| |
G4EBT > CARS 10.09.07 20:23l 156 Lines 5610 Bytes #999 (0) @ WW
BID : 346469G4EBT
Read: DL3RMU GUEST
Subj: Re: G4EBT > Reliability
Path: DB0FHN<DB0RGB<DB0PM<OE5XBL<OE6XPE<DB0RES<DK0WUE<GB7FCR
Sent: 070910/1527Z @:GB7FCR.#16.GBR.EU #:43937 [Blackpool] FBB-7.03a $:346469G4
From: G4EBT@GB7FCR.#16.GBR.EU
To : CARS@WW
Warren, KB2VXA wrote:-
> I don't know the source of your statistics but they look pretty funky
> from this angle.
I gave the links, and as I said, I made no claims for the robustness of
the survey. For market research to be of real value the sample size and
method of sampling needs to be sufficiently large enough to be confident
that the sample is representative of the whole.
Most professional market research is based on 95% confidence limits. In
other words, that the sample has a maximum error of 5% as compared to
the whole population being sampled.
As Andy, GM7HUD rightly pointed out, in surveys such as that run by What
Car (The JD Power Survey), even with a quite large sample size, the sample
isn't randomly selected but is self-selected by those car owners
who are sufficiently motivated for whatever reason, to complete the
survey.
This might include those who are especially loyal to the brand, or those
who are disgruntled. What we can be certain of, is that the sample will
not be representative of the whole.
Thus, the sample will inevitably be skewed and biased.
The only merit of the surveys I mentioned is that they are at least based
on cost of ownership, breakdowns and impressions of those who have owned
various makes of car for some time *who bothered to respond the survey*
rather than the impressions of car reviewers such as Jeremy Clarkson, who
is a journalistic "funster" - not a serious objective reviewer of cars.
>What caught my eye was the average life being somewhat less than 5 years
>when the average American car is 10 and the imports roughly 15 such as my
>old Toyota.
Yes - it must follow that whatever the make, cars will break down and the
older they get, the greater the likelihood does to ageing of hoses,
wiring, electrical components and in some climates such as the UK, road
salt
and grit in winter.
The more complex the car, the more things there are to go wrong.
I wrote:
>What Car magazine in the UK surveys its readers as to reliability and
>running costs over time. Once more, I'm not saying that these surveys
>are statistically robust, but they are at least based on owner's
>experiences - not on reviews by critics.
>For what it's worth, here's a short clip from the survey,
>covering Volvos for reliability and operating costs:
>Volvo:
>League position: 13th
>Claims per 100 vehicles: 28.5
>Average age (years): 4.43
>Average mileage: 54,838
>Average repair cost: œ322 GBP
>Average time for repair (hours): 2.4
>Volvo had two models in the top five and 13th position overall.
>The V40 and S40, both built between 1996 and 2004 are in fourth and
>fifth place respectively in the individual model league table.
>Volvo was let down by a handful of models, including the V70 (built from
>2000) in 88th, and the S80 (built from 1998) in 83rd place. Its average
>repair cost has also gone up from œ284 last year, to œ322 now.
>Sometimes, cars that are more reliable do cost more to repair when they
>go wrong. Toyota for example. Again, What Car has this to say:
>Toyota
>League position: 3rd
>Claims per 100 vehicles: 12.9
>Average age (years): 4.99
>Average mileage: 46,942
>Average repair cost: œ437GBP
Average time for repair (hours): 3.2
>Although Toyotas are very reliable their average repair cost is the
>seventh-highest in this survey and only seven makes take longer to
repair.
>The five least reliable cars in the survey in terms of claims per/100
>vehicles were:
>Model Claims/100 Vehicles
>Renault Espace 69
>Audi TT 59
>SAAB 9000 58
>BMW7 series 57
>Land Rover Freelander 55
>The survey is a bit dated now - 2005, so things might have improved.
>Anyone interested can find the full survey here:
http://www.whatcar.co.uk/news-special-report.aspx?NA=217350&EL=3142816
>A more up to date reliably index on most makes of cars, which
>also has an American reliability index, can be found at:
>http://www.reliabilityindex.co.uk/
>Again, I make no claims for the robustness or otherwise of the data and
>I've not bothered to check the sample size, but for what it's worth here
>are a few of the most and least reliable makes in the US. The lower the
>percentage, the most reliable:
>Hyundai 14.87% (Korean)
>Kia 17.39 (Korean)
>Cadillac 40.20
>Porsche 41.48
>Volvo 42.76
>Jaguar 57.84 (Worst by far).
End quote.
Locally, the Humberside Police - who service their own vehicles, have
a lot of Proton cars which are cheap and cheerful, and plenty good enough.
Contrary to popular belief, most police officers don't need high
performance cars in which to dash around playing cops and robbers -
any old thing will do for them to tootle around in.
It makes little sense to buy expensive cars with high levels of
refinement. Police cars are often used around the clock, mostly on short
journeys, at slow speeds, in urban traffic.
Due to stop-start driving and high mileages, depreciation plummets like a
stone so the cars have little residual value when they're pensioned off.
Thus, a cheap basic car which can be run into the ground and written
off over two or three years often fits the bill - hence, Protons.
Accountants should choose police cars - not cops.
For high performance cars, the local police use Subaru Imprezzas,
specially modified. (The rear seats are removed and a police dog installed
instead).
Best wishes
David, G4EBT @ GB7FCR
Cottingham, East Yorkshire.
Message timed: 14:45 on 2007-Sep-10
Message sent using WinPack-Telnet V6.70
(Registered).
Read previous mail | Read next mail
| |
