My data is normally distributed (for three groups), following one way ANOVA I get significance for two variables, post-hoc results are inexplicable based on literature over the past decade, how do I correct this or test where I am going wrong. I use statistica.