Questions and CORRECT Answers
Metameric match - CORRECT ANSWER✔✔- objects with different spectral distributions
look alike under one light source but different under another
Mesopic vision - CORRECT ANSWER✔✔- occurs at dawn or dusk and involves both rods
and cones
5 tastes - CORRECT ANSWER✔✔- sweet, sour, salty, bitter, umami
Supertasters - CORRECT ANSWER✔✔- have more papillae on the tongue, more taste buds,
and more trigeminal nerve endings
more sensitive to PTC & PROP, bitter, sweet, and salty
30 papillae sq cm vs 5
Type I Error - CORRECT ANSWER✔✔- Alpha
Falsely reject Ho (false positive)
Type II Error - CORRECT ANSWER✔✔- Beta
Falsely accept Ho (false negative)
Paired directional - CORRECT ANSWER✔✔- Picking which sample is ____ (saltier,
sweeter, etc.)
p=.5 (one-tailed)
2 serving orders
Triangle - CORRECT ANSWER✔✔- Picking the odd sample.
p=0.33 (one-tailed)
6 serving orders
,Duo-Trio - CORRECT ANSWER✔✔- Pick which is the same to a reference.
p= 0.5 (one-tailed)
Single reference = 2 serving orders
2 out of 5 test - CORRECT ANSWER✔✔- 3 cleaned and 2 spiked
P= 0.1
very fatiguing with wine
p = Probability that results could have occurred by chance alone
1‐p = Probability that a real difference was perceptible
paired preference test - CORRECT ANSWER✔✔- which do you prefer? (consumer test)
p=0.5 (2-tailed test)
Rossler tables
Types of sensory evaluation - CORRECT ANSWER✔✔- Analytical sensory evaluation-
Interested in the WINE and its attributes
Consumer sensory evaluation-Interested in CONSUMERS reactions to the wine
Power - CORRECT ANSWER✔✔- 1-beta
Probability of finding a difference (if one exists)- Depends on alpha, # of judgements, size of
the difference between the samples, directionality
(one tailed has more power)
Z-test - CORRECT ANSWER✔✔- Z=(X-np)-0.5/sqrt(npq)
n= # responses
p = probability of guessing right
q = 1-p
Duo‐trio test for new blend versus old - CORRECT ANSWER✔✔- Results: p=0.059
use the new blend since the pair test did not reject the null but H2S IS detected by some
percentage of your consumers
, Make the decision "CONSERVATIVELY"
Use more people (increase power)
Use higher Type I error level (e.g., decrease b, increase power)
Increase alpha level from p<0.05 to p<0.10
Duo‐trio test for aroma difference with more expensive and finicky yeast - CORRECT
ANSWER✔✔- Results: p=0.022
The difference is barely detectable (since 14/20 would have failed to reject the null), and the
difference in cost is LARGERecommend against itIf it is going to COST YOU MONEY to
make a change, be SURE that there really IS a difference
Use smaller values of alpha: p<0.01 or p<0.001
Minimize risk of there being truly NO REAL difference in statistical terms: Minimize Type I
error (a)
Avoiding statistical mistakes - CORRECT ANSWER✔✔- minimize Type II error (b) also
known as maximizing Power (1 - b)
Threshold value - CORRECT ANSWER✔✔- minimum amount of a substance that can be
detected, recognized or that produces a significant perceptible difference
Types of Threshold Tests - CORRECT ANSWER✔✔- detection (or absolute), recognition,
difference, and terminal
Threshold Tests System - CORRECT ANSWER✔✔- system: Any sensory system
method: Sniff bottle, wine glasses, olfactometer, GCO Test: Ascending/ Descending
concentrations (staircase)
Duo trio or paired comparison
Factors affecting thresholds - CORRECT ANSWER✔✔- wine composition, temperature,
way in which test conducted, method of calculation, and judge sensitivity at the time of the
test
Ranking (non parametric) - CORRECT ANSWER✔✔- Advantages: easy, minimal training