Define and distinguish between confounds, artifacts, and noise.

Which of these three problems is typically reduced or eliminated by:

(a) choosing a homogeneous rather than a heterogeneous sample. 

(b) using random assignment, (c) using random selection.

(d) creating experimental conditions that differ along only one meaningful dimension.

(e) maximizing experimental realism in a study? 

How do these three concepts relate to internal and external validity?