Core samples taken from ocean floors are at times analyzed for plankton content in order to infer approximate climatological conditions which existed on earth. Since many species coexist at different times (core depths), their abundance is generally correlated, and a factor analysis is frequently performed on such data. Table 7.15, represents an excerpt of data given by Imbrie and Kipp (1971) for n = 110 core depths of a single core, where each depth represents 10cm and where only p = 10 plankton species are considered,
(a) Carry out a principal component analysis of the p = 10 species, plotting component scores for the dominant component.
(b) Rotate the loadings and scores using the varimax rotation. Is there a change in the interpretation?
(c) Repeat parts a and h using a maximum likelihood factor model.
they are known as trend surface analyses-see, for example, Harbaugh and Demirmen (1964) and Cameron (1968).