Free choice profiling analysis with STATIS in Excel
This tutorial will show you how to run and interpret a Free Choice Profiling analysis in Excel using the XLSTAT statistical software.
Dataset for STATIS
The dataset used to illustrate this Free Choice Profiling analysis can be found in the library SensoMineR in R. Each of the 6 subjects of the experiment have been asked to describe several perfumes by giving them scores for multiple descriptors that they had to choose by themselves.
The goal of this tutorial is to study and visualize the proximities between perfumes as well as to determine the agreements between the subjects.
Setting up a STATIS analysis for Free Choice Profiling with XLSTAT
Select the advanced features / Sensory data analysis / STATIS. The STATIS dialog box will appear.
In the General tab, select the Configurations (a configuration corresponds to the set of descriptors chosen to evaluate the products for a subject).
Enter the number of configurations. There are 6 subjects here that correspond to the 6 configurations.
As the 6 configurations all have a different number of variables, you need to select a column containing the number of variables per configuration.
Check the Object labels option and select the column containing the product labels (in our case the perfumes).
In the Options tab, check Global scaling in order to prevent any scaling effect. Moreover, since all the variables are on the same scale within each configuration, it is not necessary to reduce the variables.
In the Charts tab, uncheck the Display charts on the first two axes box.
A new box appears to allow you to choose the axes for which the graphs should be displayed. Select axes 1 and 2 as well as axes 1 and 3 and click the Done button.
Interpreting the results of a STATIS analysis
The following graph is the main objective of STATIS which is to say to represent the observations on a 2-dimensional map, and thus identify the proximities. The first map that represents observations on axes 1 and 2 indicates that the perfumes Lolita Lempicka and Angel are perceived as close but that they are very different from the perfume Pleasures. Axes 1 and 3 indicate that the perfumes L'instant and Cinema are perceived as close and that they differ from the perfume Aromatics Elixir.
If we now focus on two subjects, it is useful to look at the RV matrix which gives the RV similarity coefficient for each pair of subjects (between 0 and 1, which increases with the proximity between the subjects).
We can here see that subject 1 has a very similar opinion to subject 2 with an RV coefficient of 0.802, but less than subject 5 with an RV coefficient of 0.528.
It can be important to evaluate the proximity of a subject to the global point of view reflected by the consensus, in order to determine if it is an atypical subject. For this purpose, we use the following bar chart which allows us to visualize the proximity to the consensus of each subject through his RV coefficient with the consensus. We observe that although subject 5 is the furthest from the consensus, we cannot say that it is an atypical subject because the RV coefficient associated to it remains high (0.706).
Finally, the following graph gives the residuals by the object, which indicate which products (here the perfumes) were rated rather the same by the subjects (low residual for the product) or rather differently (higher residual). We can observe, for example, that Coco Mademoiselle is noted in a more consensual way by the subjects than Aromatics Elixir.
Was this article useful?