Kabrich: Philly PPM Sample Falls Below Sample Target
May 5, 2008 at 6:04 PM (PT)
As if the lawsuits aren't enough, consultant and ARBITRON watchdog RANDY KABRICH has been busy analyzing the PHILADELPHIA PPM in-tabs, and he's not happy with what he sees, as the 6+ target sample of 1,530 is down to 1,524. This is the first time this has happened over the past 7 months. And, he sees other disturbing numbers as well, based on ARBITRON's attempts to bolster 25-34 in-tabs.
KABRICH told ALL ACCESS, "There were two ways to approach the 25-34 issue. First, ARBITRON could have oversampled randomly, by approximately 200 persons, and kept the current sample in place. Over time, attrition would bring the sample down to the target numbers. This would have cost ARBITRON approximately $1,000 a week.
"The other option, which they chose, was the cheap option. ARBITRON randomly deleted approximately 200 people and randomly added households that had 25-34 family members. This destroyed six months of progress that was made in boosting the 18-24 sample and has also dropped the 35-44 in-tabs to their lowest point ever.
"On top of that, ARBITRON now finds itself below its 6+ target sample of 1,530, down to 1,524, and it will cost them much more than the approximate $5,000 they saved in the last five weeks to recruit and install equipment in additional households to try and achieve their target. One can only wonder what they are thinking and shows again why we need MRC accreditation in all markets ... as someone isn't paying attention."
To see what RANDY is concerned about, just click here.
ARBITRON spokesperson JESSICA BENBOW told ALL ACCESS:
It is important to remember that ARBITRON had been over sampling in order to bring the numbers up. We are now managing the overall persons 6+ DDI closer to 100. Our benchmarks are based on MONTHLY sample performance. There will be some variation from week to week.
That said ...
* P6+ DDI is at 99.6 with 1524 average daily in-tab against a goal of 1530. We expect the P6+ sample to be close (plus or minus) to the hundred mark for the full month. Normal week to week variations may put us above the 100 mark in some weeks, under in others, but averaging out at 100.
* We are pleased to report that we are on track to deliver the 80 DDI benchmark for persons 18-34 for the full month. Until last month the improvements for this demo were based almost exclusively on growth in 18-24. Now that procedures are in place for 25-34 we will be able to deliver on the benchmark in a more balanced manner, with improved samples for 18-24, 25-34, and 18-34 inclusive.
* P18-24 DDI in week 2 is at 71; down from 73 last week. We expect this cell will stabilize in the low 70s for the next couple of weeks and then start going up again in MAY, though the lower summer response rate may impact that growth once we get into JUNE.
* P25-34 DDI in week 2 is at 85 which is slightly lower than last week's of 88 but improved vs. the March DDI of 80.
* 18-24 will stabilize in the 70s during MAY, with some fluctuations from week to week. However, we fully expect that as we have historically seen, summer numbers will be lower. This happens every year because listening habits change when people are on vacation. The important this is that longer term, we plan to further increase the 18-24 sampling rate to increase DDI in-tab levels and offset the lower in tab rates that we expect to see during the summer months.
Note: all these benchmarks are the average for the year. Some months may be higher and some may be lower.
Kabrich Responds Back...
RANDY KABRICH countered with these observations: "When the sample in-tab fell by 200, ARBITRON spin-master MOCARSKY was quick to point out for that week -- and the following two weeks -- that aRBITRON was above its 6+ target of 1,530 in-tab. Now MOCARSKY is on vacation and the sample size has fallen below the target, ARBITRON is now dismissing the sample size. This should make on question about the consistency of what is coming from various people within ARBITRON at the very least.
"This is further shown as BOB PATCHEN said that both 18-24 and 25-34 were trending up on the PPM monthly call, yet they continue to fall -- again, something ARBITRON did not address in their response. How can their Senior Research person make statements like this and yet we don't see it in the data?
"Would they please make up their mind and get a consistent story among the entire Company?
"Comparing the sample size and DDI is apples to oranges. They could miss the 1,530 target by 500 respondents and still make the DDI percentage.
"As RICK CUMMINGS pointed out during the PPM call two weeks ago, DDI is a percentage of a percentage -- and that is something that other researchers are very skeptical to look at, including ARBITRON in the past, when broadcasters looked at things that were a percentage of a percentage."