Six data matrices of bond dissociation energies ranging from a 6 × 6 matrix involving only hydrocarbon radicals to a 14 × 14 matrix involving several types of radicals are target factor analyzed. The number of factors to reproduce the data near experimental error range from two to four for the six matrices. The hydrogen and fluorine radicals are shown to be the most unique radicals. Combination sets of vectors from the data matrices reproduce the data within experimental error, with the hydrogen, fluorine, and hydroxyl radicals being the most important ones. Using target testing, a few of the properties of the radicals are identified as basic factors.
Exploratory factor analysis of large data matrices
