PRINCIPAL COMPONENT ANALYSIS AND MATHEMATICAL MODELLING FOR SENSORY AND CHEMICAL CHARACTERISATION OF MARAS TARHANA, A TRADITIONAL FOOD

The regression modelling and principal component analyses (PCA) were performed to determine the relationship between sensory scores and chemicophysical components of Maras tarhanas. Water, ash, protein and lipid contents in the tarhana samples were 70.11-120.89, 20.45-60.12, 90.86-130.78 and 10.68-40.42 g/kg, respectively; the titratable acidity was in range of 1.12-6.46 g/kg (lactic acid). The colour (Hunter Lab) values of the samples were 55.08-78.10 for L* values, (-1.37)-(+1.89) for a* values and 11.40-21.83 for b* value. The ash, lipid, colour (L* and b*) and lactic acid values of the home-made tarhana were higher than those of the commercial tarhana. The sensory tests were performed in two stages. In the first stage, the tarhana quality was determined to be affected by sourness, taste-aroma, fatness and whiteness parameters. In the second stage, the sensory scores were calculated by the test. Lipid, titratable acidity and brightness (L*) values were found to be important parameters for consumer attributes. In PCA, two components were extracted and those two components accounted for 65.53% of the total variance. In regression model, lipid, lactic acid and colour L* values were found as variable.