My data set consist of 156 individuals with fifteen variables. The variables consist of one body mass (dependent) variable and fourteen (independent) variables of different bird bone dimensions (of one type of bone). The 156 individuals can be divided over 30 bird species, where some groups of bird species contain more individuals than others. My ultimate aim is to come up with an accurate body mass estimate.
The reason I want to use iterative PCA is because I have a lot of missing data in my data set (ranging from 19,2 % - 86.5 % of the total sample size within the different bone dimensions). The other complication is that my data is grouped (as written above). My main question is, thus, can iterative PCA be applied to grouped data?
Eventually, I want use standard PCA in order to regress the principal components (because my bone dimension variables suffer from severe multicollinearity) and, subsequently, estimate body mass of a particular extinct bird species.