In certain studies of the cohesiveness of small groups the operational definition of cohesiveness are logically deficient because they do not measure the dimensions of cohesiveness as nominally defined by the investigators. These studies are empirically deficient because single measurement of cohesiveness of the same groups are not positively and highly correlated. A unitary concept of cohesiveness is unacceptable because of the incorrect assumption that different aspect of cohesiveness are highly correlated. Additional limitations are lack of concern for the importance of the negative case in scientific investigations and overgeneralization. Alternative operational and conceptual definitions are suggested.