Poll: Does/did your 2017 Raptor leak oil.

[jaz13]

I hate to tell you something else, but you have no idea what you are talking about.

But I do agree with you that a little knowledge is dangerous. And I also agree with you that your survey is not scientific nor representative. It may or may not be explatory, (I admit I dont know what that word means).

"this survey is conducting "explatory research" and is most definitely not a scientific or representative."

Since you do not understand how polls work, it might be easier for you to grasp if I just give you a couple questions to ponder.

Do you think sasquatch (who doesnt own a raptor) voted in your poll??? Do you think he voted more than once since he uses several different id's??? Does it matter to you if he voted???

Do you think that your poll results would have a different meaning if there were only 100 raptors sold total or if there were 100,000 sold???

Is this site just a hub for enthusiasts or also do customers with problems or gripes tend to be more frequent and vocal on websites?

No need to get your ******* in a ***, was just trying to make sure you took any poll results for what they are worth.

1) this poll is vulnerable to fraud just like every other survey ever given. but many respected members of this forum are complaining about oil leaks, so I'm quite confident this problem isn't the fabrication of a rogue member.

2) Population size doesnt matter at all. Thousand, million, or billion Raptors, it won't affect the results of this sample as long as the variation is the same across all the groups. Take a statistics class if you want to learn more.

3) Disgruntled customers that find this website through google write a scathing post and then never come back. Only enthusiasts visit this site daily and the vast majority of people who see and respond to the survey are regular readers of this forum. The odds of a whole bunch of disgruntled transient users finding this survey within hours of it being posted are low. Of course if tons of disgruntled people are coming to this site to complain about their oil leak, that would only happen if this problem was widespread.

Any other questions you need answered?

 

[WhatExit?]

I wonder how many times Sasbitch err Sasquatch has voted in this poll (he would if he could)

This "poll" only has value to help uncover who has/had an oil leak. Most who don't won't respond at all. Carry on

 

[df4801]

1)
2) Population size doesnt matter at all. Thousand, million, or billion Raptors, it won't affect the results of this sample as long as the variation is the same across all the groups. Take a statistics class if you want to learn more.

If you do not understand that your poll results have different meaning if the response rate is 100% vs. if the response rate is .001% then I can not help you any further.

Just keep telling yourself that population size doesnt matter, probably just like you tell yourself that ***** size doesnt matter either.

 

[jaz13]

If you do not understand that your poll results have different meaning if the response rate is 100% vs. if the response rate is .001% then I can not help you any further.

Just keep telling yourself that population size doesnt matter, probably just like you tell yourself that ***** size doesnt matter either.

I will make this as simple as I can. Lets start with three very different population sizes.

A = 1,000
B = 1,000,000
C = 1,000,000,000

All three have 10% failure rate evenly distributed through the population.

Now lets take a 100 unit sample from each population.

Sample A will have aprox 10 failures
Sample B will have aprox 10 failures
Sample C will have aprox 10 failures

Three vastly different population sizes with three nearly identical sample results. Population size doesn't matter.

Since you are so smart, explain to me why Sample A, Sample B, and Sample C will have different results because of the different population sizes.

 

[MKomertz]

3800 miles, no issues, was just under it yesterday.


Sent from my iPhone using Tapatalk

 

[df4801]

I will make this as simple as I can. Lets start with three very different population sizes.

A = 1,000
B = 1,000,000
C = 1,000,000,000

All three have 10% failure rate evenly distributed through the population.

Now lets take a 100 unit sample from each population.

Sample A will have aprox 10 failures
Sample B will have aprox 10 failures
Sample C will have aprox 10 failures

Three vastly different population sizes with three nearly identical sample results. Population size doesn't matter.

Since you are so smart, explain to me why Sample A, Sample B, and Sample C will have different results because of the different population sizes.

You are totally missing the point, so I will try to keep simple.

Say you have 100 respondents in your survey and Ford only sold 100 raptors. You would be very confident in your results of the survey. The data would be very meaningful!

Now say Ford sold 100,000 raptors and you still only had 100 respondents. Your survey is almost meaningless if you want to determine the % of oil leaks. Your data would be almost useless to predict the true rate of oil leak failure.

What if you only had 10 respondents and 4 had oil leaks? Would you conclude a 40% oil leak rate? If only 10 were sold then you may be right on. But what if 25,000 were sold???

POPULATION SIZE (raptors sold) dictates how many respondents you need to have meaningful data.

All of the above ASSUMES a properly designed and conducted poll. Which yours is NOT! So that throws things off even more.

My statistics lessons are done. Respond if you like, but I will not anymore.

 

[Dredlin]

12-12-2016 build no leaks,3263 miles


Sent from my iPhone using Tapatalk

 

[jaz13]

You are totally missing the point, so I will try to keep simple.

Say you have 100 respondents in your survey and Ford only sold 100 raptors. You would be very confident in your results of the survey. The data would be very meaningful!

Now say Ford sold 100,000 raptors and you still only had 100 respondents. Your survey is almost meaningless if you want to determine the % of oil leaks. Your data would be almost useless to predict the true rate of oil leak failure.

What if you only had 10 respondents and 4 had oil leaks? Would you conclude a 40% oil leak rate? If only 10 were sold then you may be right on. But what if 25,000 were sold???

POPULATION SIZE (raptors sold) dictates how many respondents you need to have meaningful data.

All of the above ASSUMES a properly designed and conducted poll. Which yours is NOT! So that throws things off even more.

My statistics lessons are done. Respond if you like, but I will not anymore.

Sorry, but you clearly don't understand statistics. I teach college statistics so if you want to disagree with me, that's fine. Statistics is a fascinating subject and incredibly powerful tool that shapes our everyday lives. But many parts of it are not intuitive, like population size not being relevant to sample size.

One of the interesting things you will notice on this survey is how the failure % has changed very little between 20 responses and now nearly 100. It's been in the teens the entire time and it would most likely stay there if we collected 1,000 or 10,000 responses. In surveys you know the result very early and additional data only gives you more confidence in the initial results.

Edit: The stable results between 20 responses and 90 responses also tells me the odds of fraud are low. If someone was going to mess with the results, they would most likely have done it early and over time further responses would quickly lower the average to the real level. Since this survey has shown a stable ~15% failure rate since the start, I don't think anyone is sabotaging the results.

 

[dlbb]

i have an oil leak. but not from the pan. it is a drip from the quick drain plug. is there a one use gasket? or a missing ring?

 

[df4801]

I teach college statistics

WOW, some school out there really owes some students a refund!