View Single Post
Old 09-10-2013, 09:58 AM   #42
5thWiggle
Benevolent Evil Lord
5thWiggle ought to be getting tired of karma fortunes by now.5thWiggle ought to be getting tired of karma fortunes by now.5thWiggle ought to be getting tired of karma fortunes by now.5thWiggle ought to be getting tired of karma fortunes by now.5thWiggle ought to be getting tired of karma fortunes by now.5thWiggle ought to be getting tired of karma fortunes by now.5thWiggle ought to be getting tired of karma fortunes by now.5thWiggle ought to be getting tired of karma fortunes by now.5thWiggle ought to be getting tired of karma fortunes by now.5thWiggle ought to be getting tired of karma fortunes by now.5thWiggle ought to be getting tired of karma fortunes by now.
 
5thWiggle's Avatar
 
Posts: 1,704
Karma: 48339466
Join Date: Jan 2011
Location: Evil Canada (We all have goatees!)
Device: Galaxy Note 8.0, Galaxy Note, iPad Mini, PocketEdge(retired)
Quote:
Originally Posted by fjtorres View Post
The "skepticism" isn't over the numbers but over the way the survey was made and its (apparent) intended use.

The survey as presented is a meaningless factoid that says nothing new or meaningful.
First of all, it is a survey of the general population: 25-30% of which doesn't read books at all--if you wanted it to mean something, you would limit it to actual book readers.
Second, it doesn't weigh the "votes" by reading habits: people who only read on travel, or rarely, have no use for ebook readers.
Third, as pointed out, the factoids are presented as if to deprecate ebook usage, when the numbers actually indicate significant growth in ebook popularity. Which tells us very little because they only asked about preferences, not actions.

A proper survey would begin by asking if the respondent reads books for recreation. Then it would ask how many books read in a typical year, how many read in the last year, and how many were ebook editions. Then it would repeat for actual *purchases*. For ebook users it would ask *how* the ebooks are read--phone, tablet, laptop, desktop, or dedicated reader--and it would end by asking for gender, broad age group (at least 5 ranges), broad income range (ditto), and take note of location if possible.
And, a meaningful survey would sample different locations and different times in clusters.
Then, it *might* have meaning if the questions were properly (as in clearly, unambiguously, and neutrally) phrased *and* if the sample proves to be big enough to be representative of the target population.

Polling is not for amateurs.
Doing it right is hard work.
Or they could just poll here at MR
5thWiggle is offline   Reply With Quote