Which is not unlike doubling-up at blackjack, which will get you thrown out of a casino. Start with 25 participants, then if you don't get the result you want,

Which is not unlike doubling-up at blackjack, which will get you thrown out of a casino. Start with 25 participants, then if you don't get the result you want, add 50 and try again. Carry on doubling until you get the result you want. Not a statistician here, but as each addition is essentially a new trial, shouldn't they have applied Bonferroni correction to the results?