Defining variables and avoiding bias are both parts of good experimental:
a) Design.
b) Analysis.
c) Reporting.
d) Publication.