Warranty Direct which insures the warranty on 30,000 cars per annum, have just published their findings over the past 12 months. Presumably this company would have no personal axe to grind against any particular manufacturer, so their results should be totally unbiased. They list Jeep as the worst manufacturer, and surprisingly Suburu as the second worse, followed less surprisingly by Land Rover and Alfa Romeo. Full details are published on their website :-
www.reliabilityindex.co.uk
|
Dude,
We chatted about this in Feb and my last words then can be my first this time...
You have to be so careful with statistics, particularly where the sample may be biased.
These results are not a reliability index for any particular car as such, they are a reliability index for a particular car that is covered by this warranty and relying on the claim accuracy from the garages.
It is a little like taking the results from Mondeos registered only in London or Xantias owned only by Vicars.
You get these stupid stats from time to time that, for example, folks who drink two glasses of wine a day are 5x less likely to suffer a heart attack than someone who doesn't have any. The next week it'll be those who drink more than one glass of wine a day are 5x *more* likely to suffer a heart attack....it's all in the sample profile.
Statistics can mean everything and nothing.
I must just add that in their 4x4 comments they refer to the poor showing of Land Rovers. So do they say avoid...like hell they say buy our warranty for peace of mind. What at £650 for two years on a mid 1990s Disco TDi? Better to spend that loot when you buy it on a major inspection service, timing belt, brake fluid/coolant/transmission fluids change etc.
After all that warranty certificate doesn't add one bit to the vehicle condition.
MM
|
Yup if you read the explanatory text it mentions the Subaru, suggesting it does not often go wrong but when it does it?s lots of dosh, look at the average cost of repair. This skews the impression a tad. It?s important to read the small print carefully to interpret properly.
I'm in South Africa and came at this site looking for reliability data on 4x4s as here there are no empirical dependability data available to the public here....indeed I understand certain manufacturers have squashed effort to get this going. Our motoring press are sickeningly sycophantic towards manufacturers.
We have to rely on anecdotes which can be a little unscientific.
The thing to do is to triangulate - that is to say don't rely on just one survey, see if you can spot patterns by looking at other surveys. ?Which? magazine and JD Powers spring to mind. Patterns then do emerge....Land Rover consistently being a total lemon is what springs to mind and this is congruant with my personal experience and a plethora of anecdotes I?ve heard.
|
Yes Anata, triangulation. Good point. Great word.
I think the JD Power survey is published in Which, to make it just one survey. Top Gear and Auto Express do surveys too. If you look at all these patterns do emerge.
For example Japanese manufacturers tend to do best. Of these Toyota/Lexus and Honda do consistently very well. Suzuki/Daihatsu not so good.
French and Italian cars tend to perform badly as fussy owners object to things falling off cars which won't start.
|
I think the JD Power survey is published in Which
It might be, but Which also carry out their own very detailed survey of their own readership and publish the results. I know, because I have filled in the forms!
Andy
|
Thanks Andy I haven't seen this one.
Any idea when it's published? I do like my reliability surveys.
|
|
|
JD Power are obviously in bed with ?Which? with regards owner surveys on UK data. Following on from my earlier point in this thread about triangulation it's worth expanding the data sets and looking at JD Power's US based surveys for models that are similar in the US and UK, an ever increasing trend as Detroit rusts away a bit like Layland of yore.
Of course the US has nearly 5 times the UK population so the data sets from which the results are collated will tend to be bigger and US climate covers a much wider extreme than Blighty.
Oh they still rate Discos as awful (Defenders are not sold in US anymore)!
Good source of JD Powers data in the States is the New York times...
www.nytimes.com/pages/automobiles/index.html
|
|
|
|
|
The scores for my car in the Autoexpress and JD power surveys are quite divergent. In JDP it is very highly rated. In AE it is near the bottom. How can this be of the surveys are as good as claimed by the people concerned. I've always said these surveys are of little value. Curiously the link given in this thread gives my car a brilliant rating. So I'll take that one then and feel smug.
|
It is important to realise that the JD power survey is not a reliability survey. It contains information about vehicle quality, but I know in the 2001 survey these amounted to only 34% of the overall rating of a vehicle. 30% was allocated to vehicle appeal and 17% on dealers and 19% on ownership costs.
Such information is qualitative, not quantitative, but it aims answer the question: "Did the vehicle live up to your expectations?"
|
|
|
|
|
|
Statistics are a very useful tool. As long as you know how the data was collated and it serves your needs, you can extract useful information for your benefit. If you do not believe in or have confidence in particular sets of data, you can of course ignore them.
I like all these car reliability surveys & stats; and do base my decisions on them.
|
There can only be one truly reliable way to gather true stats on vehicle reliability, and that is to take the owner/driver out of the loop. The info needs to come from a true sample of vehicles in all types of ownership across a particular type. And a return must be made for every vehicle, not just those where the user can choose to respond or not.
If I'm correct the Which sample is made up of their own subscribers. So that's a narrow group of people who are minded to pay for consumer advice and may have a highly biased view of their vehicle. Then there is the variable that they are not forced to respond, they chose to. Finally it relies on the user remembering the truth assuming the garage related that to them in the first place. Add in the possible error made by garages in the reasons for repair and the margin for error is huge.
If I remember WD rates Nissan and Saab (I'm ure they were the two I noticed) as far less than perfect... yet Which has them as the tops. And you'd tell me both surveys are unbiased and of a reasonable sample...so what gives.
Quite seriously hands on experiance of a detailed vehicle database covering the past 8yrs shows real life trends that vary a great deal from the picture painted by these simplistic survey tables.
In any case I can't think of a vehicle type (well perhaps one but there's some happy owner here just bought a new one so I'm not going there) where the repairs have ever been very significant beside routine maintenance and depreciation. In fact depreciation of a new to four year old car will cost far more than repairs on even an unreliable older one.
What do they say...there are statistics and then there are damn penguins!
MM
|
>What do they say...there are statistics and then there are damn penguins!
>MM
What I want to know has nothing to do with motoring at all, but please help me out - I used to know who it was that said "Lies, damned lies, and statistics", but have since forgotten. Now, every time I see that phrase it drives me crazy trying to remember who came up with it. Please, someone put me out of my misery!
|
Don't the Americans think it was Mark Twain and the British say Benjamin Disraeli?
I thought it was Disraeli but I'm sure someone clever will tell us.
MM
|
Don't the Americans think it was Mark Twain and the British say Benjamin Disraeli? I thought it was Disraeli but I'm sure someone clever will tell us. MM
I'm not clever; I've just looked it up in the Penguin Dictionary of Quotations:
"There are three kinds of lies - lies, damned lies and statistics" - Mark Twain, autobiography, part 5, chapter 1.
So there.
|
PS how appropriate that you got that from the PENGUIN book of quotes.
;)
|
|
|
|
problems:
1. the samples are generally self-selecting, and the n- is not disclosed
2. No confidence intervals or p-values are published;
3. Indeed, there is little or no statistical analysis of the result published
why?
perhaps - conflicts of interest?
mainly, I suspect because to make the surveys *valid* would also make them boring, or unintelligible to the vast majority of readers whose reading age is perhaps less than their shoe size.
|
Another point: do they give a statistical analysis of the form fillers? After all, it is a peculiar person who buys AutoExpress is't it. Maybe young, with a leaning to sporty cars? And as for Which magazine buyers. Can you really trust what a bunch of cheese paring tight wads think?
Think I'd better scarper before the scuds start landing ... oooh what's that whooshing sound?
|
|
|
Of course the samples are self selecting and of course there are problems with the analysis.
But the index details the releative likelyhood of having problems with a car. Consequently do not the self selection issues disappear, if the index shows that a Mondeo has a score of 30 then it will breakdown less than a Vectra with an index of 60, from roughly the same self selected population.
That is good enough to give you accurate relative figures, which is all teh index claims to do.
|
|
|
|
MiddleMan,
You do know the connection between tax inspectors and penguins?
They can both stick thier bills up thier pink fluffy dice.
Matt35
|
Matt35,
Quite appropriate as they've just pleasured me with an extra invoice!
MM
|
|
|
|
|
|
