Simply Statistics
Hypothesis Testing
Hypothesis testing is vital to the world of applied statistics. What it hopes to accomplish is the comparison of two mutually exclusive statements, the null hypothesis (Ho) and alternative hypothesis (Ha), to determine which is best supported by collected data.
The Null Hypothesis
Transcribed as Ho, the null hypothesis is the prediction that there is no difference between the samples being compared.
The Alternative Hypothesis
Transcribed as Ha, the alternative hypothesis is the prediction that a significant difference does exist between the samples being compared.
One-Tailed vs. Two-Tailed
A one-tailed or directional hypothesis is defined as a prediction in which the region of rejection only exists on one half of the sampling distribution. Essentially, the nature of the effect of the independent variable is predicted.
​
A two-tailed or non-directional hypothesis is defined as a prediction in which the region of rejection exists on both sides of the sampling distribution. It is predicted that the independent variable will have an effect, but the nature of that effect remains unknown.
Goals of Hypothesis Testing
Reject the Null
This is the goal! When you reject the null, the alternative hypothesis is accepted. This mean a difference does exist between the samples being compared.
Fail to Reject the Null
In this instance the null hypothesis is accepted. This means that there is no difference between the samples being compared.