A Sample Bias in the Evaluation of Smoked Frankfurters by the Triangle Test

SUMMARY— A taste panel was able to determine the difference between smoked and unsmoked frankfurters, at a statistically significant level, only when the smoked frankfurter was the “odd” sample; non‐significant selections were made with the unsmoked frankfurter as the “odd” sample. Positional differences of the “odd” sample in tasting the three samples did not affect its selection. Determination of difference between two food products by the triangle taste panel test can be affected by the choice of sample for the “odd” sample.