You're probably confused, because you got two answers. Be careful about your definitions. I suggest you look at wikipedia for: Fresnel equations, reflectance (reflectivity), refractive index, transmission. Also check what the spectroscopy "bible" is in your group. You also didn't tell us what field this is. Since you are working  on Solar, I assume you want a Physics explanation.

There are two points to think about. On the one hand, you have you have your absolute detector signal for your sample and reference. On the other hand you have the reflectivity of your sample and reference. Don't forget about attenuance and mass attenuation/extinction coefficient (there is a QM derivation that provides the link between the classical complex index of refraction).

You divide your signals. You may find it easy to remember the by thinking of the conservation of energy, which dictates that the total has to be one (see, e.g., Fresnel).

You may need to correct your reference, if your reference has less than 100% transmission (depends on your setup, e.g. you measure low transmission samples and you need to filter due to the dynamic range of your detector). The same is true of reflectivity. Gold is often used, but this drops off

Both can be done...

Krish Ramesh

Sometimes wavelength, etc. sets for the reference and for reflection spectra are different. Before division these sets must be the same, synchronized, using linear or polynomial interpolation.

Origin, Excel, Matlab, Octave and many others packages can be used as tools to perform such division.