anova vs. chi-square test

Hi

I've read the following example:

items(beer(3) coke(4)) sold amount in last year sold amount in
this year
3. 9. 5.
4. 7. 8.
3. 7. 3.
4. 9. 6.
.......

It want's to study if there is a significant difference between
beers&cokes sold in lastyear&thisyear.

It uses spss one-way anova.

May I group these data into 2*2 tables, then use chi-square test?

Usually I confuse about these two tests? Would somebody tell me about
their differences?

Mike

.