An experiment was conducted at the Wool of Laboratory Zoothecny Departament at Federal University of Santa Maria, RS. The objective was to examine the efficiency classification and determination of wool diameter as compared to objective system measured in laboratory. Were analysed 397 wool samples previously subjectively classified. After preparation, the samples were wool diameter determinator type "Air-flow" for an average wool determination, and wools being classified as Merina (19µ), Amerinada (21µ), Prima A (23µ), Prima B (25µ), Cruza 1 (27µ), Cruza 2 (30µ), Cruza 3 (32µ) and Cruza 4 (34µ), according to the diameter obtained. The analysis of variance and F Teste showed significant differences between the two methods of classification (F = 30.0; P < 0.01). The independency test among factors, by the Chi-square Test, showed the dependency of methods to the proportion between group of adjacent classes (Merino/Amerinada; Prima A/Prima B; Cruza 1/Cruzas 2 e 3) (P < 0.01). The subjective system, as related to the objective way, presented 20.0% of correct hits, 31.4% of light mistakes and 48.4% of hard mistakes. The subjective classification underestimated the number of Merina and Amerinada wools and overestimeted those of Primas A and B and Cruza 1 classes. It's concludes that subjective classification is not as efficient as the objective method for some wool classes.
comparison subjective/objective ways; wool diameter determination
