erinMedia Exposes Inherent Flaws in Nielsen Media Research Methodology

erinMedia Exposes Inherent Flaws in Nielsen Media Research Methodology

Unprecedented Analysis and Findings Triggered By Massive Ratings Swings in Closely-Watched D.C. African American Households; Calls For Nielsen To Release All Data To Allow Industry Scrutiny

ST. PETERSBURG, Fla., July 6 /PRNewswire/ -- Television ratings provider erinMedia today published information that exposes fundamental flaws in Nielsen Media Research's (NMR) television program ratings, based on NMR's June 23 press release, which measured African American viewership in Washington, D.C., and attached dual panel data, which compares Local People Meter (LPM) and set meter/diary ratings.

The NMR release and dual panel data purported to show that data generated by NMR's LPMs were more accurate than NMR's set meter/diary-based data, because the LPM data reflected higher viewership of African American themed programs. However, analysis conducted by Dr. Michael Vinson, erinMedia's Senior Research Scientist, using industry-accepted methods including the time- honored Student t-Test approach, reveals with a certainty in excess of 99% that at least one set of Washington, D.C. data, if not both, are "systemically flawed, and apparently not both sets of data are being derived from the same representative population." Dr. Vinson's findings and additional commentary can be found at http://www.weshouldallcount.com/.

erinMedia's detailed analysis was conducted after a simple examination of the data obtained through NMR's website revealed an average discrepancy in excess of 40% between ratings of the top 30 shows among African Americans in Washington, D.C., between April 28 and May 25, 2005, compared with set meter/diary-based results during the same timeframe. These massive and unexplained swings led to further examination, which revealed that 90% of these programs experienced a dramatic decrease in audience size when compared to LPM measurement.

"NMR's news release unwittingly reveals, with 99% certainty, that at least one set of data, if not both, is systemically flawed," said Frank Foster, president of erinMedia. "The two measurement systems are simply not pointing to the same viewing picture and it is highly unlikely that both data sets accurately represent how African Americans view television in Washington, D.C. With more data, I may come to a different conclusion. Keep in mind that we are talking about the most popular television shows; if we move down through the top 200 programs, we would expect to see even greater swings between the data sets."

"If we accept NMR's own assessment that LPMs are the 'gold standard' that provides the superior methodology, then, at least as it relates to African American viewers in highly populated African American metropolitan areas like Washington, D.C., the continued use of set meters may be severely over- reporting the results," said Frank Maggio, Chairman and Chief Executive Officer of erinMedia. "Can it really be, as Nielsen is implicitly suggesting, that a sponsor placing an advertisement to reach the African American community during a program such as 'Oprah' was actually getting almost 44% less of the audience, as shown by Nielsen's set meter/diary ratings? At a time when advertisers are re-evaluating their TV spending anyway, this discovery represents a potential loss in confidence that the medium can ill afford."

The other logical possibility supported by the analysis, Maggio noted, is that both of the NMR methodologies cited in its June 23 release are fundamentally flawed. "And remember, the data NMR released on June 23 are the data that it considers the most reassuring and helpful to its case. What does the rest of the data show? Only NMR knows for sure."

To address the massive discrepancies and statistical uncertainties exposed by NMR's own selective release of data, Maggio said, NMR needs to release all its dual panel data and allow advertisers, the television industry and expert observers to make up their own minds.

"With this new data revealing that one or both of NMR's metered methodologies are fatally flawed at accurately measuring the African American audience in our nation's capital, the only way for NMR to address the industry's concerns is to make ALL of its dual panel viewership data publicly available, immediately," Maggio said. "If it's self-regulation the industry wants, then that is the only solution. Without that level of openness and scientific scrutiny, we are essentially asking for government intervention."

Calling the limitations of NMR's data "startling," Maggio said the discrepancy provides timely and urgent evidence that a change to more inclusive methods of measurement must be immediately adopted by the industry.

"Given that the technology already exists to utilize digital set-top box data to directly measure the second-by-second viewing habits of more than 25 percent of the U.S. population, it's inexcusable that NMR is asking the industry to rely on its selective releases of data, which in turn are based on a miniscule sample of less than 7,800 households nationwide," Maggio said.

Based on this new NMR data, erinMedia and reactive cable network ReacTV also announced the launch of an ongoing "watchdog" campaign as part of their joint "WeShouldALLCount" initiative (see http://www.weshouldallcount.com/). The campaign will dig deeper into NMR's press releases, particularly as it relates to the ongoing LPM rollout across up to 210 designated market areas (DMAs). It will expose program ratings fluctuations and attempt to quantify the levels of advertising dollars potentially being misspent on viewers previously counted by NMR but later found, by NMR's newest and "best" technology, not actually to be watching. The site will also call on the industry to look beyond NMR's headlines and spin, while considering a significantly more accurate and sophisticated ratings system.

On June 16, erinMedia and ReacTV filed a landmark antitrust lawsuit against Nielsen Media Research. In their complaint, the companies cited NMR's practice of using long-term staggered contracts with broadcast networks and its attempted or announced acquisition of potential competitors, such as erinMedia or AudioAudit, respectively. For more information about this lawsuit, please visit http://www.weshouldallcount.com/.

About erinMedia:

erinMedia, LLC is a privately held media research company focused on understanding television audience behavior in all its forms -- cable, broadcast, satellite, linear, non-linear, and interactive. erinMedia's methodology combines privacy-compliant tuning data collected from advanced set-top boxes with independently gathered demographic data to produce insights into audience composition and viewing behavior. With over 135 patent-pending claims, erinMedia's proprietary technology produces accurate and timely analysis without the use of personally identifiable information. erinMedia brings experience and expertise in mathematical analysis, statistical modeling, and software development to the science of television audience research. For more information, please visit http://www.erinmedia.net/ or call 941.739.4500. erinMedia is an entity of The Frank Maggio Companies. For more information on The Frank Maggio Companies, please visit http://www.frankmaggio.com/.

About ReacTV:

ReacTV is the America's first creative media production company to develop reactive television programming. ReacTV's patented technology enables interactive gaming networks, where viewers at home compete for prizes, simply by watching and reacting to their TVs. ReacTV has exclusively licensed the patented CRAV Reactive Ad System, which allows mass media broadcasters, publishers and sponsors to create a new level of consumer awareness through reactive advertising. ReacTV is owned and founded by new media and real estate magnate Frank Maggio. For more information, please visit http://www.reactv.com/ or call 727.828.3688.

REPORT #1

Top 30 Programs Washington DC Among African American Households April 28, 2005 to May 25, 2005 Mon-Sun (7:00a-1:00a) As Reported by Nielsen Media Research, June 23, 2005

erinMedia Analysis of NMR Data: Query: Were the Top 30 Set Meter Programs being Under-reported or

Over-reported vs. LPM? Note: Under-reported programs show a positive %, and over-reported show as a

negative %

Set Set LPM LPM Meter RATING Meter Rank Program RATING RATING +/- Rank

2 American Idol-TU-FOX 23.3 26.4 -13.30% 1 1 American Idol-WE-FOX 27.3 25.4 6.96% 2 4 House-FOX 18.4 25.3 -37.50% 3 11 CSI-CBS 12.2 23.0 -88.52% 4 20 Without a Trace-CBS 10.4 19.4 -86.54% 5 25 CSI Sunday Special-CBS 10.1 19.2 -90.10% 6 8 Law & Order:Crim Intent-CBS 13.3 18.0 -35.34% 7 24 CSI:Miami-CBS 10.2 17.8 -74.51% 8 6 Desperate Housewives-ABC 16.7 17.7 -5.99% 9 3 Primetime Live Spec (Am Idol) 19.2 17.4 9.38% 10 n/a CSI: Miami Sunday Spec.-CBS 9.8 * 17.3 -76.53% 11 n/a Extreme Home Ed.-8P-ABC 9.8 * 17.2 -75.51% 12 5 NBA Playoffs-ABC-SA3 17.6 16.9 3.98% 13 14 NBA Playoffs-ABC-SU1 11.6 16.4 -41.38% 14 10 Law&Order: Spec Vic Unit-NBC 12.3 15.7 -27.64% 15 9 Girlfriends-UPN 12.9 15.7 -21.71% 15 n/a Disney Special-ABC 9.8 * 15.6 -59.18% 17 17 Law & Ord: CI-NBC 11.4 15.0 -31.58% 18 n/a Wheel of Fortune 9.8 * 14.5 -47.96% 19 27 FOX5 News-TEN 10.0 14.5 -45.00% 19 30 Oprah Winfrey 9.8 14.1 -43.88% 21 n/a Crossing Jordan-NBC 9.8 * 14.1 -43.88% 21 19 Half & Half UPN 10.6 14.1 -33.02% 21 n/a Extreme Home Ed-5/22 9.8 * 13.9 -41.84% 24 n/a Jeopardy 9.8 * 13.8 -40.82% 25 n/a Cold Case-CBS 9.8 * 13.8 -40.82% 25 n/a Lost-ABC 9.8 * 13.4 -36.73% 27 16 NBA Playoffs-ABC-TH1 11.5 13.4 -16.52% 27 n/a Law & Order-SA-R-NBC 9.8 * 13.3 -35.71% 29 n/a FOX Morning News 9.8 * 13.3 -35.71% 29

Note: * Programs that do not qualify for Top 30 LPM programs are assigned a best case scenario Set Meter Rating of 9.8; actual results will likely illustrate higher levels of SMR over- reporting vs. LPM.

Rank the former Top 30 Set Meter Ranked Programs based on over/ under-reporting vs LPM

Set Set LPM LPM Meter RATING Meter Rank Program RATING RATING +/- Rank

25 CSI Sunday Special-CBS 10.1 19.2 -90.10% 6 11 CSI-CBS 12.2 23.0 -88.52% 4 20 Without a Trace-CBS 10.4 19.4 -86.54% 5 n/a CSI:Miami Sunday Spec.-CBS 9.8 * 17.3 -76.53% 11 n/a Extreme Home Ed.-8P-ABC 9.8 * 17.2 -75.51% 12 24 CSI: Miami-CBS 10.2 17.8 -74.51% 8 n/a Disney Special-ABC 9.8 * 15.6 -59.18% 17 n/a Wheel of Fortune 9.8 * 14.5 -47.96% 19 27 FOX5 News-TEN 10.0 14.5 -45.00% 19 30 Oprah Winfrey 9.8 14.1 -43.88% 21 n/a Crossing Jordan-NBC 9.8 * 14.1 -43.88% 21 n/a Extreme Home Ed-5/22 9.8 * 13.9 -41.84% 24 14 NBA Playoffs-ABC-SU1 11.6 16.4 -41.38% 14 n/a Jeopardy 9.8 * 13.8 -40.82% 25 n/a Cold Case-CBS 9.8 * 13.8 -40.82% 25 4 House-FOX 18.4 25.3 -37.50% 3 n/a Lost-ABC 9.8 * 13.4 -36.73% 27 n/a Law & Order-SA-R-NBC 9.8 * 13.3 -35.71% 29 n/a FOX Morning News 9.8 * 13.3 -35.71% 29 8 Law & Order: Crim Intent-CBS 13.3 18.0 -35.34% 7 19 Half & Half-UPN 10.6 14.1 -33.02% 21 17 Law & Ord: CI-NBC 11.4 15.0 -31.58% 18 10 Law&Order: Spec Vic Unit-NBC 12.3 15.7 -27.64% 15 9 Girlfriends-UPN 12.9 15.7 -21.71% 15 16 NBA Playoffs -ABC-TH1 11.5 13.4 -16.52% 27 2 American Idol TU-FOX 23.3 26.4 -13.30% 1 6 Desperate Housewives-ABC 16.7 17.7 -5.99% 9 5 NBA Playoffs-ABC-SA3 17.6 16.9 3.98% 13 1 American Idol-WE-FOX 27.3 25.4 6.96% 2 3 Primetime Live Spec (Am Idol) 19.2 17.4 9.38% 10

Findings: Number of Set Meter Top 30 Shows with higher LPM ratings: 3 Percent: 10.00% Number of Set Meter Top 30 Shows with lower LPM ratings: 27 Percent: 90.00% Avg. level of over-reporting for Top 30 SM shows w/ lower LPM ratings: -45.45% Avg. level of under-reporting for Top 30 SM shows w/ higher LPM ratings: 6.77% Avg. level of over or under reporting for ALL Top 30 SM Shows: -40.23%

Note: * Programs that do not qualify for Top 30 LPM programs are assigned a best case scenario Set Meter Rating of 9.8; actual results will likely illustrate higher levels of SMR over-reporting, and therefore the average level of overall over-reporting vs. LPM for Top #30 programs is likely to be greater than calculated.

Source: erinMedia, LLC

CONTACT: Kristin Celauro, for erinMedia, LLC, +1-212-850-5652

Web site: http://www.weshouldallcount.com/ http://www.erinmedia.net/ http://www.reactv.com/

------- Profile: Ent