Confounds

Variables whose effects are so intermixed that they cannot be measured separately, making design of the experiment internally invalid and its results impossible to interpret.

To introduce bias into a research study.