[rmh47]

I think they mean four animals per test condition, i.e., each combination of humidity and temperature, which would require several dozen animals total.

That's still a very small sample size, but valid for a first investigational study, i.e., enough to justify further research.

[Right Wing Assault]

It’s a start, but each sample of 4 g-pigs would contain ones with different resistance to the flu. If one sample had one with high resistance, he alone would skew the results by 25%. And if another group had one with a low resistance, the same would happen in the other direction. In other words, maybe the conditions have nothing to do with it but the varying resistances of the individual pigs do.

A larger sample of dozens would be less skewed by individual pigs picked by chance with high or low resistance.

If I picked 4 kids at random and tried to sell them drugs, maybe 1 will buy. Then on Sunday, I pick 4 more kids at random and 2 buy. Headlines say, “Kids more likely to buy drugs on Sundays!”