n Tue, Nov 24, 2015 at 9:50 AM, Tom Wolper <[email protected]> wrote:
> I took my time writing my first response so that I would not get the > statistics ideas screwed up and yet I still did. I made the remark about > sample size which is generally part of a tirade against Nielsen not polling > enough people to get an accurate number and I did not mean that. > > I will give a hypothetical to illustrate what I meant. I think Carson used > to have a steady 30 million viewers. And post-1995 they broke up to 17 > million for Leno and 13 million for Letterman. If 500,000 viewers decided > to switch to Letterman the numbers would not move enough to justify a > column about how things are changing in late night. The latest numbers for > late night from TV by the Numbers shows Fallon 3.79, Colbert 3.17, and > Kimmel 2.54 (season to date). If 500,000 viewers change the channel now it > becomes a significant change and the columns are written about how it must > somehow be political. > Right - this is the basic point, and is pretty much what you and Melissa and I mean when we say it is too soon to judge whether the changes in viewership are just noise, or reflect real changes. It is a point worth reminding ourselves of whenever looking at TV ratings - this data is typically not reported with the information needed to judge if changes are statistically significant (by "significant" here we mean not important or meaningful, but consistent or reliable). The size of each sample is part of what we would need, in addition to looking at the size of the difference relative to the absolute value of each group (which is what you are referring to above). We would also need to know the standard deviation of each group. Since we can not make reasonable inferences about the reliability of these changes over the short term, we need to wait for more data points. Having said that though, I think the specific example you give actually suggests a different conclusion than the one you are trying to support. It is true that a 500,000 viewer change from one late night program now is a larger fraction of each program's total viewership than it was for Dave or Carson, but that fact itself is one reason that such a change would be more likely to signal real changes in the underlying population (but again, we have to know some other things too before reaching that conclusion). Another way of saying that last point is that in a more fractured viewing environment, each viewer becomes more valuable. -- -- TV or Not TV .... The Smartest (TV) People! You received this message because you are subscribed to the Google Groups "TV or Not TV" group. To post to this group, send email to [email protected] To unsubscribe from this group, send email to [email protected] For more options, visit this group at http://groups.google.com/group/tvornottv?hl=en --- You received this message because you are subscribed to the Google Groups "TVorNotTV" group. To unsubscribe from this group and stop receiving emails from it, send an email to [email protected]. For more options, visit https://groups.google.com/d/optout.
