Question

1. A researcher chooses to use ? = .01 rather than ?. = 05. What effect does this have on the probability of making Type I and Type II errors?

2. In most research situations, we use a two-tailed test instead of a one-tailed test (even in the case of a directional hypothesis). Why?

Answer 1

(1) We know that probability of making type I error is equal to alpha level

it is given that alpha level is reduced to 0.01 from 0.05

So, probability of making type I error is also reduced to 0.01

Relationship between probability of type I and type II error is this that these two probabilities are inversely proportional. When probability of making type I decreases, the probability of making type II error increases and vice versa.

So, in this case, the probability of making type II error will increase as the probability of making type I error is reduced.

(2) In most research situations, we use a two-tailed test instead of a one-tailed test (even in the case of a directional hypothesis) because a one tailed hypothesis is not as conservative as two tailed hypothesis. One tailed hypothesis used only direction to make the final conclusion whereas the two tailed hypothesis uses two directional system to make the final conclusion. So, a two tailed is more conservative as compared to one tailed hypothesis, that is why two tailed hypothesis is used most of the time.