One difference between the A-Shell ISAM-A and the AMOS ISAMPLUS is that ISAM-A files grow automatically as data is added to them. This can create a bit of confusion due to the fact that when the file is first created, regardless of how many records you asked for, the size of the DAT file will be zero, and the number of records allocated according to the stats will be 100 million (which is just a dummy number, intended to assure your application that it didn't need to worry about file expansion). The IDX file will start out with a few blocks to accommodate the index structure and the meta information about the index.
Note that although the files will expand themselves as you add data to them, they will not automatically contract themselves as you delete records. Instead, deleted records are marked with a special control byte, and re-used for subsequent add operations. If you delete a large number of records at one time, you may want to compress the file by dumping it and reloading it.