The Auditory Inference Span Test
Developing a test for cognitive aspects of listening effort for speech comprehension
Niklas Rönnberg 1, 3, Stefan Stenfelt 1, 3, Mary Rudner 2, 3
Technical Audiology, Department of Clinical and Experimental Medicine, Linköping university, Linköping, Sweden
2
Department of Behavioural Sciences and Learning, Linköping university, Linköping, Sweden
3
Linnaeus Centre HEAD and Swedish Institute for Disability Research
1
Introduction
Most hearing aid fittings today are almost solely based on the patient’s
audiogram. However, more individual parameters of the patient’s hearing
thresholds together with the patient’s cognitive abilities to process the
auditory signal are required [1-2]. Hearing aid fitting may be seen as a
process aimed to improve the patient’s hearing thresholds rather than to
improve communication ability. Another way to think about hearing aid
fitting is to ease the patien’s listening effort [3]. However, it is not clear
how to measure the listening effort.
The Auditory Inference Span Test (AIST) is a dual task hearing in noise
test, that combines auditory and memory processing and is well suited
as a clinical test for listening effort.
Pilot testing the AIST
The test was performed with 41 subjects with a mean age of 32 years
(SD 6.8), 44% women and 56% men.
Sentences and
The test used Hagerman sentences [4-5] in subquestions, 1 - 3
lists of three sentences. After each Hagerman
sentence there was a question to verify speech
Info. question 1
perception, termed sentence question.
Subsequent to the three sentences there were
three information questions, that were divided in
three different memory load levels. After these
questions the subjects were asked to rate their
listening effort.
same for all memory load levels. This was controlled by comparing the
rated listening efforts within each subject, as shown in Figure 4. These
data show small differences in perceived listening effort between
different memory load levels for each subject, which indicates that the
subjects were consistent when stating the perceived listening effort.
Discussion
The presented study has shown that AIST probes the cognitive capacity
of an individual. The subjects rated their listening effort similarly between
the different levels of memory load. This indicates that the effort was
affected by the listening condition and not by answering the questions.
However, there were large variations of perceived listening effort
between subjects. This indicates that subjects experience the same SNR
differently demanding, that can be related to both auditory and cognitive
capacities.
The new test setup
In a future study, AIST will be evaluated with measurements of the
subject’s cognitive capacity as well as the subject’s hearing thresholds.
Info. question 2
Audiogram
Audiogram to determine the test subjects hearing
thresholds.
Info. question 3
Reading span
Cognitive tests to discern the test subjects
cognitive capacity: reading span, letter memory.
Listening effort
Letter memory
Hagerman test
Hagerman test to assess the test subjects general
ability to hear in noise, using three SNRs: -6dB,
-4dB, -2dB.
Results
Effort test
Listening effort test using Hagerman sentences in
the above three SNRs.
Two analyses of variance were performed, one for the accuracy and
one for the reaction time. For both there were main effects of memory
load level (accuracy: F(2,120) = 22.4, p < 0.001; reaction time:
F(2,120) = 19.5, p < 0.001).
AIST
The AIST, in the above three SNRs, without the
question about listening effort.
10
Reaction time (seconds)
6
Mean score
5
4
3
2
0
It is expected that the test will show correlations between SNR of the
speech material, number of correct answers, the reaction time, and the
perceived effort. This relation was also shown in a previous study [6].
8
6
4
When the signal to noise ratio gets worse the score also gets worse,
the reaction time longer, and the perceived effort greater.
2
1
Level 1
Level 2
Memory load level
0
Level 3
Error Bars: 95% CI
Figure 1. The relation between correct answers and memory load levels. (n = 41)
Level 1
Level 2
Memory load level
Level 3
Error Bars: 95% CI
Figure 2. The relation between reaction time
and memory load levels. (n = 41)
Figure 1 shows that the greater the load on the working memory the
poorer the score. Figure 2 shows that there were significantly longer
reaction times for the memory load level 3 questions. This indicates that
a greater amount of processing of the information was required before
an answer could be given.
It is also expected that the score of the AIST will correlate to the test
subjects’ cognitive capacity.
A test subject with poorer cognitive capacity will score worse, have
longer reaction time, and perceive greater effort, than a test subject
with greater cognitive capacity.
It is expected that this difference will be greater when the signal to noise
ratio gets worse, a relation also shown in a previous study [7].
2
Intra subject listening effort
8
Listeing effort
Expected findings
6
4
2
0
References
1
0
-1
-2
Level 1
Level 2
Memory load level
Level 3
Figure 3. Box-plot of listening effort and
memory load level. (n = 41)
Level 1
Level 2
Memory load level
Level 3
Figure 4. The variation of perceived listening
effort and memory load level. (n = 41)
As can be seen in Figure 3, the perceived effort differed between
subjects but was on average similar for all levels. Since the sentences
were all presented at the same signal-to-noise ratio (an SNR
of 0dB), the perceived listening effort should be the
1. Stenfelt S, Rönnberg J (2009). Background and basic processes - The Signal-Cognition interface: Interactions between degraded auditory signals and cognitive processes. Scandinavian Journal of Psychology, 50,
385-93.
2. Edwards B (2007). The future of hearing aid technology. Trends in Amplification, 1, volume 11, 31-45.
3. Sarampalis A, Kalluri S, Edwards B, Hafter E (2009). Objective measures of listening effort: effects of background noise and noise reduction. Journal of Speech, Language and Hearing Research, 52 (suppl. 5), 1230-40.
4. Hagerman B (1982). Sentences for testing speech intelligibility in noise. Scandinavian Audiology, 11, 79-87.
5. Hagerman B, Kinnefors C (1995). Efficient adaptive methods for measuring speech reception threshold in
quiet and in noise. Scandinavian Audiology, 24, 71-7.
6. Larsby B, Hällgren M, Lyxell B, Arlinger S (2005). Cognitive performance and perceived effort in speech
processing tasks: effects of different noise backgrounds in normal-hearing and hearing-impaired subjects.
International Journal of Audiology, 44 (suppl. 3), 131-43.
7. Gatehouse S, Naylor G, Elberling C (2003). Benefits from hearing aids in relation to the interaction
between the user and the environment. International Journal of Audiology, 42 (suppl. 1), 77-85.