Originally Posted by InMyArms:
“Youtube dislikes and likes of his performances
Week 1: 4860 likes, 300 dislikes. (5.81% dislikes)
Week 2: 2551 likes, 683 dislikes (21.11% dislikes)
Week 3: 1657 likes,1197 dislikes (41.94% dislikes)
Week 4: 205 likes, 230 dislikes (52.87% dislikes)”
Does not prove Christopher is any less popular if you know anything about statistics in general....
Statistics are more accurate with a greater number of samples, and as above in week-2 there are approx one third less samples as in week-1
In week-1 you have 4860 likes +300 dislikes = 5160 votes
In week-2 you have 2551 likes +683 dislikes = 3234 votes
You would need the same number of samples as week-1 in all the following weeks to get the true picture.
Average dislikes over 4-weeks = 25% which means 75% of voters like Christopher.
You can quote statistics anyway you like to get them to show what you want.
This is what The Government and Speed Camera groups do in order to try and boost their campaigns,
but as you rarely see the raw data you can't see they have cherry picked the data in their favour.