Ranking test was applied to questions of sensory intensities. Three
samples were tested with a panel of 20 untrained panelists. The
panelists received the samples (coded with three digit numbers) in
balanced, random order. Panelists ranked three samples in
ascending order of any related property. Water was provided for
mouth rinsing during the ranking of individual samples. Because
the rank scale used was 1 ¼ preferred least and 3 ¼ preferred most,
the highest rank sum meant that the product was the most
preferred. The ranking datawas analyzed by using the critical value
tables (Lawless & Heymann, 2010; Newell & MacFarlane, 1987). For
each property, the ranks were summed and the differences between
the sums were compared to critical value in this table. The
number where the two points cross was the critical value for the
difference; in our case it was 15. If the difference of sum of their
rank between each pair of samples was greater than the critical
value, it was considered that the samples were significantly
different in the related property or in overall acceptability
(Marshall, 2006). On the other hand, to determine which samples
were significantly different, the multiple comparison test (Duncan)
method was also performed. Significant differences between fermented
beverages having different apple juice concentrations were
evaluated for each sensory property, individually.