ABSTRACT
In developmental studies, we often sample data from numerous subjects on numerous occasions. We hypothesize for a model of certain attributes and attempt to test its validity empirically (Diggle, Liang, & Zeger, 1994; Kessler & Greenberg, 1981; Lindsey, 1993). In fitting and evaluating these models, we pay attention to several issues such as serial dependency (Jones, 1991; Jones & Ackerson, 1990; Jones & Boadi-Boateng, 1991; Nunez-Anton & Woodworth, 1994; Rosner & Nunoz, 1988), deterministic models (Tuma & Hannan, 1984), and individual variations in developmental processes (Baltes & Nesselroade, 1973; Nesselroade & Baltes, 1979; Nesselroade, 1991; Nesselroade & Boker, 1994; Wohlwill, 1973). Dynamical system models deal with serial autodependency and cross-variable dependency among variables across time (Brown, 1988; Coleman, 1964; Huckfeldt, Kohfeld, & Likens, 1982; Newell & Molenaar, 1998; Sheinerman, 1996; Tuma & Hannan, 1984; Vallicher & Nowak, 1994). However, most dynamic system models do not deal effectively with individual variations in a developmental process (Allison, 1990; Arminger, 1986; Brown, 1988, 1995; Coleman, 1964, 1968; cf. Nesselroade & Boker, 1994; Tuma & Hannan, 1984).
