Influência de cafés de gôsto rio em ligas com cafés de bebida mole

Abstract

The authors study the influence of Rio coffee in mixture with Brasilian coffee of Soft taste. Increasing percentages of Rio coffee were tried: 0,0; 0,5; 1,0; 1,5; 2,0; 2,5; 3,0; 3,5; 4,0; 4,5; 5,0; 7,5; 10,0; 12,5; 15,0; 20,0;25,0; 30,0; 35,0; 40,0; 50,0. Two experiments were carried out, in a balanced incomplete block design, with t = 21 treatments (those just mentioned), k = 3 plots per block; r = 10 replications; b = 70 blocks, L = 1. Each plot had 3 standard cups, and was tasted by 3 tasters, who gave only one opinion for each plot. Data thus collected were, therefore, 630 for each experiment (210 plots x 3 tasters). But each trial had indeed only 210 plots, to every plot corresponding the average of 3 opinions. The experiments gave results in good agreement, which led to the following conclusions: a) Transformation of the data is necessary, since variances corresponding to different treatments are rather different. b) The square root transformation gave satisfactory results. c) Rio coffee injures a lot Soft coffee taste, for percentages from 2,0% upwards. d) For percentages of 4,5% and over the mixture has taste Rioy or Rio. e) The regression obtained is not strictly linear, but the straight line gives a reasonable approximation. f) Taken in account only percentages 0,0 through 10,0%, the regression equation is: T = 1,7045 - 0,1278 X, where X is the percentage of Rio coffee, and T gives the square root of the score corresponding to tasting categories. g) For treatments 0,0 through 10,0% the regression equation for the original data is: Y = 3,0997 - 0,3281 X, that is, there is a decrease of 0,3281 in the score corresponding to coffee taste for every unity of percentage of Rio coffee.