SEDL / STP / STP594-EB / STP30033S

General Guidelines for Selecting Objective Tests and Multiple Regression Applications

Kramer, A
Professor, University of Maryland, College Park, Md.

Pages: 8    Published: Jan 1976

Download this paper for $25 PDF (140K)          View License Agreement
Abstract

The objective assessment of any sensory quality parameter must be based on its agreement with subjective “panel” evaluation. First, each sensory parameter must be accurately defined by a profile-type panel and results on a limited number of specimens correlated with several proposed objective tests. The selected procedure (r = 0.9<) is then correlated with difference-type panel scores on an adequate number of specimens (40+). The objective test should then be refined to improve precision and calibration. For scaling purposes, a consumer preference-type panel must be used, and the quality levels (grades) must be determined from the regression equations where panel scores are the dependent (y) variable and the objective test data are the independent (x) variable(s).