David, If a 4.9 = 4.2 because both are in the wobble, then everything the Arbitron/Nielsen critics have been saying is true. As someone who has a few statistics classes under my belt, I completely understand the concept of "wobble", but this was over three reporting periods on a straight downward slide - that's not a wobble, that is a trend. I don't believe for a minute that at this point in time you would tell your clients that this represents a "wobble" and you would stand pat until the next month comes in at a 4.0 or worse.
I also know you wouldn't be going with that stupid vacation possibility either. This book covers the period AFTER the Labor Day weekend through the beginning of October, a period when few people are actually on vacation and listening (and TV viewing patterns) tend to return to their norms coming off of their various summer vacations and activities.
ChannelFlipper:
What you're overlooking is that, under PPM, we're getting monthly books that used to comprise one-third of what we used to call the winter, spring, summer and fall books. August, September and October average out to a 4.5. I've lost the 6+ numbers for May and June, but they were, if I recall, lower than July's 4.7. So the three-month average there is likely a 4.5 or lower.
Additionally, I have never considered anything less than a full point to be cause for concern. Wobble is very real.
Finally, all this is over the meaningless 6+ numbers. 25-54 is what matters.