Originally Posted by
CD93:
“Well the viewing figures take in to account 5,100 households, so it has never been all that comprehensive 
”
Well, it's the only economically viable system of measurement. Anything larger would cost a fortune to run.
One vital aspect is that household members are required to log-in on the BARB remote. So, on a representative sample of 5,100 households / 11,500 people, the measurement of specific groups is taken in account.
Even if we eventually reach a point where every TV in the land is connected to the Internet, the totals would only show the total number of screens connected to iPlayer. A sample group would still be needed to show what type of people are watching the programmes.
Bear in mind that the margin of error is very low on a sample of 11,500. Just like it's very low on political opinion polls. (The election polls only
seemed more inaccurate than the actual margin of error because our archaic "first past the post" system throws up anomalies. If we had the superior proportional representation system, the actual votes cast were very close to the polls.)
Originally Posted by komentaightor:
“Have these famous "viewing figures" ever taken into account:
1) the number of viewers who record the show to watch it later when more convenient;
2) the number of viewers ditto who go straight to iPlayer for the recorded view;
3) the millions of people who get BBC on cable overseas and who can watch it live, like me - or do expats and foreign Dr Who fanatics not count?
I bet the viewing figures are way higher than those extrapolated from a few set-top boxes in middle class suburbia.”
I like how you put "viewing figures" in inverted commas, to make them seem less credible. As if the basis of most recommissioning decisions and advertising sales on commercial channels is somehow based on a myth.