Hi Jonathan,
Firstly congratulations on getting as far as you have with ISIS. Although it is very powerful, particularly for producing research grade spectra from more advanced spectrographs than the Star Analyser, it can be tough to master and most people use different software with the SA like Visual Spec, RSpec, BASS project which can be easier to use with the SA. (I tend to use mixture of VisualSpec and ISIS, mainly because that was what was around then). I think you actually are ok though with what you have done and I suspect the flat line is just because the resulting spectrum has some really high and low values at the edges outside the wavelength range where there is good data and ISIS has autoscaled to include them. If you rescale manually or crop the bad edges (using the crop tool in the profile display) hopefully you should be able to see the result.
There are also some other resources on the website here that you might find useful
https://britastro.org/section_information_/equipment-and-techniques-section-overview/spectroscopy
David Boyd’s “Scientific analysis of amateur spectra” at the top is an excellent overview of the whole subject.
Particularly for the Star Analyser, my slides for the tutorial “Low Resolution Slitless Spectroscopy” downloadable here
https://britastro.org/document_folder/baa-document-store/past-talks/spectroscopy-workshop-at-the-nlo-oct-2015
Also the document on my website here specifically tackles the subject of response correction (flux calibration)
http://www.threehillsobservatory.co.uk/astro/Relative_flux_calibration_20221222.pdf
Cheers
Robin