Statistical analysis for improving data precision in the SPME GC–MS analysis of blackberry (Rubus ulmifolius Schott) volatiles

Statistical analysis has been used for the first time to evaluate the dispersion of quantitative data in the solid-phase microextraction (SPME) followed by gas chromatography–mass spectrometry (GC–MS) analysis of blackberry (Rubus ulmifolius Schott) volatiles with the aim of improving their precision. Experimental and randomly simulated data were compared using different statistical parameters (correlation coefficients, Principal Component Analysis loadings and eigenvalues). Non-random factors were shown to significantly contribute to total dispersion; groups of volatile compounds could be associated with these factors. A significant improvement of precision was achieved when considering percent concentration ratios, rather than percent values, among those blackberry volatiles with a similar dispersion behavior. As novelty over previous references, and to complement this main objective, the presence of non-random dispersion trends in data from simple blackberry model systems was evidenced. Although the influence of the type of matrix on data precision was proved, the possibility of a better understanding of the dispersion patterns in real samples was not possible from model systems. The approach here used was validated for the first time through the multicomponent characterization of Italian blackberries from different harvest years. [Display omitted] •Precision of SPME GC–MS data for blackberry volatiles was studied by statistical analysis.•Contribution of non-random factors to dispersion of experimental data was proved by statistics.•Blackberry model systems were used to evaluate the matrix effect on precision of SPME data.•Precision was significantly improved by using percent concentration ratios among correlated volatiles.•A precise multicomponent characterization of Italian blackberries was achieved by this approach.