ABSTRACT
This chapter describes what is meant by successful measurement and why it is particularly important for understanding development. It also describes some methods for establishing successful measurement. The chapter emphasizes the need for pragmatism as well as progress in measurement in longitudinal studies, examines Jack McArdle's contributions on one of their coauthored articles, and offers recommendations for further efforts on measurement in developmental studies. It briefly reviews three relatively well-known approaches which highlights some advantages and challenges: mechanistic modeling, latent variable modeling, and Rasch measurement. Given the need for valid measurement in order to make valid conclusions, particularly in developmental studies, it is necessary to identify methods that allow for strong arguments of representational measurement, especially interval measurement. Mechanistic modeling is predicated on the standard theory-to-model approach most common in statistical analysis, particularly in fields like cognitive psychology that aim to describe mechanisms or processes that lead to observed outcomes.
