Hi, I have just finished a one-way ANOVA on a dataset looking at data across four regions over 3 years. The Levene's statistics are all fine (all over 0.05 by a large amount) so I have stated we don't reject H0 as there is strong evidence (with the ANOVA data included) to say that there is no difference between regions over the 3 year sets. The thing is, I also did a post-hoc (LSD) and found a statistically significant result in one year with 2 regions having a p-value=0.019 i.e. a comparison of West and South regions in 2009 had a p-0.019). So, is it correct to still retain the H0 when Levene's and ANOVA both strongly support this despite the one finding in the post-hoc, and therefore allow me to just say that whilst there is a difference in that region comparison on its own, the hypothesis still stands? Or do I need to overall reject H0 because the post hoc did find a significant difference <0.05? Thank you, Vanessa