Recently, we proposed quantum language (or, measurement theory), which is characterized as the linguistic turn of the Copenhagen interpretation of quantum mechanics. We believe that this language has a great powet of description, and therefore, even statistics can be described by quantum language. In this paper, we show that ANOVA (analysis of variance (one-way and two-way)) can be formulated in quantum language. Since quantum language is suited for theoretical arguments, we believe that our results are visible and understandable. For example, we can answer the question "What kind of role does Kolmogorov's probability theory play in ANOVA?" That is, the readers find that Kolmogorov's probability theory is merely used in order to calculate multi-dimenstional Gauss integrals, and thus, they can avoid to confuse the relation between Kolmogorov's probability theory and statistics.