You are wrong.Statistical conclusions can be different depending upon the sample.
REP: I said can be.
What is wrong with that. They differ over a period of time.In case of statistics the time difference is sometimes less than a year.

Solve the following statistically:
1.Smoking is Injurious to health.
2.On 13th there are more accidents.
3.If you wear Red your chances of winning increases.
4.Not enough exposure to Real World reduces the capability of immune system.
5.Those who drink Wine are healthier than non drinkers.
6.Those who are Black are less likely to get a job and are therefore less intelligent(given the equal availability of opportunity).
7.Those who shave daily are less like to get a heart attack.
The list goes on and on...
And then there are brilliant exceptions.
Exceptions which make the Statistic look like a stupid logical formulation of thought for practical use for strong action.
The statistical truths are known to vary over sample space and time.
The conclusions are not like the Physical Truths.

Every year the truth changes depending upon the sample.