In scientific experimentation, a variable is best described as:
an adjustable condition in an experiment.
A comparative condition in an experiment in which nothing is introduced.
Something that must be consistently reproducible.
Subject to chage and revision based on evidence.