Boxster Failure Survey Form

[Gundo]

I'm cross posting this from another board. LorenS has promised to capture as much information as possible about Boxster engine failure modes.

Read on:

As promised last week, and due to a significant amount of interest, I have created a "Boxster Failure Survey Form" for everyone to download. Instructions are on the form. I will be using the collected data to create a Statistical Failure Analysis for Boxsters. This efort will take me 3-4 months (weekends)depending upon the number or respondants.

Please note: It is very important that I get a significant number (hundreds) of responses. I want to hear from everyone, whether or not you have had a failure. I regularly read this board but not others, so if someone can provide a link for Babblers and any other forums, I would appreciate it. Deputies, can I please get a banner link to this form? It's in all our best interrests.

The collected data will be used to compile a database for present, previous, and future Porsche Boxster owners; and only pertaining to engine/transmission failures.

After I receive a significant number of responses, I will be correlating this data into a computer analysis matrix, looking for patterns, and attempting to create a statistically accurate “Boxster Failure Prediction Formula” for all of our use. All collected data will be made public upon completion of my work. The level of accuracy depends on the amount of responses and me and my computer finding correlation and patterns within the data.

I have no motives for this endeavor other that wanting some hard data on this long, ongoing PPBB discussion issue. I am not employed by PCNA, and have no association with anything related to the auto industry. Please get everyone you know who has had a Boxster to fill out a form.

To download: Turn off any download accellerators, click on the link, go to the bottom of the page, click on the free button, and click on the file to download.

Survey Form Boxster Failure

Edited April 25, 2005 by Gundo

[LVDell]

Not to burst your bubble but this is not going to be anything other than alot of just "obtained" invalid data. You have already started off on the wrong plane to obtain a "statictically accurate" result.

Think about the people you are surveying and who they represent. A very small % of Porsche owners and more specifically those with the desire to post on the boards. Not to mention you never get somebody posting that their engine is great. You see a large number of failure posts b/c this is the place to post them. Nowhere else. PCNA, PAG don't care and your letters to them usually go ignored...ESPECIALLY if outside the warranty period.

The only reason I bring this up is unless you have a significant amount of time and want to take this on as a hobby, nothing valid can come out of it.

Message boards do not represent at all the "population" of Porsche owners and for you to even have a chance a obtaining a valid result you would need to sample from THAT population with everyone being given an equal chance to be selected for response......NOT voluntary reporting which is going to severely bias your results.

You failure formula is not something you want to everyone to use since it is nothing more than your view based on inaccurate results and can only scare owners into a panic about anything obtained since your "failure formula" said so.

You really need to think about what you are trying to do here. This is not empirical in any way, shape, or form.

While I admire your endeavour, I just want to make sure you understand the reality of this from somebody who does this for a living.

FYI......my Ph.D. is in reserach methods and statistics.

[99-Boxster-MK]

I second that, as a scientist you are already biased by calling a failure form. You will get people with faliures to vent by filling out the form, maybe even in duplicate and you will probably get >50% failure rate, which is bogus.

Not to sound to critical, its a great idea you have. Why not rename the form to something neutral, take it to the local porsche dealerships and have them give it out to boxster owners. This way over a year you can get people coming for maintenance and such and get semi-accurate results, but still slightly biased as the chance of someone going to the dealership for a problem is greater then someone without any problems :), but it would be significantly more unbiased then an internet based survery.

Mike

Not to burst your bubble but this is not going to be anything other than alot of just "obtained" invalid data.  You have already started off on the wrong plane to obtain a "statictically accurate" result. 

Think about the people you are surveying and who they represent.  A very small % of Porsche owners and more specifically those with the desire to post on the boards.  Not to mention you never get somebody posting that their engine is great.  You see a large number of failure posts b/c this is the place to post them.  Nowhere else.  PCNA, PAG don't care and your letters to them usually go ignored...ESPECIALLY if outside the warranty period.

The only reason I bring this up is unless you have a significant amount of time and want to take this on as a hobby, nothing valid can come out of it.

Message boards do not represent at all the "population" of Porsche owners and for you to even have a chance a obtaining a valid result you would need to sample from THAT population with everyone being given an equal chance to be selected for response......NOT voluntary reporting which is going to severely bias your results.

You failure formula is not something you want to everyone to use since it is nothing more than your view based on inaccurate results and can only scare owners into a panic about anything obtained since your "failure formula" said so.

You really need to think about what you are trying to do here.  This is not empirical in any way, shape, or form.

While I admire your endeavour, I just want to make sure you understand the reality of this from somebody who does this for a living.

FYI......my Ph.D. is in reserach methods and statistics.

<{POST_SNAPBACK}>