Cog & Speech Perception Flashcards
Explain how the TEN test is able to identify a cochlear dead region.
The TEN test is designed to identify on-frequency and off-frequency listening. On-frequency listening occurs when the area on the basilar membrane that is responding is the same as what is being tested. Off-frequency listening occurs when the response moves away from the region of peak basilar membrane vibration and areas along the cochlear dead region are responding. The TEN test uses a calibrated signal to determine which is responding. This noise will mask any off-frequency listening that might be occurring. For example, if noise is presented and the noise masks the response, this indicates an off-frequency listening because the tone must be louder than the noise to be heard. If the noise is presented and response is still heard, this is on-frequency listening.
stream of acoustic information is interpreted as speech and not just random sounds
perceptual grouping of speech sounds
ability to understand a speaker with previously unheard foreign accent especially with a repeated exposure to that accent
perceptual learning
ability to identify words in a stream of speech (if one is familiar with the languae)
Segmentation of speech into meaningful units
perceiving either one phoneme or another when VOT is either increased or decreased
Categorical perception
Speech production and speech perception are linked. What are some examples of this? Select all that apply.
someone learning a new language will speak with an accent
interrupting auditory feedback causes speaker to change their speech
babbling during speech development
Lombard effect
Stenger effect
interrupting auditory feedback causes speaker to change their speech
babbling during speech development
Lombard effect
Which of the following is/are considered part of cognition? Select all that apply
understanding information
storing and retrieving information (memory and recall)
mental activities or processes
all
All information necessary to recognize sounds is contained in the acoustic signal and analyzed in the auditory pathway
bottom up processing
Higher level cognitive operations are involved in identifying and analyzing speech sounds
Top down processing
Prior knowledge and expectations are involved in speech perception
Top down processing
measures the number of words that can be recalled while also identifying if a word is a noun or verb
Working memory
Provide a short definition for each of the cognitive abilities
Attention
Processing Speed
Inhibition
Short-term memory
Working memory
Executive function (as commonly applied in cognitive hearing science)
Semantic knowledge
Attention is the ability to attend to a desired target in the presence of a distractor. For example, listening in noise.
Processing speed is the amount of time it takes to process a stimulus. For example, reaction time.
Inhibition is the ability to inhibit extraneous information when responding to a target stimulus. For example, the stroop test.
Short term memory is memory that is stored for short periods of time. For example, how many numbers you are able to remember.
Working memory is task specific and simulateneously stores and processes the information. For example, being able to repeat numbers that were presented but backwards.
Executive function is the control of cognitive processes and includes planning and adapting and monitoring behavior. For example, trail making task where you have to follow the numbers in order.
Semantic knowledge is accumulated through lived experiences. For example, picutre naming and being able to put things into categories.
Stroop task
Inhibition
Sort pictures into categories
Semantic knowledge
measures the number of letters that can be recalled accurately and in order
Short term memory
trail-making task
Executive function
reaction time task
Processing speed
task is to respond to all the black letters; participant sees imagees containing letters and numbers; the task is to press a space bar
Attention
List three ways a speech signal could be distorted or difficult to understand prior to reaching a listener’s auditory system.
One way a speech signal can be distorted before reaching the listener’s auditory system is with background noise. Another way it can be distorted is if they have an accent or different dialect than the listener and they are a new talker. The third way it can be distorted is through reverberations in the room.
As used in the Ease of Language Understanding model, how do implicit and explicit processing differ?
2) Give one example where implicit processing is likely occurring and a second example where explicit processing is likely occurring. For each of your examples, explain how those situations exemplify implicit or explicit processing.
For each of your two total examples, explain the scenario and how the listener would likely be processing the speech. For example, “The listener has hearing loss and the room is noisy. They hear ‘I’m going to the [inaudible] to get some groceries’ and they figure out the person is saying the word ‘store’. This is an example of (either implicit or explicit) processing because _________. “
Implicit processing is automatic and occurs during bottom-up processing. Predictive processes is also a part of implicit. Explicit processing takes effort and uses our working memory to understand. Explicit is a part of top-down, as well as postdivitive processing.
During ease of language understanding model, we use implicit processing to look for a match in our long term memory of the multimodel input that we received. This occurs automatically and is fast. If a mismatch occurs, or the brain can not find a match in long term memory, then it goes on to explicit processing where it will actively pull information to try and understand.
For example, in implicit, the listener has hearing loss and is sitting in an auditorium watching a musical. They hear the person say “Now, we are going to _____ God Bless America.” The listener figures out that the person said “sing.” This is implicit because it was automatic and used the information from the rest of the sentence to determine what was said.
For example, in explicit, the listener has hearing loss and is sitting on a busy bus during rush hour. This is the first time they have had to use the bus on this route. The intercom comes on and says “The next stop is at ______”. They have to look around at signs for their location and use the bus map to figure out the next stop. This is an example of explicit because the situation is not familiar to the person and it requires them to use their working memory to understand what the missing piece was that was said.
`
Explain the predictive and postdictive role of working memory in speech/language processing. Give an example of each.
Predictive listening is an example of implicit listening because it occurs automatically and fast. Predictive listening determines what might be said before they comprehend it in their system. It uses phonologic and semantic information to process. An example of predictive listening would be having a conversation with a familiar talker that you are used to their dialect, accent and suprasegmentals. If you don’t understand something, you can predict what it is they might have said based off of previous knowledge of conversations with that person you are familiar with.
Postdictive listening is an example of explicit listening because it is effortful and is slow and deliberate. Postdictive listening occurs after mismatch has occurred and they didn’t understand something. Postdictive listening is when we do not understand something so we store it to process it and pull from our working memory in order to understand it after it has been said. For example, postdictive listening would occur if you were listening to an unfamiliar talker and they were speaking to you in a noisy environment.
For a dual-task paradigm, the primary task must always be a speech task (if measuring listening effort).
Which of the following could be used as a secondary task when measuring listening effort? Select all that apply
recall task
speech recognition threshold in noise task
reaction time task
tracking task
attention task
recall task
reaction time task
tracking task
attention task
Why might a listener with hearing loss who also has a large working memory capacity be able to derive benefit from fast-acting compression (the example used in class) where someone with the same hearing loss but a smaller working memory capacity might not derive benefit from this technology? Select all that apply.
Larger working memory span correlates with greater processing speed, which allows for better understanding of the distorted signal.
Larger working memory span correlates with a larger vocabulary, which allows for a match to be made more readily
Larger working memory span allows a listener to hold more information in memory while trying to match the distorted signal to what’s in long-term memory storage.
Larger working memory span allows for greater use of the cognitive system’s predictive abilities during speech processing.
Larger working memory span allows a listener to hold more information in memory while trying to match the distorted signal to what’s in long-term memory storage.
Larger working memory span allows for greater use of the cognitive system’s predictive abilities during speech processing.
What are the three categories of measuring listening effort?
For each category, give an example of a test that would fall under that category.
For this question, you only need to provide the test example, you do not have to describe or define the test.
The first category is physiologic measures for listening effort. An example of this would be measuring pupil response during listening or an fMRI during functional testing.
The second category is self-reporting measures. An example of this would be a questionnaire.
The third category is behavioral measures and an example of this would be the dual-task paradigm.
reduces listening effort or the amount of working memory capacity required to process the signal
amplify soft sounds so they are audible to a listener
increases listening effort or the amount of working memory capcity required to process the signal
alters the temporal envelope resulting in reduced variability of amplitude fluctuations
For someone with hearing loss, knowing the topic of conversation can facilitate speech understanding. In the context of what we’ve learned in this class so far, why does knowing the topic of conversation facilitate speech understanding?
Knowing the topic of conversation for someone with hearing loss allows the listener to use top down, processing.
Which value of d’ (d-prime) indicates a more sensitive test, a higher value or a lower value?
higher
What is the advantage of using d’ as a measure of test sensitivity?
d-prime is independent of pass-fail criteria
What is the premise underlying the TEN test, which is to identify cochlear dead regions?
The TEN noise is designed to mask off-frequency listening
Which of the following exemplifies or defines top-down processing?
all of the information necessary to process speech is contained in the acoustic signal
a listener uses prior knowledge to process speech
knowing the topic of conversation helps a listener process speech
a listener uses context to process speech
a listener uses expectations to process speech
a listener uses prior knowledge to process speech
knowing the topic of conversation helps a listener process speech
a listener uses context to process speech
a listener uses expectations to process speech
slow deliberate effortful
explicit
rapid, automatic, relatively effortless
implicit
Hearing aids can potentially both increase AND decrease the amount of cognitive effort required to process speech.
t
survey question
self repor
dual task paradigm
behavioral
pupillometry
jphysiologic
shown to work i a controlled ideal setting
efficacy
shown to work in real word settings
effectiveness
Which of the following are lost or reduced with outer hair cell cochlear hearing loss? Select all that apply.
semantic knowledge
two-tone suppression
top-down processing
pure tone sensitivity
cochlear tonotopicity
distortion products
two-tone suppression
pure tone sensitivity
cochlear tonotopicity
distortion products
We discussed a type of “hidden hearing loss” in this class that is sometimes referred to as cochlear synaptopathy. While we do not yet know for sure how manifests in humans, the mouse model suggests it is a result of damaged cochlear inner hair cell synapses, resulting in a loss of information relayed to the auditory nerve fiber. Which type of auditory nerve fiber is affected?
those responsible for higher-lever inputs
Select all the statements below that are true regarding the type of “hidden hearing loss” referred to as cochlear synaptopathy. Assume everything else in the auditory-cognitive system is intact and functioning typically.
pure tone thresholds are unaffected
affects the ability to process complex auditory information at suprathreshold levels
can be addressed using wide dynamic range compression
is correlated with middle ear pathology
speech in quiet is not significantly impacted
affects speech in noise processing
pure tone thresholds are unaffected
affects the ability to process complex auditory information at suprathreshold levels
speech in quiet is not significantly impacted
affects speech in noise processing
Speech in quiet ability typically correlates with speech in noise ability.
f
A speech in noise test has been developed and normed on young adult listeners with hearing thresholds no poorer than 15 dB at any frequency. The typically-hearing listeners were able to achieve 50% correct at an SNR of 4 dB.
You test a patient using this test and they need a more favorable SNR. They are able to achieve 50% correct at an SNR of 9 dB.
What is the SNR loss in dB?
5
can compare aided vs. unaided
sound field
can compare monaural vs. binaural
headphones/inserts
this transducer type is more commonly available
headphones/inserts
calibration/standardization is easier
headphones/inserts
Which of the following speech characteristics have an associated visual cue? Select the option if there is a visual cue at least some of the time.
syllable information
voice onset time
voice fundamental frequency
speaker emotion
place of articulation
syllable information
speaker emotion
place of articulation
What is the TEN noise designed to do?
confirm air conduction masked thresholds
mask a response at the best frequency
account for variability in ear canal size
mask an off-frequency response
mask an off-frequency response
What is the advantage of using d’ as a measure of test sensitivity?
d-prime is unaffected by “noise” in the test
d-prime is independent of pass-fail criteria
d-prime is the most stable measure
d-prime is independent of pass-fail criteria
All information necessary to recognize speech is contained in the acoustic signal and analyzed in the auditory pathway
bottom up processing
Higher level cognitive operations are involved in identifying and analyzing speech sounds - prior knowledge and expectations
top down processing
Prior knowledge and expectations are involved in speech perception
top down processing
Explain how listening can be effortful, according to the Ease of Language Understanding model (3-5 sentences). Give 2-3 examples of what might create effortful listening situations.
Draw a general schematic of the Ease of Language Understanding model to go along with your explanation. Since you are drawing this from memory, do not worry about drawing exactly the same as the figure in the reading. Just convey the concepts of the model, beginning with multimodal sensory input and ending with understanding.
When multimodal information is received it goes through implicit processing, which is fast, automatic and pretty effortless. When going through implicit processing, the information is compared to what is stored in long-term memory. If there is a match in long term memory, then understanding occurs. If there is not a match in long-term memory, it moves on to explicit processing which is effortful and slow. This is where previous knowledge and experience comes into play to understand the information. Once this happens, understanding can occur.
One situation that might create an effortful listening situation is going to dinner at a noisy restaurant with multiple talkers around the table. Another situation is a classroom that is very reverberrant and can get noisy at times. A third situation that could create an effortful listening situation is having a conversation with an unfamiliar talker that has an unfamiliar accent.
may occur after extended periods of difficult listening situations
listening-related fatigue
is thought to occur when explicit processing is required to understand speech
listening effort
Explain why a listener with threshold independent deficits associated with cochlear synaptopathy (one type of “hidden hearing loss”) would be able to understand the speech in quiet with high levels of accuracy but would have difficulty with the speech in noise.
Cochlear synaptopathy causes distortion in high intensity signals but not in low intensity signals. This is why in speech in quiet, there is no distortion because all aspects of the signal are still available to the listener in order for understanding. With speech in noise, the speech signal is mostly covered by the noise which causes distortion an in turn causes worse performances because what is stored in long-term memory no longer matches and this leads to explicit and more effortful listening in order to understand. The peaks of the signal are drowned out by the noise.
Describe two-tone suppression and distortion products. How are these affected by cochlear outer hair cell loss? How does this affect speech processing?
Two tone suppression is when one tone close in frequency to another suppresses that second tone. In speech processing with loss of outer hair cells, it is no longer able to suppress the tone and in turn, it affects vowel perception because the peaks are lost and everything becomes flat.
Distortion products happen when two tones are presented and a third tone is created. This third tone is the distortion product. This affects speech processing because with loss of outer hair cells, they are no longer able to create this distortion product. This causes speech to have to be louder in order for the individual to understand, as well as distortion in the system that makes it harder to hear in noise.
Speech in noise testing can be carried out in two different ways, each yielding a different type of score.
Fixed = the SNR does not change throughout the test
Adaptive = the SNR changes throughout the test (either automatically or is changed by the tester)
What type of score does each testing type give?
Yields a score in SNR
adaptive
Yields a percent correct score
fixed
Yields a score in SNR
adaptive
Yields a percent correct score
fixed
All things being equal, it is easier to hear speech with a fluctuating masker rather than a steady-state masker. Why? Select all that apply.
A fluctuating masker does not cause energetic masking.
A fluctuating masker does not contain linguistic content.
A fluctuating masker has brief moments where the SNR is favorable to the listener
A fluctuating masker has brief moments where the SNR is favorable to the listener
masker evokes stronger neural excitation than target speech
energetic masker
linguistic content interferes with speech perception
informational masker
takes place in the peripheral auditory system
energetic masking
takes place at the auditory processing or cognitive level
informational masking
Select all the statements below that are true regarding the type of “hidden hearing loss” referred to as cochlear synaptopathy. Assume everything else in the auditory-cognitive system is intact and functioning typically.
affects the ability to process complex auditory information at suprathreshold levels
pure tone thresholds are unaffected
can be addressed using wide dynamic range compression
affects speech in noise processing
is correlated with middle ear pathology
speech in quiet is not significantly impacted
affects the ability to process complex auditory information at suprathreshold levels
pure tone thresholds are unaffected
affects speech in noise processing
speech in quiet is not significantly impacted
refers to audibility of the signal
sensitivity
refers to changes in the listener’s auditory-cognitive system
distortion
automatic
implicit
deliberate or intentional
explicit
predictive
implicit
fast
predictive
component of RAMBPHO (initial processor of multimodal sensory information)
predictive
explicit process
postdictive
thought to be in play after a mismatch has already occurred
postdictive
slow, deliberate
post
part of implicit processing
pre
part of explicit processing
post
You are demonstrating a simulated hearing loss with filters applied to a speech recording to simulate severe and moderate hearing loss. You create three different speech samples. Each sample is of the same passage read by the same talker and only differ in the amount of sound that is filtered out. The samples are:
“no filter” (simulating normal hearing)
“moderate filtering” (simulating a moderate hearing loss)
“highly filtered” (simulating a profound hearing loss)
Your goal with the hearing loss simulator is to demonstrate the effects of loss of audibility to someone with typical hearing who doesn’t understand hearing loss. After considering what you’ve learned in this class, you (correctly) decide to present in the following order:
the highly-filtered speech sample first;
followed by the moderately filtered;
ending with the not-filtered speech sample
Why is this the correct order to give the full effect of reduced audibility to a listener with typical hearing thresholds? Why would the experience for the normal-hearing listener be different starting with the highly-filtered sample (severe hearing loss example) vs. starting with the full sample (not filtered, no hearing loss example)?
The reason as to why this is the correct order to give the full effect of reduced audibility is so that the normal listener doesn’t have any context as to what the stimulus is saying. If you started with no or moderate filtering, they would already have pre-dictive knowledge as to what is being said and would therefore not get the full benefit of hearing highly filtered speech or notice as big of a difference because they can figure out what is being said from background knowledge of the content. By giving the highly filtered speech sample first, it shows real-world scenarios as to how these listeners feel and navigate in their lives from day to day. By cutting out most of the speech first, it shows how hard it is to piece together the information with no knowledge of what is being said beforehand.
Imagine that in your first job as an audiologist, you are tasked with developing a basic evaluation protocol. You decide that speech in noise testing should be a part of every basic audiologic assessment where a patient indicates difficulty hearing speech in noise. The office manager, who is in charge of how much time is allotted for each appointment and which test materials are purchased, doesn’t know about audiology but trusts you. How would you explain to them why speech in noise testing should be part of most hearing evaluations, or a part of every hearing aid evaluation at a minimum? What points would you use to convince them that it is worth the time and expense?
I would explain that speech in noise testing should be included because assessing this during the evaluation creates a baseline for our patients, we can monitor their performance, and we can better counsel and make decisions for amplification that better fit the patients needs. It also shows that we are listening to our patients concerns and makes them feel like their concerns are valid and heard and that we take them seriously. In the long run, I would explain that by doing this we will have more satisfied patients because they feel we hear their concerns and take them seriously, as well as most likely less amplification returns because there is a mutual understanding of what they are for and how to use them to get the most benefit for that patients main concern.
For someone with hearing loss, knowing the topic of conversation can facilitate speech understanding. In the context of what we’ve learned in this class so far, why does knowing the topic of conversation facilitate speech understanding
Knowing the topic of conversation can facilitate speech understanding because it helps the listener with top-down processing. In top-down processing, the listener uses previous knowledge in order to understand. If the person already knows the topic of conversation, they already have this background knowledge in order to make quicker matches and reduce listening effort in order for them to understand. It can reduce their listening related fatigue and reduce the amount of listening effort it takes for them to follow and contribute in the conversation that is occuring.
