Ten beats Nine with shows enjoying Tuesday ratings rise

even though charlie pickering refuses to grow sideburns, i want 7pm project to succeed, glad to see its on the up

Top Gear doesn’t have anywhere near 1.2 Million viewers.

Top Gear was on for all four time periods in only 66 of the homes which have Oztam ratings boxes.

If 3 people in each of those 66 metered homes were watching Top Gear, then the total guaranteed audience was 198 people. Extrapolated data is guesswork at best.

Oztam, Neilsen and McNair Anderson always admitted that the methodology and sampling variance affected numbers and when questioned directly admitted that (following from B&T:) “no one really knows how many people are actually watching anything. Based on the known demographics, we can only take a guess”.

Coles and McDonalds have what is known as “ASR” Accurate Statistical Reporting.
McD ASR shows how many Cheeseburgers (for example) were sold across all stores in Australia during the previous 24 hours. It is accurate because money was paid for each item. Each cheeseburger was registered. The category total for that item for the 24th Jan 2009 was 367,600 including 15,439 that were sold as part of a Happy Meal.

If Oztam did the data, it would only have boxes in 90 McDonalds stores. Oztam would show that, based on extrapolated data sampling of 90 stores, 943,700 cheeseburger were sold that day, when we know that it was closer to 367,000

No one can say that 1.2m people watched TopGear or any other program. The only thing that anyone can say for sure is that possibly 198 people were tuned in and registering their viewing via the Oztam system. But then again, if each of those metered houses only had 1 person in it, maybe there were only 66 people watching TopGear.

How many of those 66 people will go out and buy that nice new Audi they saw in the third break wedged between Harvey Norman and KFC.

Well said AJ. Similar to last week’s ep, it had been shown on SBS 3 times, and released in Australia on DVD, a yet over a million people out of the 5 metro’s still watched it??? Hard to believe. The Oztam stats are not fact, the more boxes that are put in homes = a more actuate reading, especially when we are using 400 odd boxes to represent 15 million people living in metro cities around Australia.

The Oztam Peoplemeter Set-top boxes report at 2am each day to the Tam Interface.

Wednesday at 2am (the morning after the Top Gear episode), 2300 meters reported.

Of the 2300 meters that were online, 66 of those showed that those households were tuned to Nine at 19:30, 19:45, 20:00, 20:15 and 20:30. The PSD database entry for those times, for that date, records that Nine was showing Top Gear.

The demographic data held by AC Neilsen and OzTam for all homes which have peoplemeters in them is accessed as part of the data transer compilation process.

It works out that based on the information collected at the time that the Oztam household interview was completed and any changes to the knowledgebase data (new baby, adult children moving back in with parents etc) from that they can tell the statistical breakdown and compositions of the 66 homes had that program on.

It can also tell you that in those 66 homes, based on what they know, 17 of those people were aged 54+ (or just under 17% of the viewing audience. ABS census data showsthat the Australian population aged 54+ is 3,959,392 (or around 5.41% of the Total Australian population)

Oztam extrapolates that data to show that their 11 people aged 54+ represent 14.6% of the population.

It’s also incorrect to assume that the 66 homes that actualy were watching Top Gear or the other 2234 homes that were watching something else or not watching at all represent the viweing habits of 21,431,781 people which make up the current estimated population of Australia or the 21,413,081 who actually get a TV signal.

http://www.maths.usyd.edu.au/
This email may contain confidential information frrom the statan lab statistical analysis at the Faculty of Mathematics at the University of Sydney t If you believe that you have received this email in error, please immediately delete this email and any attachments, and inform via http://www.maths.usyd.edu.au/reply e-mail.

Thanks Harry,

There’s a lot of other factors involved and litterally columns of numbers. Oztam also doesn’t like anyone over 85. the ABS counts everyone over 75 as being in the same group. (i.e they only have a general idea of how many people are aged 91 for example). It’s hard to explain in great depth.

I worked on sampling and population numbers for similar organisations in the US and Canada through the University of Minnesota.

Somehow Oztam (and they’re not the bad guys here – Nielsen, AGBMcNair, they all do the same thing), they can somehow extrapolate numbers and turn 11 people watching Danoz Direct Home shopping into 45,000 people – are they kidding? 45,000 people watching 125 minutes of Pro-Activ skin treatment and knives that can carve through bricks?

You said it Harry: “I don’t get it “. And that’s the point. No one is supposed to get it. As long as Toyota believe that 1.2 million people saw their Land Cruiser ad – that’s all that matters.

Oh what a feeling.

AJ & JG,

I think your missing the point here – there’ll never be a box in every home to accurately record the exact numbers of what people are watching (barring significant future developments). The industry needs a common currency so airtime can be priced, bought and sold efficiently and that is what Oztam provides. i don’t think anyone thinks the numbers are 100% accurate, however at this stage they are the best we’ve got and everyone agrees on the trading metric they provide. What alternative would you suggest? The networks, already cash strapped, fork out to put a box in every Australian home?