Research Modules 4: Null Hypothesis Significance Testing

Experimental Hypothesis (H1)

A predictive statement that anticipates effects, which could be a difference or association; also known as an alternative hypothesis.

Null Hypothesis (H0)

A hypothesis that posits no effect, relationship, or differences to be observed; is the alternative to the experimental hypothesis.

Null Hypothesis Significance Testing

A process involving proposing a null hypothesis about a population parameter, assuming it is true, and evaluating the probability of sample mean occurrence under this assumption to make inferences.

NHST: if sample mean is typical

then it does not provide evidence for a difference (does not reject null hypothesis)

NHST: if sample mean is extreme

the it does provide evidence for a difference (does reject null hypothesis)

Alpha Level (level of significance)

The probability value used to define the concept of 'very unlikely' in a hypothesis test.

Critical Region

Comprised of extreme sample values that are highly improbable (as per the alpha level) to be obtained if the null hypothesis holds true; sample data within this region lead to rejecting the null hypothesis.

if sample data falls in critical region

then null hypothesis is rejected