Survey: How long were you in the cult vs what year you were baptized
by ILoveTTATT2

ILoveTTATT2
I am giving this lots of thought.
The people who are out now and came in the 2000's are coming out faster, and here's why I claim that:
Let's say I get 100 responses from people that were baptized in 1970, and 100 responses of people baptized in the 2000's. Then we can compare. If the average of the 1970 group is 20 years (i.e. most left in the 90's), and the average of the 2000's group is 10, meaning most left in 2010, then the statement that the exit is speeding up is valid.
Given that I am discounting the ones that are in automatically, what this study shows is that,
"The current exJW community who is actively online shows a pattern of decreasing time spent inside the borg."

Oubliette
ILoveTTATT2: If the average of the 1970 group is 20 years (i.e. most left in the 90's), and the average of the 2000's group is 10, meaning most left in 2010, then the statement that the exit is speeding up is valid.
No it is not. Your conclusion does NOT follow from your premise.
If you would have somehow been able to conduct a similar survey of the 1970 group in 1985 you would have gotten an average much lower than 20 years due to the fact that you were surveying a 15 year time period. It could not have been longer than the period under consideration.
I suspect the relatively large percentage of those leaving after the 1975 debacle would actually have made the average Length of Time a JW for the 1970 Group much lower. But you are asking people forty years later. It's no coincidence that the average time a JW now is 20 years:
 (2015  1975) / 2 = 20 years

It would be more telling to try and determine what percentage of people leave the religion each year and has that trend changed.
Again, please see my reference on the previous page to Paul Grundy's work and read his detailed study. You'll find it most informative.
A few years ago, the Pew Forum posted a study that showed that JWs had the worst retention rate of any religion in the United States. But it did not account for converts, only bornins.
Pew Forum survey  Key Findings and Statistics on Religion in America (note this link goes to an updated analysis of the 2007 report referenced below):
Jehovah's Witnesses have the lowest retention rate of any religious tradition. Only 37% of all those who say they were raised as Jehovah's Witnesses still identify themselves as Jehovah's Witnesses.  (From a nationwide survey conducted from May 8 to Aug. 13, 2007)
 (2015  1975) / 2 = 20 years

Oubliette
BTW, I would not be at all surprised to find that the rate of defection from the JW religion has increased as a result of the internet and the corresponding access to disconfirming information. This is true for religion in general.
I am just pointing out that your calculations do not support that hypothesis.
You might find this article in Psychology Today informative:

gmatired2
1. Born in/raised
2.Baptized 1962 14
3 Left 1988
Stuck it out 40 years

Athanasius
1. Born in. My mother converted to JWism before I was born, my father was not religious.
2. 1955 Baptised age 9.
3. 1984 woke up and left the JWs the same year.
29 years from baptism to exit.
Internet played no role in my awakening. Reading Crisis Of Conscience caused me to examine JWism and escape the Organizx.

ILoveTTATT2
No Oubliette. If I had made a similar survey in 1985 regarding the people baptized in the 1970's, it is possible that the average response would be closer to the 15 years difference. Since 2000 we have had 15 years, and the average response is less than 10 years from baptism.
The ones baptized in the 90's had an average of 15 years... And it's been 2515 (depending on how you see the numbers).
There is going to be a "triangle", and that "triangle" is reducing.
To completely confirm my hypothesis, a repeat study including all of the ones who exited from 20152020 would be necessary. But I predict that it will be the same average or lower (i.e those baptized in 2010 will start showing up around 2017 or so).
If I am proven right, then the data will be a wonderful confirmation of the power of the availability of information.

Oubliette
ILoveTTATT2: If I had made a similar survey in 1985 regarding the people baptized in the 1970's, it is possible that the average response would be closer to the 15 years difference.
Sure it's "possible," if they all left in the early '80s. This is not only logically absurd, it ignores the historical facts what actually happened to the membership of the JW religion post1975. Not only is it NOT what happened, your response
Do you not know how to calculate a mathematical average and (more importantly) what it means? There are three kinds of averages: mean, mode and median. They are different and have corresponding applications depending on what one is trying to measure and report.
That notwithstanding: gathering, collating and correctly interpreting statistical data is significantly more complex than the informal survey you are conducting here.
I was trying to be a helpful, but your response shows a lack of interest in approaching this in a scientific manner.
Have you heard? There are three kinds of people in the world: those that are good at math and those that aren't.

Oubliette
ILoveTTATT2: If the average of the 1970 group is 20 years (i.e. most left in the 90's), and the average of the 2000's group is 10, meaning most left in 2010, then the statement that the exit is speeding up is valid.
Again, your conclusion does NOT follow from your premise.
You have presented this argument:
 If (2015  1975) / 2 = 20, and (2015  1999) / 2 <= 10, then "the exit is speeding up."

It is the logical equivalent of this:
 If Obama is President of the US and Anthony Morris III has a thing about tight pants and colored socks, then my dog is a cat.
 Barack Obama IS President of the US, and Anthony Morris III DOES have a thing about tight pants and colored socks, therefore: my dog MUST be a cat.

The significant difference is thatwhile I think there might actually be something to your conclusion even though the scientific data supporting it is sketchy* your conclusion does NOT follow from your premise.
                  

Simon
The people who are out now and came in the 2000's are coming out faster, and here's why I claim that
You can claim it, but unfortunately your statistics don't definitively prove it.
Not because your approach is fundamentally flawed or should be dismissed, just that there are other factors likely to influence the counts so you can't hold it up as "certain proof of XYZ". Any claims have to include the assumptions made when reaching the conclusion.
There will always be more people who have "just left" the religion when measured because the longer the gap between leaving and being measured, the less chance there is of being measured.
Surveys are interesting though as long as we realize they measure the people who are still hanging around. For more raw turnover stats I think the baptism + publisher counts are what will give the churn rate. But then those don't show the time factor which I think is the interesting bit you are trying to highlight.
It's still very interesting, you just need to include the footnotes to the conclusions to make it less sensational / harder to challenge.

ILoveTTATT2
I am interested in doing this as best as possible.
An average is not necessarily halfway in a data set, it all depends on its distribution.
If something has a 5050 chance, like a flip of a coin, the law of large numbers says that if you flip a coin enough times, you will have a 5050 chance of getting heads or tail.
If you are right with your math, calculating exactly half for the time difference, then that means that JW's going out of the religion is purely random and has a normal distribution. This would be really discouraging if found to be true, since there would be no factor that influences the exit of JW's. But our intuition tells us otherwise, and I hope to show it with data.I have thought about it, and just the formula was wrong. The formula should be: