Sensory Science Research

SDT Assistant

Michael has written, and made available for free use, an SDT Assistant program.

SDT Assistant is software that undertakes analyses based on Signal Detection Theory (SDT). Currently the software is capable of fitting Receiver Operating Characteristic (ROC) curves based on several different models for a number of psychophysical tasks (or sensory evaluation difference tests) that use the rating response format. The fits are based on maximum likelihood estimation.

The tasks and models available include:

  1. A-Not A or Yes-No: likelihood ratio (β) decision strategy; equal or unequal variance perceptual distributions;

  2. Two-alternative forced choice (2AFC): differencing (τ) or β decision strategies;

  3. Same-different: τ or β decision strategies;

  4. A-Not A or Yes-No with Reminder: τ or β decision strategies;

  5. 2AFC with Reminder: τ, β, or comparison of differences (COD) decision strategies.

Models for 2, 3, 4, and 5 are based on equal-variance perceptual distributions. The model for 5, using the COD decision strategy, is appropriate for the fixed-reference duo-trio procedure.

Note that SDT Assistant now has two additional modules: 1) calculates d' and its standard error for binary response data obtained using most of the methods listed above; 2) generates data that can be used to plot ROC curves for most of the methods listed above.

Download (for Windows only): ~ 59 MB