Saturn VUE reliability

[mkaresh]

Reliability information could be much more helpful than it is today. So a year ago I started collecting my own data. My focus will be on concrete, readily comparable stats like "times in the shop."

I haven't started collecting data on the VUE yet because not enough are signed up. I start when I have 25 for a model year (less for 2007s).

All results free of charge to those who help provide the data.

Details: Vehicle reliability research

Comments and questions very welcome--they make it much easier to improve the process.

[AzBlondeRocket]

I've seen many of your posts at GmInsideNews.com. You've got lots of good information. Hopefully you'll get a good response from the multitudes of SaturnFans.

[kodak_jack]

Just out of curiosity, are you a Reliability Engineer or statistician? I'd be interested in Mean Time to Repair and Mean Time Between Failures.

[mkaresh]

I've had some training in statistics, including a few years at the National Opinion Research Center while working on my Ph.D. at the University of Chicago, but I'm having someone who does this stuff for a living handle it for me. It's been a few years and I don't have time to relearn it all. Going to start fairly simple, but you can count on more sophisticated analyses as time and budget permits.

My wife is currently studying to be an actuary. Might be interesting what can be applied from that trade.

[kodak_jack]

Originally Posted by mkaresh

I've had some training in statistics, including a few years at the National Opinion Research Center while working on my Ph.D. at the University of Chicago, but I'm having someone who does this stuff for a living handle it for me. It's been a few years and I don't have time to relearn it all. Going to start fairly simple, but you can count on more sophisticated analyses as time and budget permits.

My wife is currently studying to be an actuary. Might be interesting what can be applied from that trade.

Mean Time to Repair is an indication of how long a certain fix takes. Having to order in parts because they aren't stocked ties up your vehicle. Taking a long time to repair is also something affected by the design of the vehicle (design for maintainability). If half of the vehicle needs to be taken apart to make a simple fix, that's a lousy design.
Mean Time Between Failures also refects on the design. If you can't keep the thing running because the same part keeps failing, it will get you frustrated in a hurry. From what I've read, some outside manufacturer needs to be designing new bushings for the Vue! Putting the same junk in there is doing us no good.

[mkaresh]

I've thought about MTBF. First things first, though. Need to get up and running. Later, the more data I have the more I plan to do with it.

Up to 18 2006 VUEs. Need just a few more to start collecting data on that year.

[mkaresh]

Getting there, but not there yet.

[CanNorthVue]

Great idea...just signed up. Hope you get 25 soon.

[mellissam]

Originally Posted by mkaresh

Comments and questions very welcome--they make it much easier to improve the process.

I think it is great someone is taking the time to this, but the method by which you get 'participants' may not be truly random...plus your survey sample size of 25 isn't large enough, for all intents and purposes. For fun, using a survey size calculator, I got 20% margin of error with 25 participants, and a 95% confidence level....
I certainly don't want to knock your efforts, just pointing out some flaws... (ex PR person with Cert. P.R. and B.Sc. (physics)). Realistically, you need to get a truly random sample (maybe via VINs?), get about 130 of them, then you'd have your typical 5% error and 95% confidence....
Well, good luck nonetheless....
ps - I am one of those pesky 06 VUE owners too tired to bother...and I don't have any problems, so why bother..???

[pdough]

I agree with Mellisam in that a lot of people are on this forum specifically looking for fixes to their problems. You may get a group of 25 people to join but how can you be sure it is a random sample since the group from which you are drawing has a higher percentage of people with problems?

The same probably holds true for the folks on the other manufacturers forums. You can get an idea of what types of problems people are having in general, but drawing reliability data from a group of people who are most likely having troubles of some sort seems risky to me. What about the other 200,000 people who have bought Vues and are not having trouble, or do not join forums...

[mkaresh]

It appears that neither of you has looked into the methods I'm using.

First off, I start collecting data with 25 owners signed up. I intend to have far more eventually. This is just to get started. This is not just a one time survey, but a continuous longitudinal survey.

Using the sample size as a reason not to participate...well, I won't comment on that.

You are correct about the confidence intervals. If anything they are wider than you suggest.

Second, with the exception of the first month I only collect data on problems going forward. The main survey does not collect repair histories (though a second survey does, and I state that it cannot be used to determine repair rates). So when people sign up neither they nor anyone else can say how many problems they'll be reporting. This is a big difference from the other reliability surveys, where people know what they'll be reporting at the time they decide to participate.

Third, the survey clearly defines what counts as a repair, providing as little leeway as possible for subjectivity. This further reduces the problem of not having a random sample. Random samples are most necessary when surveying opinions.

The proof is in the results. Even with very small sample sizes they've been consistent through two sets of results and with other such studies. Latest results:

TrueDelta Vehicle Reliability Survey results

[mkaresh]

Current counts:
2007: 9
2006: 17
2006: 8
2005: 15
2004: 13

Start collecting data at 25. (12 for 2007 because model year only partly over)

[binkybaby]

After a year and 21,000 miles on my 2006 Vue i have no problems whatsoever; i change oil every 3000 miles (some may see this as a little excessive) and i'm due for another one soon.

[mkaresh]

I'm finding that your odds of getting a problem-free car are higher than many people think.

Still looking for more VUE owners.

[mkaresh]

Now just four from the minimum for the 2006 and six short for the 2007. Hopefully I can start collecting data on these next month.

2003 and 2004 also more than halfway.

[russellhome]

I signed up my 2005 FWD V6 Vue. I have had no repairs in the first 3 years and wish I could report that in the first survey - but I understand the reasoning for excluding the dated information.

My biggest problem with CR reliability reports is that they are for vehicles that don't exist. In 2004 - the Vue had 3 different transmission options and 2 engines. Some combinations were more problematic than others, yet you don't get reliability ratings for each option. You get some sort of summary, I guess. So, for example -- if the VTi tranny would rate a black DOT and the 5 speed auto & stick rate a full or partial red dot - how accurate (or helpful to consumers) is it to report a white dot or white/black dot for transmission reliablity? The '04 Vue was not sold with a combo stick/VTi/auto transmission.

The CR mileage info and rating also bothered me. They only test drove the AWD V6 - the lowest mileage version of the Vue available. The info on the Vue detail page I don't have a problem with it because there it explains what model they tested. But on comparison pages of mileage with other SUVs, there is no detail about powertrain options on models tested -- so the value of the data for accurate comparison is suspect at best.

[mkaresh]

My guess is that most of their owners had the CVT.

The site has a separate Repair History Survey for people who want to report their prior experience. I created this survey because people kept sending me their repair histories.

The results of this survey are immediately posted as-is to the site. I don't analyze them.

This survey is currently available for nearly every U.S.-market 2003 or newer model, no minimum sample required.

Vehicle Repair History Survey

[BobbyP]

mkaresh - I added my 2006 VUE RL data.

One question under repair, it sounded like you wanted to enter the sale price of the vehicle as a repair?

[mkaresh]

I leave that up to each person, whether they want to report the purchase and sale prices.

I do want the month purchased and the current month recorded, just so people can tell the start/end dates.

I probably should automate the end date (another project onto the pile).

It's easy to delete an end date later on if you make a new entry. Just enter the month/year again, leaving the rest blank.

[06VuePcola]

Just added my 06'. Hope it helps