read very lightly (1-2 books a year) is significantly greater.
I don't think it's much of a stretch to see how most people may be non/light-readers and will drag the average down. Whether it's made up or based on the logic pdurrant suggests, it does at least feel like a reasonable number to me.
|
Joe's first paragraph illustrates my last point. The second shows his bias [1]. If we assume that most people are in the non-reader category, then it's those of us in the avid reader category that are in fact dragging the average *up*.
[1] I don't mean that in a bad way - we all have bias, we just have to try and be aware of how it affects our interpretation of results.
