Round table: Sensory analysis and its importance to determine VOO quality for comercial markets

Introducing the estimation of uncertainty in results analysis

The metrological approach of the concept of uncertainty, defined as “an estimate attached to a certified value of a quantity which characterizes the range of values within which the true value is asserted to lie with a stated level of confidence” according to the ISO Guide 30:2015, has not been explored with sufficient detail thus far. This concept was developed in the early 20th century, at a time when the organoleptic evaluation of certain food products was conducted only in compliance with the technical requirements set out in official standards.

When putting the uncertainty in the spotlight, two metrological concepts always come into play: (1) the accuracy of measure methods and results and (2) the conformity assessment.

• The accuracy of measure methods is influenced, among other factors, by the level of precision of the methodology used. According to the ISO standard 5725-1:2023, the accuracy results from combining (i) the trueness, defined as “the closeness of agreement between the arithmetic mean of a large number of test results and the true or accepted reference value”, and (ii) the precision, understood as “the closeness of agreement between different test results”. The relationship between uncertainty and precision is shown as follows:

× ±𝑅 ≠ 𝑥 ± 𝑈

Where x is the estimated value of the measurement, R is the reproducibility, together with repeatability the two extremes of precision, and U is the uncertainty.

• The conformity assessment is used to determine whether or not a product complies with the requirements set out in the applicable standards. Within this context, the probability of error is known as risk.

At the end of the 20th century, the incorporation of quality assurance systems in the field of chemical and biological analytics in sensory laboratories represented a turning point on how the organoleptic evaluation of food products was to be performed from then on, but not without difficulties.

The challenge of adapting the estimation of uncertainty to organoleptic assessments

Measuring involves comparisons with official standards traceable to the International System of Units (SI). In physical measurements, the standards used are traceable to the National Metrology Centres, and the certification of their real value is indicated by the correction of the nominal value and its uncertainty.

As usually happens in chemical laboratories, analytical variables do not have a direct relationship with the measurands, and there is no relationship that allows the uncertainty of certain variables to be transferred to the variation of the measurands.

The sensory analysis of virgin olive oil also presents certain peculiarities that require the application of uncertainty estimation techniques, mainly:

• Type A uncertainty: the estimate of the measurement is made from a certain number of repeated and independent observations.
• Type B uncertainty: the estimate of the measurement comes from other means and is determined taking into account the information available. The traceability of calibration certificates is of great importance, although not unique.
  The sensory analysis of virgin olive oil is carried out by a panel of experts constituted by 8- 12 members on the basis of the following assumptions:

R=f(E,F, Psc)

In case that F and Psc are constant  R=f(E)

Where R is the physical response to a stimulus, E is the stimulus, F is the Physical factor, and Psc is the psychophysiological factors.

Following the ISO/IEC 17025:2017 standard, the estimation of uncertainty in organoleptic analysis requires the implementation of a quality assurance system at laboratory stage, through which control systems are implemented to provide the necessary data for a correct estimation.

The sensory analysis of virgin olive oil is based on (i) the identification of sensory attributes and (ii) the estimation of the intensity of these attributes. Both processes have a joint objective: the classification of virgin olive oil by carrying out an assessment of conformity.

As previously stated, uncertainty is part of the result of measurements, thus acquiring a great significance when affecting the assessment of conformity. A distinction can therefore be made between:

1. Uncertainty in the identification of attributes: This is of great relevance due to the system of evaluation of the sensory responses of the analytical method. Thus, it is possible in certain situations, that all tasters detect negative attributes, but the robust coefficient of variation for each of the identified attributes is higher than 20. The estimation of these cases must be performed on the basis of the sensory analysis itself and would be based on the calculation of the intensity in which the attributes must be found, and which should correspond to the interval between the so-called limit of sensory perception and the limit of identification.
2. Uncertainty in the intensity of attributes: Main difficulties in estimating this uncertainty lie (a) in the lack of an established system of references with which to compare, (b) in the number of sensory attributes defined for virgin olive oil and (c) in the interactions between them.
   The estimation of the precision, for instance through a set of reference sensory laboratories and Inter laboratories, complies with the same parameters as other analytical methods: precision, internal control and/or interlaboratory standard deviations associated with the resolution of intensity measurements.

It may be concluded that, when uncertainty is well estimated, it is the most effective approach to measure the metrological quality of a sensory laboratory and, thus, of the organoleptic assessment of virgin olive oil.