ABSTRACT
The purpose of research is to allow scientists to draw valid conclusions about relationships between variables. In some cases, these relationships are hypothesized in advance of data collection on the basis of theory, previous empirical findings, or experience in applied settings. In other cases, interesting relationships emerge unexpectedly from data. In either type of situation, researchers must have confidence that they are drawing accurate—that is, valid—conclusions from the data. Cook and Campbell (1979) described four types of validity to which researchers should attend: internal validity, construct validity, statistical conclusion validity, and external validity. Two other forms of validity can be added to this list: theoretical validity (Cook, Gruder, Hennigan, & Flay, 1979) and hypothesis validity (Wampold, Davis, & Good, 1990). Table 4.1 defines these types of validity, listing them in the order in which I discuss them.
