Quiz 2 Flashcards

Learn (60 cards)

1
Q

Signals are always detected in background of _____, which varies _______

A

noise
randomly

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
2
Q

The separation between the two distributions is a measure of _______ known as _______

A

sensitivity
d prime

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
3
Q

Many non-sensory variables affect detection, including
1.
2.
The influence of these non-sensory factors is called _____ ______

A
  1. probability of stimulus occurrence
    Consequences of response

Response vias

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
4
Q

When a subject has a high expectation of stimulus occurrence, the probability of saying “yes” is _______-p

A

increased

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
5
Q

What are the four outcomes ?

A

Hit (yes when stimulus is presented)

Miss (no, when stimulus is presented)

False alarm ( yes, when no stimulus present)

Correct rejection (no, when no stimulus present)

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
6
Q

ROC is the relationship between the probability of “___” when stimulus is present (hit rate) and “___” when stimulus is absent (false alarm rate) this is plotted on a chart called ________ _______ _______

A

yes
yes

receiver operating characteristic

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
7
Q

The ROC curve shows how _____ rate changes as a function of changed in the ______ _____ rate

A

hit

false alarm

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
8
Q

Each ROC curve is a ______ stimulus intensity

A

single

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
9
Q

“______” for detections will influence responses

A

Payoffs

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
10
Q

I the payoff for correct detection is high, subject will say “___” frequently and the hit rate and false alarm rate will ______

If the payoff for correct rejections and the hit and false alarm rate will _______

A

yes
increase
decrease

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
11
Q

An ROC curve can be created by changing the _______ and ______ of detection while maintaining stimulus intensity and probability

A

values
costs

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
12
Q

The combinations of costs and values will produce the _____ on the ROC curve for a given stimulus intensity

A

points

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
13
Q

As the separation between the N and SN distributions increases (detectability increases) the ROC curve _______ more rapidly

A

rises

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
14
Q

Controlling response bias can be done through ________ training of subjects, but is often accomplished through the use of “______ _____” in which no stimuli are presented

A

systematic
catch trials

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
15
Q

In TSD, subjects make a decision about whether the sample can from the _ distribution or the __ distribution, based on a _________ ratio

A

N
SN
likelihood

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
16
Q

When the likelihood ratio exceeds 1.0, it is more likely that the sample came from the __ distribution and when <1.0 is more likely that sample came from the _ distribution

A

SN
N

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
17
Q

TSD criterion assumes that the observer sets a particular likelihood ratio as a criterion (β), as that when the sample is at or above β, the observer chooses __ (and vice-versa_

A

SN

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
18
Q

High (strict not to much yesing) criteria will _______ false alarms and ______ correct rejections while low criteria will ______ misses and _____ hits

A

minimize x2
maximize x2

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
19
Q

The optimal criterion depends on the ________ probabilities and the _____ and _____ of response

A

stimulus
costs
values

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
20
Q

The finding that threshold measure are affected by changes in the _______ criterion throws _____ on the viability of the classical procedures

A

listeners
doubt

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
21
Q

The methods of TSD attempt to independently measure ________ and ______ ______

A

sensitivity
criterion location

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
22
Q

TSD asserts that d prime is equal to the difference of the ______ of the __ and _ distributions divided by the _______ ______ of the N distribution

A

difference
N
SN
standard deviation

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
23
Q

d’ can be determined from the _______ ______ and _____ rates

A

false alarm
hit

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
24
Q

The criterion location on the noise distributions is determined by subtracting the _______ of _____ ______

A

probability
false alarms

1.0 - p (FA) = ZN (z score)

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
25
The location on the SN distribution is 1.0 - _ (_) = ZSN also d' = ZN - ZSN
p (H)
26
Once the ROC curve is determined, the listener's criterion can be found, if the point is at the bottom of the curve, the criterion is ____ and if the point is at the top of the curve the criterion is _____
high low
27
The ordinate of the N distribution at criterion is the ordinate value (O) corresponding to 1.0 - __
FA
28
The ordinate of the SN distribution is the ordinate value corresponding to 1.0 - _
H
29
Criterion (β) is O subSN/O subN (β) Beta value is at 1 where the coordinated cross (more than 1 moving to _____ less than 1 is moving to the _____)
right left
30
Calculating d' and β is a way to examine the effects of a variable on _______ and/or criterion and a way to examine the outcomes of different experimental conditions.
detectability
31
Ex. given a d’ of 1, what is the probability of a false alarm if the probability of hits is a .5
.15
32
Given a d’ of 2, what is the probability of a correct rejection if the probability of a hit is .5?
.02
33
You have tested the hearing of an individual whose hit and false alarm rates are .84 and .18 respectively. Assuming that the N and SN distributions are normal with equal variance, calculate the d’, B, C’ for this person and explain what each number means
Method A 1-.18 = .82 then find Z value on chart which is Zn= .92 - -.99(Zsn) = high d’ 1.921
34
What are the TSD procedures? which one is important in psychoacoustics*
Yes-No Forced choice* Confidence rating Same-Different Adaptive methods: Forced -choice tracking Up-down transformed response Pest Maximum-likelihood
35
Yes-No describe
-Long series of SN and N trials are conducted (with a stated proportion of SN and stated values and costs for each session) -Payoffs and signal probabilities are constant for a session -Experimental data for different signal probabilities and payoffs yield expected ROC curves, and measurements of sensitivity are stable under all conditions -If only 1 session is conducted, the value of d’ and β can be estimated *but an ROC curve cannot be plotted.
36
Forced Choice
-This method is unaffected by changes in criterion. -For each trial, 2 or more intervals are presented and the observer reports which interval contained the signal. -The proportion of correct responses—p (C)— is converted to d’ from a table (depending on the number of intervals involved). -Forced-choice procedures are also useful in measuring difference thresholds.
37
Confidence Rating?
-This method is very efficient for collecting data in a single session. -The observer makes a confidence rating for each yes-no judgment made (e.g., 5 if very sure, 4 if fairly sure, etc.). -Proportions of responses for each rating are determined and an ROC curve is derived (see handout).
38
Same - Different?
-As the name states, the observer’s task in this procedure is to report whether two stimuli in a trial are the same or different. In a discrimination of A and B, there are, therefore, 4 possible outcomes on each trial—2 same (AA, BB) and 2 different (AB, BA). -The hit rate is the proportion of “different” responses when the stimuli differ, and the false alarm rate is the proportion of “different” responses when the stimuli are the same; d’ calculated from these, although some adjustments may have to be made, depending on the decision rule used by the subject. -Same-different can also be used to measure the difference threshold.
39
Oddity?
-In this procedure, the observer is asked to discriminate between different stimuli. -There are 3 or more intervals with one stimulus in all but one of the intervals and the other stimulus in the remaining interval. -d’ depends on p (c), and is determined through the use of a table.
40
The threshold is a "_____" meaning the _____ stimulus level that elicits a response on a specified proportion of presentations ___% is a typical performance criterion
statistic lowest 50%
41
minimum audibility curves assess _____ ______ (detectability) for a large number of young, otologically normal listeners across the range of audible frequencies in the free field and under earphones
absolute thesholds
42
Minimum audibility curves plot the obtained ______ (central tendencies) on a graph to obtain the _______ audbility curve
thehsolds minimum
43
Lower frequency limit of audibility is usually defined as __ Hz responses < __ Hz require _____ intensity levels, are not described as _____ and are more ______ than ______
20 20 high tonal tactile auditory
44
Upper frequency limit is defined as __K Hz, since higher frequencies elicit a tactile sensation
20
45
"best" sensitivity is senn in the _____ Hz to _______ Hz region
2000 - 5000
46
From 500 to 5000 Hz, sensitivity is around __ dB SPL
10
47
Below 500 Hz and above 5000 Hz, hearing becomes increasingly ____ sensitive, reflecting the transmission characteristics of the __ (and __)
less Me OE
48
Conditions under which thresholds are obtained influence results: (6)
Ambient Noise Method Subjects Calibration / Transducer type Sound field (MAF) Earphones (MAP)
49
What are the two methods of measurement for minimum audibility
Minimum audible field (MAF) -Sound-treated or anechoic chamber with listener seated 1m at 00 azimuth from speaker -Measurement made after listener removed, with microphone at center of head Minimim audible pressure (MAP) -MAP/MAP-C -Measurement made using probe mic in EAM (close to TM) or coupler; sound usually delivered via earphone
50
Which is more sensitive MAF or MAP
MAF is slightly more sensitive -MAF is a binaural measure -MAP influenced by masking effects of physiological and other noise -Head and body diffraction effects -Differences in how measurements are made
51
Intense sound (130 dB SPL) of any frequency will elicit a _______ sensation, which is called the "_______ of ______"
tactile threshold of feeling
52
Further increase in intensity elicit a ______ sensation (140 dB SPL), this threshold is also known as the "_______ threshold" and it does not cary with frequency
pain terminal
53
The area between minimum audbility curve and threshold of feeling is called the ______ _____ area
auditory response area OR Dynamic range of hearing
54
As duration decreases below about ____ ms (critical duration), power must be _______ for detection to occur (______ = ______ / _____)
200 increased Power = Energy / Time
55
For each 10-fold decrease in time (below about 200 ms), a __ dB increase in level is needed, and for each 2-fold decrease in time (<200 ms) a _ dB increase is needed
10 dB 3 dB
56
Integrating the power of a tonal signal takes place over approximately ____-___ ms; differs for other stimuli
200-300
57
Temporal integration thresholds improves with a _______ stimulus because the increased time provides more opportunities to detect the stimulus through repeated sampling
longer
58
Temporal resolution is the ________ detectable _____ separation between signals
smallest time
59
The ability to resolve fine temporal differences is critical for ______ ______. A common method of assessing this is through ____ detection threshold tests
speech recognition gap
60