Dodge Caravan 2002 Transmission Problem

On Tue, 25 Oct 2005, Whoever wrote:

> Another issue with the CR data is how can one possibly compare reports
> of failure rates in 2004 vehicles with failure rates in 1997 vehicles?

The only ways to do so with statistical validity are to elucidate vehicle
mileage explicitly or incorporate it as a part of the rating algorithm. An
example of the first possibility is e.g. "1997 transmissions have
such-and-such a failure rate. Sample population had a mean mileage of so,
a median mileage of such, and the mode mileage is thus." An example of the
second possibility is e.g. "1997 transmissions have a calculated failure
rate as shown in the following line graph. The X-axis gives vehicle
mileage, and the Y-axis gives number of reported nontrivial transmission
problems per 100 vehicles at specific vehicle mileage. Sample population
had a mean mileage of so, a median mileage of such, and the mode mileage
is thus. Portions of the graph for which sample population was
insufficiently large for validity are highlighted in yellow."

Neither is completely satisfactory, but either would be a damn sight
better than what CR puts out, which is "Here's the failure rate. Trust
us."

DS

On Tue, 25 Oct 2005, David wrote:

>> If it wasn't about breakdowns, why did they issue a bulletin warning
>> Ultradrive owners to ignore the DEXRON marking stamped into the
>> dipstick?
>
> Because it affected the performance and shifting. As ATF+ has different
> properties then Dextron.

True, but there is no "t" in DEXRON.

On Tue, 25 Oct 2005, larry moe 'n curly wrote:

> How can you rule out changes in manufacturing unless you're an insider
> who've very familiar with what goes on with the company?

That would be difficult indeed. You are assuming those who post the
statements you're objecting to couldn't possibly have inside connections.

> outright false (i.e., twins with different brands from the same factory
> getting very different quality ratings).

That particular one is quite true. Be careful; you're falling into the
"I've never seen it, therefore it doesn't exist" trap.

> There have even been recalls for newer vehicles

True and irrelevant.

> Also how do you explain why the CR surveys show very similar results for
> mechanically identical vehicles, such as Dodge and Chrysler minivans,
> even though their owners are surveyed separately?

What's your point with this question? It's nonresponsive to anything
posted so far in this thread. The Toyota Corolla/Geo Prizm dichotomy
was primarily based on CR's own, er, "tests".

Daniel J. Stern wrote:
> On Tue, 25 Oct 2005, larry moe 'n curly wrote:

> > How can you rule out changes in manufacturing unless you're an insider
> > who've very familiar with what goes on with the company?
>
> That would be difficult indeed. You are assuming those who post the
> statements you're objecting to couldn't possibly have inside connections.

I'm assuming no such thing and have said nothing to imply that.

> > outright false (i.e., twins with different brands from the same factory
> > getting very different quality ratings).
>
> That particular one is quite true. Be careful; you're falling into the
> "I've never seen it, therefore it doesn't exist" trap.

I've looked for it in their reliability tables but haven't found it.
Cite two models and a year.

> > There have even been recalls for newer vehicles
>
> True and irrelevant.

Not irrelevant. I'd expect an older design to have its bugs worked out
and its reliability to stay the same or decrease the longer it's
produced, unless changes are made.

> > Also how do you explain why the CR surveys show very similar results for
> > mechanically identical vehicles, such as Dodge and Chrysler minivans,
> > even though their owners are surveyed separately?
>
> What's your point with this question? It's nonresponsive to anything
> posted so far in this thread. The Toyota Corolla/Geo Prizm dichotomy
> was primarily based on CR's own, er, "tests".

What Corolla-Prizm dichotomy? Those cars have always ranked closely,
both in CR's judgement and their surveys. Show me where they didn't,
either by model year or CR issue, or admit that you don't have proof.

The fact that brand twins have shown very similar CR reliability
ratings indicates that CR's surveys are very repeatable, something you
wouldn't expect if they were useless garbage.

Daniel J. Stern wrote:
> Your argument here amounts to "OK, CR might be full of ****, but what
> choice do we have but to believe them?". I reject that argument as...well,
> stupid. Yes, it's a shame CR isn't what it once was -- go to the public
> library and take a look at what they were in the '60s --

It was back in the mid- to late-60's that they were making television
repairmen scads of money by rating the worst sets highest and the best
sets lowest. So even back then they sucked.

--
If John McCain gets the 2008 Republican Presidential nomination,
my vote for President will be a write-in for Jiang Zemin.

larry moe 'n curly wrote:

> Matt Whiting wrote:
>
>>larry moe 'n curly wrote:
>>
>>
>>>Again, all CR is doing is reporting the reliability year
>>>by year, so how do the exact versions of the transmissions matter?
>
>
>>Because the exact same transmission design is used in several of the
>>years listed, yet the reliability ratings are dramatically different
>>year to year. The tranmissions don't know what year they are so they
>>have know way to know that they should break if made in one year, but
>>not break if made the year after.
>>
>>If the data was at all accurate, you would expect to see step changes in
>>reliability only if a step change was made in the transmission design,
>>materials or assembly processes. The reality is that none of these have
>>happened at every major reliability change listed in the CR table.
>>Thus the data is very highly suspect at best.
>
>
> How can you rule out changes in manufacturing unless you're an insider
> who've very familiar with what goes on with the company? After all
> there have been recalls for newer vehicles where the design had been in
> production for years.

How can you rule out that I'm an insider?


> I'm not saying that CR is anywhere close to perfect, but some of the
> criticisms expressed here about their surveys don't seem valid and
> sometimes even outright false (i.e., twins with different brands from
> the same factory getting very different quality ratings).
> There have even been recalls for newer vehicles Also how do you
> explain why the CR surveys show very similar results for mechanically
> identical vehicles, such as Dodge and Chrysler minivans, even though
> their owners are surveyed separately?

As I said earlier, I'm a CR subscriber and have been for more than 20
years now, however, their biases on import vs. domestic car brands is
just so blatant that it is ridiculous.

They should ALWAYS show similar results for identical vehicles. The
trouble is, they often don't. The fact that they do sometimes doesn't
negate the fact that many times they don't, when in fact they should
NEVER show different results for identical vehicles.

As I also said earlier, I'm not necessarily placing the blame on CR as
much as on their subscribers who submit the surveys. However, I do
believe that the biases that CR publishes have affected the perceptions
of their readers and these perceptions then in turn affect the survey
results. I've seen enough study on human perception to know how
unreliable it is and how easily many people can be swayed.


Matt

On Tue, 25 Oct 2005, larry moe 'n curly wrote:

>>> outright false (i.e., twins with different brands from the same
>>> factory getting very different quality ratings).
>>
>> That particular one is quite true. Be careful; you're falling into the
>> "I've never seen it, therefore it doesn't exist" trap.
>
> I've looked for it in their reliability tables but haven't found it.

Keep a-lookin'.

> Cite two models and a year.

Asked and answered. It's not for me to do your homework; hit your local
well-stocked public library.

On Tue, 25 Oct 2005, Matt Whiting wrote:

> I've seen enough study on human perception to know how unreliable it is
> and how easily many people can be swayed.

Amen. *clink*

..
..
=========
=========
MaceFace asked:

So why did Chrysler respec the fluid 3-4 times since introducing their
1st 4-spd auto? Was it because they had no breakdown problems?
==========
==========
ks (Steve)
incorrectly corrected with:

****The fluid was never about "preventing breakdowns," it was to
allow the torque-convertor clutch to be used in a partial-lock mode that
no other transmission had done before. The fluid was re-specced to
improve its lifetime and performance (reduction of shudder) not because
it was causing failures.
=========
=========

Steve,
yer incorrect.

Fluid types have not changed because of
lockup strategy.

Partial lockup has been around for over a decade.......and definately
not specific to
Mopop trannies.

GM has partial lockup......it doesn't require
Mopop fluid.

neither does Ford

neither does Mercedes

and lets not even get into not sticking Honda
fluid in a Honda.....whewww....now there's a
shudder.


MarshMonster
~sips his crownroyal~

"Matt Whiting" > wrote in message
...
>
> As I said earlier, I'm a CR subscriber and have been for more than 20
> years now, however, their biases on import vs. domestic car brands is
> just so blatant that it is ridiculous.
>

Keep in mind that while CR may not cater to advertisers they have to
sell magazines off the magazine rack to continue to replace subscribers
lost through attrition with new subscribers.

Since there's more domestic cars sold than imports in the US, (even
today) if you gore the ox owned by most people, rather than the ox
owned by fewer people, you will get more attention.

Ted