Sony Tops Online Poll of ‘Best Brands’

A few problems with this from a statistical point of view.

1. Online surveys are open to manipulation at collection and are biased towards the demographic that has access to the internet.

2. With this is mind, note that the people questioned were from the U.S. which may not be representative of other populations.

3. 2,351 adults may seem like a lot but results can be distorted, generally the bigger the better (e.g. 10,000 is a nice number).

4. Stats can tell you anything you want to know, it depends on the tests you use and how stringent they are (generally the cut off point is 5% (1 in 20)
that the results aren't due to chance).

Being the most recognised brand isn't necessarily a good thing (would anyone consider Enron for their energy needs these days?) but I don't doubt that Sony is up there. I am especially impressed that they beat out Coke.

I am suprised that MS didn't get a look in given the ubiquitous nature of their software.

Sony have lost every single major consumer electronics market other than videogames to other manufacturers. PS2 is notorious for it's poor build quality. Something smells a bit whiffy to me.

It depends what you mean by "lost". Do you mean they have been forced to abandon those markets altogether in light of overwhelming dominance by other companies... no. What you mean is they are not number one in those markets. But they are top ten in nearly every market...



Notorious mainly amongst trolls, Nintendo/Sega/Microsoft fanboys and people who whing a lot in Internet forums. Your average consumer thinks they are fine.

In sales, maybe, but the products just don't stack up. VAIOs are completely crap PCs; theres about a million mp3 makers you'd consider before even thinking of getting a Sony; aside from Bravia (which is ridiculously overpriced) their TVs are pretty outmatched by those from Toshiba and LG, nevermind the fact that they're not supporting SED. They're not the top or even close 2nd manufacturer in any market other than videogames, so either all the people that voted are fanboys or something dodgy is going on.




Utter bull. Look at any single stress test made for all 3 consoles, PS2 always dies out several rounds before GC or Xbox. I've seen much more anecdotal evidence showing that you have to replace the system much more often, too.

If we're talking anecdotal then I've seen many more stories about having to replace your 360 due to it locking up and dying than I have about PS2.

I have seen one story of a guy who is on his seventh 360. In a maximum of nine months.

Well obviously in recent times we're going to hear more about 360s breaking, but when PS2 launched there was a lot of press coverage of PS2s breaking.

Hell, Sony had a class-action lawsuit filed against them for their dodgy systems. Anyone who thinks PS2s are just as reliable as any other system is an idiot.

That's a reasonable and mature response. The only problem is that a respected opinion survey company has just completed a major poll that indicates that, in fact, most people think Sony is the top brand, above Microsoft, above Nintendo, in fact above any of the ones you mention.

They have empirical evidence, and you have anecdotal.

And despite Thane Jaw's asserion that 10,000 is a "nice number", 2,300 is a statistically reliable sample, and considerably larger than the ones that are often quoted in the press.

How on earth is 2300 reliable thats not even .01% of the planets population !!

I suggest you go read a textbook on statistical analysis, and specifically 'poll sample sizes', 'confidence intervals' and 'confidence levels' before you go posting about something you clearly do not understand.

While you are at it, you may also like to get a textbook on grammar. The apostrophe is used both to indicate possession and also contraction. A question mark is traditionally used at the end of a question.

The prescence of Kraft Foods in this poll leads us to assume this was a US poll, since that company's market presence and reputation is not so great in other territories.

On a population of 295,734,134 (The US population) to achieve a confidence interval of 2, you need to poll 2,400 people. So it seems that Harris was aiming for almost exactly that confidence interval.

A confidence interval of 2 means that in a poll of 2,400 people, if 50% choose a given answer, if you polled the entire population, between 48 and 52% would give the same answer.

A confidence interval of 1 would require approximately 9,000 respondents.

The confidence level is the rate of assumption you make of the conformance of the general population to the confidence interval. If you wanted to use a higher confidence level, you would require a larger sample, a 100% confidence level, combined with a 0 confidence interval would require you to interview every memeber of the public. Clearly this would be impractical.

Most researchers use a 95% confidence level, and that was why I said that 2,358 (or whatever) was a statistically significant sample.

And when PS3 launches we'll hear about it breaking, we'll probably hear some stories about Wii breaking too. First run hardware has faults. Fact! PS2s no longer have read errors, XBox 360s will not overheat in future and by March or April there will be very few stories about PS3s and Wiis failing.

Stop taking a single moment in time and using it to define your whole world view.


Well, thanks for that, so I'm an idiot. I'm obviously such an idiot that I think that my bought on day one PS2 was used as the SPOnG test machine for two years without any issues and only started having problems this time last year. Thanks for correcting me on that, it's obvious now you point it out that I've actually had three or four PS2s since the launch.

In fairness I have to point out that our test 360 has had no problems and our day one Japanese PSP has had no defects with either the square button or the drive bay. Maybe we are just a lot more careful with delicate electronics than the average person?

No! NO. No.

We're just SOOOOO not!

Damn! I missed out the <irony> tags!