DriverPower survey

[Dieselman]

Interesting reading this year.

The Top 100 | Driver Power | Car Reviews | Auto Express

 

[finisterre]

Is it marked by boringness?

 

[Dieselman]

Is it marked by boringness?

You obviously couldn't be bothered to read the categories before replying.
Reliability, build quality, practicality, ease of driving, performance, comfort, handling, braking, cost to run, braking, ride quality.

I don't see boringness there...

 

[corned]

Interesting!

I am more than somewhat surprised that there is no category for estate cars, given their relative prevalence on today's roads.

And it is interesting that the survey appears to support the thinking behind my next car choice.

It is also interesting that none of the cars mentioned in 'Sports Cars' are what I regard as sports cars...

 

[Adam230K]

reading that article makes me feel smug for buying my mum a Honda Jazz...

 

[verytalldave]

I am firmly of the opinion that real top-quality cars like Mercedes will never do well in surveys such as these purely because typical owners cant be bothered to do these type of surveys. They generally have better things to do with their time than fill in meaningless surveys.
If that survey is to be believed, then 75% of us should be driving around in Skodas.
Skodas may very well be alot better than they used to be.
Then again, that's not exactly difficult. Only one step up from Wartburgs and Yugos.
It wouldnt worry me if Mercedes came bottom.
I KNOW what I would rather drive - and why.

 

[st13phil]

^ Some valid points from VTD.

I always read surveys like this with a slightly jaundiced eye, not least because the respondents are self selecting. Want to justify your purchase in the eyes of others? Fill in a survey and give it a glowing report. Angry with your purchase for whatever reason? Fill in a survey and tell everyone it's rubbish...

That said, it's gratifying to see that BMW manage a best of 35th place

 

[Dieselman]

I always read surveys like this with a slightly jaundiced eye, not least because the respondents are self selecting. Want to justify your purchase in the eyes of others? Fill in a survey and give it a glowing report. Angry with your purchase for whatever reason? Fill in a survey and tell everyone it's rubbish...

So all of one manufacturers are happy, so give glowing reports and another manufacturers are unhappy so give a lower score.

Errm.. isn't that the idea of the survey?

As it's based on average scores it doesn't matter that more of one manufacturers customers report or not...that's how averages work..

 

[st13phil]

As it's based on average scores it doesn't matter that more of one manufacturers customers report or not...that's how averages work..

Eh? The whole point is that it isn't an average it's a statistically biased result precisely because the sample is self-selecting

 

[bpsorrel]

Good to see Mercedes beating it's rivals in all segments! BMW haven't done well have they!

And, oddly neither have Lexus/Toyota, who normally lead these things...

But, IMO, surveys are fairly meaningless as only those who can be bothered do them. The vast majority of happy owners just don't think about it.

 

[Dieselman]

Eh? The whole point is that it isn't an average it's a statistically biased result precisely because the sample is self-selecting

And as a result is open to all comers.

If there isn't a sufficient sample of a model it doesn't get included, so seemingly enough (in this case) Mercedes customers did complete the survey..as we can see..

It appears that Jaguar customers also completed the survey...maybe Jag isn't a luxury brand...

 

[Charles Morgan]

Statistically all of this is pretty much garbage. 23,000 self selecting respondents (so most probably either very happy or very ****** off, the rest not caring) giving opinions on 100s of cars means that unless they happen to be owners of multiple cars, each make has a tiny sub-sample of voters. Which leads on to the next problem that is the means by which you measure satisfaction - it is entirely relative. So if you have had a crappy car before, that is the measure by which you judge the next. If you have only ever dealt with a Mercedes dealer, a Skoda one might come across as helpful and competent etc.

However, given their current build quality and ride I would actively consider owning a Skoda Superb if I had need for a new large estate without the eyewatering price of a Merc or BMW.

 

[Palfrem]

Perhaps on the recent Census they should have asked in a bit more detail about our cars and how happy with them we were?

 

[Dieselman]

Statistically all of this is pretty much garbage. 23,000 self selecting respondents (so most probably either very happy or very ****** off, the rest not caring) giving opinions on 100s of cars means that unless they happen to be owners of multiple cars,

No it doesn't.

The questions are specifically geared to elicit factual answers, not opinion.

When asked has your car broken down in the last 12 months there is only two possible answers, then when asked how many times, only a set number again, then when asked the nature of the fault from specific categories and failure types, again only certain answers.

This isn't just a load of people in a Pub bar having an argument about their car being better because they think so, it is very specific.

As usual, we are playing the game of not letting facts spoil a good story..

I didn't complete this years survey but when I did I listed every fault the car in question had had in it's warranty life.
BOTH faults were very minor and sorted at service time...subsequent re-occurrences and new faults are so far NIL....in 7 years..

Maybe that's why VTD says Mercedes owners won't do such surveys...they haven't got the time to list all the faults and failings...

 

[st13phil]

And as a result is open to all comers.

While the survey is open to all comers as you put it, Charles describes exactly why only a small subset of people who have a particular axe to grind actually do participate. That's what's called "self selection bias" and is why the survey is statistically worthless, it's nothing to do with whether or not factual questions are asked, it's all about whether or not a particular owner chooses to respond.

 

[Dieselman]

While the survey is open to all comers as you put it, Charles describes exactly why only a small subset of people who have a particular axe to grind actually do participate. That's what's called "self selection bias" and is why the survey is statistically worthless, it's nothing to do with whether or not factual questions are asked, it's all about whether or not a particular owner chooses to respond.

But we can see that all the models in the survey had enough respondents to be included.
From what you are saying, Mercedes drivers have an axe to grind....why would they do that if they are happy with the car...???

So it appears that some people like their cars and score them highly..because they are pleased..and others score theirs low because they aren't pleased.

Isn't that the whole point of the survey...to find out what people think of their car...?

 

[rhud]

I think the Which? surveys - and road tests - are more scientific and thorough - and there can be no suspicion that they are tainted by commercial considerations.

 

[Sp!ke]

I think the Which? surveys - and road tests - are more scientific and thorough - and there can be no suspicion that they are tainted by commercial considerations.

What?

From experience which usually get everything totally about face. I dont know if sweetners are involved but something is most definitely not right about most of their content.

 

[st13phil]

From what you are saying, Mercedes drivers have an axe to grind....why would they do that if they are happy with the car...???

You've perfectly demonstrated my point. Someone who is particularly satisfied or dissatisfied is much more likely to respond to a survey like this than they would if their view was less polarised. Add in the fact that someone who buys a premium brand probably has a higher expectation level than someone who buys a "value for money" brand and you have created the perfect scenario for statistical bias.

 

[Mr E]

What does the %age score represent? Percentage of what? How are a wide variety of quantative answers formulated into a single score?

I actually find it more interesting that the "Top 100" cars are divided only by 16%, and the Top 10 by 3.6% - but without knowing what the percentage means, then it's meaningless.