Chi-square

I want to see if the mean frequency of using irony by teachers is statistically significant when elementary, intermediate and advanced classes are compared. So I suppose language proficiency (elementary, intermediate, advanced) is my independent variable and the mean frequency of irony use my dependent variable. In practice, I want to see if these numbers 0.5, 1, and 1.75 (mean frequencies of irony for elementary, intermediate, advanced classes)are significantly different to say language proficiency significantly affects irony use by teachers or not. I thought of chi-square but chi-square needs at least two columns. Also, I'm not sure if anova is OK because my data (the frequencies) are not continuous data. What do you suggest? please help.Thanks.