05/09/2013 12:53:20 if_in_doubt Well I definitely think it should be a lot bigger than 0.06% of the total number of households in the country. It can't seriously be only 1,000 houses can it? There's no way you could a good spread of demographics with a number that small. 1000 households is a big enough number to cover the nation as all the demographics are covered. A whole range of individual and household types are used as controls to ensure that the panel at all times represents the whole population.
05/09/2013 12:57:17 if_in_doubt Hill I'd be absolutely amazed if over one million tv's were tuned into the match last Sunday.
As far as the accuracy of the figures goes there's definitely scope for criticism of the system used. Does it include people watching online instead of on a tv? Or people watching through apps on a phone?
I've never met anybody who keeps a diary about what they watch, I'd guess the sample demographic used might need to be bigger, especially now with a variety of channels / cables subscriptions availableI think the system is fine. Read tamireland.ie for more info on how the system works.
1000 households is a big enough number to cover the nation as all the demographics are covered. A whole range of individual and household types are used as controls to ensure that the panel at all times represents the whole population.
Ormo we're a much more diverse country than ever before, tastes and preferences are changing constantly, especially when it comes to tv viewing when you factor in new digital packages, internet viewing, on the go apps, on demand tv and things like Netflix.
I just don't think 1,000 people would give a good sample base for the whole country. It works out at less than 40 homes per county. I genuinely don't believe that's a large enough base to get a wide and varied enough sample to accurately reflect viewing figures.
I think the system is fine. Read tamireland.ie for more info on how the system works.
You see from the off I think there is blatant flaw in the system, the sample is much too small.
The mathematical systems they use might make sense and might work well for what they do but I don't think measuring less than 0.1% of the population gives an accurate level of data.
05/09/2013 14:10:46 if_in_doubt Ormond we're a much more diverse country than ever before, tastes and preferences are changing constantly, especially when it comes to tv viewing when you factor in new digital packages, internet viewing, on the go apps, on demand tv and things like Netflix. You see from the off I think there is blatant flaw in the system, the sample is much too small.
The mathematical systems they use might make sense and might work well for what they do but I don't think measuring less than 0.1% of the population gives an accurate level of data. Fixed Your Post But the experts clearly know otherwise. They also do thousands of interviews to update the panel so that it is constantly changing to ensure it represents the population. How much of the population do you think should be accounted for to make a more accurate figure when the experts think otherwise?
I don't like the word "expert" Ormond. Im outdoors during the day and the "experts" recently told me there would be a heatwave on a particular week. I spent that week getting soaked. Just because someone is (or claims to be) an expert doesn't mean we should automatically assume they are correct in everything they say. All the same Id say the viewing figures were massive, and lets hope so, as it was a great advertisement for football.
But the experts clearly know otherwise. They also do thousands of interviews to update the panel so that it is constantly changing to ensure it represents the population. How much of the population do you think should be accounted for to make a more accurate figure when the experts think otherwise?
Experts aren't immune from criticism.
Experts aren't immune from being wrong either.
Do you honestly think that by monitoring 0.06% of households a good spread of what is a vast array of demographics is covered?
To be honest I would have thought you'd want to get at least around 5% to give a decent view of the population and viewing patterns and trends.
I mean the census doesn't just take sample of 30 odd houses per county to gather stats every few years does it?
05/09/2013 16:30:45 joncarter I don't like the word "expert" Ormond. Im outdoors during the day and the "experts" recently told me there would be a heatwave on a particular week. I spent that week getting soaked. Just because someone is (or claims to be) an expert doesn't mean we should automatically assume they are correct in everything they say. All the same Id say the viewing figures were massive, and lets hope so, as it was a great advertisement for football. No but people without proper knowledge on the situation cannot knock the people who have the knowledge. The people who carry out the stats and work everything out know what theyre doing and if they think the sample size is enough it clearly is as otherwise they would increase/decrease the sample size queried on their viewing habits
05/09/2013 20:48:36 if_in_doubt Experts aren't immune from criticism. Experts aren't immune from being wrong either Do you honestly think that by monitoring 0.06% of households a good spread of what is a vast array of demographics is covered? To be honest I would have thought you'd want to get at least around 5% to give a decent view of the population and viewing patterns and trends. I mean the census doesn't just take sample of 30 odd houses per county to gather stats every few years does it? The census something completely different. Experts are not immune to criticism but by monitoring a wide enough range of households of various different types is what works. By getting 5% of households requires 10 times the work and means there is more chance of things not working out and its easier to do and clearly works. If there was an issue with the sample size used the people in charge would change.
Decided to carry out my own survey yesterday evening asking around the locality who was at the All Ireland in '89 in attempt to guage just how many Antrim spectators were in attendance.
Now, in the interest of getting a good sample I decided that I would need to survey at least 370 people or, 0.06% of the Co Antrim population. Of the 370 people surveyed all 370 swore they were there that day with one guy in the pub adamant he was there twice so 371. This means that of the 80,000 patrons in Croke park that day, 618,333 of them were from Co Antrim.