Biology Reference
In-Depth Information
5.1 Limiting factors in studying growth
using morphometric approaches
The methods discussed in this chapter use landmark coordinates to
study the way in which component structures of an organism rear-
range relative to one another as the organism increases in size. This
approach differs from more traditional growth curve analyses (see
Zeger and Harlow, 1987) where measures that span gross anatomical
regions (e.g., crown-rump length, head circumference) or that repre-
sent overall body size (e.g., body weight, stature) are plotted across
time as curves that summarize growth patterns. When studying
growth, the limitations of landmark data need to be recognized.
Beyond the limiting factors presented in Chapter 4 , we need to
acknowledge that growth is a very complex process. The choice of ana-
lyzing landmark data as collected from the morphological
consequences of this process places limits on our perceptions. For these
reasons, we add two more limiting factors to our list:
Identifiability of the locus of growth. We cannot infer about
the way in which the material between landmarks changes
during growth from the information given by the landmark
coordinates. If chosen intelligently, the landmark data may
capture the evidence of growth processes adequately, but our
answers will be limited to changes local to landmarks.
Identifiability of the timing of growth events. Information
obtained on timing and rate of growth is totally dependent
upon the time points at which we sample data. We can infer
little about the timing of specific growth events if we don't
have data that window that event adequately.
5.2 Longitudinal vs. cross-sectional data
There are two major data types that can be used to study growth: lon-
gitudinal and cross-sectional. Since growth occurs continuously over
time, data points are usually sampled from this continuum for analy-
sis. When working with laboratory animals, the increments can be
chosen to fit the research design. In this case, the observer's schedule
may be the only complicating factor. If a study of human growth is
prospective, as in the case of growth data being collected to establish
Search WWH ::

Custom Search