Arbitron claims PPM recruiting success in Houston
No ratings data is yet being compiled and released in Houston, but Arbitron is claiming success in its Portable People Meter (PPM) recruitment. In the first wave of recruiting panel participants, Arbitron achieved a 52.7 household sample performance indicator (SPI). That means that of the households randomly selected as targets for participation, 52.7% agreed to participate.
With Nielsen not participating as a partner in the Houston test, Arbitron is trumpeting that it was able to achieve a high response rate on its own, using the recruitment techniques developed after initial low response rates in Philadelphia. Arbitron has traditionally relied exclusively on phone contacts, while Nielsen sends recruiters door to door. What's being used in Houston is a hybrid, with phone contact attempted first, followed by in-person visits to households that don't respond. The test is also using only Arbitron-branded materials, while the previous test in Philadelphia was co-branded with Nielsen.
"Arbitron recruiters, using Arbitron-only branded materials, have delivered an initial PPM panel that we believe will stand up to the scrutiny of the radio and television industry in terms of its quality," said Arbitron Vice President of Research Standards and Practices Bob Patchen.
In contrast to Nielsen's Local People Meters, where fault rates have tended to be higher among minority households, Arbitron says the household default rates in Houston so far have been well under 10% "and are consistent by race and ethnicity." Company officials believe that's because PPM is a passive, portable device that isn't disabled if someone attaches another device to a television set.
The first wave of recruitment was completed in Houston at the end of November, resulting in 554 people being outfitted with pager-like PPM devices. Wave two, now underway, has recruited 486. Two more waves, beginning after January 1st, are to bring the panel to 2,100 by the end of April. That will give Arbitron the participants needed to run the test with two parallel panels of 1,000 each.
Arbitron plans to begin collecting and evaluating data at the total persons level and in July, after recruitment is complete, to begin delivering station-level ratings data.
Additional Panel Metrics from Wave 1 of the Houston PPM Market Trial
"Compliance Qualified" Response Rate
|
|
Houston
Wave 1
|
2003 Response Rate Test
|
In-person
|
Multi-mode
|
Household
|
40.1
|
41.2
|
42.7
|
(n)
|
(239)
|
(121)
|
(120)
|
Persons
|
29.4
|
32.2
|
33.6
|
(n)
|
(633)
|
(326)
|
(316)
|
Compared to a Sample Performance Indicator (SPI), the "Compliance Qualified" Household Response Rate adds a daily compliance edit of at least one person in the household to determine the numerator used in the calculation. This is the calculation for that was used in the 2003-2004 PPM Response Rate Tests.
Houston data Aug.-Nov, 2004; Response Rate Test data April-July 2003.
2003 Response Rate Test data are shown for the equivalent duration of the current state of the Houston panel.
Household Cooperation Rate
|
|
Houston
Wave 1
|
2003 Response Rate Test
|
In-person
|
Multi-mode
|
Households
|
69.1
|
62.7
|
63.7
|
Household Cooperation Rate: Total Installed Basics/Total Installed Sampling Points
Houston data August-November 2004; Response Rate Test data April-July 2003
Houston Preliminary Installed Household Proportionality,
by Householder Race/Ethnicity
|
Wave 1
|
Universe
|
Installed
|
Index
|
Black
|
16.5%
|
20.0%
|
121
|
Hispanic
|
24.0%
|
22.0%
|
92
|
Other
|
59.5%
|
58.0%
|
97
|
Index = (Installed %/Universe %) x 100
Houston Preliminary Household Fault Rates,
by Race/Ethnicity
|
Wave 1
|
In-Tab
(avg.%)
|
Faulted
(avg.%)
|
Total HH
|
92.8
|
7.2
|
Black HH
|
92.8
|
7.2
|
Hispanic HH
|
95.1
|
4.9
|
Other HH
|
91.9
|
8.1
|
(Houston Wave 1, August-November 2004)
Source: Arbitron
RBR observation:
This is why you test. Response rates were initially a problem for PPM in Philadelphia, so a solution had to be found. If the initial stats hold up, it looks like Arbitron has made real progress. Another improvement in the Houston test will be a modification to the PPM device that will differentiate in-home and out-of-home media use, based on proximity to the docking/recharging unit. TV guys, who've never had out-of-home viewing measured by Nielsen, are salivating over that. But cost remains a big concern for many radio companies, along with how PPM affects morning drive ratings.