ABSTRACT
The frame of reference developed by Campbell and Stanley (1963) for the evaluation of internal validity contains three sources of error that are particularly relevant to descriptive developmental research and need additional emphasis: testing, instrumentation, and statistical regression (see also Chapter Five). These sources of error affect the internal validity of designs and, in interaction with age or cohort, external validity as well. Basic to these effects is the general issue of establishing measurement equivalence across occasions and persons (see also Chapters Seven and Twelve).
