Please find below a summary of the implemented improvements in the Panel and Panellists Performance (new) analysis.
Repeatability is the new default option, rather than reproducibility
- Repeatability is a measure of the consistency of the assessors in evaluating the same products, verifying they can accurately and reliably detect the same sensory attributes across multiple sessions.
- Poor repeatability highlights a specific need for further training or calibration for that individual panellist.
- Reproducibility measures how closely a panellist’s scores align with the overall panel average.
There is the option to show Pairwise Performance Plots within the panel performance analysis, rather than having to perform a separate analysis.
Added a table called Panellist Contribution to Interaction.
- The panellists contribution to the interaction is calculated by breaking down the assessor by product interaction component in the ANOVA table into separate contributions for each assessor.
- It gives information on how consistently panellists perceive and score differences between different products and highlights where certain assessors are interpreting specific products differently.
Other changes
- Ensured the axes labels on all plots are more explanatory
- Made colour coding consistent across all tables
- Stop calculating correlations if there are fewer than 5 products. If there are fewer than 5 products, correlations are not a reliable metric.
- Renamed Consistency as Agreement
- Added the Mean row to the Panellist Summary table which show the mean score of Discrimination, Agreement and Repeatability per panellist.