Original article Common components and specific weights analysis for the discrimination and evaluation of vegetable oil quality Maresa Cust odio Molinari Ferreira, 1 Poliana Macedo dos Santos, 2 Val eria Rampazzo, 3 Evandro Bona, 1 Jorge Leonardo Sanchez, 1 Giselle Maria Maciel 2 & Charles Windson Isidoro Haminiuk 1 * 1 Post-graduation Program of Food Technology (PPGTA), Federal University of Technology Paran a, Campo Mour~ ao, 87301-899 PR, Brazil 2 Chemistry and Biology Department, Federal University of Technology Paran a, Curitiba, 81280-340 PR, Brazil 3 Post-graduation Program in Food Engineering (PPGEAL), Federal University of Paran a, Curitiba, 91531-980 PR, Brazil (Received 8 March 2017; Accepted in revised form 27 March 2017) Summary The aim of this study was to assess the quality of different vegetable oils by applying common compo- nents and specific weights analysis as a tool for the evaluation and discrimination of chromatography, spectral and physicochemical data. This multiblock method of data analysis divided the data into three common components, corresponding to 56.44%, 34.74% and 8.77% of variance, and it was influenced mostly by chromatography, physicochemical and spectral data, respectively. Gas chromatography, which was used for discrimination of the botanical origin of oil and the groups of saturated, monounsaturated and polyunsaturated fatty acids, was situated in the first common dimension; physicochemical analysis, which was applied to evaluate quality parameters such as acid and saponification value and determine the stability of the product, was situated in the second common dimension. FTIR analysis, by exerting a minor influence on the common dimensions, was considered dispensable in evaluating the quality of veg- etable oils by common components and specific weights analysis. Therefore, multiblock analysis could effi- ciently discriminate vegetable oils. Keywords Chemometrics, Fourier transform infrared spectroscopy, gas chromatography, multiblock analysis, physicochemical. Introduction Vegetable oils are widely consumed by the world pop- ulation. They represent an important source of energy and essential fatty acids and can be a source of pheno- lic compounds and components associated with a reduced risk of cardiovascular disease (Grossi et al., 2013). Thus, the attestation of quality and authenticity of these products has great importance and industrial interest. The main aspect associated with oil adulter- ation is the addition of vegetable oils of low commer- cial value to expensive and raw oils, such as soya bean oil added to extra virgin olive oil. The adulteration leads to changes in the original fatty acid composition of the oil, which are not easily detected (Zhang et al., 2012). The quality of vegetable oils is directly related to the absence of adulterant agents and the stability dur- ing the oil’s shelf life. Factors such as hydrolytic and oxidative rancidity promote the development of free fatty acids, free radicals and other co-products of these degradation reactions (Damodaran et al., 2010). In addition to the fatty acid composition, other vegetable oil information can be obtained from its characterisation using instrumental analyses such as gas chromatography, Fourier transform infrared spectroscopy and titrimetric analyses such as acid and saponification values. With advances in technol- ogy and the instrumentation used in food analyses, the identification of highly complex information has been increasingly demanded. Chemometrics fits these aims because of the possibility of processing chemi- cal data with mathematical and statistical methods to obtain the required information (Bosque-Sendra et al., 2012). Several methods have been proposed by researchers for the evaluation of quality and fraud detection in food. However, there are no studies related to the use of multiblock analysis for the discrimination and eval- uation of vegetable oils. In this context, common com- ponents and specific weights analysis (CCSWA) is a method of analysis that simplifies comparison between data tables and verifies relationships in analytical *Correspondent: E-mail: haminiuk@utfpr.edu.br International Journal of Food Science and Technology 2017 doi:10.1111/ijfs.13473 © 2017 Institute of Food Science and Technology 1