Reliability?

[itguy08]

Originally Posted By: eljefino

Their subscribers are a subset of the population that chooses to be told what TV to buy, because they fail at such things themselves.


Then what would they be doing buying the unreliable vehicles in the first place? They wouldn't touch those vehicles down on the list and yet CR has data on them too...

Think about that for a minute. I'm so incapable of making a decision that I use CR to base my decision for me. That alone means I'm not going to look at the things at the bottom. So there should be little to no data for them. Yet there is plenty of data for them. Where is that data coming from? Made up surveys? Or perhaps accurate assessments of people's vehicles?

And when that data correlates to what other, non CR outlets says, maybe they are on to something.

Just because you or I had a great experience with something doesn't mean they all do. Wife got rid of her 03 Escape with 159k. It was a great, reliable car needing few repairs over its life. Original transmission and engine. The data shows the CD4E was a weak point in those vehicles. I'd not suggest the CD4E was a reliable unit - there are plenty of them that didn't make it very far. But her example was. I'm not refuting the mountains of data that the CD4E was not great - our example was fine but to suggest all were is crazy.

 

[KGMtech]

I find the CR reviews to be biased. They overlook faults of a Honda / Toyota and point out the same type faults on GM/Ford/others.

Truedelta is building a good database for reliability, but until they are bigger, the CR owner reported data is the gold standard.

 

[IndyIan]

Of all the cars in our family, CR tends to right. The Toyota and Honda's have been very reliable and cheap to own. The Mazda's and Fords haven't been quite as good, but nothing major.
The Chrysler minivan did need a couple transmissions, and some other major work. My Neon did need a new water pump and headgasket but then became ultra reliable.
So they are better than random guessing in terms of reported problems in my experience.
I'm not a big fan of "predicted reliability" ratings though. But I guess they need something to lure a new car buyer into reading.

 

[stephen9666]

My experience has been that CR reliability surveys have been pretty much spot on. Their surveys have largely mirrored the trouble spots I've had in my vehicles.

I check their reliability ratings when shopping for cars, but they aren't the be-all-end-all to me as I consider more than just reliability when purchasing a car.

They provide useful testing that I consider when making purchases. Why some people get their undies in such a wad over CR is beyond me. If you don't want to read them, don't read them.

 

[gathermewool]

Originally Posted By: dishdude
Originally Posted By: Quattro Pete
Originally Posted By: gathermewool
I don't think you get it. An Aveo has the same parts and quality control as any Caddie. This is a fact.

OK, so that covers GM. What does that have to do with Toyota?


It's not fact, it's the ramblings of a moron with a keyboard.


Might wanna turn up your sarcasm meters, guy. Does everyone need to be spoon-fed emoticons so as not to turn nasty?

Seriously, 'same parts' didn't give it away???

 

[clinebarger]

Originally Posted By: gathermewool

I don't think you get it. An Aveo has the same parts and quality control as any Caddie. This is a fact.


Aveo's are built in Korea, I have worked on them & They share nothing with Domestic GM products.

Toyota is always going to be the favorite even though I hate my Toyota with a passion.

 

[gathermewool]

Originally Posted By: clinebarger
Originally Posted By: gathermewool

I don't think you get it. An Aveo has the same parts and quality control as any Caddie. This is a fact.


Aveo's are built in Korea, I have worked on them & They share nothing with Domestic GM products.

Toyota is always going to be the favorite even though I hate my Toyota with a passion.


I don't think that's right. I think the Aveo at least shares some suspension components and ornaments with most Cadillacs. I'm actually 90% positive.

 

[clinebarger]

Originally Posted By: gathermewool

I don't think that's right. I think the Aveo at least shares some suspension components and ornaments with most Cadillacs. I'm actually 90% positive.


The Faded Lines.... HERE

 

[gathermewool]

Originally Posted By: clinebarger
Originally Posted By: gathermewool

I don't think that's right. I think the Aveo at least shares some suspension components and ornaments with most Cadillacs. I'm actually 90% positive.


The Faded Lines.... HERE


A message of hopelessness? How appropriate.

 

[Gasbuggy]

CU is smarter than most car buyers.

 

[VNTS]

Waste of time.

Their data is not quality data, it is a survey of their subscribers, hence it is not random data and thus is garbage because the data is skewed, any one with a quality back ground in manufacturing laughs at it.

 

[440Magnum]

You can read my opinion of what fine bird-cage liner Consumer Reports data makes all over this forum, but I'll summarize briefly:

Their data set is deeply and profoundly flawed by self-selection. The people who report data back to CR are not a normal distribution ("normal" in both the vernacular and rigorous mathematical senses) of the range of consumer satisfaction or of vehicle problem rates. They self-polarize into two groups: 1) those that have been burned and want to rip a particular model/manufacturer/dealer a new one, 2) fanatics of a model/manufacturer who love to crow about how it can do no wrong.


The whole range of information around the mean of the distribution is depleted because people who have near-average experiences with their vehicles and don't have any particular emotional investment in a brand simply don't have any reason to report. The tails of the distribution are excessively weighted. The result: nonsense conclusions.


And secondly, the WHOLE set of people reporting data are a tainted dataset before they even start "skewing" themselves into a modal distribution as described above. They're people who don't necessarily do a lot of reserarch, but rely on CR to do it and then they accept the results. Otherwise, they wouldn't SUBSCRIBE, they'd just buy a copy (or look online) when they come to a particular buying decision, as well as checking other resources. But no, they use it so consistently that they are predisposed to have faith in CR, so there's a feedback loop where previous CR reports are tainting the current evaluations.

 

[hattaresguy]

You get a similar affect with internet amplification.

If X manufacture has an issue with a small percentage of vehicles, than a few of those owners get [censored] off and start posting everywhere about it. Before you know it the problem is way over blown and the vast majority of owners who never have an issue never even know about it.

The Toyota sludge thing people get all worked up about on this site is a perfect example of that. It probably affected less than 1% of their motors but not if you follow this site. I still see a large number of those vehicles on the road to this day and they are 15 years old now...

The reality is that all modern vehicles are more or less pretty much the same and more or less pretty reliable. Some have more issues tha others due to their nature, ie a BMW 750IL will experience more issues than say a Civic because it has many times the systems to fail, and the systems it has are more advanced and less tried and true. But generally if you compare like models, lets say mid sized 4 door sedans, they are all about the same and they will all yield similar cost per mile over their service life.

Trucks are a great area for this as well, all the big three are pretty much the same. You could do far worse than to simply buy whichever will give you the best deal at the time.

 

[Nate1979]

Originally Posted By: 440Magnum
You can read my opinion of what fine bird-cage liner Consumer Reports data makes all over this forum, but I'll summarize briefly:

Their data set is deeply and profoundly flawed by self-selection. The people who report data back to CR are not a normal distribution ("normal" in both the vernacular and rigorous mathematical senses) of the range of consumer satisfaction or of vehicle problem rates. They self-polarize into two groups: 1) those that have been burned and want to rip a particular model/manufacturer/dealer a new one, 2) fanatics of a model/manufacturer who love to crow about how it can do no wrong.


The whole range of information around the mean of the distribution is depleted because people who have near-average experiences with their vehicles and don't have any particular emotional investment in a brand simply don't have any reason to report. The tails of the distribution are excessively weighted. The result: nonsense conclusions.


And secondly, the WHOLE set of people reporting data are a tainted dataset before they even start "skewing" themselves into a modal distribution as described above. They're people who don't necessarily do a lot of reserarch, but rely on CR to do it and then they accept the results. Otherwise, they wouldn't SUBSCRIBE, they'd just buy a copy (or look online) when they come to a particular buying decision, as well as checking other resources. But no, they use it so consistently that they are predisposed to have faith in CR, so there's a feedback loop where previous CR reports are tainting the current evaluations.




What you say may be true about the CR data set but that is so obvious the most basic statistician would be able to see that. So, either CR hires the dumbest statistician ever or they are accounting for that in their data handling.

 

[andrewg]

For the consumer to get an accurate reliability question answered about a particular vehicle they might be interested in, Consumer Reports is but one source of info. One can start with them...but then move on to other reviews and assessments done by at least five to ten other organizations. Then it would be advisable to go to message boards focused on the car. After doing all that, a person would then get a good idea if a vehicle is/will be reliable for them.

 

[MCompact]

Originally Posted By: 440Magnum
You can read my opinion of what fine bird-cage liner Consumer Reports data makes all over this forum, but I'll summarize briefly:

Their data set is deeply and profoundly flawed by self-selection. The people who report data back to CR are not a normal distribution ("normal" in both the vernacular and rigorous mathematical senses) of the range of consumer satisfaction or of vehicle problem rates. They self-polarize into two groups: 1) those that have been burned and want to rip a particular model/manufacturer/dealer a new one, 2) fanatics of a model/manufacturer who love to crow about how it can do no wrong.


The whole range of information around the mean of the distribution is depleted because people who have near-average experiences with their vehicles and don't have any particular emotional investment in a brand simply don't have any reason to report. The tails of the distribution are excessively weighted. The result: nonsense conclusions.


And secondly, the WHOLE set of people reporting data are a tainted dataset before they even start "skewing" themselves into a modal distribution as described above. They're people who don't necessarily do a lot of reserarch, but rely on CR to do it and then they accept the results. Otherwise, they wouldn't SUBSCRIBE, they'd just buy a copy (or look online) when they come to a particular buying decision, as well as checking other resources. But no, they use it so consistently that they are predisposed to have faith in CR, so there's a feedback loop where previous CR reports are tainting the current evaluations.




You nailed it.

 

[itguy08]

Funny to see the excuses here for the "bad data" form CR. How do you explain that their data seems to be in agreement with JD Power who does a more random approach? Or TrueDelta who has a more "enthusiast" slant? Or the other places that track this sort of thing?

You're making a whole lot of ASSumptions about the CR reader which may or may not be true. If what you write is true they would simply blindly buy the top rated brand. Then there would be no data for the low rated brands as they are not recommended and therefore bought. Yet they do say when they have no data and do have data on low rated brands. How could that be if they blindly bought what CR said was best?

 

[Eddie]

I agree with itguy08. Ed

 

[eljefino]

Originally Posted By: andrewg
For the consumer to get an accurate reliability question answered about a particular vehicle they might be interested in, Consumer Reports is but one source of info. One can start with them...but then move on to other reviews and assessments done by at least five to ten other organizations. Then it would be advisable to go to message boards focused on the car. After doing all that, a person would then get a good idea if a vehicle is/will be reliable for them.


I read my MILs CR a couple years back and the "other resources" they offered were just more CR published items. The idea to go to a forum for your car for another data point, they won't dare publish. Despite any forum being way more up to date than a printed monthly mag. CR is a nonprofit and it should be in their mission statement to be helpful, even if the helpfulness is a referral to sources outside their control.

Forums are at least peer-reviewed, not in the formal sense, but in the "this guy knows his stuff, listen to him" sense. CRs lab doors and methodologies are closed to the public. They'd probably say it's so manufacturers can't cheat their tests (like VW cheated EPA!) but who knows how strong they are in any one category they test?

 

[itguy08]

I'd say car forums are heavily biased on the negative. Same for any other forum. They may be good fir figuring out what are particular issues with a car (like common issues) but really don't tell much about overall reliability.

Why? Cause most post only when there are issues. How many posts on any forum are "my car (or widget) is awesome"? Very few. Most are of the "my car (or widget) is doing xxx how do I fix it.

A blended approach is best but when all the data says roughly the same things you can't say it's all bad. Unless it doesn't favor your favorite brand.