A researcher asks herself the following question: ""Can effects within my experiment be attributed to variables that I intentionally manipulated?"" What property of behavioral measurement is the researcher addressing?