TABLE 1.

Summary of What Works Clearinghouse criteria for experimental designs.

Design element	Meets standards	Meets standards, but with reservations	Does not meet standards
Independent variable(s)	Actively manipulated by researcher	—	Researcher does not control changes to conditions
Dependent variable(s)	Measured systematically over time	—	No systematic measurement (e.g., anecdotal case study)
	Measured by more than one assessor	—	Only one assessor
	Includes interassessor agreement on at least 20% of data points in each phase	—	No interassessor agreement, only in some phases, or in less than 20% of data points
	Interassessor agreement meets minimal thresholds	—	Poor interassessor agreement
Length of phases	At least 5 data points per phase	3–4 points per phase	< 3 points per phase
Replication of effect	General:	—	< 3 replications
	3 attempts to demonstrate the effect at three points in time or with three phase repetitions
	Reversal/withdrawal:	—	< 4 phases (e.g., AB, ABA, BAB)
	4 phases per case (e.g., ABAB)
	Changing-criterion:	—	< 3 criteria
	3 criteria
	Multiple-baseline/multiple-probe:	—	< 3 cases; < 6 phases
	6 phases across at least three cases
	Alternating treatments:	Only 4 repetitions	2 treatments without baseline
	2 treatments compared to baseline or 3 treatments compared to each other
	5 repetitions of each condition		< 4 repetitions