JD Power-Car reliability

[dmclone]

No real big surprises here. I like this one because it focuses on 3 years of ownership. The things I don't like is that you really don't know how big the problems are.

Chrysler never does real well but having the bottom 4 is embarrassing. GM is always all over the map. Toyota is boring as hell but 3 of the top 5 is impressive. Ford is doing it right.



Lexus Back on Top of 2012 J.D. Power Dependability Study, Chrysler Last - WOT on Motor Trend



Brand, Problems per 100 Vehicles – Industry average 132 PP100
1. Lexus, 86
2. Porsche, 98
3. Cadillac, 104
3. Toyota, 104
4. Scion, 111
5. Mercedes-Benz, 112
6. Lincoln, 116
7. Ford, 124
8. Buick, 125
8. Hyundai, 125
9. Acura, 129
10. Honda, 131
11. Chevrolet, 135
12. Volvo, 143
13. Audi, 148
13. Audi, 148
13. Smart, 148
14. Subaru, 149
15. Nissan, 152
16. Mitsubishi, 153
17. BMW, 154
18. GMC, 158
19. Mini, 161
20. Mazda, 163
21. Suzuki, 167
22. Kia, 169
22. Volkswagen, 169
23. Infiniti, 172
23. Jaguar, 172
24. Ram, 174
25. Jeep, 179
26. Dodge, 183
27. Chrysler, 192

[Bobber]

No real big surprises here. I like this one because it focuses on 3 years of ownership. The things I don't like is that you really don't know how big the problems are.

Chrysler never does real well but having the bottom 4 is embarrassing. GM is always all over the map. Toyota is boring as hell but 3 of the top 5 is impressive. Ford is doing it right.



Lexus Back on Top of 2012 J.D. Power Dependability Study, Chrysler Last - WOT on Motor Trend



Brand, Problems per 100 Vehicles – Industry average 132 PP100
1. Lexus, 86
2. Porsche, 98
3. Cadillac, 104
3. Toyota, 104
4. Scion, 111
5. Mercedes-Benz, 112
6. Lincoln, 116
7. Ford, 124
8. Buick, 125
8. Hyundai, 125
9. Acura, 129
10. Honda, 131
11. Chevrolet, 135
12. Volvo, 143
13. Audi, 148
13. Audi, 148
13. Smart, 148
14. Subaru, 149
15. Nissan, 152
16. Mitsubishi, 153
17. BMW, 154
18. GMC, 158
19. Mini, 161
20. Mazda, 163
21. Suzuki, 167
22. Kia, 169
22. Volkswagen, 169
23. Infiniti, 172
23. Jaguar, 172
24. Ram, 174
25. Jeep, 179
26. Dodge, 183
27. Chrysler, 192

Mercedes has made some strides. Honda is slipping.

VW should be embarrased. Have never quite figured out why some consumers are fixated on European cars. They never score that great and are expensive to fix.

[ISU_phoria]

That's interesting.....I wonder why GM has such a wide range of reliability between its car lines? To be honest, I'm really surprised to see Cadillac that high.

Guess that shows the general belief of foreign cars being more reliable can't really be substantiated. (Chrysler brands excepted)

[clone52]

Mercedes has made some strides. Honda is slipping.

VW should be embarrased. Have never quite figured out why some consumers are fixated on European cars. They never score that great and are expensive to fix.

Wonder if the rankings take into account the severity of the issues. Having to deal with engine, transmission or computer problems is more of a pain that cosmetic issues.

[besserheimerphat]

#1 Lexus reported 86 problems per 100 cars over 3 years - that's 0.29 problems per car per year.

#27 Chrysler reported 192 problems per 100 cars over 3 years - that's 0.64 problems per year per car.

Not much difference across the 27 brands. By reporting "per 100 cars" it makes the numbers look larger and they appear to carry more weight. But I don't know anyone who owns 100 cars. My concern is how often should "I" expect to take my car in for unscheduled repairs during the warranty period. It rounds up to once for the most reliable brand, and twice for the least reliable brand - so the difference over three years is one unscheduled trip to the dealership. For all the attention this receives, would you be really upset if you had, on average, less than two unplanned trips to the dealership in 3 years? Remember that this is during the warranty period too (assuming you don't drive a lot). Everybody makes a pretty reliable car anymore. And even the "most reliable" brands have a lemon slip through from time to time. I wonder if any of those differeces are actually statistically significant?

Edit: I do not work for or drive a Chrysler, though I can see how my comment might imply that. But I do work with product reliability so this kind of report interests me.

[Krug]

Wonder if the rankings take into account the severity of the issues. Having to deal with engine, transmission or computer problems is more of a pain that cosmetic issues.

I would also like to know what the severity of the issues are. I would hope a bad transmission doesnt count for the exact same as a blown fuse or burned out headlight.

[dmclone]

#1 Lexus reported 86 problems per 100 cars over 3 years - that's 0.29 problems per car per year.

#27 Chrysler reported 192 problems per 100 cars over 3 years - that's 0.64 problems per year per car.

Not much difference across the 27 brands. By reporting "per 100 cars" it makes the numbers look larger and they appear to carry more weight. But I don't know anyone who owns 100 cars. My concern is how often should "I" expect to take my car in for unscheduled repairs during the warranty period. It rounds up to once for the most reliable brand, and twice for the least reliable brand - so the difference over three years is one unscheduled trip to the dealership. For all the attention this receives, would you be really upset if you had, on average, less than two unplanned trips to the dealership in 3 years? Remember that this is during the warranty period too (assuming you don't drive a lot). Everybody makes a pretty reliable car anymore. And even the "most reliable" brands have a lemon slip through from time to time. I wonder if any of those differeces are actually statistically significant?

Edit: I do not work for or drive a Chrysler, though I can see how my comment might imply that. But I do work with product reliability so this kind of report interests me.

Over the years I think this survey has been a good indication of overall quality and whether companies are improving. A few things to think about.

#1 Companies like Porsche are difficult to judge because they only have 3 models and low volume. New models tend to perform poorly so if Porsche has major changes to 2 of their models they could really be hurt. On the other hand, a company like Ford has a ton of models so one car shouldn't kill their ratings.

#2 This gives a good indication of long term reliability (more than 3 years). If you have double the amount of problems with a car in the first 3 years, when you shouldn't have any problems, there is a good chance that is a trend.

#3 This gives a good indication where a company is headed. 15 years ago Hyundai was handing out near the bottom. In the last 5 years or so they have skyrocketed to the top. They are now making quality cars and being rewarded.

#4 Your correct that the scores have improved over the years.

#5 I don't think this survey should steer anyone away from a certain make but if I was considering one of these makes I would defiantly do the research. If they were at the bottom year after year, I would be concerned. On the other hand, you could have some comfort in knowing that if you bought a Lexus the chances of you having problems is pretty slim. It's been in the top 3 since 1989 when it was introduced(I think).

[besserheimerphat]

#2 This gives a good indication of long term reliability (more than 3 years). If you have double the amount of problems with a car in the first 3 years, when you shouldn't have any problems, there is a good chance that is a trend.

This is the only part of what you said that I have a slight disagreement with - without having the actual data, it's difficult to say with any confidence when certain things will fail. The following is speculation based on my work as a reliability engineer, but here it goes anyway...

The automakers probably know with pretty high confidence (I'm talking statistical confidence from testing and field data, not "gut feeling") when certain parts should fail and I doubt any of those parts are known to have significant failure rates during the warranty period. Failures during the warranty period cost the company money, so they design around the warranty period to minimize that expense. Failures during the warranty period are likely random rather than wear-out, which means they are caused either by either environmental factors (ie catching a rock or bug in just the right spot, things that the manufacturer can't control) or unknown variations in part quality (which the manufacturer can sometimes control, but costs increase as quality control increases). And random failures can't be used to predict life. So while it's possible that the few warranty period failures can be an indication of overall life of the car, I'd say it's unlikely.

Ultimately, there isn't enough information in the article to draw any conclusions about any cars other than those used for the study - the data could be there behind the study, but the article didn't give the necessary details. Even just adding the sample size for each brand and the stdev to go along with the mean would do a lot to help determine if any of the differences are real or just a product of random sampling. And there is no data there that can truely predict the long-term maintenance costs of any of those brands.

[ISUAgronomist]

24. Ram, 174
25. Jeep, 179
26. Dodge, 183
27. Chrysler, 192

Good work Chrysler Group.

[vortec22]

One other thing I have noticed from reading car and truck forums. It seems that a lot of say BMW drivers for example will overlook some problems as "BMW quirks" whereas say a Ford or Chevy driver will complain to no end about the same thing. I think the European makes get a lot of leeway from their owners, and the domestics not so much.

Just my observations and opinions.

[CloneAggie]

FYI ... here is the JD Power press release along with some graphs and additional information. I absolutely agree with besserheimerphat here. Reporting the raw number doesn't really provide a lot of information. I would much rather see an error bar chart (which takes into account standard deviation and sample size rather than just the mean) for each brand rather than the bar chart based on the raw number. That way one could easily pick out the significant differences between brands.

2012 U.S. Vehicle Dependability Study | JDPower.com

[Bobber]

This is the only part of what you said that I have a slight disagreement with - without having the actual data, it's difficult to say with any confidence when certain things will fail. The following is speculation based on my work as a reliability engineer, but here it goes anyway...

The automakers probably know with pretty high confidence (I'm talking statistical confidence from testing and field data, not "gut feeling") when certain parts should fail and I doubt any of those parts are known to have significant failure rates during the warranty period. Failures during the warranty period cost the company money, so they design around the warranty period to minimize that expense. Failures during the warranty period are likely random rather than wear-out, which means they are caused either by either environmental factors (ie catching a rock or bug in just the right spot, things that the manufacturer can't control) or unknown variations in part quality (which the manufacturer can sometimes control, but costs increase as quality control increases). And random failures can't be used to predict life. So while it's possible that the few warranty period failures can be an indication of overall life of the car, I'd say it's unlikely.

Ultimately, there isn't enough information in the article to draw any conclusions about any cars other than those used for the study - the data could be there behind the study, but the article didn't give the necessary details. Even just adding the sample size for each brand and the stdev to go along with the mean would do a lot to help determine if any of the differences are real or just a product of random sampling. And there is no data there that can truely predict the long-term maintenance costs of any of those brands.

So Consumer Report data may be a better indicator since it covers a longer time frame?

[dmclone]

This is the only part of what you said that I have a slight disagreement with - without having the actual data, it's difficult to say with any confidence when certain things will fail. The following is speculation based on my work as a reliability engineer, but here it goes anyway...

The automakers probably know with pretty high confidence (I'm talking statistical confidence from testing and field data, not "gut feeling") when certain parts should fail and I doubt any of those parts are known to have significant failure rates during the warranty period. Failures during the warranty period cost the company money, so they design around the warranty period to minimize that expense. Failures during the warranty period are likely random rather than wear-out, which means they are caused either by either environmental factors (ie catching a rock or bug in just the right spot, things that the manufacturer can't control) or unknown variations in part quality (which the manufacturer can sometimes control, but costs increase as quality control increases). And random failures can't be used to predict life. So while it's possible that the few warranty period failures can be an indication of overall life of the car, I'd say it's unlikely.

Ultimately, there isn't enough information in the article to draw any conclusions about any cars other than those used for the study - the data could be there behind the study, but the article didn't give the necessary details. Even just adding the sample size for each brand and the stdev to go along with the mean would do a lot to help determine if any of the differences are real or just a product of random sampling. And there is no data there that can truely predict the long-term maintenance costs of any of those brands.

The test involves over 43k owners.

JD power also does short term tests and those almost always back up the long term data. If a car does horrible during the first 90 days it's almost garunteed to do bad in the long term as well. Like you said, no test is perfect but this one has a pretty good record. Studies like this have really brought to light when manufacturers like Honda have an issue. When Honda was having issues with their transmissions around 2000 it really hurt their ratings. It would be nice if they pointed out the problem areas for each model. I guess that's what consumer reports is for. I'd be willing to bet that if you look at the consumer reports for cars in 2009, Chrysler doesn't score well and Lexus is bulletproof.

[dmclone]

One other thing I have noticed from reading car and truck forums. It seems that a lot of say BMW drivers for example will overlook some problems as "BMW quirks" whereas say a Ford or Chevy driver will complain to no end about the same thing. I think the European makes get a lot of leeway from their owners, and the domestics not so much.

Just my observations and opinions.

It's funny that you mention this because I've heard just the opposite. Is someone in a Ford focus going to complain more about a rattle than someone who just paid 80k for a Lexus?

[vortec22]

It's funny that you mention this because I've heard just the opposite. Is someone in a Ford focus going to complain more about a rattle than someone who just paid 80k for a Lexus?

I don't really have any facts, just what I have read. I see people complain about a speedometer on a ford being off 3 mph while the bmw even states in the service manual it can be 10% + 2.4 mph off and everyone says "it always has been this way". Or the chevy fuel gauge reads low gets ******* about but the Volvo that reads high is not a big deal because its a Volvo thing. Just my personal observations, probably goes both ways.