Statistics Help: Does increasing N reduce the risk of making a Type I error?

[webnewland]

Type 1 error: Rejecting the hypothesis incorrectly

so does increasing N reduce the risk of making a Type I error? I'm thinking the alpha level we set (normally .05) is the same no matter how many N we use, so the more leptokurtotic graph still have .05 chance of getting type I error? Is this incorrect?
Thanks

 

[Vortex22]

Doesn't the alpha level = chance of Type I error?

 

[webnewland]

yup, so I thought if we decide alpha level at the very beginning, no matter how many N we test, our chance of rejecting incorrectly will always be the alpha level. But many of my friends seem to say that increasing N will reduced the risk of making a Type I error.

 

[Spoooon]

It's a stretch, but you could make an argument for how increasing N reduces the risk of Type 1 error.

 

[amnesiac]

Err..doesn't alpha itself distinguish the percent chance of making a Type I error?
If alpha = .05 = 5% chance of Type I error, right?

 

[GoingUp]

yes

 

[webnewland]

could you guys elaborate on how can increasing N reduced risk of making Type I error please?
Thanks