Growth curve models refer to a class of techniques
that analyze trajectories of cases over time. As such,
they apply to longitudinal or PANEL data, where the
same cases are repeatedly observed. John Wishart
(1938) provided an early application of the growth
curve model in which he fit polynomial growth curves
to analyze the weight gain of individual pigs. Zvi
Griliches (1957) looked at the growth of hybrid corn
across various regions in the United States and modeled
predictors of these growth parameters. Since
these early works, growth curve models have
increased in their generality and have spread across
numerous application areas.
The word growth in growth curve models reflects
the origin of these procedures in the biological
sciences, whereby the organisms studied typically
grew over time, and a separate growth trajectory could
be fit to each organism. With the spread of these techniques
to the social and behavioral sciences, the term
growth seems less appropriate, and there is some tendency
to refer to these models as latent curve models