Second, the table of evaluation designs omits the nonequivalent control group design, perhaps the most common of the quasiexperimental designs. Instead, it is relegated to a subhead of randomized control group designs called "incomplete randomization." The problem goes beyond using nonstandard terminology, as the nonequivalent control group design is not an incomplete randomized design because it is not a randomized design at all. One could imagine other designs being called incomplete randomized designs—mixed designs, or designs in which randomization is in force at only some times or for only some groups or persons within groups—but including this design here only adds to the potential confusion.