How do I fix mis-called size standards and improve automatic calling?

You have to select any peaks in the size standard that it missed and unselect the extra ones. You can do that by right clicking on them in the electropherogram or in the list of peaks. This must be done when no locus is selected. This is the state it is in when you first open a run or you can click on the gray area above the loci to clear a selection.

What you really want is to have STRand do this automatically. For that you may need to tune up a few things. First is the size standard definition. For GS 500, we use:

35;50;75!;100!;139!;150!;160!;200!;250~;300!;340!;350!;400;450;490;500 (30%/10000)

The ! means the peak is required. This is an optional thing for you and depends on where you start and cut off the trace either by the settings in STRand when you load it or by your run parameters.

The ~ means not to use that peak for calculating sizing. ABI says the 250 peak moves around.

Most important is the bit at the end. The 30% ignores all size std peaks less than 30% of the tallest peak in that color. The 10000 ignores peaks that are under a peak in another color that is over 10000 tall as it is probably bleed-through ("pull-up").

You can also add another / and a number at the end in the parenthesis to indicate the scan line where you would like STRand to start looking for size standards.

The other things to do to improve size calling are to optimize how much you cut off the beginning of the data so that you are past the primer junk and to optimize your reaction so that it produces clean data.