E-modules/knowledge clips week 2 Flashcards
(18 cards)
definition discrimination tests
used to determine whether samples are perceptibly different
limits:
- only with subtle differences possible
- only 2 samples, more is laborious and error sensitive
- no direction of the differences
duo-trio testing
3 samples presented at the same time
- indicate which of 2 is similar to reference material
ref + A + B
A-not_a test
difference between 2 samples
- first sample A, then a second sample: same or different?
A | X
triangle test
difference between 2
- 3 samples at the same time ( 2 identical, 1 different
A + B + (A or B)
same-different test
difference between 2
- 2 blind coded samples at the sa,e
- indicate whether same or different
(A or B) + (A or B)
reversed duo-trio test (ABX discrimination)
- presented with A&B
- then sample x
- indicate if x looks like A or B
A + B | X
Tetrad test
differences between 2
- 4 samples presented
- cluster them in 2 groups
A + B + A + B
Two out of five
- presented with 5 samples
- devide 2 groups on similarity
difference from control
- control + codes sample presented
- determine if there’s a difference in magnitude on a scale
paired comparison
- 2 samples given
- asked which one has greatest intensity on 1 attribute
three alternative forced choice
3 blind coded samples, 2 same, 1 different
- aksen which sample has greatest intensity
harris-kalmus test
8 samples
- divide into 2 groups
- 4 times done for 2 samples
signal detection theory
= method used to determine the ability of a subject to detect a very weak stimulus among rise
- noise or sois + signal
- detection of stimulus is a matter of decision making based on:
1. sensory info (signal and noise)
2. criterion of the subject (when subject says yes)
difference threshold
minimum amount in which 2 stimuli should differ to be able to discriminate them
- increase needed to produce a noticeable increase in sensation
absolute threshold
level below which no sensation would be produced by a stimulus and above which a sensation would reach consciousness
value of difference
difference between means of signal+noise & noise distribution
signal/no signal & yes/no
true + is hit
false + is false alarm
false - is miss
true - is correct rejection
ROC curve
curve with the false alarm rate & hit rate
- d prime: no discrimination between stimulus and noise
- the more curved, the better the participant is