Practical sensory evaluation considerations

  1. Tasters should taste the wine ‘blind’.
    The identification of the wines to be tasted should not be known to the taster(s). Wines should be presented in a different, randomised order for each taster, with no clues as to their identity. This ensures that the biases of all tasters are minimised, if not eliminated.
  2. Have at least two independent tasters
    Quality control assessments, such as wine additive taint screening or cork taint checking, require at least two tasters who have strengths in that type of assessment, (e.g. cork taint recognition) to evaluate the wine. If the two tasters do not agree, more rigorous testing may need to be applied.Knowledge of winemakers (and other staff members) sensory strengths and weaknesses is important for this type of testing. Variation among tasters in their ability to perceive different aroma and flavour compounds can be quite large. For example, some wine tasters might have a high threshold for Brett flavour compounds, but be very sensitive to cork taint or oxidation. To evaluate your staff members sensory strengths and weaknesses, sensory fault kits are available through Affinity Labs or through other sources such as the Australian Society of Viticulture and Oenology. These kits will allow individual staff members to be tested for their sensitivity to common taints and faults, and are also useful for training in taint and fault recognition and identification.It is important to note that sensory testing doesnt have to be limited to winemakers. Any company staff member including administration and cellar door staff can potentially be used for sensory analysis provided they are familiar with the type of test, and their individual strengths and weaknesses have been evaluated. It is therefore strongly recommended that the cellar floor staff members are trained in sensory evaluation. This has two benefits: it will increase the number of tasters available for sensory evaluation and will also make the cellar floor staff members more aware of taints and faults, which is an important skill for people working with your product every day.
  3. Repeat the tasting
    When performing a difference test a single tasting by each taster might not provide the most accurate information about a wine due to the chance of tasters guessing the correct answer. Having tasters repeat the tasting exercise can decrease this chance of guessing. Difference tests also require a certain number of answers or responses to determine statistical significance and for this, the greater the number of responses the better. An easy way to increase the number of responses without increasing the number of tasters is to have each taster repeat the tasting exercise. This is simple to do as tastings can be organised so that the same sample comparison is presented twice, with the wines presented in a different order each time.
  4. Minimise presentation effects
    Fatigue, adaptation, suppression/masking of flavours and visual biases are all effects that can be decreased with correct presentation of the samples. Ideally samples should be pre-poured at a constant tasting volume (30 mL) and temperature (approx. 20ºC), into covered glasses, preferably coded with three digit random numbers. The samples should be presented in a random order, which differs for each taster. Tasters should taste within a set period (e.g. 1 hour), and if this is not possible, the samples should be repoured (but not by the taster). This is standard practice in scientific sensory assessments and also should be practiced in commercial tastings.
  5. Minimise talking during tasting
    To prevent tasters influencing each others judgements, tasters should not communicate until they have made, and written down, their judgement. To ensure tasters do not communicate during tasting, tasters should taste in isolation, either at different times, i.e. one person goes into the sensory lab as one goes out, or they could taste in different physical areas. If this is not possible, tasters should at least face away from each other and avoid eye contact and talking during the tasting. Use of tasting sheets is also suggested as they make the taster write down a response, and enable tasters to taste and record their results in a standardised format each time. Tasting sheets also enable easy collation of results, and can be filed so there is a record of all tastings. Proformas of tasting sheets used for several sensory techniques are available on the AWRI website (here).
  6. Reduce physiological effects
    Fatigue, degree of tiredness, hunger and other issues of emotional state will affect taster performance. Generally, it is recommended to carry out assessments in the morning, with no tasting held for at least half an hour after smoking, eating or drinking. To reduce effects of fatigue and adaptation, ideally a maximum of six to seven wines should be presented at any one session, with tasters having a short rest if more samples are to be assessed.
  7. Establish if a difference exists before deciding on preference
    Before considering preference testing, establish if there is a significant sensory difference with a difference test. Preferences are an important part of sensory testing and a winemaker will often need to state their preference to aid decision making. Before doing this though, it is essential to ensure that a real difference actually exists between the wines. If there is no sensory difference, or if personnel cannot reliably and repeatedly detect a difference between samples, their preferences are meaningless, and probably due to random choice.

Note: For more detailed information on conducting sensory tests there are a number of Australian Standards available. Please contact the AWRI library for more information.