01-02-2021, 01:33 PM
I'm afraid this is getting off-topic again ...
One cannot apply statistics in that way.
The more samples one takes, the younger the 'most recent' sample becomes. This is simply a consequence of the fact that the measurement error has some error distribution.
One cannot just take a number of samples, and then concentrate only on the extreme case (either way). This is a selective, or biased approach.
All samples play a role. They are all related to the creation time. Even if it took 10 years to create the MS, and the sheets were bought little by little, the sources that went into the MS contents would not have appeared at the very end of this time frame.
One cannot apply statistics in that way.
The more samples one takes, the younger the 'most recent' sample becomes. This is simply a consequence of the fact that the measurement error has some error distribution.
One cannot just take a number of samples, and then concentrate only on the extreme case (either way). This is a selective, or biased approach.
All samples play a role. They are all related to the creation time. Even if it took 10 years to create the MS, and the sheets were bought little by little, the sources that went into the MS contents would not have appeared at the very end of this time frame.