Person1: But what about my choice
Person2: Science cannot measure that - so it is irrelevant
Preferences may be observed and measured in the context of which you talk about. Person 2 is possibly incorrect in the observation and any conclusion drawn from that can be doubted..
A fellow called Paul Samuelson proposed something about revealed preferences. It has taken on a few forms, none of which are without criticism (and that's fine), but its nevertheless illustrative.
I won’t go into details, but Wiki provides its usual barely friendly word salad.