Temporal Dominance of Sensations (TDS)
Temporal Dominance of Sensations are used to detect attributes that dominate at different moments while testing products. Available in Excel using XLSTAT.
What is Temporal Dominance of Sensations (TDS)?
Temporal Dominance of Sensations (TDS) is a temporal multidimensional sensory method (Pineau, Cordelle & Schlich, 2003). Panellists are presented with a list of attributes and asked to choose the dominant ones over consumption of the product. A dominant attribute is the most striking perception at a time, not necessarily the most intense one (Pineau et al., 2009).
The TDS tool of XLSTAT allows visualizing dominant attributes for a set of products. It is available under the Sensory data analysis menu.
Two data formats are accepted:
- Dominance format: XLSTAT expects that you have in Excel, a table with as many rows as there are combinations of products, panellists, attributes and potentially sessions, and as many columns as time points. Data must be binary, with 1 if the attribute was selected as dominant at a given time, 0 otherwise. For each combination of panellist*product*session, each attribute must appear only once.
- Citation format: XLSTAT expects that you have in Excel, a table with one row per attribute selected by a panellist for a given product and potentially given session and one column containing the time of selection. For that format of data, there must be a “start” and a “stop” attribute. For each combination of panellist/product/session, each attribute may appear as many time as it was selected as dominant.
Temporal Dominance of Sensations options in XLSTAT
- None: Activate this option to keep the data as is (no standardization).
- Right: Activate this option to standardize time in order to bring all end times to the same scale. Standardized times are reduced between 0 and 1. for each panelist*product*session times are divided by the maximum time, that is the time of end of evaluation.
- Left-Right: Activate this option to standardize time in order to bring all start and end times to the same scale. Standardized times are reduced between 0 and 1. for each panelist*product*session times are reduced by the minimum time, that is the time the first attribute was selected, and divided by the maximum time, that is the time of end of evaluation.
Significance level (%): Enter the significance level for the test.
Temporal Dominance of Sensations charts options in XLSTAT
- Smoothing: Activate this option to smooth curves.
- Chance limit: Activate this option to display the chance limit of dominance. The chance limit is the dominance rate that an attribute can obtain by chance and is defined by P0=1/K, K being the number of attributes.
- Significance limit: Activate this option to display the significance limit of dominance. The significance limit is the minimum value the dominance rate should equal to be considered as significantly higher than P0.
- Bands (Yes/No): Activate this option to display all significantly dominant attributes in one band.
- Bands by attribute: Activate this option to display a 2 dimensions band-plot. For each period during which a given attribute is significantly dominant, a band with height relative to the mean dominance rate over the given period is drawn.
Dominance per product per attribute: The tables of dominance per product and attribute show the dominance rate per product and attribute for each time point.
TDS curves: For each product, a plot showing the dominance rate for each attribute plotted against time is displayed. Dominance rates are computed for each time point as the proportion of evaluations (panellist*session) for which the given attribute was assessed as dominant (Pineau et al., 2009). If asked by user, dominance rates are smoothed using cubic spline. If asked by user, the chance limit and significance limit are displayed.
Yes/No bands: For each product, a plot showing the significant dominant attributes as a single band is displayed. The band is composed with stacked colored rectangles (Monterymard et al., 2010). The total height of the band is constant. The x-axis represents time.
Bands by attribute: For each product, a plot showing the significant dominant attributes as bands is displayed. The x-axis represents time and the y-axis shows the different descriptors. The heights of the band are proportional to the mean dominance rates, allowing the user to estimate the importance of each attribute (Galmarini et al., 2016). For each given time period the height of the band is calculated as the mean dominance rate over the period divided by the maximum dominance rate for the given product.
Galmarini, M. V., Visalli, M., & Schlich, P. (2016). Advances in representation and analysis of mono and multi-intake Temporal Dominance of Sensations data. Food Quality and Preference.
Monterymard, C., Visalli, M., & Schlich, P. (2010). The TDS-band plot: A new graphical tool for temporal dominance of sensations data. In 2nd conference of the society of sensory professionals (pp. 27–29).
Pineau, N., Cordelle, S., & Schlich, P. (2003). Temporal dominance of sensations: A new technique to record several sensory attributes simultaneously over time. In 5th Pangborn symposium (p. 121).
Pineau, N., Schlich, P., Cordelle, S., Mathonnière, C., Issanchou, S., Imbert, A., Rogeaux, M., Etiévant, P. and Köster, E. (2009). Temporal Dominance of Sensations: Construction of the TDS curves and comparison with time–intensity. Food Quality and Preference, 20(6), pp.450-455.