S. Jackson and S. Jacobs' article entitled "Generalizing about Messages" represents an important methodological innovation in communication research. The article urges procedural change by recommending the use of multiple-message designs because of a weakness in single-message designs; it recommends analyzing message-replication data with a random-effects analysis because such a procedure provides the appropriate basis for the generalizations that are wanted. That is, "Generalizing about Messages" proposes both a certain sort of data-gathering procedural change (i.e., message replications) and a data-analytic procedural change (i.e., random-effects analysis). Indeed, "Generalizing about Messages" has significantly changed customary research procedures in communication. It is much more common now to see multiple message designs. "Generalizing about Messages" is unique in its exploration of the interconnection of procedure and substance--it does not separate issues of methodology from issues of substance. Its implications are far-reaching. For instance, the same arguments underwriting the use of random-effects analyses in primary research designs with multiple messages also underwrite the use of random-effects analyses in meta-analysis. Broadly put, the relevant general principle is that replications should be treated as random whenever the underlying interest is in generalization. The article also invites consideration of the variability of effect on a phenomenon. (Contains eight references.) (TB)