To accurately infer cause and effect, experimenters should use:

A. naturalistic observation
B. standard deviations
C. scatterplots
D. random assignment