To expand on what Ken wrote:
https://literature.rockwellautomation.com/idc/groups/literature/documents/um/1734-um001_-en-p.pdf
N.B. likely typos pp. 102 and 104: 100V should probably be 10V.
Yes, the High and Low Engineering values' NumericUpDown controls in that dialog window expect 16-bit signed integer values (-32768 to +32767).
I suspect the only requirement is that the high and low values cannot be the same.
The greater the difference, the greater the resolution (I think you meant resolution when you wrote "accurate reading"; accuracy is another thing entirely).
Whatever high and low values you specify will determine the range and scaling of integer values that will be read from this card by whatever entity (PLC input scan?) reads this card. If your input pressure sensor/transducer outputs 0V at 0bar and 10V at 16bar, then the [Low Engineering] value will be read from the -IE2V at 0bar and the [High Engineering] value will be read at 16bar.
That is why Ken suggested e.g. Low=0 and High=16 x 10**N where N=1 or 2 or 3 etc.; with e.g. N=3 and High=16000, the 16-bit signed integer reading will be mbar, which makes converting the
integral "engineering" values to
actual engineering units a trivial exercise.
Another option would be to leave the values at the defaults i.e. 0 and 10000: the inputs to the PLC are then mV. The advantage there would be for techs diagnosing possible problems with the -IE2V, because they could directly compare voltmeter readings of the card's analog input to the digital "engineering" value output by the card.
For maximum resolution, put the max and min values in: -32768 and +32767; although the actual resolution of the -IE2V over 0-10V is only 15 bits i.e. 10V/32767 = 305uV (the 1734-in027 document has a likely typo that says 320uA).
The -IE2V stated accuracy is 0.1% (10mV?), so any "accuracy" inferred from High values greater than 1k to 10k or so is an illusion.
Whatever you use, you are going to have rescale it to bar (unless you use N=0 i.e. Low=0 and High=16 and are satisfied with a resolution of 1bar
)
Whatever you do, if at all possible add it to the comments of your PLC program (or whatever) near where you scale the value to actual engineering units, which I suspect will be a floating point value in bar units.