FCC investigates PPM

[radioprofessor]

In my humble opinion this is overdue. Pittsburgh just blew up three urban stations, multiple Smooth Jazz, hispanic and other formats are going away because of the People Meter. Does this ratings parameter have flaws and does it hurt diversity through bias? These are big questions in my humble view. If this were just a diary or phone call-out the FCC would have little jurisdiction but the PPM is different since it uses signals on the government airwaves. The FCC is well within its right to ban the encoded signals until they are proven they don't harm the radio industry. The FCC has ultimate control on what happens with PPM because they have ultimate control over what is broadcast over public airwaves including encoded signals. While one can debate whether stations are truly free in a major market to not subscribe to arbitron or not encode, but in reality to not encode in protest is suicide. Arbitron is a monopoly and you must take their service to play in the big leagues in major markets in my view. Small markets have some other choices, but not major markets. The FCC in its letter of Inquiry is looking at using its power to stop encoding which it can do in the public interest. In a major market, shouldn't the service that decides your fate be accredited and approved by the FCC, before its encoded signal is allowed on public airwaves? Seems the FCC will make that decision soon.

 

[radiojomo]

The PPM station is obviously blowing up fine radio stations from the airwaves (Indie 103.1, FM Talk etc.) but you can't say it's a bad way to measure ratings. The PPM is more accurate than the Diary ratings system will ever be. While people say some people would not carry around the device, how many people carry around a diary with them with a pen in hand and list every station they flip it on to? When a PPM device is on, it records almost every second of radio listening that a person makes unlike the diary, where people usually record their favorite stations in every box. The PPM's challenge from the FCC will be worthy, but it will not pay off in the end because people only see flaws in the system when they are effected by them.

 

[Buckethead]

You miss a major point Jomo. Yes the PPM device accurately measures what a person is listening to better than the dairy. Nobody disputes that.
What is at issue is WHO is carrying the device, and is that a sample representative of all of the various subcultures within a metropolitan society.
The sample size is much much smaller than the diary was and therefore much less accurate in measuring the listening habits of the varied groups that make up the patchwork of a city's multi-ethnic and muti-cultural society.

Smaller subcultures within each ethnic group, whether they be within the Latin, African American, Caucasian, or Asian communities etc, will be less likely to get accurately measured when there is a much smaller sample size.

If you have a station that caters to a minority subculture (and yes your Indie 103 was catering to a minority subculture withing the Caucasian society) and you are hoping Arbitron will find your subculture within a metro area of 12 million people, the chances decrease exponentially when the sample size changes from 3000 to 1200. The mathematical odds are that a PPM device will land in the hands of the more mainstream groups withing the society, leaving the minority subcultures under represented.

Also the PPM panelists are more permanently fixed than the monthly diary samples and therefore will not change significantly from month to month. If your station doesn't show up on the PPM there's little hope you will get a more fortunate placement in the next months and therefore the decision to blowup a station come more quickly.

Also, with PPM every person in the household needs to participate and being that Arbitron was rushed to meet their internal goals they were
more motivated to sign up multi-person households rather than people who live alone. That too affects the type of listener they are measuring.

So yes on the surface the PPM is a highly accurate device, but the sampling method does not serve to accomplish the goal of accurately mearuring listening habits within a community. That is a diservice to the community and to the broadcasters.

 

[ChannelFlipper]

You miss a major point Jomo. Yes the PPM device accurately measures what a person is listening to better than the dairy. Nobody disputes that.
What is at issue is WHO is carrying the device, and is that a sample representative of all of the various subcultures within a metropolitan society.
The sample size is much much smaller than the diary was and therefore much less accurate in measuring the listening habits of the varied groups that make up the patchwork of a city's multi-ethnic and muti-cultural society.

Smaller subcultures within each ethnic group, whether they be within the Latin, African American, Caucasian, or Asian communities etc, will be less likely to get accurately measured when there is a much smaller sample size.

If you have a station that caters to a minority subculture (and yes your Indie 103 was catering to a minority subculture withing the Caucasian society) and you are hoping Arbitron will find your subculture within a metro area of 12 million people, the chances decrease exponentially when the sample size changes from 3000 to 1200. The mathematical odds are that a PPM device will land in the hands of the more mainstream groups withing the society, leaving the minority subcultures under represented.

.If you have a station that caters to a minority subculture (and yes your Indie 103 was catering to a minority subculture withing the Caucasian society) and you are hoping Arbitron will find your subculture within a metro area of 12 million people, the chances decrease exponentially when the sample size changes from 3000 to 1200. The mathematical odds are that a PPM device will land in the hands of the more mainstream groups withing the society, leaving the minority subcultures under represented.


And you have the detailed understanding of statistics to make this claim? If not, how many statisticians have you contacted.

The idea that this has anything to do with statistics is absurd on its face. What this is is a political response to a (demogaphic) factual issue. The fact is that "subcultures" by their very definition, are well, small, and not likely to be counted in high numbers in ANY sample. Don't need to be a math major to figure that one out. If you want your station to be counted in the PPM, try the approach that CBS used when they switched to AMP. Don't see any problems with minorities, young adults without landlines, and whatever else garbage excuse their coming up with at AMP.

If you cater to small audiences you will get a small PPM number. And Newsflash, people don't listen to urban and subcultural stations as much as they say they do. Sorry.

 

[Buckethead]

Yes I do.
Channel Flipper, let me put this in a way that is simpler to understand.

You have 1000 balloons taped to a large dartboard and they are all yellow except for 10, which are blue. (those are your listeners, for a station that would be profitable with a 1.0).

If you throw 3 darts, you have a 3% chance of hitting a blue balloon. If you throw one dart, you have a 1% chance of hitting a blue balloon.


The darts are 3 diaries as opposed to 1 PPM.

 

[K6JHU]

But you need to normalize for the fact that you have thrown 3 darts and not 1. Odds on each dart is still one out of 100 (except what makes it worse is when you throw the second dart, one balloon is already broken, which changes the odds). So until you have enough balloons broken to change the odds, the answer is still 1/100 no matter how many darts you throw.

 

[ChannelFlipper]

Yes I do.
Channel Flipper, let me put this in a way that is simpler to understand.

You have 1000 balloons taped to a large dartboard and they are all yellow except for 10, which are blue. (those are your listeners).

If you throw 3 darts, you have a 3% chance of hitting a blue balloon. If you throw one dart, you have a 1% chance of hitting a blue balloon.

Thank you for proving my point. Regardless of the sampling methodology used, the fact of the matter is you don't have nearly enough listeners to make a difference. You also inadvertently made another point. The skin color of the person, and the color of the balloons, have nothing to do with the sampling methodology.

The results have proven it out time and time again. Winners like KROQ, KIIS, KLSX(AMP format), KYSR (Rock format) have all been successfully maneuvered in the PPM world. KJLH and others have not, because they didn't have the listeners to begin with. What you had before was a mirage, based on what people wanted to say they listened to, not what they actually did listen to. Like all mirages, when the facts became known, all you had was a fist full of dirt.

I can't wait for the Obama administration FCC to start forcing recommending to Arbitron to oversample minority households until they get the "right" desired results sampling distribution.

 

[Buckethead]

For the sake of argument these three darts are being thrown simultaneously.

I am not referring to skin color AT ALL.

The 10 blue balloons represent the audience of a station that would be quite content with a 1.0 share.
However with 3 times less darts being thrown with PPM, you are 3 times less likely to hit a blue balloon.

Nothing can change that fact.

 

[radioprofessor]

The points above are interesting and why the FCC should use its ability to stop the use of encoders on the public airwaves until this matter is sorted out. This is well within the rights of the FCC, to limit what is broadcast over public airwaves (encoding), especially if it might hurt diversity and radio formats that serve the public interest. There are three issues at play in my humble view;

1. Is the PPM panel representative of the market? Arbitron says yes, MRC has agreed in just two markets. Until they agree in all markets we are not sure the panel is representative and therefore stations may be injured.

2. Is the panel large enough to allow smaller diverse formats representation? In the diary methodology the panel, if you will, changed each month meaning thousands more participated in surveys over a year. Over a four book period the smaller stations did show up in "the book" from time to time and therefore an audience level could be extraped. With the size of the PPM sample it is unlikely that smaller stations will show up. Size matters,in this case.

3. Cell-phones and 18-34. These demos have not been sufficiently addressed. In the diary methodology they were weighted. In this methodology that is not possible or desirable (partly due to small sample size compared to a four book in the diary). The undersampling creates severe issues especially in younger formats. The PPM panel must reflect the market or the panel is flawed.

Clearly the FCC needs to stop encoding thereby forcing arbitron to "get it right" and get accredited before releasing data that destroys stations, diversity and formats in my humble view. Would you allow a surgeon to cut you open, without a medical degree. This is an easy decision for the FCC and requires no court action. They have every right to institute a mandatory stop on encoding until arbitrion is accredited and the methodogy is fixed. Who does it hurt to wait? Stay with diaries, stations save money when they need it most and start the PPM when it passes the FCC parameters.

 

[davideduardo]

The sample size is much much smaller than the diary was and therefore much less accurate in measuring the listening habits of the varied groups that make up the patchwork of a city's multi-ethnic and muti-cultural society.

Actually, the daily, weekly and monthly sample is larger than in the diary survey. But that is not the real issue...

Smaller subcultures within each ethnic group, whether they be within the Latin, African American, Caucasian, or Asian communities etc, will be less likely to get accurately measured when there is a much smaller sample size.

Wrong. The diary was a random probability sample, and a new one each week. The PPM uses a panel, which is balanced to reflect the population in geder, age groups, ethnicity, geography, language usage, etc. A panel is, by definition, a miniature model of the universe, exact in every detail except size.

The mathematical odds are that a PPM device will land in the hands of the more mainstream groups withing the society, leaving the minority subcultures under represented.

In LA, the Hispanic part of the panel has been overrepresented, and had to be weighted down. A panel, again, is built by seeking out a group of households and household members that mirror the universe almost perfectly.

Also the PPM panelists are more permanently fixed than the monthly diary samples and therefore will not change significantly from month to month.

The diary survey does a weekly sample. And the PPM panel turns about 9% per month.

Also, with PPM every person in the household needs to participate and being that Arbitron was rushed to meet their internal goals they were
more motivated to sign up multi-person households rather than people who live alone. That too affects the type of listener they are measuring.

Arbitron needed to get a representative panel. Households not meeting sample goals were not invited to participate. And the PPM uses cell only households, often one or two people only.

So yes on the surface the PPM is a highly accurate device, but the sampling method does not serve to accomplish the goal of accurately mearuring listening habits within a community. That is a diservice to the community and to the broadcasters.

It's a panel, not a sample. A sample is used to build the panel, but a panel is as near 100% representative as is possible.

 

[davideduardo]

1. Is the PPM panel representative of the market? Arbitron says yes, MRC has agreed in just two markets. Until they agree in all markets we are not sure the panel is representative and therefore stations may be injured.

The MRC looks at methodology and implementation. They have not said the lack of accreditation has to do with the panel being representative, nor have they said it is not. There is more to it than just the panel.

. Is the panel large enough to allow smaller diverse formats representation? In the diary methodology the panel, if you will, changed each month meaning thousands more participated in surveys over a year. Over a four book period the smaller stations did show up in "the book" from time to time and therefore an audience level could be extraped. With the size of the PPM sample it is unlikely that smaller stations will show up. Size matters,in this case.

More stations, many more, "make the book" in PPM than in the diary. That is because the PPM picks up light and occasional listening and hearing, and registers it.

3. Cell-phones and 18-34. These demos have not been sufficiently addressed. In the diary methodology they were weighted. In this methodology that is not possible or desirable (partly due to small sample size compared to a four book in the diary). The undersampling creates severe issues especially in younger formats. The PPM panel must reflect the market or the panel is flawed.

Both cell only and 18-34 are within the targets, and any stratification variable that is under or over proportionality is weighted up or down.

Clearly the FCC needs to stop encoding thereby forcing arbitron to "get it right" and get accredited before releasing data that destroys stations, diversity and formats in my humble view.

The FCC has asked for comments, in what is called an Inquiry. The deadline is 30 days, I believe. Until then, there will be no action at all.

Arbitron has not been accredited for 10 years in onbe top-15 market... there is and never has been a requirement for accreditation, just a desire for the guarantees it offers.

There is even an issue of whether the FCC has jurisdiction over this matter.

Do I think the PPM is perfect, or as good as it could be? No. But it will not be fixed by stopping it, as Arbitron can not afford to do many markets with no subscribers while they improve it... much of the improvement requires fine tuning as the survey is being done, not in preparation for startup.

 

[Buckethead]

David, I use the word sample the same as panel. That's beside the question.

What's debatable here is whether having less PPM's in tab than were used in the diary method makes the panel less likely to find the niche format listeners. I say that having fewer people in tab on a panel makes the monitoring of the fringe formats less accurate.

By fringe formats I mean formats that KNOW they have less listeners than a top 40 station and are ok with that but just would like to be showing up on the PPM where they know they have listeners and Arbitron tells them they have none.

Clearly there is an issue here or the FCC wouldn't be investigating it.

 

[davideduardo]

David, I use the word sample the same as panel. That's beside th equestion.

What's debatable here is whether having less PPM's in tab then were used in the diary method makes the panel less likely to find the niche format listeners.

A properly proportional panel should, with far fewer participants, find everything of significance and be quite stable in the reporting of them.

Remember, the weekly in tab for the diary was a hair over 600; the proportionality and balance in each week was all over the place. Until you got month's worth, there was no reliability, and the wobbles were far greater than in the PPM. There are other issues, like language enumeration by proxy and so on that are bing worked on or getting fixed... nothing like this starts perfectly, but the key is seeing how consistently even the low rated stations perform.

 

[Buckethead]

Should is a key word. My understanding is that this issue is less about race and more about diversity of the airwaves.
We understand that there are many diverse subgroups within each ethnic group and some stations are not shooting for the lowest common denominator. They need to be given a chance to survive.

The statement about the stability of smaller stations makes sense if the panels were changing frequently but they are not.

 

[nmoore6676]

It all comes down to confidence. Much like the banking system that has been in the news a lot lately. We all know that the money we deposit is the money banks use to create loans and there is not at any time funds representing the total value of the depositors' accounts. It works so long as we believe that at any time we can access any or all of our individual account balance. When we don't then we get an IndyMac situation.

So getting back to ppm or the diary. So long as whatever method has credibility with those who use it (advertisers and programmers) then it works. If the confidence is broken and people quit using it then it goes away. Statistical analysis is always a point of debate because those who like the results will praise the method. Those who don't like the outcome will be critical. There are many ways of selecting representative groups and in one way or another we are all exposed to polls and surveys every day. Politicians rely on them, social scientists and other researchers use them.

My own ax to grind is that in either method of ratings my group (the aging WASP population) is under represented so as I get older the radio programming drifts away from what I like. Is that fair, probably not but I am learning to live with it. The same goes for any other person or group like the Indie 103 fans so unless we have the money to buy a station and program it without being concerned about profits that is how it is going to be.

I do not like the government getting involved in dictating content by any means. It takes away the freedom of speech first amendment rights and even if we want it to protect a certain format that caters to our tastes it can has undesirable unintended results and needs to be approached very cautiously.

 

[Buckethead]

The government is not dictating content. It is looking out for the little guy and making sure that he is getting a fair shake.
Is that anti free speech?
No, the point is to ensure diversity.

 

[nmoore6676]

The government is not dictating content. It is looking out for the little guy and making sure that he is getting a fair shake.
Is that anti free speech?
No, the point is to ensure diversity.

If the government intervenes in any way to regulate content or who is allowed to broadcast that is nothing less than dictating content. How could you call it anything else?

The little guy has as much chance as anyone, at least before Clear Channel became the largest owner of broadcast stations. However the government allowed that to happen with only mild grumbling from the public and broadcast industry people. Now if they try to undo that by regulation it would amount to seizure of property which our founding fathers definitely frowned upon.

The Arbitron and other ratings systems are private businesses and are used by other private businesses. Unless we want to turn this into the United Socialist States of America let us leave it that way. Let the users decide if they like the system and that is the way free enterprise is supposed to work.

 

[davideduardo]

My own ax to grind is that in either method of ratings my group (the aging WASP population) is under represented so as I get older the radio programming drifts away from what I like.

That's not true at all. 55-64 and 65+ are proportional to their percentage in the population in the PPM panel, and are either proportional or weighted to proportionality in the diary markets.

But stations look at 18-49 and 25-54 and their subsets almost exclusively because advertisers look at that .... they don't look at 55+ because there is limited or no interest in advertising to that demos. The numbers and tables are there... but there is no ad money.

 

[davideduardo]

Should is a key word. My understanding is that this issue is less about race and more about diversity of the airwaves.
We understand that there are many diverse subgroups within each ethnic group and some stations are not shooting for the lowest common denominator. They need to be given a chance to survive.

The only issues are about ethnic group representation and the effecct on stations that program for them. There is absolutely no issue about "protecting" formats or music genres. And the issue is not about "race" since, while one ethnic group is also a race, the other, Hispanic, is not a race but a cultural grouping.

There have been no issues about one or another Black or Spanish language format being "hurt" because all the comments about potential issues have been about ethnic groups as a whole (and the only ones that receive ethnic controls are Black and Hispanic).

And there has been concern from NABOB and SRA about measurement of Blacks and Hispanics, but there have been no comments about small or niche stations being "hurt" by the PPM because, if there is a proportional sample all stations will have equal opportunity to demonstrate their audience.

The only way to improve accuracy is by adding sample... the PPM cost 60% more than the diary, and to improve accuracy by reducing by one half of a standard error, the sample would have to be quadrupled.... and no station could afford that.

So there is little that can be done, other than slowly improving the techniques employed to recruit panelists, etc. We have had the diary since 1965, and it was being improved constantly still... the PPM is new, so it will also improve over time.

 

[Buckethead]

"The only way to improve accuracy is by adding sample... the PPM cost 60% more than the diary, and to improve accuracy by reducing by one half of a standard error, the sample would have to be quadrupled.... and no station could afford that. "


So we agree David that the accuracy can be improved by a larger sample.

PPM doesn't cost Arbitron 60% more, it costs the stations 60% more. For that much of a rate hike you would think Arbitron would take the extra steps to ensure accuracy for ALL it's clients.