Statistical null hypothesis (H_{0}) is a statement which we want to check. It is usually an ordinary boring hypothesis which we want to reject. For example, if we are going to compare effect of essential oils alone and in the presence of gatifloxacin, null hypothesis will state that there are no differences in the activity of oils. Alternative hypothesis (H_{1}) will state that there are differences.

Therefore, H_{0} states that there are no differences between activity of oils alone and in the presence of gatifloxacin. H_{1} states that there are differences in activity of oils alone and in the presence of gatifloxacin. When the null hypothesis is rejected, the alternative hypothesis is accepted.

P-values and significance levels

The P-value is defined as the probability of getting the observed result, or a more extreme result, in a case if the null hypothesis is true. For example, the null hypothesis states that there are no differences in the activity of oils with and without addition of gatifloxacin, and the P-value is 0.02. In this case, the probability that differences are indeed absent is 0.02, or 2%. Therefore, we can readily reject the null hypothesis and state that the differences present with the probability of 98%.

The significance level is defined as a P-value at which we can reject null hypothesis without significant risk of accepting wrong decision. The convention in majority of researches is to use the significance level of 0.05. If the P-value is less than 0.05, the null hypothesis can be rejected. In our example, P = 0.02, it means that we can reject null hypothesis about absence of differences in the activity of oils alone and in the presence of gatifloxacin, and we can conclude that activity of oils is significantly different. Sometimes lower significance level is accepted, such as 0.01 or even 0.001, and differences are considered as significant only if P<0.01 or P<0.001. It depends on a level of desired accuracy.