Even before you get into the definition of a 'smartphone' it is

So, while they waited for the iPhone 4S, they bought Androids to keep them going?

Android has been increasing market share, which means that people are using more Android devices, not that Apple owners are scrapping their current devices and waiting for the iPhone 4S...

But, as others have said, the figures are spurious anyway.

And there are still a lot of people like my gf, who want a phone that they can talk on, send the odd SMS, recharge it once a week and DOESN'T have internet access, because that is an extra expense they don't want to have.

Why is that?

There's bugger all causality between the first and second halves of the statement, yet they are presented as though the first half implies the second.

I'm not a statistician, but it seems to me that we can reasonably infer that purchasing of smartphones causes ownership of smartphones. If it's known that X percent of the population doesn't own a smartphone, and if you are able to determine from sales figures which percentage of smartphone sales are going to brand new owners, then knowing how many smartphones are sold gives a directly causes a change to the percentage of people who now do and do not own such phones. Have I missed something*?


* I am open to the possibility, my maths is not good.

These reports are written for people with a reasonable knowledge of the problem and I don't think such a person would interpret it that way (I certain wouldn't). I think some (obvious) details and assumption are missing here (from that particular short sentense) for that same reason.

Maybe if it went like this it would be more clear to a lay person:
Smartphones made up 69.1% of sales over the 12 weeks and after taking into account other various factors* we estimate that 43.8% of the GB population now own a smartphone.
*e.g. ratio of smartphone sold to previous smartphone users and to user ugrading from a feature phone

I think you assume that Kantar don't know what they are doing and you have spotted and obvious error in their methodology.
On the other hand assume that DO know what they are doing and they just haven't bother with including some obvious details.

43.8% of the UK population (over 62m but we'll say 62,000,000 for sake of argument) is 27,156,000

That is an awful lot of people, especially when you consider that a physical survey said that approximately 11 million are in use, that is more than double that

Ofcom states that 27% of the adult population claim to own one
http://www.eyefortravel.com/news/europe ... s-research

The Ofcom survey also estimates that there is a greater than 1:1 ration of mobile phones to people
http://articles.timesofindia.indiatimes ... blackberry (another article that cites the same survey

Depending how you define an adult (say over 24 based on the demographic measurement) there are 12 million at a very generous guess under 18 (12 million being the total up to 24), that leaves 50 million over that age, now if 27% of this Ofcom survey own a phone then that means that 12.5 million adults (approximately) own a smartphone

that means every single boy and girl under this age owns a smartphonem I don't see many infants playing on their iPhone.........


Stats can be used to show anything, pulling a random survey or report from a company that does these sort of forecasts is pretty much a waste of time. If you want to know how badly wrong a forcast can be: Michael fish told the UK there wasn't a hurricane coming in October 1987, he was 100% wrong http://en.wikipedia.org/wiki/Great_Storm_of_1987

Justin Beiber walks into HMV and buys 62 million copies of his latest single

It CANNOT be inferred or implied as a result that the entire UK population bought it

Actually, I suspect far more "children" own smartphones than "adults" - especially if your idea of "adult" is the over 24!!

I know quite a few people that own several smart phones; it's not unusual for someone who uses a Blackberry for work to have an iPhone as a toy. This skews the results terribly; as does the fact that so many people have half a dozen Zombie phones in their sock draw...

Estimating share of population owning a smartphone at the current time is NOT a FORECAST. Forecast is predicting FUTURE. Kantar are estimating PRESENT. Fisher was predicting FUTURE. So your argument is pointless.

If you can't see a simple thing like this then I am not sure how much time I want to spend discussing statistics this with you.

As an aside, I was watching a show last night about whether we are going to have a harsh winter or not and they were interviewing the guy who uses weather predictions and trends to control stock retention and buying at Morrison's (obviously, we buy different food depending on the weather) . He stated - on national television no less - that he was pretty sure the east coast and Scotland would have snow in the final week of this month, that December would see wintry weather but no major snowfall and that we would get a big freezein the early new year.

He's either very brave, very stupid or should be earning much more than Morrison's are probably paying him.

Actually Mr Fish was 100% correct as it wasn't a hurricane.

Only because of the definition of a hurricane also refers to the humidity levels and region and behaviour, the winds were of hurricane level forces though

I was pulling 24 as that was the way the data was represented in the population breakdown by age/gender, the age group was 14-24 so even rounding up the data didn't match up



They are both estimating, unless you can survey every single person and find out what they do the numbers do not add up

I can (and do see) and with experience dealing with analytical data I'm perfectly aware that data can be represented to skew a perception to an incorrect conclusion. The surveys like this do not account for people that already own smartphones, buy replacement phones due to being lost/damaged or stolen or use multiple phones for various reasons such as work/personal. Correlation is not causation and you simply cannot attempt to read between the lines between these sorts of guesses. You can have statistical information based on the UK census that means sweet f.a. when you realise that survey size, proportion, region, how representative it is etc all discredit the data before they are even started. Show me a survey of greater than 10% and I'll begin to consider it even vaguely accurate, asking 1 million people only needs approx 394 people to be 95% accurate with 5% margin for error on the data, that can't be used as a definitive answer.

Predictions are forecasts, which in turn are estimates and estimates are at best educated guesses.

You seem to believe that whatever these places come out with is gospel, it's far from it. I think you will be hard pushed to find one place of these overpaid suits that saw the iPhone 4s coming in and didn't predict the iPhone 5, as a result the Apple stock dipped

This is what I have problem with. You haven't seen the report, you don't know methods applied, assumptions made and sources used but you know for sure they got it wrong and you got it right. I really find it amazing.

I don't take it as a gospel. I think they are more likely to be right than you. If Kantar was making useless market intel. reports they wouldn't be in business because nobody would be bying reports from them. But they are in business so it would seem that other bussineses see the value in the service they provide. But again, you think they are a fraud.

What research have you done to back up your numbers? And what are your numbers anyway? Let's see them...