OK, I am still confused by this short/float... integers/fractions issue.

You say that short datasets may appear as integers only (say, when I look at the value of a particular time point in the afni viewer graph window), but that a scaling factor is stored within the file to preserve the precision of the measurement (i.e. a value may show up as 1234, but the fact that the actual number is 1234.1234 is still preserved by the scaling factor in the associated file).

So, presumably, I could do something like use 3dcalc to multiply a dataset by .2789 and store the output dataset in short format:
3dcalc –a TS+orig –expr “a*.2789” –datum short –prefix TS_2+orig

Then, I could take that short output file, do another 3dcalc operation on it (this time * 1), but this time, I will force it to be float:
3dcalc –a TS_2+orig –expr “a*1” –datum float –prefix TS_2_1+orig

Now, if the short file (TS_2+orig) has preserved at least some of the fractional parts created by multiplying by .2789, then I should be able to see some fractions now, in the TS_2_1+orig, b/c I’ve essentially just changed it from short to float (in order to reveal the decimal value that I crave), right?

When I actually do these calculations, I do NOT produce any visible fractional parts in the TS_2_1+orig file. What am I missing here? Please straighten me out.