OK, so you have no idea how statistics work. quoting arbitrary percentages like this with no calculation to back it up.
What to make %? just take 22000/ 1.25 billion
See the above.
Are you sure? SO the chance of me getting my hand on the people who took part in the survey ....Say indians were miniature humans and I was randomly picking from 1.25 billion....what is the chance I would select 1 of the people surveyed?
Obviously you have never looked at how confidence intervals change from say picking 10 from 10,000 to say 1000 from 1,000,000.
According to you the sample ratios are the same therefore the confidence intervals are the same right?
And you claim I failed statistics.
Try read up on normal distribution and the effect of bell curve density on the fundamental theory of statistical modelling.
deals with the
existence of uncertainty either in observations or in the noise that drives the evolution of the system
Then why failed to apply it?
Was the knowledge just to be put on paper and not practical?
Oh it was very practical. Lets just say whenever you fly in an aircraft with a newer generation of PW engine in it....it will most likely use some of the results we derived both from numerical simulation and physical validation....specifically regarding better CFD modelling for the fuel combustion.
Simply taking the layman wiki definition of stochastic with cheap remarks is not going to cut it I'm afraid.
so you say that less than 0.01% is an ACCURATE sample of 1.25 billion?
With perfect random selection yes. In fact it can be much lower given the already large size of the sample (22,000). The overall population size is not such a massive factor after a certain size in the sample is reached. If you have handled the most basic of normal distribution data series, you would know this.
Thats why I am saying you can only come across this from a bias angle of the sample itself, not its size.
So RSS or anything extremist in india can be extrapolated to all of indians? awesome!
That would be a clear case of totally non-random sampling and a very high input bias. So short answer is no.
do tell me in which experiment on this planet do you take a samples size of less than 1% and say it is a CLEAR REPRESENTATIVE of the whole 100% ?
Again it is quite apparent you have never picked up a book on normal distribution or run some numbers beyond what you read in the Z-tables.