Investigation of operationally more powerful duo-trio test protocols: Effects of different reference schemes

The effects of reference schemes (reference mode – constant vs. balanced-, number, position, and nature – stronger vs. weaker) on the performance of replicated duo-trio tests were investigated using four versions of duo-trio protocols (DTF, DTM, DTFR, and DTFM), using two different pairs of tomato juices with different sodium content. Examination of proportion of correct answers (Pc) and d′ estimates indicated that when the cognitive strategy was fixed to the comparison of distances (COD) strategy, the constant-reference mode showed the superior test performance over the balanced-reference mode, only for DTFM performed using the stronger reference. Among the tests with the balanced-reference mode, for DTF, estimates of d′ of the tests using the stronger-reference (W-odd) were significantly higher than tests using the weaker-reference (S-odd), whereas the opposite was true for DTM, revealing the operational influence of the reference position. In general, the DTF or DTFM using the stronger (saltier) product as a constant-reference is recommended as an operationally more powerful protocol to discriminate products having high sodium contents.

► Roving experimental design using two different sample pairs allowed comparisons in d′ based on the same model.
► Duo-trio test using saltier-reference showed the higher test performance.
► Reference position in a test showed operational effects on the test performance.
► DTFM with constant-stronger-reference tended to show the highest test performance.