Phrasing of Hypotheses

When you are phrasing your hypotheses, make sure that you term them in group differences language (i.e., couples were expected to have higher levels of sexual dissatisfaction after three years of marriage than after one and a half years of marriage).

Do not phrase your hypotheses in regression-type language (i.e., length of marriage was expected to be associated with sexual dissatisfaction, such that longer duration of marriage would be associated with higher sexual dissatisfaction).

That is, remember that ANOVA compares distinct groups, not dimensional levels.