1) All free testosterone tests are not the same. There are a lot of at-home tests that do not use reliable and valid collection or analytical methods (e.g. Everlywell). Even tests collected at a lab, if using an ELISA assay, can be inaccurate (our advisor, Dr. Rand McClain, joking calls them "Free Wrong T"), unless they're using a special analytical method (Mass Spectrometry/Equilibrium Dialysis). Our advisors in decades of experience find that rather than measuring Free T directly, calculating Free Testosterone is a reliable and accurate methodology.
2) You cannot compare a test result to another test result unless it was done with the same method at the same exact time (it's not "apples to apples"). Testosterone levels not only vary within a day, but between days, if affected by other factors (e.g. how much sleep you had, if you ate before, how recently you had sex, your stress/cortisol levels, etc.). It is possible for testosterone levels to vary significantly day to day if affected by these third-party variables, which could also explain your results if not accounted for by #1.