Let's say you're putting together a mail merge listing the fees people in the county have to pay for getting their new cat license. You'd put together the spreadsheet data kind of like this, then create a database based on that spreadsheet.
Your mail merge document, with the various prompts and the field names, would look kind of like this:
Fee for license: <Fee>
So far so good. But let's say that sometimes there's no fee, perhaps organizations don't need to pay for cat licenses. So your data might look like this:
BUT that will not work because mail merges as of 3.2 (possibly earlier) don't like you to mix your types of data in a column. You can have all text, like N/A, or all numbers like 12, but not both.
Here's an example of some test and screen shots I took.
Here's the spreadsheet.
Here's what is in the database, and therefore what a mail merge would have access to.
So: in the scenario I outlined you could:
just put in 0, or just blank, instead of N/A
OR do this, which is workable but a bit more complicated.
Name Fee Applicable or Not
Bob $12 [leave blank]
Jean $13 [leave blank]
Library [leave blank] N/A
Then for your mail merge put both the fee and the applicable or not fields next to the Fee prompt. If there's no data the field won't take up any room, and only the correct field will show.
Fee for license: <Fee><Applicable or Not>
Note: This change in 3.2 really is a proper approach to databases; having mixed types is not proper database structure. But it is probably causing a few headaches.