About converting .sav to Stata
Moving a dataset from SPSS to Stata is the one conversion where a spreadsheet detour hurts most. Go through CSV and you lose every label, every declared missing value, and the date encoding. This converter writes a real Stata .dta file (format 118, readable by Stata 14 and later, and by pandas and haven), so the metadata survives: variable labels are carried across, value labels become Stata label sets, and dates are re-based from SPSS's 1582 epoch to Stata's 1960 epoch — a shift of 12 billion seconds that silently corrupts dates when done by hand.
The two formats are not identical, so a few things are adjusted along the way and reported to you after the conversion. Stata variable names allow at most 32 characters of letters, digits, and underscores, so other names are sanitized (the original name is kept as the variable label if there was none). Stata attaches value labels to integers only, so labels on non-integer codes are noted and dropped. Strings longer than 2,045 bytes — Stata's fixed-width limit — are truncated. System-missing values become Stata's "." missing.
The alternative route is R's haven package or Python's pyreadstat, which are excellent but require a working installation and a few lines of code. This page produces the same kind of file from a drag-and-drop, entirely offline in your browser — appropriate for restricted-use data that must not touch a third-party server.