IFPrograms® Rating/Ranking Data Analyses

As well as that of ranking data, this module of IFPrograms permits the Thurstonian analysis of three types of rating data.  All data must be categorical.

            - Intensity, degree of agreement, purchase interest, etc.: The scale must involve a range of values going from low to high (Note: degree of difference data should be analyzed with the degree of difference option).  The analysis reports whether differences exist between the products that are being compared and returns d’ values and their associated variances.  The relative locations of the scale boundaries are also provided.

            - Degree of difference (different from control): Each rating corresponds to the degree of difference between two unknown samples or a sample and a control.  The categorical scale ranges from identical to some level of difference (e.g., ‘Very different’, ‘Extremely different’).  As for the intensity rating scale, the analysis summarizes whether significant differences exist between the samples, and provides d’ values and their variances.  Scale boundaries locations are also provided.

            - Relative to reference: Ratings are given on a categorical scale relatively to a reference (e.g., ‘Much less sweet’, ‘Less sweet’, ‘Same as reference’, ‘Sweeter’, ‘Much sweeter’).  Like for the other scale analyses, analysis results include statistical differences among samples and between each sample and the reference, d’ values, d’ variances and scale boundaries.

            - Ranking: IFPrograms also allows the analysis of rank data based on complete or incomplete designs.  Rank orders, d’ values, d’ variances and inter-product differences are provided.

Multivariate modeling is available in the Tools, Standard, Enhanced and Professional versions.

Sample output/results windows generated when performing these tests are shown below. For more examples, please see the IFPrograms® User Manual


Scale Analysis

Degree of Difference


The analysis also generates text output:

Discrimination          Hedonics          Multivariate          Combinatorial