A frequent problem is how the intensity values and how they are congested by the software. You frequently provide to your software 2theta-Intensity data. That works well if your Intensity data is true intensities in "counts". Then, and only then error^2 = Intensity.

If for whatever reason the intensity has been scaled, e.g. the software simply provides 2theta-intensity/time data and the progam does not know, the caclulated errors (which go into the weights) are incorrect, and the whole meaning of Rwp and Rexp gets lost. In that case an appropriate format is 2theta-intensity/time-error, where you have to provide the correct error. The same can be used if you use variable counting time and if you use some detector where the output is some intensity value, but the intensity is not error^2.

Note than in any case, the whole classical conting statistics considers that the difference between I_obs and I_fitted is due to random counting errors and not due to systematic errors of your fitting. But the better your measured date get, the more likely (ensured) it is that your difference is dominated by systematic errors in your model.

Agree with Andreas Leineweber, could you provide the screenshot of your refinement? The best practice the peak-to-background ratio rather than the signal-to-noise ratio. Below is the example (refer to the attached paper):

Thank you everyone for your responses. Andreas Leineweber your answer was most helpful, and I closely examined my input files. It turns out that I was using XY files (which contain 2 columns: 2theta and counts/second). I was missing the third column of data which, as you said, is the error in the counting rate. Thus, I think GSAS was making up its own value for this error, giving me a high Rexp. I tried using an XYE file as my input file (XYE contains 2theta, counts/second, and error) and my Rexp values immediately dropped from 53 to 3 for the same data set. Thanks again, and I will make sure to use XYE files moving forward.

2theta and counts/second in association with .xy file: exactly that happened recently in our group a few months ago.... I think this issue should be repeated on a yearly basis in connection with the radiation safety instructions.... :-)

Andreas Leineweber agreed!!

How do convert xy data format into xye format? or Do i need to perform xrd scan again?

Naresh Prajapati the “e” is the error in your “y” (intensity), and should be collected by your diffractometer. Try exporting/saving your data from the instrument as an xye, rather than as an xy. You cannot convert xy to xye because the “e” is an extra, collected data value.