The null hypothesis is that the population means measured different levels of the experimental factor are all equal. The alternative is that they are not all equal, or the factor is significant. The statistical significance level cut-off was chosen as P 0.05