The usefulnes of research on the attributes of innovations is mainly to predict their future rate of adoption. Most past research, however, has been postdiction instead of prediction. That is, the at- tributes of innovations are considered independent variables in ex- plaining variance in the dependent variable of rate of adoption of in- novations. The dependent variable is measured in the recent past, and the independent variables are measured in the present; so attributes
are hardly predictors of the rate of adoption in past research. Generalizations, however, about such attributes as relative advantage or compatibility to explain rate of adoption have been derived from past research, and these generalizations can be used to predict the rate of adoption for innovations in the future.