Chi square tests for deviations in frequencies from expected frequencies. ANOVA tests for differences between means of numerical variables. They are quite different. Can you provide an example of your data? If you only have 3 samples (n = 3), I doubt you will get any significant results however.
Chi Square test is non-parametric test and suitable for categorical data. One way ANOVA is parametric test for measuring differences between mean values. It totally depends on the nature of data.
Chi square may help in comparing standard deviations of two different groups while anova study compare mean, these are quite different. I would suggest ANOVA study for you. It is well recognised and may also help in determining significant figures.