In: Statistics and Probability
1. The minimum probability of rejecting the null hypothesis that is most acceptable in the field of science is .05. Why are researchers generally less willing to raise the probability of rejecting the null hypothesis to levels that are greater than .05?
The level of significance is , that is defined
as the probability of discarding the null hypothesis given the fact
that it was actually correct. It is also known as Type-1 Error and
this value is always fixed in advance in case of hypothesis testing
and interval estimation. Mathematically:
When , this
implies that there is 5% chance of rejecting a correct null
hypothesis. The type-1 errors are to be dealt with utmost care
because it implies the probability of declining a true null
hypothesis and hence making incorrect inferences about the
population. Thus, if the researchers raise the value of
, it
would mean that a greater percentage of correct null hypothesis are
being rejected and hence a greater number of incorrect inferences
about the population are being stated.
Hence, increasing the value of type-1 error, that is, is
not favorable.