Conferences

The economy does not cease to exist in between observations nor does it function in discrete time jumps corresponding to the observations, remarked Rex Bergstrom, the pioneer of continuous time modeling in econometrics. In the same vein, opinions and attitudes do not stop to influence and to be influenced between interview waves. Especially for the longer intervals of one or more years usually chosen in panel studies, standard discrete time modeling is a simplification and often a distortion of reality. Discrete time analysis gets into extreme trouble, when observation intervals within and/or between studies are unequal. Autoregressions and cross-effects between variables from different intervals become incomparable. Continuous time analysis is needed to enable correct comparisons and to avoid the contradictory results easily produced by discrete time analysis.

The first part of the course will go into detail about the problems of discrete time analysis, illustrated by the behavior of reciprocal effects in the cross-lagged panel design. Monotone and nonmonotone autoregression and cross-lagged effect functions will be central in this part. Next the basics of the linear stochastic differential equation model will be dealt with. The use of LISREL for the estimation of the ADM (approximate discrete model) and Mx for estimation of the EDM (exact discrete model) will be explained. It will also be explained how panel data collected with different intervals (from different regions, different countries) can be correctly compared and combined. In the third and final part a real-life application from an election study will be presented, in which the relationships between Individualism, Ethnocentrism and Nationalism are analyzed by the ADM and the EDM.

Literature:

Oud, J.H.L. (2007). Continuous time modeling of reciprocal relationships in the cross-lagged panel design. In S.M. Boker & M.J. Wenger (Eds.), Data analytic techniques for dynamical systems (pp. 87-129). Mahwah, NJ: Lawrence Erlbaum Associates.

Toharudin, T., Oud, J.H.L., & Billiet, J.B. (2008). Assessing the relationships between Nationalism, Ethnocentrism, and Individualism in Flanders using Bergstrom’s approximate discrete model. Statistica Neerlandica, 62, 83-109.

Oud, J. H. L., & Jansen, R. A. R. G. (2000). Continuous time state space modeling of panel data by means of SEM. Psychometrika, 65, 199-215.