It is still a survey.

YMMV,
MidCow3

 

Of course it is completely unbias of thousand of owners. Ludicrous to dismiss results from thousands of owners participating on their own. Again your premise that only satisfied owners is totally unbased. Please tell us all how you know only happy owners fill out these surveys. If I or most owners got a survey from J D Power after three year of ownership I and others would be much more likely to fill out the survey if we had problems as opposed to being happy. Regardless that doesn't matter even if comparing happy owners with happy owners or unhappy with unhappy owners. The only way these surveys would not be strongly correlated with dependability would be if they surveyed happy owners of NX's and unhappy owners of X3's. if these surveys are useless by a Land Rover drive it for five years and report back to us all about your dependability.

 

My opinion, you take all available data points (Consumer Reports, JD Powers, forums & owners groups, youtube reviews etc.) in as data when you buy a vehicle. IMO NONE of them is perfect. Some are better than others in terms of their independence and survey approach (not printing survey results if not enough respondents for a specific vehicle for example).

Midcow, exactly what do you think is the "perfect source" you recommend based on your methods? You seem to like sniping at surveys without providing a best case alternative, one that exists today not pie in the sky.

PS Consumer Reports surveys all their subscribers about their vehicles (both new and used / older cars) for their published results, you're incorrect about their approach. For example, they've got survey results on mine when they are new, as well as after owning for 5 years. Also why would you survey "non-car owners" to get data about a car they never owned? That doesn't like it would be useful.

 

From a statistical standpoint, the perfect sampling is to pick a large enough set of random responses that will represent the entire data spectrum you are attempting to analyze. In the case of car surveys, for example Consumer Reports, you are only sending out the surveys to Consumer Reports subscribers. My contention is that the Consumer Reports subscribers is not a representative set of all car owners and is therefore biased. The true data spectrum for car analysis for example should include all people, non-car owners ( did they have a car in the past, do they use public transportation now, etc.), new car owners that subscribe to Consumer Reports, old car owners that subscribe to Consumer Reports, new and old car owners that were previous Consumer Reports subscribers ,car owners that subscribe to Consumer Reports, car owners that do not subscribe to Consumer Reports, other people that may or may not be ready to buy their first car, and all others. The entire data spectrum representing cars is really everyone, not just those who subscribe to Consumer Reports. You specifically pointed out "non-car owners". What is they owned that brand of car in the past? What if they test drove the car and found faults compared to their current car brand? The valid data, or at least one criteria would be how many of total population buy a certain car or specific car brand? Yes, is you are evaluating one specific car brand, or a specific car within that brand, you could limit only you that specific car brand for a more specific analysis: previous owners( why did they leave brand?, etc.), new owners, existing owners,

Again, look at the universe you are trying to analyze, get a representative set ( big enough to eliminate extremes, standard deviation, mean, average), the representative set should not be restricted to only subscribers of a certain magazine or further restricted to only subscribers who respond; a whole subset analyses could be performed on who respond to surveys and who doesn't , and why responded response reason: like, hate, other ?))

Surveys and polls are only as good as the accuracy of data received and how well the data are analyzed. Take all surveys and polls with a grain of salt.

YMMV,
MidCow3

 

Notice MC just keeps saying the same thing over and over..I guess it will stop when everyone gets all this right and agrees.

 

Sorry @romer, just trying to explain statistics and sampling in laymen terms and bring it out of "Oz"

Peace to all


YMMV,
MidCow3

 

Didn't answer my core question, exactly where does that exist today? Can you share what the magic statistical source is that you're espousing? Or are you just blue skying in order to put down existing surveys who don't live up to an unrealistic standards? I agree with you CR is not statistically perfect, but it's one of the better sources that exists today (at least until you tell me where this mystery perfect sampling company is!)

So you've now explained what you mean by 'non-car owners', probably better called 'former owners'. I agree its good to get input from former owners (like here on this forum), but somebody who test drove one for 15-30 minutes is an impression at best IMO. I wouldn't want that mixed in with actual owner statistics or REMOTELY valued as highly as an actual owner who has lived with that vehicle for months or years.

 

[QUOTE=DaveGS4;11806554]Didn't answer my core question, exactly where does that exist today? Can you share what the magic statistical source is that you're espousing? Or are you just blue skying in order to put down existing surveys who don't live up to an unrealistic standards? I agree with you CR is not statistically perfect, but it's one of the better sources that exists today (at least until you tell me where this mystery perfect sampling company is!)

Dave GS4--great question. One that several others have asked previously and the response you'll get will be the same as before. Crickets.

 

In fantasy land he will have perfect statistical data. Meanwhile in the real world CR has the best data.
I live in the real world.
¯\_(ツ)_/¯

 

@DaveGS4 , sorry I looked back and you didn't ever specifically ask "exactly where does that exist today?" You asked what do I think is the "perfect source" which I answered from a pure statistical standpoint.

Your solution is a good method to evaluate current cars, by looking at and assimilating all car surveys. The assimilation process is somewhat nebulous especially if two different surveys has dramatically different answers. Again reading car review helps better in understanding. Motor Trend Reviews and new car review are helpful. of course they are not infallible, as Motor Trend named the Renault Alliance as Car of the Year in 1983 ( yes I have been looking at cars for a long time)

What do I do? Sure I look at CR JD Powers, etc,. I also read extensively the new reviews and do a lot of analysis on my own. I also typically build a Kepner Tregoe model to more objectively quantify my thoughts. Kepner Trego, you basically list all the attributes you want to evaluate , then you weight each attribute, and then you value each candidate against each attribute and provide a value score. The value score is multiplied by the weight for an attribute score. all attribute scores are assed and the highest overall score wins. In most cases I have hit winners with no regrets,

For new cars, which I typically bypass most of the surveys because they are not completely helpful because they talk about past/older cars. But, yes, they give an indicator and I quickly review them.

Thanks for you responses and comments.


YMMV,
MidCow3

 

I work in manufacturing engineering. Most companies collect engineering and manufacturing error data through root cause analysis. I’m assuming this is the information MC is talking about. This data is for internal consumption and never released to the public.

 

I’m a retired Chartered Accountant. In the 1980’s I had to study Statistics for my exams and I hated it. This thread reminds me why I hated it🤪…. But I have weirdly really enjoyed reading it every day!😬

 

Unfortunately, some posters here need to be right and only their opinion matters. When you call them on their "stuff" they do the dance, sort of like politicians running for office who wont answer questions directly, they provide a large volume of words w little substance . Personally I ascribe to Dave's method of evaluation in post #33. I've had Lean/Six Sigma training and certification but that doesnt mean I use it to pick a wife or choose a car..... just my not so humble opinion....

 

Looking at my Daughter’s ex-husband it is a crying shame she didn’t use Lean/Six Sima to choose a husband!😂

 

I am the original poster of this thread. I came across this article and forward it to NX owners because I thought it would make them feel good about the vehicle they purchased. Number four in most satisfied car is pretty good stuff. After all this is the NX fourm it's not a Tesla or BMW forum. It is hard to believe the controversy that this post has caused. I have read all these post this morning and find it hilarious where this thread has gone. Also find it sad and for most a complete waste of time. Remember what the original post was about. There's one member and I think we all know who I'm talking about has to get in a pissing contest with every thread. He has an NX and he can't let it go that it's the fourth most satisfied car. He owns this vehicle. But again he does this with every thread. The one about oil change interval he's telling people not recommending not passing on information about interval changes but saying you must change your oil at 5,000 miles. Dave there's some websites where you can mute members. Meaning their post don't show. Is that available on this website? Probably better he can mute me and will not see any comment or article praising his purchase. Again there's something wrong with someone who reads something about a vehicle that they purchased an expensive investment that positively reinforces their purchase and they get into again a pissing contest about its wrong. Imagine the fury this post would have received from him if it showed that the NX was at the bottom 10% of satisfied car owners.