You do have to look at these in a bit more detail though.
In the which survey Honda did indeed achieve the most overall reliable brand but it's Civic was joint bottom with the C4 in it's class.
It also performed badly last year and at the time I was on this site posting to people how it will all be different next year because Honda listens to it's customers and will not stand for such poor reliability. If I remember correctly I had a bit of a dig at stuartli on the subject as he raised this first.
Oh how wrong I was ! Honda have clearly not done enough to sort the Civic out quickly enough.
Edited by Pendlebury on 20/06/2008 at 21:49
|
I also suspect the 'type' of customer who buys certain cars has an impact. I suspect age.
I was involved in some research in the 90's that looked at housing satisfaction and we looked at lots of variables (multivariate analysis) independently and older people were more satisfied with their home than younger people - even after accounting for type of house, area, income etc etc. Older people seemed emperically happier with their lot.
I've often thought this might explain the high results of Jaguar, Honda etc. A simple way of testing if age has an impact is to take the Ford Focus results and split the satisfaction scores in to age groups (18-24, 25-34, 35-44, 45-54, 55-65 etc). Compare this with the same split on Honda Civics. You need large sample sizes though.
The which honda civic results shows though that this age related higher satisfaction is on average - its still possible to get poor satisfaction from your core buyers (who I would say Which members probably are) if you try hard enough!
Phill
|
You are right, Pendlebury - the Civic comes bottom of it's class, which they put down to the diesel being unreliable. It's quite hard to get anything conclusive from these surveys, I find - often the sample is small and satsfaction varies between models.
I have also been reading the Top Gear Test Drive mag and it's interesting to note the differences in reviews. TG seem to be about how a car looks and how it drives, Which? is alll about reliability and safety. eg. Which? like the Daihatsu Sirion, TG hate it.
|
I didn't actually read the detail about the Civic but to support your point, Which place it as bottom of class due to problems with the diesel engine, but you read any 'car' mag or website and the best thing about Hondas these days is the fantastic diesel engine. I would sugget that the Honda diesel owners on here may support the Which findings. There is also the management of expectations as pointed out above.
When I owned VW's I expected the car to be returned under warranty at least 4 or 5 times - and they never let me down in that expectation - but I considered it to be ok - it was only a VW afterall. My Honda let me down once, (clutch master cylinder) and I am more aggrieved - because I expected it to be better than that.
|
|
|
|
>>I had a bit of a dig at stuartli on the subject as he raised this first.>>
To be fair, the latest Which? does point out that whilst Honda is the most reliable car maker for the second year in a row, adds: "But it (Honda) may have to work hard to maintain this crown, as the latest Civic still hasn't shaken off its early reliability problems - it's no better than last year."
One other particular fact that emerged yet again is the excellent standing reliability wise of Skoda; the publication comments that "Skoda outperforms Seat and VW itself - so much for being the budget arm of the VW Group."
Even more puzzling is that VW Group marques' model share so many common parts and components, yet there are different outcomes.
Edited by Stuartli on 22/06/2008 at 13:56
|
|
|
|
I have bought the Which? Car Guide today. Again the reliability results are interesting. Daihatsu come third in the reliability index - but this is based on a sample size of only 195 responses. Honda came top but the sample size is 8 764. Porsche are 10th with a sample size of 216. Proton are 17th just above Mercedes with a sample size of 55.
I'm sorry, but I don't see how that can be at all a fair comparison, you can't compare a sample of a couple of hundred (or less!!) fairly against that of several thousand and this is the major problem I have with Which's findings when compared with the likes of JD Power or the more recent TG surveys whic spcified a minimum number of cars before they would be included in the results...
I'm afraid that I will continue believe the findings of the larger surveys over Which's for exactly that reason...
|
What are the problems with the Honda diesel engines?
As far as I know the only problem has been the manifold at about 90k. I have an Accord Tourer diesel bought new in 2004, and it has behaved perfectly-touch wood.
Fuel consumption is excellent. It was poor for the first 6,000 miles but improved tremendously.
|
|
|
I'm sorry but I don't see how that can be at all a fair comparison; you can't compare a sample of a couple of hundred (or less!!) fairly against that of several thousand
Why not? It isn't the total number of faults, but the proportion that have faults.
>>and this is the major problem I have with Which's findings whencompared with the likes of JD Power or the more recent TG surveys whic spcified a minimum number of cars before they would be included in the results...
"Which?" also specify a certain number of cars before they are included in the results.
Edited by tyro on 21/06/2008 at 12:44
|
For *reliability* you don't need large sample sizes as cars are relativey homogenous particularly if you limit the survey to only cars a couple of years old so its the same generation of say a civic. I would have thought 50 -100 cars per model would be adequate for this - the survey company should know how many on the basis of previous surveys.
Where you do need large sample sizes is for satisfaction, as this is subjective and will vary by the background of the person responding. You need a good cross section of people for this to be represnetative. The which? sample sizes look small to me in this regard - not sure about JD Power as they rarely say.
The usual way of dealing with small samples is to quote a confidence interval. I suspect though if this was done, very few of the differences between cars would be statistically significant for satisfaction, which would make a poor report.
The best designed survey was a JD Power survey a few years ago where they sampled from DVLA records - this is great as you can control for age of car and get an accurate random sample of owners. AFAIK they didn't manage to repeat this.
Phill
|
>>For *reliability* you don't need large sample sizes as cars are relativey homogenous
I think that is true if you assume that reliability isn't affected by the owner, and specifically, the way they treat the vehicle. So, while the vehicles themselves may be relatively homogeneous, owing to the inhomogeneity of their treatment, I don't think you can have great confidence about their inherent reliability statistics with a small sample.
|
|
|
I do wonder with the satisfaction angle on these surveys whether there is a skew in favour of the brands that a more niche population of "knowing" people tend to favour - people who research minutely the offerings of the market and make choices that they consciously think are "better" than the common herd and are prepared to make a more original purchasing decision and stand by it - Skoda, Daihatsu, Subaru, Honda etc. Does it then translate that when these people fill in surveys (and perhaps they may seek out such surveys to demonstrate their "knowing" market intelligence) that they rate themselves/their clever choice/their ownership experience alot higher than someone whose decision was much more mainstream and unremarkable and less important to them.
We all know people who when you explain to them the range of interesting choices out there in different product sectors just choose the one that they've always bought or they recognise or is just an easy choice based on it's nearest to them or available in their local high street. Does this then mean that such people are quicker to criticise their choice if it's not good because they have less "invested" in their choice of product?
|
Fair points UKB
Owners who buy the makes you mention may also maintain their vehicles with far more care than someone who just sees a set of wheels to be used with no attention or maintenance or care given.
They're probably meticulous about fuelling and cleaning off salt etc as well, and it all helps to keep your investment reliable.
There will always be exceptions though, but for example its not often you see a honda jazz anything but spotless and driven very carefully, bound to have posititve effects over time, apart from an owners pride in caring for their vehicle.
|
Owners who buy the makes you mention may also maintain their vehicles with far more care than someone who just sees a set of wheels to be used with no attention or maintenance or care given.
That's got me thinking. Perhaps . . .
People who buy cars for reliability are more likely to maintain them carefully.
People who buy their cars for appearance are more likely to wash them
People who buy their cars because they are cheap are more likely to skimp on maintenance
People who buy their cars for performance are more likely to thrash them
(etc, etc. Feel free to add to the list!)
|
|
|
|
Where you do need large sample sizes is for satisfaction as this is subjective and will vary by the background of the person responding. You need a good cross section of people for this to be represnetative.
This is an important point, and is the real problem with the satisfaction surveys - like JD Power and AutoExpress.
Asking people what they think of the ride, or what they think of the handling or what they think of the performance and then saying "Car X is the best car for ride" or for handling or for performance or whatever is misleading, and virtually (though not quite) meaningless.
Satisfied people will generally be satisfied with their cars, and people who are satisfied with their cars will say that they are satisfied with the ride, or the handling, or whatever, even if it is not very good. And of course the converse is also true.
|
|
What about reliability index by Direct Warranty? According to Direct Warranty''Based on data gathered from millions of pounds worth of claims handled by Warranty Direct every year, the Reliability Index brings you the most comprehensive overview of cars on the road today''. The top 10 most reliable cars are Honda Accord, Subaru Forester, Mazda MX-5, Mitsubishi Carisma etc in descending order. This seems to agree with other surveys that put the Japanese built cars as the most reliable. I think this is probably true.
|
|
|
The full table is available here
tinyurl.com/44r74u
|
|
|
Sorry, chaps, but you've not convinced me, I still cannot see that surveying 55 cars (to quote the original post) can give you a true representation of the several thousand of that model of car sold... its also interesting to see the comments about the JDP/TG "satisfaction" surveys concerning the people who take part in them and the "subjectiveness" - exactly the same could be said of the people who take part in the Which survey, noting also that there are a lot less owners taking part...
At the end of the day all these surveys can do is give an "indication" of what you can expect, though for my part I will still believe more in the larger surveys for the obvious reason that they include more cars....
Its the same old chestnut about "percieved" issues with certain cars because of people complaining about them on forums such as this one - at the end of the day we only represent a minute number of the total drivers and have a vested interest in a particular make, thus slewing our comments. The only true survey would be one which surveyed evry single car for every single fault and I doubt the car makers would let that happen!
|
Some valid points, b308.
At the end of the day all these surveys can do is give an "indication" of what you can expect
Yup.
Which survey noting also that there are a lot less owners taking part... I will still believe more in the larger surveys for the obvious reason that they include more cars....
"Which?" say "We sent the Which? Car survey to nearly half a million "Which?" members in January 2008. From this we acquired data for 89,762 cars"
JD Power survey for 2008 acquired data for 16,301.
In JD Power's favour, they only test only cars of the same age (two years old, I think), whereas "Which?" states "Cars are up to eight years old."
|
|
|
|
|
For *reliability* you don't need large sample sizes as cars are relativey homogenous particularly if
Phill
A large sample size is always better than a small one for statistical reasons, regardless of what you are measuring. You are trying to extrapolate the results of your sample to the whole population, and this introduces an error which is proportional to the inverse of the square of the sample size.
Say you want to know if Car A is more reliable than Car B. You sample 100 of each car type and find that 1 Car A had a fault and 2 Car Bs had faults. Car A is more reliable you think. In actual fact if you do the correct statistical test you find there is an almost 50% chance that Car B is in fact as or more reliable than Car A from these data.
If you sample 1000 cars, and find 10 and 20 faults (the same percentage as before), you can be 95% certain that Car A is more reliable than Car B. So 100 is likely too small a sample to rely upon.
Lies, damn lies, and statistics.....
Martin
|
And then it also depends what the seriousness of the fault is - a door handle coming loose is a pain, but a faulty wheel bearing could be dangerous and a real pain!...
As said, lies, damned lies, etc...
|
If you do a 'survey of surveys' then a pattern emerges. Look at all the surveys (What Car, Which, JD Power, Warranty Direct etc and others that you can read online - like the German ADAC and an Australian one) and a pattern emerges. Namely that the Japanese cars are invariably toward to top of the table.
This also coincides with my own anecdotal evidence - people I've known who've had Renaults and Fiats have had a lot of problems, whereas the people who've owned Japanese have had very few problems.
|
It would be in interesting if you could as an experiment, swap owners on taking delivery of the cars.
On the F forum, lads are chipping their GP`s and Pandas, putting neat cooking oil into common rail and generally thrashing some of them to death. One chap seemed to think the rev limiter was there to be used all the time even from cold.
( I`m not judging them, their own cars and they are young as we once were and so on;)
But If these Fiats were allocated to typical Toyota aygo and Honda jazz owners instead it would be interesting to see how they fared.
Even more interesting to see how the Japanese brands would fare with some of the Fiat lads.
Sure, non driver related problems would be there. But when did you last see an Aygo or Jazz or Honda diesel being chipped and ragged?
|
My thinking too, Oilrag. It could also help to explain the presence of Skoda at the top of the charts, with their, ah, mature driver profile.
Nevertheless it's a bit worrying that my extremely sober Saab still languishes near the bottom of all these surveys!
|
I don't buy into the owner argument at all. I don't think Skoda are particularly favoured by "ah, mature" drivers these days. Most of the ones I see are family owned and of course there are the VRS models too.
I think most people buy a new car, drive it, and have it serviced on annual basis. I am not sure what else the owner can do to make it more or less reliable. The unreliable VW we had (2003 model) had faults that were completely unrelated to servicing or driving style (coil pack, heater unit, lamda sensor, parts of electrical system and premature wear of droplinks) all common problems if you read VW forums.
I don't think Renaults are driven by boy-racers either - yet they always come out badly. Similarly Chrysler.
My sister has owned several Fiats (Uno then Punto) and she is as careful a driver as you could ever meet, car is always serviced on time. She's had trouble with them all, not 'rattle and creak' type trouble, but 'cut out on the M1' type trouble. She keeps buying the things because she lives just around the corner from the dealer - not that that would be a good enough reason for me! She now has a 2008 Panda diesel, and yes, its got problems. Gearbox is crunching like the clutch is not disengaging. Its gone in for a strip-down. Apaprently there is a service bulletin on this.
One of the reasons I bought an Impreza is that my neighbour had a series of turbo models since about 1994. He drove them hard as hell and they just had the annual service. I spoke to him and he told me he never had a single warranty job in all that time. I reckoned if they were OK in his hands then a non-turbo would be OK in mine. Touchwood all is OK so far!
|
I don't think Skoda are particularly favoured by "ah, mature" drivers these days. Most of the ones I see are family owned and of course there are the VRS models too.>>
I've pointed out on more than one occasion that the excellent taxi companies in the resort where I live have had a very large proportion of their fleets made up of Skoda Octavias for a number of years - the drivers, almost to a man or woman, praise them highly.
Ironically there's no Skoda dealership in the town, but most of the drivers seem to find the best deals in Glasgow, some 200 miles away..:-)
|
|
Not that this conundrum about reliability looks like being resolved by our collective speculations - but I'm well aware of the Skoda taxi fleet, and would assume that their drivers' maintenance regime would surely be "mature"...?
|
I don't buy into the owner argument at all.
I suspect that you are correct, though, as I suggested above, I think it would be interesting to try to test the hypothesis.
Mind you, I'm surprised that no-one has suggested that the reason that the new Honda Civic is doing poorly in reliability surveys is that it is being purchased by the sort of people who do not usually buy Hondas - and that this just goes to prove that the reason Honda did so well in these surveys over the years is that they were bought by old people. As soon as Honda targets a model at a younger age group, and younger people start buying them, the reliability statistics take a nose dive. :-)
|
|
As its late here's one to nod off to...
In the real world bigger is not always better:
a) 'research design effects' - if your sample is biased, sampling more does not make it less biased. I see lots of surveys where this mistake is made and all it results in is an organisation spending lots of money when it didn't need to.
So the Which survey might be no better than the JD Power survey despite being significantly bigger.
b) if the population is relatively homogeneous - large sample sizes are unecessary as you hardly alter the std error as you sample more (that square root bit). I think car reliability is likely to be like this but I don't know for certain.
I love Oilrag's idea of getting young Fiat drivers to be forced to drive Hondas or Skodas - its a randomised controlled experiment! Well nearly..
Maybe we should get Lexus service people to work at a Fiat or Jeep dealer for a bit too? Would the dealer improve or the Lexus peeps lose it?
Phill
|
Maybe we should get Lexus service people to work at a Fiat or Jeep dealer for a bit too? Would the dealer improve or the Lexus peeps lose it?
Fantastic idea, they both need the practice. Let them do a wife swap thingy for a month or two.
Lexus mechanics only do servicing, as the cars don't go wrong., they probably scratch their heads bald if they have to diagnose anything, would be good for them.
(Lexus dealers i've been to don't seem to have many mechanics, tells me something)
I commend the idea to the house..;)
|
|
Still would like to see some method of unscrambling the time v usage equation. For example a car which has covered 15k miles over three years may not ( and one would hope not ) have had any problems. Conversely a car which has covered 100k over the same period may also not have had any issues. Under the current measurements the two vehicles would be regarded as equally reliable. I would be interested to see stats which analyse reliability as a function of mileage rather than time.
|
Here's a thing. Just did a little experiment using the Autotrader website. Took three reasonably random mainstream brands. Ford, Honda and Skoda. If you check the percentage of each brands cars currently for sale today which have more than 100k miles on them you get equal scores at 4% for Honda and Ford. Skoda however, only has 2.5%.
One of many theories which could be postulated from this is that Fords and Hondas are equally capable of being run viably to high mileages and that Skodas seem less so. And before anyone points it out I do know there is far more to it than that !
Can't be bothered to check all the other brands but it might just help to tell us a little more about which cars actually do stand up better to the rigours of high usage ?
|
If you note which cars are used as mini cabs - you then have a good idea which are capable high milers.
It does include some French cars like Peugeot 406 etc.
|
One of the most interesting questions, as far as I am concerned, is: How reliable are Skodas anyway?
Skodas do very well in satisfaction surveys like JD Power.
However, in reliability surveys, they tend to come out as fairly average.
The latest "Which?" survey puts them at 78% (along with 9 other brands), with 10 brands above, and 18 below. That works out as average, or a little above.
The latest ADAC survey, looking at cars from 2002 to 2007 shows the Fabia at joint 19th out of 27 among small cars, and the Octavia as joint 4th out of 12 (behind the Passat, but ahead of the Avensis).
Skoda do pretty well on the Warranty Direct Index, though I must confess to finding it rather odd. (Citroen comes just behind Toyota and ahead of Lexus.)
|
I would like to see these surveys balanced on the mileage covered by those who take part.
After all, the 25,000 per year driver of a Civic (for example) will have a greater experience of 'reliability' that the person who does 5,000
|
I would like to see these surveys balanced on the mileage covered by those who take part. After all the 25 000 per year driver of a Civic (for example) will have a greater experience of 'reliability' that the person who does 5 000
These surveys aren't designed to be just about reliability they are customer satisfaction surveys one is not the same as the other.
|
|
Yes. To understand the data fully you would have to know what kind of people bought the car and what they expected of it. Better listen to people of your own cast of mind, or something like it.
|
|
Be careful about the mini cab analogy, Skoda are the only manufacturer as I understand it to warrant their cars for cab use. I was in a 214K & it sounded like a bucket of bolts. It also had wind up windows no a/c etc. It's easy to be reliable when you have no kit to go wrong!
|
|
If they really do, tr7, it is very gung-ho of them. One way to make a motor old before its time is to minicab it.
|
Maybe there is already some sort of cross driver/ Marque, Fiat-Toyota experiment going on. Scroll to the bottom of this thread and don`t let your respectfully driven `almost new` Toyota see it ;)
tinyurl.com/5myln4
Edited by oilrag on 23/06/2008 at 22:09
|
Good heavens OR, why was a boy allowed out unaccompanied.
The comment about not doing it to his own vehicle says it all, what possible pleasure could thrashing that engine have given, bring back the birch i say.;)
I don't think i'll browse any more of that forum, and as for thought of anyone considering an ex demo, well perish that.
I do hope you've thoroughly chastised the offender, if you use some of your more cryptic phrases, a reply may be some time coming...
As an aside Oily, have been delivering some Cit nemo and Pug bipper vans, a little smaller than the partner/berlingo, but i could have sworn the ignition plipper is Fiat, and the vehicle feels Fiat like, have they got another collaboration going, like the expert/dispatch/scudo.
Not hijacking the thread, just a query.
GB
|
have they got another collaboration going like the expert/dispatch/scudo?
According to boxsterboy*: "I understand the Berlingo/Partner replacement is a partnership with Fiat to replace the Doblo as well."
*see www.honestjohn.co.uk/forum/post/index.htm?t=41654
|
Tyro, i've been under the new partner/berlingo and its deffo the new C4 picasso underpinnings, whether Fiat's new doblo is similar i don't know, i havent seen one yet.
I can't see anything familiar pug wise about the nemo/bipper, it seems to be a whole new model.
It will be interesting to see what fiat bring out in their small van/mpv range.
no sign of auto on the French models, i wonder if Fiat will have the courage to try them out, i think they'd sell well (proper auto, not that other thing).
|
Hi GB, Yes, I have seen the Fiat version, the Fiorino at the local fiat dealers. For anyone who has not seen it.
tinyurl.com/5zp8wk
Regarding the `cough`auto, there is a thread on FF running into the hundreds about experiences with that too, on the GP that is.
I had a Doblo, but it was devastatingly unreliable and faulty. Great to drive and so on, but i got a bit spooked when the Sat Nav started to go on the blink again just out of warranty.
It seems the newer bigger engined Multijets (not 1.3s) are not getting anyway near the fuel consumption of the older 1.9 JTDs too.
Can`t wait for your eagle eyes to scrutinise the new KA (500/KA) when it starts being transported to dealers. You`re probably going to be better placed than anyone to spot whether its actually galvanised like its 500 sibling. Or whether Uncle Henry has saved a few bob and gone the allegedly `traditional` route. ;)
That said, with our current Mk2 Punto diesels being so trouble free (and worth so little) we won`t be making any changes for a long time.
;)
|
The Which report has some interesting points:-
uk.reuters.com/article/motoringNews/idUKNOA7331702...e
I wouldn't disagree with any of their bottom 10 - they've got that dead right, I reckon.
|
A survey that doesn't have the Scenic II in the bottom 10? Well I never! ;-)
Cheers
DP
|
A survey that doesn't have the Scenic II in the bottom 10? Well I never!
It's stranger than that. A British car made the top ten! :-0
|
>>>As soon as Honda targets a model at a younger age group, and younger people start buying them, the reliability statistics take a nose dive.<<<
Interesting observation, Tyro, but if you have a look at this site:
www.hondabeat.com/highmiles.php
you will find many Honda owners who swear by the reliability of their vehicles.
I read their posts regularly and get the impression that the owners are quite a cross-section of the population. However, I suspect that some of them have had their cars for so long that they, as well as their vehicles, are getting on a bit so may come into the older owner category :o)
Someone will no doubt make the point that Yanks do not have diesels so that brings us back to why Civic diesels seem to have a problem.
|
All I can remember about Honda diesels is this advert that surely had the motor industry green with envy?
tinyurl.com/6ao398
|
The Which? survey is covering the MK1 Scenic.
The MK2 Scenic gets a fairly good slating in the What Car Survey: "Scenic and Grand Scenic owners faced massive reliability problems - the Renault had some of the worst results across the board". Whoops...
|
Hmmmm....
"Ford Scorpio owners should be wary, over half of those on Warranty Direct?s books needed warranty work last year."
Well, considering the youngest Scorpio is now 10 years old, is this really a shock?
|
|
|
|
|
|
|
