Among living plants a wide range of growth forms from herbs to shrubs and large-bodied trees and lianas can be characterized by identifying different mechanical signals based on changes of structural Young\s modulus (Estr) and flexural stiffness (EI) during ontogeny. Over recent years this `ontogenetic approach\ has been adopted to infer growth forms of fossil plants by treating the fossil stems and their tissues as composite materials and using appropriate centrisymmetric models to calculate values of EI and Estr. One of the main difficulties facing mechanical reconstructions of fossil plants and a confident assessment of growth form result from disarticulated stem parts. Two approaches of viewing mechanical data from both living and fossil plants include: (1) observing mechanical data based on different ontogenetic stages as `general data\~independently of relative position on the plant body (applicable to both living plants and especially isolated fossil stem parts); (2) viewing the mechanical data as `positional data\ where relative positioning of stem parts and their mechanical parameters may be observed (rare in fossil plants). A comparison of general and positional data sets among extant plants indicates that both are informative for large-bodied growth forms including self-supporting trees and non-self-supporting lianas that show appreciable changes in stem diameter during ontogeny. Among the small-bodied plants tested such as Lycopodiella cernua an informative mechanical signal was generated only from the positional data set. The conclusion reached in extending this approach to fossil plants is that large-bodied fossil plants showing a wide range of stem diameter-based ontogenetic stages during growth may generate an informative pattern from isolated fragments but small-bodied plants may require interconnected positional data for an unequivocal mechanical growth form signal.