Rationale for blindness to treatment assignment in between-group experiments testing generalized communication and language

Abstract Blindness to participants’ assigned treatment group is an essential protection against detection bias in between-group experimental designs testing treatment effects on generalized communication and language competence. Blindness to treatment group cannot be achieved when the other people in the communication sample (e.g. parents or peers) or the reporters of children’s communication or language behavior (e.g. parents or teachers) are also the implementers of the treatment. Internally valid between-group experiments measure generalized communication and language in ways that are constant across treatment groups by examiners who are blind to treatment assignment. When language or communication samples are used to measure generalized skills or developmental level, staff members who are blind to treatment assignment must conduct the data reduction methods (e.g. transcription, verification, and coding). Otherwise, the children’s skill level can be inadvertently over-estimated in the treatment group relative to those in the control group.