Arbitron PPM overview tonight

[radi0chik]

Since I've seen quite a few discussions about PPMs coming into this market, here's a good opportunity to have your questions answered by "the guys" themselves.

Society of Broadcast Engineers Chapter 67's meeting tonight will be all about that.

Sam Brown of Arbitron will present a demo of the system, examples of the equipment, and plenty of time for Q & A, which should make for an informative and entertaining meeting. The meeting is entitled "Encoding Big D"

Apparently they're (CC) already running the "watermark" on Lonestar.

Details:

Dinner Begins: 6:00 P. M.
Meeting Starts at 7:00 P. M.

Crystals Pizza and Spaghetti
930 W. Airport Freeway
Irving, Texas
(Southside of Airport Freeway & MacArthur near Office Depot)

If you go, look me up. I'll be probably the only female there.

 

[VERITAS DE VOCE]

Hopefully, you won't be the only PERSON there. ;D

I know I'm interested.

 

[davideduardo]

Hopefully, you won't be the only PERSON there. ;D

I know I'm interested.

These presentations for SBE and station groups are technical and don't have to do with the survey itself. The encoder is explained and demonstrated, along with discussion of the method of detection and failsafes for encoding... if you are not encoded, you are as good as off the air to the survey.

 

[Paul Boomer]

How'd it go, radi0chik?

I really wanted to stop by (I'm mostly interested in the technical aspects), but had to work late. In addition, I enjoyed a three-hour drive home thanks to the wreck on 635.

 

[radi0chik]

It was informative in my opinion at least. Wasn't a LOT of technical stuff there; there were the obligatory "how can it tell if I'm sleeping so it wont log music then" questions. So they explained that those things have motion sensors in them, and if you're still for a long period of time, it doesn't log stuff.

Then came the various other questions about what you could or couldn't hook it up to - so that it'd move (err vibrate??) even if it wasn't attached to a human. Considering it was mostly geeks at the meeting, I'll leave it to your imagination how they discussed faking out the motion sensor.

 

[Robert Bass]

Motion sensors on PPM meters ???

Oh boy, another aspect of PPM that could potentially stir up controversy.

R

 

[davideduardo]

Motion sensors on PPM meters ???

Oh boy, another aspect of PPM that could potentially stir up controversy.

The motion senser simply detects, when the meter is undocked, whether the meter moves occasionally. An algorithm, applied to the data when downloaded to Arbitron, determines if the movement indicates carriage or not. It's been that way since the meter was invented more than a decade ago.

The meter does not have to be in constant motion. Just enough to indicate the person has it with them. For example, a woman who clipped it to their purse would have many hours durning the workday when it does not move at all. The meter itself makes no decisions and is always detecting. The software that processes the data at Arbitron decides what is valid and not valid in a set of MRC approved "edit rules" such as "do not tabulate detections after one hour following docking at nighte" or "accept all detections after 4 AM."

The meter has no GPS capability, so we don't know where it is although it knows when it is at home and when it is away based on proximity detection to the base.

 

[radi0chik]

And then there was the discussion of if a person left the Dallas market, for example, to go to Houston for the weekend: what would the PPM do?

Apparently it adds the Houston stations the person listened to, to the Dallas book, which I thought was interesting. It doesn't add it to the Houston book. If we were to have something like hurricane Rita again, where everyone in Houston came to Dallas, there'd be a LOT of Dallas stations showing up in the Houston book that timeframe.

 

[Robert Bass]

Motion sensors on PPM meters ???

Oh boy, another aspect of PPM that could potentially stir up controversy.

The motion senser simply detects, when the meter is undocked, whether the meter moves occasionally. An algorithm, applied to the data when downloaded to Arbitron, determines if the movement indicates carriage or not. It's been that way since the meter was invented more than a decade ago.

The meter does not have to be in constant motion. Just enough to indicate the person has it with them. For example, a woman who clipped it to their purse would have many hours durning the workday when it does not move at all. The meter itself makes no decisions and is always detecting. The software that processes the data at Arbitron decides what is valid and not valid in a set of MRC approved "edit rules" such as "do not tabulate detections after one hour following docking at nighte" or "accept all detections after 4 AM."

The meter has no GPS capability, so we don't know where it is although it knows when it is at home and when it is away based on proximity detection to the base.

Sorry David but that can be controversial. Motion sensors can be fooled, as Adam and Jamie showed us on Mythbusters. This is the first time I have ever heard the meters have motion sensors. Had I known this 10 years ago, I'd have said the same thing then as I am saying now.

R

 

[davideduardo]

Sorry David but that can be controversial. Motion sensors can be fooled, as Adam and Jamie showed us on Mythbusters. This is the first time I have ever heard the meters have motion sensors. Had I known this 10 years ago, I'd have said the same thing then as I am saying now.

Of all the aspects of the PPM, the motion sensing is one of those that has not been questioned. Since it is not required to have constant motion, and there is little for the panelist to gain by creating constant motion, why would anyone do this? The algorithms that interpret the motion sensing and determine usability of data would likely spot repetitive and unnatural motion in any such attempt, anyway.

 

[Robert Bass]

I don't know that someone would intentionally try to fool the motion sensor on a PPM meter (although stranger things have happened). But the fact remains that motion sensors are not foolproof, and even one instance of false data can skew things up.

One of the motion sensor myths Adam and Jamie's show visited, was the myth that motion sensors are not foolproof. In fact they are, and quite easily! So even if someone wearing a PPM doesn't intentionally try to fool the sensor, it can be done unknowingly.

R

 

[little1]

But there's obviously a difference between your typical motion sensor, and whatever the PPM has, right?

I mean, I'm pretty sure the PPM isn't going to flip my outside lights on for me if it senses somebody in the yard...

But okay, let's say in CAN be fooled...What is going to be gained by it? I guess if someone unethical got ahold of a PPM, they could rig it up to 'hear' the station they want to give credit to for all day, but Arbitron already questions 'unnatural' listening patterns, (Really> you listened from 6a-mid, 7 days straight/ REALLY?) seems like they would continue to do this under PPM...

 

[davideduardo]

But there's obviously a difference between your typical motion sensor, and whatever the PPM has, right?

I mean, I'm pretty sure the PPM isn't going to flip my outside lights on for me if it senses somebody in the yard...

But okay, let's say in CAN be fooled...What is going to be gained by it? I guess if someone unethical got ahold of a PPM, they could rig it up to 'hear' the station they want to give credit to for all day, but Arbitron already questions 'unnatural' listening patterns, (Really> you listened from 6a-mid, 7 days straight/ REALLY?) seems like they would continue to do this under PPM...

The motion sensor simply determines if the meter, when undocked, has moved. The circuit, I understand, is very much like those on laptop motherboards that determine motion and decide whether to park the hard drive to avoid damage. In the PPM, motion is recorded in internal memory, just as detections are, and Arbitron gets a daily dataset from each meter when it docks. The edit rules determine if the motion indicates usage. Edit rules also call into review unusual listening and the panelist is asked about it on the phone. Panelists have frequent contact with Arbitron. The meter also "talks" to the base wirelessly and if in range (at home) records data on in home or out of home for listening location information.

 

[radi0chik]

wirelessly? They never said anything about that last night. The only way they said it would "download" the data to the docking station/mothership was when it was actually docked at night for charging/uploading. I looked at one of the devices and there was no FCC type-acceptance sticker on the device. What band/freq(s) do these devices use?

 

[Robert Bass]

wirelessly? They never said anything about that last night. The only way they said it would "download" the data to the docking station/mothership was when it was actually docked at night for charging/uploading. I looked at one of the devices and there was no FCC type-acceptance sticker on the device. What band/freq(s) do these devices use?

Once again, David is wrong.

On the Arbitron website, it states:

The Household Hub
Data from each base unit are transmitted to the household data collector (or "hub") through the household wiring. No extra wiring is required. The hub receives data from all of the base units in the household and passes these data to the central computer system over the household telephone line. Both the base unit and the hub have been designed for ease of installation and ease of use by participants. The hub has an LCD screen for simple instructions and problem diagnosis.

R

 

[davideduardo]

wirelessly? They never said anything about that last night. The only way they said it would "download" the data to the docking station/mothership was when it was actually docked at night for charging/uploading. I looked at one of the devices and there was no FCC type-acceptance sticker on the device. What band/freq(s) do these devices use?

I did not say the data is transmitted wirelessly. The data is only sent from the dock.

However, the meter, when undocked, knows it it is in the home but not docked because it detects a wireless signal from the base, reception of which it records as part of the detection process. If the PPM can not "hear" the base, it is assumed it is "away from home."

I have no idea what frequencies the base uses to tell the PPM it is at home. The PPM itself receives the signal, so if there had to be an FCC sticker, it would be on the base. Each meter and base have unique IDs so that family members do not mistakenly put the wrong meter in a base, or carry sombody else's meter.

 

[davideduardo]

wirelessly? They never said anything about that last night. The only way they said it would "download" the data to the docking station/mothership was when it was actually docked at night for charging/uploading. I looked at one of the devices and there was no FCC type-acceptance sticker on the device. What band/freq(s) do these devices use?

Once again, David is wrong.

On the Arbitron website, it states:

The Household Hub
Data from each base unit are transmitted to the household data collector (or "hub") through the household wiring. No extra wiring is required. The hub receives data from all of the base units in the household and passes these data to the central computer system over the household telephone line. Both the base unit and the hub have been designed for ease of installation and ease of use by participants. The hub has an LCD screen for simple instructions and problem diagnosis.

R

Try re-reading, Robert. I am talking about the way the meter knows if it is at home and not away when not docked.

Data is only downloaded when each PPM is docked in its own base. However, to know if a PPM is at home, but undocked, the base communicates with the PPM wirelessly... reception of the base signal tells the meter to code each detection as "at home." If out of range, the PPM data will be considered away from home as it does not contain the "at home out of dock" signal.

Again, I am talking about how the PPM "knows" it is at home or away. I said nothing about data transmission (detections) being sent wirelessly to the dock. Those are two separate issues. I said, "The meter also "talks" to the base wirelessly and if in range (at home) records data on in home or out of home for listening location information" and I said that said communication is only to tell if the meter is at home and undocked. I said nothing about data uploads.

 

[radi0chik]

I did not say the data is transmitted wirelessly. The data is only sent from the dock.

However, the meter, when undocked, knows it it is in the home but not docked because it detects a wireless signal from the base, reception of which it records as part of the detection process. If the PPM can not "hear" the base, it is assumed it is "away from home."

I have no idea what frequencies the base uses to tell the PPM it is at home. The PPM itself receives the signal, so if there had to be an FCC sticker, it would be on the base. Each meter and base have unique IDs so that family members do not mistakenly put the wrong meter in a base, or carry sombody else's meter.

Thanks for the clarification. I didn't look that closely at the docking station to see if it had an FCC ID on it, besides I think it was a prototype.. the hole cut out for the 9 pin serial connector looked like a 9 year old had made it That particular unit may have been pre-type-acceptance or something.

So it sounds like the bases are sending out SOME kind of wireless signal to the units, I'd be curious as to how that works exactly - but what would the advantage of that be - whether you were "at home" or "at your office" for example?

 

[davideduardo]

At home and in car and at work and other are the 4 traditional listening locations. The PPM does not at present detect a difference between away at work, away at play and away in car. It will be able to distinguish, by sometime in '08, the difference between in car away and not in car away via a device placed in each person's car.

Knowing listening location is important to advertisers (in car is nearer to point of purchase, for example) and stations. Programmers know that behaviour is different in each situation, and knowing where listening happens is critical for designing everything from music sweeps and stopset placement to rotations.

 

[Robert Bass]

The only information I can find on Arbitron's website regarding PPM, pertaining to wireless transmissions, is the ability to wirelessly monitor internet streams. David, the way you wrote you statement could have been worded better.

R