Citation
Multi-Dimensional Outcome Measures of FM System Usage in Places of Worship

Material Information

Title:
Multi-Dimensional Outcome Measures of FM System Usage in Places of Worship
Copyright Date:
2008

Subjects

Subjects / Keywords:
Auditory perception ( jstor )
Disabilities ( jstor )
Ears ( jstor )
Hearing aids ( jstor )
Hearing loss ( jstor )
Listening ( jstor )
Microphones ( jstor )
Signals ( jstor )
Statistical significance ( jstor )
Statistics ( jstor )

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Embargo Date:
7/12/2007

Downloads

This item has the following downloads:


Full Text





MULTI-DIMENSIONAL OUTCOME MEASURES OF
FM SYSTEM USAGE IN PLACES OF WORSHIP


















By

JAMES P. SHEEHAN JR.


A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY

UNIVERSITY OF FLORIDA

2007


































2007 James P. Sheehan Jr.



























To my parents, Jim and Val, for their endless love and support.









ACKNOWLEDGMENTS

I would like to sincerely thank Dr. Scott Griffiths for serving as my supervisory chair,

close friend, and career mentor. He supported me ever since I toured the CSD department 4

years ago. I immediately saw how deeply devoted he was to his students. He taught me two

graduate courses one-on-one just so I could catch up on the first semester that I had missed while

completing my master's degree. He helped me through the ups and downs of teaching Fundies,

the nervousness of presenting at the American Academy of Audiology in Washington, DC and

the life adjustments that surfaced during my brief time at UF. He always challenged me and

patiently dealt with my attitude. I hope to exhibit a fraction of his example.

I would also like to thank my supervisory committee for their guidance. I thank Dr. Alice

Holmes and Dr. Jay Hall for introducing and encouraging me to the possibility of a research

degree when the clinical version offered a better view of the light at the end of the tunnel. I

thank Dr. Monika Ardelt for constructive contributions during the initial study design. I thank

Dr. Dave Fabry for motivating me with his tremendous enthusiasm.

Special thanks go to Helmut Ermann of Phonak Hearing Systems of Warrenville, IL for

providing the grant support for this study and the career opportunity that I have been very

fortunate to have.

Thanks also to Dana Ulmer for her emotional support during my last year. I thank Idella

King for her logistical support and for allowing me to vent with her in the CSD office. I thank

Rachel McNeal and Cassie Eiffert for their assistance with the data collection and entry

Lastly, I would like to acknowledge Sharon Powell and Stacey Rebello for their close

friendship. I will never think of my graduate work in Gainesville, FL without our memories.

Finally, I sincerely appreciate the loving support of Kristen Klimek during my final steps of this

educational process and eagerly look forward to the start of our relationship in Chicago.










TABLE OF CONTENTS

A CKN OW LED GM EN TS ..................................................................

LIST OF TABLES ................... ............................................... ................. ........ .7

LIST OF FIGURES ........................................................................ 10

L IST O F A B B R E V IA T IO N S .............................................................................................12

A B STR A C T .................. .......... .............. ....................................................... 13

CHAPTER

1 IN TRODUCTION ............... ...................... ............................ ........ 15

2 REVIEW OF THE LITERATURE ............................................... ...............21

Speech Perception in Children with FM System s ...................................... ........... ....21
Speech Perception in Adults with FM Systems......... ............... ................ 29
Electroacoustic Performance Factors in FM Systems .................................. ....38
Guidelines for Fitting FM Systems............... ......... ...... ........41
FM System Usage and Satisfaction .................... ......... ..... ...............45
Outcome M measures .................................... .... .. ...................... 51
Abbreviated Profile of Hearing Aid Benefit (APHAB) ................................................53
Glasgow Hearing Aid Benefit Profile (GHABP) .................................. ....55
Glasgow Benefit Inventory (GBI) ............................................ ............... 57
Hearing Handicap Inventories (HHI) ....................................................59
Psychosocial Impact of Assistive Technology Scale (PIADS) .....................................62
Spiritual W ell-Being Scale (SW B S) ........................................................ 65
Research Question .......................................................67
Pilot Investigation........................... ...... ........... .67
Main Investigation....................... .......... ... .......70
FM receiver selection ............................................... ... ....75
E nrollm ent by place of w orship ............................................................................. 77
H hearing aid adjustm ents ................... ............................................ ............... 79
Quality control protocol for verifying large area FM ......................................... 80
Q questionnaire booklets .................................... ..................... .......... .. ...
Adaptations to outcomes measures .............................. ............... 83
C om pensation/O their B benefit ...................................................................................... 85
Analyses ...................................................... ........85
Hypotheses ................................................86

3 RESULTS ......................................................... ................. ........101

Between-Subject Comparisons on Audiometric Data at Baseline .............. .............101
Within-Subject Comparisons on Outcomes Measures across Time Intervals......................101
Glasgow Hearing Aid Benefit Profile (GHABP)................. ................ ............101










Glasgow Benefit Inventory (GBI) ................................................ .... ............... 108
Psychosocial Impact of Assistive Devices Scale (PIADS) .........................................109
Abbreviated Profile of Hearing Aid Benefit (APHAB) ................................................110
H hearing H handicap Inventories (H H I) ................... ......................... ....... .....................111
Spiritual Well Being Scale (SWBS).......................................................113
Between-Subject Comparisons on Selected Outcome Measures ............... .................113
Glasgow Hearing Aid Benefit Profile (GHABP).................................... .................114
Psychosocial Impact of Assistive Devices Scale (PIADS) ............... ........ .......115
R religious Item s .................................................................................. 116

4 D ISCU SSIO N ........................................................................ 133

Glasgow Hearing Aid Benefit Profile (GHABP) ..........................................134
G lasgow B benefit Inventory (GB I) ....................................................... 136
Psychosocial Impact of Assistive Devices Scale (PIADS).................... ...............137
Abbreviated Profile of Hearing Aid Benefit (APHAB) .......................................................139
Hearing Handicap Inventories (HHI) ..............................................139
Spiritual W ell-Being Scale (SW BS).............................................. ........ 140
Between-Subject Comparisons on Selected Outcome Measures .............. ...........141
Study Lim itations.................................................................... ................... 143
Summ ary ................ ................................................. ......... 144

APPENDIX

A DECISION MATRIX FOR SELECTION FM RECEIVERS..............................................146

B BEHAVIOR MATRIX FOR MICROPHONE ACTIVATION................. ...............147

C QUESTIONNAIRE BOOKLET FOR THE FM GROUP...............................148

Abbreviated Profile of Hearing Aid Benefit (APHAB) .......................................................148
Glasgow Hearing Aid Benefit Profile (GHABP) ..........................................150
Glasgow Benefit Inventory (GBI) ............................................... ............... 152
H hearing H handicap Inventory (H H I) .............................. ............................................ 153
Psychosocial Impaact of Assistive Devices Scale (PIADS)............................................... 155
Spiritual W ell B being Scale (SW B S) ........................................................................... 156

D TEST STATISTIC SUM M ARIES .............................................. ............... 157

Friedman Analysis of Variance Tests............................................................................. 157
Wilcoxon Signed Ranks Tests ................. ............................. 163
M ann W hitney U Tests ......................................................................... ...... ........................... 167

E IMAGES OF FM EQUIPMENT USED IN STUDY..................................178

L IST O F R E F E R E N C E S ................................................................................................... 183

BIOGRAPHICAL SKETCH ......................................................................192









LIST OF TABLES


Table page

3-1 Pilot investigation data summary: pearson inter-item chi-square values........................87

3-2 Main investigation enrollment summary: participant selection of FM receivers and
transmitters.......................................................... ................ ........ 88

3-3 Individual participant data for the 23 participants in the FM group: amplification
history and audiom etrics. .............................................................89

3-4 Individual participant data for 6 participants in the comparison group: amplification
history and audiom etrics. .............................................................90

4-1 Satisfaction scale results of the Glasgow Hearing Aid Benefit Profile (GHABP) for
the FM group: mean scores and standard errors (SE).................................................... 118

4-2 Reported Benefit scale results of the Glasgow Hearing Aid Benefit Profile
(GHABP) for the FM group: mean scores and standard errors (SE). ...........................119

4-3 Derived Benefit scale results of the Glasgow Hearing Aid Benefit Profile (GHABP)
for the FM Group: mean scores and standard errors (SE). ...........................................120

4-4 Use scale results of the Glasgow Hearing Aid Benefit Profile (GHABP) for the FM
group: mean scores and standard errors (SE). .........................................121

4-5 Handicap scale results of the Glasgow Hearing Aid Benefit Profile (GHABP) for the
FM group: mean scores and standard errors (SE)................................ ........... ....122

4-6 Glasgow Benefit Inventory (GBI) results for the FM group: mean scores and
standard errors (SE). ................................................ .. ...... 123

4-7 Psychosocial Impact of Assistive Devices Scale (PIADS) results for the FM group:
mean scores and standard errors (SE). ................................ ..................... ............ 124

4-8 Abbreviated Profile of Hearing Aid Benefit (APHAB) results for the FM group:
m ean scores and standard errors (SE). .................................................................. 125

4-9 Hearing Handicap Inventory (HHI) results for the FM group: mean scores and
standard errors (SE). ................................................ .. ...... 126

4-10 Spiritual Well Being Scale (SWBS) results for the FM group: mean scores and
standard errors (SE). ................................................ .. ...... 127

D-1 Friedman Analysis of Variance (ANOVA) statistics for Worship Specific Items on
the Glasgow Hearing Aid Benefit Profile (GHABP)......................................................157









D-2 Friedman Analysis of Variance (ANOVA) statistics for the Glasgow Benefit
Inventory (GBI). ....................................................... 158

D-3 Friedman Analysis of Variance (ANOVA) statistics for the Psychosocial Impact of
Assistive Devices Scale (PIADS). ............. ...............................159

D-4 Friedman Analysis of Variance (ANOVA) statistics for the Abbreviated Profile of
Hearing Aid Benefit (APHAB). ...........................................................................................160

D-5 Friedman Analysis of Variance (ANOVA) statistics for the Hearing Handicap
Inventory (HHI). .......................................................161

D-6 Friedman Analysis of Variance (ANOVA) statistics for the Spiritual Well Being
Scale (SWBS). ........................................................ 162

D-7 Wilcoxon Signed Ranks test statistics for worship specific items that exhibited a
significant main effect (a = .05) on Analysis of Variance (ANOVA) in the Glasgow
Hearing Aid Benefit Profile (GHABP)........................ ......................................... 163

D-8 Wilcoxon Signed Ranks test statistics which exhibited a significant main effect(a =
.05) on Analysis of Variance (ANOVA) in the Psychosocial Impact of Assistive
Devices Scale (PIADS)................... .... .................................... ... ............ 164

D-9 Wilcoxon Signed Ranks test statistics for items that exhibited a significant main
effect (a = .05) on the Analysis of Variance (ANOVA) in the Glasgow Benefit
Inventory (GBI). ....................................................... 165

D-10 Wilcoxon Signed Ranks test statistics for items that exhibited a significant main
effect (0 0 =.05)zn the Analysis of Variance (ANOVA) in the Abbreviated Profiled
of H hearing A id B benefit (A PH AB ). ................... ...................1..........................6

D- 1 Mann Whitney U test statistics for comparison of audiometrics between the FM
group and com prison group. ................................................. ........ 167

D-12 Mann Whitney U test statistics for comparison between FM group and comparison
group on the Glasgow Hearing Aid Benefit Profile (GHABP). .....................................168

D-13 Mann Whitney U statistics for comparison between Ear level receivers and Belt level
receivers within the FM Group on the Glasgow Hearing Aid Benefit Profile
(GHABP). ........................................................... 170

D-14 Mann Whitney U Statistics for comparison between Personal FM Users and Worship
only FM users on the Glasgow Hearing Aid Benefit Profile (GHABP)..........................172

D-15 Mann Whitney U statistics for comparison between participants that scored below
20% and above 20% on word recognition in noise within the FM Group in the
Glasgow Hearing Aid Benefit Profile (GHABP).............................................................174









D-16 Mann Whitney U statistics for all four between subject comparisons on the
Psychosocial Impact of Assistive Devices Scale (PIADS)..............................................176

D-17 Mann Whitney U statistics for the Religious Items (Item S11 on the Hearing
Handicap Inventory for the Elderly (HHIE) and Items 18 and 21 on the Abbreviated
Profile of Hearing Aid Benefit (APHAB). .... ..................... ...............177









LIST OF FIGURES


Figure page

3-1 Left ear pure tone air thresholds (dB HL) by frequency (Hz) for the 23 participants in
the FM group: means (+/- 2 standard errors). .......................................91

3-2 Right ear pure tone air thresholds (dB HL) by frequency (Hz) for the 23 participants
in the FM group: means (+/- 2 standard errors). ............................... ...........92

3-3 Left ear pure tone air thresholds (dB HL) by frequency (Hz) for the 6 participants in
the comparison group: means (+/- 2 standard errors). ...................................... 93

3-4 Right ear pure tone air thresholds (dB HL) by frequency (Hz) for the 6 participants in
the comparison group: means (+/- 2 standard errors). ...................................... 94

3-5 Word recognition performance in percent correct scores on Northwestern University
word list #6 (NU-6) for the 23 participants in the FM group. .......................................95

3-6 Word recognition performance in percent correct Scores on Northwestern University
word list #6 (NU-6) for the 6 participants in the comparison group. .............................96

3-7 Quality control measures for electroacoustic verification: average acoustic output
(dB SPL) by frequency (Hz) of earbuds at volume 3 .....................................................97

3-8 Quality control measures for electroacoustic verification: average acoustic output
(dB SPL) by frequency (Hz) of earbuds at volume 6. ........................................... 98

3-9 Quality control measures for electroacoustic verification: average acoustic output of
(dB SPL) by frequency (Hz) of headphones at volume 3..........................................99

3-10 Quality control measures for electroacoustic verification, average acoustic output of
(dB SPL) by frequency (Hz) of headphones at volume 6.............................................100

4-1 Satisfaction scores from items 5-9 of the Glasgow Hearing Aid Benefit Profile
(GHABP) for the 23 participants in the FM group. ............. ....................128

4-2 Reported Benefit scores from items 5-9 of the Glasgow Hearing Aid Benefit Profile
(GHABP) for the 23 participants in the FM group.................... ....................129

4-3 Derived Benefit scores from items 5-9 of the Glasgow Hearing Aid Benefit Profile
(GHABP) for the 23 participants in the FM group. ............. ....................130

4-4 Reported Use scores from items 5-9 of the Glasgow Hearing Aid Benefit Profile
(GHABP) for the 23 participants in the FM group. ............. ....................131

4-5 Handicap scores from items 5-9 of the Glasgow Hearing Aid Benefit Profile
(GHABP), Items 5-9 for the 23 participants in the FM group......................................132









E-1 Phonak TX-300V large area transmitter. Note: ............. ..................... .....178

E-2 Phonak M L9S ear level receiver ................................. ........ ..............................179

E-3 Phonak M L8S ear level receiver .................................. ........ ..............................180

E-4 Phonak MLxS universal ear level receiver .............. ..................181

E-5 Phonak M icrovox belt level receiver. ............................... ... ...............................182









LIST OF ABBREVIATIONS

ALD Assistive Listening Device

ANOVA Analysis of Variance

ANSI American National Standards Institute

ASHA American Speech, Language and Hearing Association

BTE Behind-the-Ear

dB Decibel

DAI Direct Audio Input

DSL Desired Sensation Level

NAL National Acoustics Laboratories

NIDCD National Institute of Deafness and other Communication Disorders

FM Frequency Modulation

HA Hearing Aid

HL Hearing Level

PW Place of Worship

SE Standard Error

SNHL Sensorineural Hearing Loss

SNR Signal-to-Noise Ratio

SPL Sound Pressure Level

SL Sensation Level









Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy

MULTI-DIMENSIONAL OUTCOME MEASURES OF
FM SYSTEM USAGE IN PLACES OF WORSHIP

By

James P. Sheehan Jr.

May 2007

Chair: Scott K. Griffiths
Major: Communication Sciences and Disorders

Places of worship are commonly reported environments where people with hearing

impairment still have listening difficulty despite wearing hearing aids. The provision of an

assistive listening device such as a Frequency Modulation (FM) system has been documented to

improve speech perception in these adverse listening situations. However, research on the

application of such devices in the real-world has been scarce. This study sought to determine if

the device would impact a sample of 29 experienced hearing aid wearing adults who regularly

attend religious services in various outcome measures.

Of the 29 enrolled, 23 elected to participate in the experimental group and were fit with an

FM system and 6 decided to wear only their current hearing aids. All participants were

administered the following outcome measures at baseline, 12 weeks and 24 weeks: Abbreviated

Profile of Hearing Aid Benefit (APHAB) for assessment of disability; Glasgow Hearing Aid

Benefit Profile (GHABP) with specified worship situations for assessment of disability and

handicap; Glasgow Benefit Inventory (GBI) for assessment of health related quality of life,

Hearing Handicap Inventories for the Elderly (HHIE) and for Adults, (HHIA) for assessment of









handicap; the Psychosocial Impact of Assistive Devices Scale (PIADS) and, the Spiritual Well-

Being Scale (SWBS).

The participants in the FM experimental group had statistically greater satisfaction,

reported benefit and derived benefit as measured by the GHABP worship specific items across

time intervals. These effects were found for the main presenter, other presenters and

music/lyrics and at 12 weeks and 24 weeks. The participants in the FM experimental group also

exhibited a significant main effect for improvement in the Adaptability scale of the PIADS.

Within the FM Group, the 6 personal FM users scored significantly greater on the

Competence and Adaptability scales at baseline, 12 weeks and 24 weeks than the FM users that

were only exposed to the FM signal during the worship service.

The findings of this study indicate the FM systems in conjunction with hearing aids

resulted in significantly greater outcomes than hearing aid alone. The adapted GHABP and

PIADS measures were most sensitive to these differences.









CHAPTER 1
INTRODUCTION

Approximately 29 million Americans have a hearing impairment (National Institute of

Deafness and other Communication Disorders [NIDCD], 2004). A major consequence of

sensorineural hearing loss (SNHL) is difficulty communicating, especially in adverse listening

situations such as those with a considerable distance between the speaker and listener or

environments with excessive amounts of noise and reverberation (Duquesnoy & Plomp, 1980;

Nabelek & Mason, 1981; Plomp & Duquesnoy, 1982; Dubno, Dirks & Morgan, 1984; Hawkins

& Yacullo, 1984; Needleman & Crandell, 1995; Crandell & Smaldino, 2000; Crandell &

Smaldino, 2002). The difficulty communicating may often lead to a reduced quality of life

(Ringdahl & Grimby, 2000; Pugh & Crandell, 2002), but may be partially restored with the

provision of hearing aids (Mulrow et al., 1990; Yueh et al., 2001).

Hearing aids have been successful in restoring the audibility of speech signals for listeners

in quiet environments (Humes, 1991; Humes, & Christopherson, 1996). Hearing aids have also

been successful in normalizing the limited dynamic range in those with SNHL. The dynamic

range refers to the range at which a frequency or pitch is just noticeable and extends to a level at

which the same frequency or pitch becomes intolerable. Those with SNHL have a reduced

dynamic range. Therefore, a sound must be more intense to be perceived and then quickly

becomes uncomfortable when the intensity is increased. However, restoring audibility and

maintaining comfort are only part of rehabilitating hearing loss. Individuals with SNHL may

also have other distortion factors such as reduced frequency resolution and reduced temporal

resolution (Needleman & Crandell, 1995). These deficits show those with SNHL cannot

distinguish frequencies and timing differences as well as those with normal hearing. Presumably

for these reasons, hearing aids do not share the same success in improving speech perception for









listeners in noisy and reverberant environments. To overcome these deleterious effects, the

signal-to-noise ratio (SNR) must be increased. The SNR is the level at which the desired sound

(i.e. speech) is presented above the undesired sound (i.e. noise). The critical SNR is the decibel

(dB) value where 50% of speech is understood. Individuals with normal hearing may perceive

speech when noise and speech signals are equal (0 dB SNR) or even negative (noise exceeds

speech signal) but a person with SNHL needs a much greater advantage (higher SNR) to achieve

the same level of speech perception performance.

Unfortunately, traditional hearing aids provide little or no improvement in the SNR

because the environmental distortions have altered the signal by the time it is received at the

level of the hearing aid microphone. Some hearing aids are equipped with directional

microphone technology. With directional microphones, sound to the rear and sides of the

hearing aid wearer are attenuated. By eliminating some processing of the competing noise, the

directional microphone can enhance speech perception directly in front of the listener. This

technology has shown improvement in speech perception in noise by 6-8 dB (Hawkins &

Yacullo, 1984; Gravel, Fausel, Liskow, & Chobot, 1999; Preves, Sammeth, & Wynne; 1999;

Ricketts, Henry, & Gnewikow, 2003). However, individuals with SNHL require the speech

signal to be 4 to 12 dB louder than the background noise (Killion, 1997; Moore, 1997) and

additional 3-6 dB louder in reverberation (Hawkins & Yacullo, 1984).

Assistive listening devices (ALDs) are devices that assist individuals with hearing loss to

sounds they would otherwise not perceive or understand. A hearing aid can be considered an

assistive listening device in the strict sense of the word, but typically the term 'ALD' is reserved

for two types: listening for alerts and listening to speech. An alerting type of ALD would signal

the user by tactile/vibratory motion or visual/flashing lights to a specific stimulus (e.g. doorbell).









An ALD designed for speech perception could be a stationary (e.g. amplified telephone) or

remote (e.g. FM) system.

One specific remote ALD that is well recognized for improving speech perception in

adverse listening environments is frequency modulation (FM) technology. With a personal FM

system, a wireless microphone picks up the voice of the speaker near his or her mouth, where the

affects of noise and reverberation are minimal. The acoustic signal is converted into a high

frequency radio waveform that is transmitted via FM to a receiver that is operating on the same

frequency band. If the FM system is designed for use by one user, it may be considered a

'personal FM system' because the sound is intended for one individual. It may also be

considered a personal FM system because the user may own both parts: the transmitter and

receiverss. Even with a personal FM system, the transmitted sound may be received by other

listeners if they are tuned into the same frequency of transmission. When the signal is intended

for a number of listeners at the same venue, a large area FM transmitter may be used to

intentionally relay the signal over a larger area to several FM receivers operating on the same

frequency band. Typically, these users borrow FM receivers from the venue during a particular

event but some FM users may use their own advanced equipment to scan and lock on to the

transmitting FM frequency (e.g. Phonak Smartlink FM microphone/transmitter also serves as a

synchronization device to access available channels with synthesized receivers).

When the signal is received at the receiver, it is transduced back into the acoustic

waveform and presented at the level of the ear through any one of the various coupling strategies

such as headphones or by connecting directly to a hearing aid. This process brings the relative

sound that is desired closer to the listener, therefore minimizing the effects of distance (noise and

reverberation). By eliminating these deleterious effects, the listener is able to perceive speech









more effectively. Specifically, FM systems have been shown to improve speech perception in

persons with hearing impairment by as much as 10-20 dB in the SNR over unaided conditions

(Crandell & Smaldino, 2000) and 10-18 dB over hearing aid alone (Hawkins, 1984; Fabry, 1994;

Pittman, D.E. Lewis, Hoover & Stelmachowicz, 1999; Crandell & Smaldino, 2000; M.S. Lewis,

Crandell, Valente & Horn, 2004).

Traditional, body-worn, or box style FM systems consist of relatively large transmitters

and receivers. Their size approximates an early 1990's era pager to a 1970's era garage door

opener. Both boxes are clipped to a belt, placed in a pocket, or held in the hand of the user.

Each box has a wire. The person speaking wears the transmitter box with a wire that extends to a

microphone clipped on an article of clothing near the mouth (i.e. lapel or collar). The person

listening wears the receiver box in similar fashion, but the wire extends to the ear(s) through any

of a number of coupling devices (e.g. headphones or ear buds).

Among the advances in FM technology is the ability to receive the signal in a wireless,

miniaturized unit. One wireless strategy is the incorporation of FM system receivers into

behind-the-ear (BTE) hearing aids. The mini receiver may be ordered as a feature inside the

hearing aid (Phonak iLink, AVR Sonovation Logicom, Phonic Ear Sprite) or it may be connected

via an audio boot (Phonak Microlink or Phonic Ear Lexis). One manufacturer has streamlined

the connection of the mini receiver and hearing aid to appear as a single unit (Phonak Claro,

Perseo and Savia models represent first, second and third generations of development

respectively). The remote microphone/transmitter worn by the person speaking may also be

operated in a miniaturized, wireless fashion rather than a wired connection to the wired box unit.

The small microphone/transmitter may be clipped to the lapel directly or worn around the neck

with a lanyard. Behind-the-ear FM systems (BTE-FMs) provide similar benefit in speech









perception as traditional body-worn FM systems (Boothroyd and Inglehart, 1998) and may be

cosmetically more appealing than the traditional systems. This is particularly advantageous for

many children (Madell, 1992) and adults who are concerned with the associated stigma of large

electronic devices and noticeable wires. Behind-the-ear FM systems may also facilitate

management of the units (e.g. storing in desk or locker) because the receiver is directly

connected to the hearing aid rather than in a storage space. Furthermore, the electroacoustic

variations (changes in the frequency response and electromagnetic interference) of FM systems

when coupled to different devices (direct audio input, silhouette inductor, telecoil neck loop, etc)

are markedly reduced. Because the FM signal is routed though the BTE hearing aid, it allows for

finer amplification adjustments and overall electroacoustic flexibility (Crandell & Smaldino,

2002). Achieving acoustic transparency is important in fitting BTE-FMs because the addition of

the FM receiver should be independent of the processing parameters and frequency response of

the hearing aid. There are procedures for verifying this independence (discussed later).

When referring to the body-worn FM, the local microphone on the receiver box is often

referred to as the environmental microphone, even though the system may be coupled to a

hearing aid in various ways. When referring to the BTE-FM, the hearing aid microphone serves

as the environmental microphone and the terms are often used interchangeably. In both BTE-

FM and traditional body-worn FM systems, there are three typical microphone configurations: 1)

FM-only, used to listen to an individual sound source such as a talker far away (the

environmental microphone is attenuated); 2) FM/EM, used to listen to multiple sound sources

such as talkers near and far (both the FM microphone and environmental microphones are used);

and 3) EM-only, used to listen to a sound source such as a talker nearby (only the environmental

microphone is used). Research has shown that speech perception performance from a favorable









SNR is best in the FM-only mode, least in EM only and with the FM/EM, performance is

somewhere in between (Hawkins, 1984; Fabry, 1994; Pittman et al., 1999; M.S. Lewis, Crandell

& Kreisman, 2004). Because noise and reverberation are included in the environmental

microphone, the overall benefit of increased speech perception via FM is reduced. However,

listeners may prefer to have the environmental microphone activated since they feel increasingly

detached as their local microphone becomes less dominant. Therefore, it may be preferable to

include the option that allows the user to select between FM, FM/EM and EM only. It may also

be preferable to have adjustable gain on the FM and EM microphones to allow for changes to the

FM advantage (the gain applied to the FM microphone vs. the EM microphone). The default

setting on most Phonak Microlink receivers is a +10 dB. Synthesized FM technology allows for

manipulation of the FM advantage from -6 to +24 dB if the user is experiencing difficulty with

speech (increase FM gain) or comfort (decrease FM gain). The FM advantage can also be

adjusted by decreasing the gain in the environmental microphone in the hearing aid fitting

software. In the hearing aid only mode, the gain is actually reduced to a level below threshold

instead of attenuating completely.

The development of FM technology has been quite remarkable and the audiological benefit

of improving the acoustic signal is well documented. However, the acceptance of wireless

technology has not been widespread (Ermann, 2006). Some postulated reasons include: cost;

cosmetics; complexity; reduced clinical competence in fitting FM with hearing aids among

professional; and, lack of sufficient counseling and coaching on appropriate usage to the FM use

(Ermann, 2006). This study seeks to determine the impact that FM usage has on an experienced

hearing aid wearer in their place of worship, a commonly reported environment in which users

report listening difficulty despite wearing their hearing aids.









CHAPTER 2
REVIEW OF THE LITERATURE

Speech Perception in Children with FM Systems

The nature of the educational environment demands that students hear what is being said

before the lessons can be learned. Classroom acoustics are critical variables to consider in the

achievement of students with hearing impairment (Finitzo-Hieber & Tillman, 1978). For this

reason, it is no surprise that pediatric fittings have dominated 80% of the FM market (Ermann,

2005). The benefit of increased speech perception performance for children using FM has been

thoroughly validated (Hawkins, 1984; Moeller, Donaghy, Beauchaine, D.E. Lewis and

Stelmachowicz 1996; Boothroyd & Englehardt, 1998; Pittman, D.E. Lewis, Hoover and

Stelmachowicz, 1999).

Hawkins (1984) evaluated the effects of various hearing aid and FM systems on speech

perception in noise. Subjects included 9 children ranging from 8 to 13 years of age with mild-to-

moderate SNHL. The subjects wore Phonic Ear hearing aids (805CD Behind-the-ear, BTE) and

a Phonic Ear FM system (441T transmitters, 445R receivers). Speech perception was assessed

with spondaic words and phonetically balanced (PB-K) words presented at 65 decibels sound

pressure level (dB SPL) at the subject's head in a classroom with a reverberation time of 0.6

seconds. Speech stimuli were presented via a loudspeaker at 2 meters from the subject at 0

degrees azimuth. Noise stimuli were presented at 4 meters from the subject at 180 degrees

azimuth. In the conditions that used a frequency modulated (FM) system, the transmitter

microphone was located 6 inches from the loudspeaker with the speech stimuli. An adaptive

procedure was used where the level of the speech stimuli remained constant while the noise level

was adjusted in 2 dB steps to a signal-to-noise (SNR) that resulted in 50% performance. Speech

perception was assessed in the following conditions: 1) monaural hearing aid in omnidirectional









microphone mode; 2) monaural hearing aid in directional microphone mode; 3) binaural hearing

aids in omnidirectional microphone modes; 4) binaural hearing aids in directional microphone

modes; 5) FM only in directional mode to monaural hearing aid via neck loop; 6) FM only in

directional mode to monaural hearing aid via silhouette inductor; 7) FM only in directional mode

to monaural hearing aid via direct audio input; 8) FM only in omnidirectional mode to monaural

hearing aid via direct audio input; 9) both FM microphone in directional mode and hearing aid

microphone in omnidirectional mode routed monaurally via direct audio input (DAI); 10) both

FM microphone in directional mode and hearing aid microphone in omnidirectional mode routed

binaurally via DAI and 11) both FM microphone in directional mode and hearing aid

microphone in directional mode routed binaurally via DAI. Results revealed that directional

microphones on the hearing aids provided significantly speech perception scores than the

omnidirectional microphones (2.6 to 3.5 dB). There were no differences between binaural and

monaural hearing aid fittings. The FM only condition provided significantly better speech

perception scores than all hearing aid only conditions (11.8 to 18.4 dB, mean 15.3 dB). The 'FM

only' condition provided significantly better speech perception scores than 'FM plus hearing aid'

conditions (7.9 to 16.9 dB). There were no differences in speech perception performance

between the coupling methods. On the FM transmitter, the directional microphone mode

provided significantly better speech perception than the omnidirectional microphone mode (3.3

dB). Taken together, these results led Hawkins to recommend the option to switch between

modes of FM microphone only, FM and hearing aid (FM+HA) microphones active

simultaneously and hearing aid microphone only. He also recommended that when in the FM+

HA mode that a directional hearing aid microphone for an optimal classroom amplification.









The hearing aid fitting in the Hawkins (1984) study used the half-gain rule. This may

provide too much amplification of the hearing aid microphone in the FM+HA mode, obscuring

the benefit of the SNR. Also, noise was presented at 180 degrees azimuth. This may not reflect

a realistic listening environment. The improved speech perception when the hearing aid was in

the directional mode compared to the omnidirectional mode is not surprising because the target

speech is originating from the FM microphone. The attenuation provided by directional

microphones proximal to the hearing aid helps to improve the SNR but total attenuation (FM

only) is the best way to preserve the advantage. The FM only configurations were superior to

both configurations that incorporated the hearing aid microphone (HA only or FM+HA). If the

goal were to perceive multiple speech signals simultaneously through the FM and hearing aid

(e.g. an alerting situation such as a baby monitor), omnidirectional microphones at both levels

may be beneficial to maximize the amount of acoustic input. However, if only two speech

sources are desired, one near the FM microphone and one near the hearing aid microphoness, it

would appear that directional microphones at both ends would be most beneficial because it

attenuates competing noise from the sides. Therefore, the polar plots (the graphic representation

of microphone sensitivity) of both microphones should be considered with regard to the desired

and competing sound sources.

Although improved speech perception ability in the classroom was found for students

using FM systems, the effects of an improved signal over a longer period of time did not show

similar benefit. Moeller et al., (1996) evaluated the effects FM system usage on language

development in nonacademic settings. Subjects included ten children ranging in age from 2 to 4

(at the beginning of the study) with mild-to-severe sensorineural hearing loss (SNHL) who

attended an aural/oral preschool where speech therapy was provided in the school setting only.









All subjects wore hearing aids and Phonic Ear (471 or 475) FM systems during school. Half of

the subjects wore a Telex TDR-6 FM as often as possible outside the classroom setting while the

other half wore their hearing aids only. Issues such as intermittent FM interference, multiple

caregivers, family dynamics and device bulkiness made it difficult to expect full-time use. All

subjects were followed for 2 years. The FM microphone was set 5 dB higher than the

environmental microphone. Language was assessed with at least 200 play based conversational

turns every 6 months. The conversations were recorded and scored with Systematic Analysis of

Language Transcripts (SALT) and Developmental Sentence Scoring (DSS) procedures. Results

did not reveal any differences in language development between the children who used the FM

system outside of the school setting compared to those who did not. Parents reported that the

child preferred to use the FM while listening to TV or story tapes and group situations with a

primary talker. Parents also reported that they preferred to use the FM system with their children

when visiting places such as stores, parks and the zoo because it facilitated parental control or

discipline. Younger subjects also reported a greater sense of security when caretakers were not

visible. Parents also reported negative aspects of using the FM system: it was bulky and

cumbersome; interference; communicative delays and/or inabilities in passing and/or

unwillingness to pass the transmitter microphone; and the wearer may serve as a messenger of

private information to peers. The authors also point out the potential for violating pragmatically

accepted distances for communicating (e.g. talking from another room). Despite training, the

parents sometimes failed to activate the FM and environmental microphone setting in the

presence of a third party because the unintentionally forgot or because they intentionally used the

system to control their child's' behavior.









If body-worn systems are seen as bulky and cumbersome, the miniaturization of FM

systems may be needed before a greater number of individuals can appreciate the benefit.

Boothroyd and Inglehardt (1998) examined speech perception benefit of body-worn and ear-

level FM systems and the effect of the level of the FM microphone sensitivity relative to hearing

aid microphone sensitivity. Subjects included 13 teenagers ranging from 15 to 17 years of age

(mean = 15.7 years) with severe to profound SNHL who attended an oral school for the deaf. All

subjects were experienced users of personal and classroom FM systems. The Phonic Ear 471 T

served as the transmitter, which was located at 12 inches from the speaker's mouth. Two

transmitters settings were used: the standard setting which had a compression threshold of 75 dB

SPL and the modified setting which had 15 dB less sensitivity than the standard unit, a 90 dB

SPL compression threshold. Receivers included the Phonic Ear 471 R body-worn and the

Phonic Ear Free Ear ear-level receivers. All subjects used both receivers and all but one subject

were fit bilaterally. The FM systems were adjusted for equal gain via the remote FM

microphone and built-in hearing aid microphones using 65 dB SPL sinusoidal inputs to both.

Speech perception was assessed in a classroom designed for the education of those with hearing

impairment. Speech stimuli included Arthur Boothroyd's isophonemic word lists, which were

scored as a percentage of correct phonemes identified, in quiet and noise. Multi-talker babble

served as the noise competition and was presented from four loudspeakers located 4 feet from

and facing each of the corners of the room at a height of 30 inches. The uniform sound field of

noise was adjusted to reach 55 dB SPL at the location of the teacher and student. Speech was

presented via monitored live voice without visual cues with an intensity level measured at 75 dB

SPL at the FM microphone and 60 dB SPL at the student's hearing aid microphone. Responses

were written and scored by percent correct for initial consonants, vowels, final consonants and









all phonemes. Phoneme omissions, substitutions and additions were counted as errors. The

authors concluded that the addition of the remote FM microphone was equivalent, on average, to

doubling the number of independent channels of information provided by the hearing aid alone.

The result, on average, was a 25 percentage point improvement in phoneme recognition with the

FM system in students who scored 40-60% in the aid-alone condition. This improvement was

reduced for students with poorer scores in hearing aid alone condition. Subjects performed

significantly better in quiet than in noise. However, the benefit was greater when the FM system

was worn in noise (30%) than when it was worn in quiet (20%). There was no significant

advantage of the body-worm FM system over the ear-level FM system. There were no

significant differences in speech perception between the modified and standard FM

microphone/transmitters.

Boothroyd and Inglehardt (1998) stated that although students with the most severe losses

have the greatest need for FM, they would obtain the least help with the addition of FM, at least

when expressed in terms of a percentage point increase in phoneme recognition. They found that

proportional benefit of FM is independent of aid-alone phoneme recognition performance. There

was greater benefit in the noise condition compared to the quiet condition because there was a

greater need and ability to overcome the adverse effects associated with noise. The 20 dB SNR

at the remote microphone (e.g. teacher) might not have been enough to eliminate interfering

effects at the level of the environmental microphone in the aid + FM setting. However, the

authors suggested that the 75dB SPL compression threshold of the FM microphone was probably

exceeded because speech stimuli were used. Speech has a higher crest factor than pure tone

stimuli (13 vs. 3 dB respectively) so it is possible that the amplitude peaks of the speech

spectrum, while not reflected in the root mean square (rms) average, probably activated the









compression algorithm. The FM advantage was virtually eliminated when the FM microphone

was reduced to 15 dB to equalize the FM and HA outputs as recommended by the American

Speech, Language and Hearing Association (ASHA, 1994). In this study, the equal output

rationale was not supported for those with profound hearing loss. Since the audibility of

conversational speech via the hearing aid microphone is limited, further reduction of the hearing

aid microphone to achieve a 15 dB difference below the FM microphone would be detrimental.

The ASHA guidelines now include a revised recommendation for FM fittings, stating that the

FM microphone should be 10 dB greater than the hearing aid microphone (ASHA, 2000).

Investigations on FM system benefit in speech perception performance have utilized the

same amplification scheme in both ears. To determine if the benefit of asymmetrical

amplification schemes, Pittman, et al. (1999) assessed speech perception in four system

configurations of hearing aid and FM microphones. Subjects included 11 children (mean age =

10.5) with bilateral moderate-to-severe SNHL and 8 children with normal hearing. All subjects

with hearing impairment were experienced users of hearing aids and FM systems and were

mainstreamed (oral communication). The children were fit with two different FM personal FM

systems, the Phonak Microvox and the Phonic Ear Solaris used in conjunction with Phonak Pico-

Forte C2 BTE hearing aids. Fitting was verified with the Desired Sensation Level (DSL) fitting

formula. The children with normal hearing were tested without any amplification. Speech

stimuli included four subsets of the Nonsense Syllable Test (NST) (Dubno & Dirks, 1982). Five

loudspeakers were setup in a classroom with a reverberation time of 0.6 seconds. Speech stimuli

were presented through three loudspeakers at 0, 90 and 270 degrees measured 66 dB SPL at a

location of 2 meters from the subject. The FM microphone was positioned 20 centimeters from

the front speaker, where the long term average speech spectrum (LTASS) measured 80 dB SPL.









Competing speech weighted noise was presented through four different loudspeaker positioned

at 90, 135, 225 and 270 degrees azimuth to represent students within the classroom. The side

loudspeakers presented the noise at 60 dB SPL at a distance of 1.7 meters. The loudspeakers

located behind and to the sides of the subjects presented noise at 59 dB SPL at a distance of 1.9

meters. The hearing aid and FM microphones were activated in the following configurations: 1)

FM and hearing aid binaurally; 2) FM and hearing aid binaurally simultaneously with a system

giving precedence to the FM when the input exceeds 72 dB SPL; 3) FM and hearing aid

monaurally and FM only monaurally; and 4) FM only monaurally and hearing aid only

monaurally. The conventional FM system (no precedence) gave a 10 dB system advantage by

setting the FM microphone output 5 dB above the DSL target and the hearing aid microphone

output 5 dB below the DSL target. The FM precedence system automatically reduced the

hearing aid microphone to 10 dB less than the FM microphone when it was activated. Results

revealed significantly better speech perception when speech stimuli were presented via the front

loudspeaker (with FM) than either of the side loudspeakers for both groups. There were no

significant differences between the different HA/FM configurations. Regardless of the HA/FM

condition, speech perception via the FM microphone (75%) was significantly greater than via the

environmental microphone (55%). When the speech stimuli were presented to those with

hearing impairment, initial consonants were perceived better than final consonants in both front

and side loudspeaker conditions. The authors suggest that a variety of HA/FM system

configurations may be used to provide speech perception benefit in noise, at least for a similar

group of students with moderate-to-severe SNHL. Pittman et al. (1999) also suggested that an

increase of 5 dB to the FM over the HA microphone was not large enough to improve speech









perception performance when activated simultaneously; that the output levels were already

sufficiently high to maximize performance in these children.

Speech Perception in Adults with FM Systems

Although children may use FM systems more often than adults, there is little reason to

believe that adults could not benefit from an improved SNR. Speech perception benefit has also

been documented in the adult population (Fabry, 1994; Jerger, Chmiel, Florin, Pirozzolo and N.

Wilson, 1996; Boothroyd, 2004; M.S. Lewis, Crandell, Valente and Horn, 2004).

In addition to the fair SNR benefit in directional microphones and excellent SNR benefit in

FM systems, attempts have been made to provide SNR benefit within a single microphone. Such

methods rely on changes at the microchip processing level of the hearing aid. When the resultant

frequency response is visualized, the reduction in low frequency gain is apparent; the stimulus is

high pass filtered. To determine if high pass filtering of the hearing aid microphone would

improve the SNR already provided by an FM system, Fabry (1994) assessed speech perception

ability in listeners with hearing impairment using FM integrated hearing aids that were modified

with this rationale. Subjects included five adults, ranging from 33 to 65 years of age, with

moderate-to-severe SNHL. Each subject was fitted binaurally with Phonic Ear 471 hearing aids.

Each subject was also fit monaurally with a Phonic Ear 471 FM transmitter and AT 595 Earmic

environmental microphone receiver. Real-ear measures confirmed that hearing aid fittings were

within 5 dB of National Acoustic Laboratories Revised (NAL-R) targets for the frequency

range 200-4000 hertz (Hz) with 60 db SPL composite noise. Saturation Sound Pressure Levels

at 90 dB (SSPL-90) did not exceed unaided loudness discomfort levels (LDLs) at 750 and 2000

Hz, verifying that intensity levels were not uncomfortable. Speech perception was assessed with

an adaptive speech reception threshold (SRT) procedure utilizing the Hearing in Noise Test

(HINT) sentences (Nilsson, Sullivan & Soli, 1994) in five conditions: 1) FM only (standard FM









system); 2) environmental microphone/hearing aid only with a high-pass filter (EM-HP); 3)

environmental microphone/hearing aid only with a standard frequency response (EM-S); 4) both

FM and environmental microphone/hearing aid microphones with a high-pass filter (FM/EM-

HP) and 5) both FM and environmental microphone/hearing aid with a standard frequency

response (FM/EM-S). Target speech and multi-talker noise were presented though loudspeakers

at 0 and 180 degrees azimuth, respectively. The noise was presented at 56 dB and 66 dB on the

A-weighted noise scale measured at the location of the subject. The subject was seated in the

middle of sound suite measuring 10 square meters. The FM microphone was positioned 8 inches

from the loudspeaker though which speech was presented. Results for both 56 dB A and 66 dB

A noise levels were approximately the same. The results were plotted as SNRs corresponding to

speech reception thresholds (SRTs) of 50% performance: 16 dB in the FM only mode; 13 dB in

the FM/EM-HP; 9 dB for FM/EM-S; and about 6 dB for the EM-HP and EM-S modes. There

were no significant differences in speech scores between the standard and the high-pass filtered

in the EM (hearing aid only) conditions. However, when the EM was high pass filtered for the

FM/EM conditions, speech perception scores were significantly greater (4 dB) compared to the

standard EM of the FM/EM condition.

In the Fabry (1994) study, it was hypothesized that because the ambient environmental

noise contains predominantly lower frequencies that the attenuation of the EM in this range

would improve speech perception in the FM+EM mode over traditional FM+EM mode. This

method has been commonly used as a noise reduction strategy in regular hearing aids with

questionable success. This may cause a noticeable difference in listening comfort to the hearing

aid user, but has yet to prove beneficial by increasing speech perception scores (Fabry & Van

Tassell, 1990; Tyler & Kuk, 1989). The rationale is to differentially apply spectral changes to









desired sound (speech) compared to compromised (noisy and reverberant) signals. The Fabry

article stated that in addition to replicating these results with high-pass filtering or other spectral

manipulations, that it is important to identify the appropriate level of attenuation of the EM of

the FM/EM configuration for maximal speech perception in noise from both microphones.

Although the high-pass filter looks promising because it maintains the SNR advantage for the

FM microphone while allowing the occasional (environmental) inputs via the hearing aid to be

perceived, it may be detrimental if those infrequent inputs are speech signals.

Jerger, Chmiel, Florin, Pirozzolo and N. Wilson (1996) compared the relative impact of

adding a traditional monaural FM system to a conventional monaural hearing aid on

audiological, neurophysiological and sociological outcome measures. Participants included 180

elderly adults, one hundred of which have previously worn hearing aids (average duration = 10.2

years). Ninety-four of the previous users had binaural fittings. The inclusion criteria included:

1) age greater than 60 years; 2) high frequency sensitivity loss, quantified as the average of the

pure-tone average of threshold levels at 1000, 2000, and 4000 Hz greater than 15 dB in both ears

3) normal middle ear status by immittance audiometry; 4) average score of 3 or less on a self-

report measure of physical health based on current health status, eyesight, extent to which daily

life was limited by the subject's state of health, how many days the subject had been sick, and

how often the subject had visited the doctor during the 6 months before the study; 5) a normal

score (24 or more) on the Mini Mental State Exam (Folstein, Folstein & McHugh, 1975) and 6)

no previous history of neurologic of psychiatric disorder. The experimental groups included: 1)

hearing aid only; 2) FM only, 3) hearing aid and FM, and 4) no amplification. Each treatment

condition lasted for 6 weeks. The order of treatments and the ear that received amplification

were both randomized. The hearing aids were assigned according to degree of low frequency









(LF) hearing loss. Those with mild LF hearing loss received 3M Memory Mate and those with a

greater LF hearing loss received Siemens Triton 3000. A Comtek personal FM system was used.

In the FM only condition, an earbud was used. In the hearing aid with FM condition, a neckloop

was used with the hearing aid telecoil activated. There were six test sessions. Session 1

measured pure-tone, speech and immittance audiometry and a series of screening measures of

vision, physical health and mental status to determine if the inclusion criteria were met. If so,

earmold impressions were taken and the 5 remaining sessions were scheduled. In session 2, a

battery of tests was performed to measure central auditory status, neurophysiologic status and

self-assessed hearing handicap. The first group of tests was referred to as subject descriptors:

NEO Five-factor Inventory for personality (Costa & McCrae, 1986), California Verbal Learning

Test for memory (Delis, Kramer, Kaplan & Ober, 1987); Time Sustained Attention Test for

attention (Mahurin & Pirozzolo, 1986); Purdue Pegboard test for manual dexterity; Duke Social

Support Index (DSSD; George, Blazer, Hughes, & Fowler, 1989). The second group of tests

were referred to as the outcome measures: Hearing Handicap Inventory for the Elderly (HHIE,

Ventry & Weinstein, 1982); Speech Perception in Noise Test (Kalikow, Stevens & Elliott, 1977);

Brief Symptom Inventory (BSI; Derogatis, Rickels & Rock, 1976); Social Activity Scale

(Graney & Graney, 1974); Life Satisfaction in the Elderly Scale (LSES; Salamon & Conte,

1986); Affect Balance Scale (Bradburn, 1969); and a use of amplification scale pertaining to the

past 6 weeks. The outcome measures were repeated at sessions 3, 4, 5 and 6. Each session was

scheduled 6 weeks apart. There was significant improvement in speech understanding in noise

in the two FM conditions. No significant differences were noted in hearing handicap between

the three amplification conditions. Despite the superiority of the FM conditions in noise and

anecdotal report of improved sound quality with FM, participants overwhelmingly choose the









conventional hearing aids as the amplification system they would use in daily life. Only five

participants choose the ALD over the hearing aid. Interestingly, the participants who chose the

hearing aid only perceived themselves to be more handicapped, tended to be less well adjusted

psychological and were less satisfied with their quality of life. They also showed larger left ear

deficits in dichotic listening ability, suggesting greater difficulty in the central processing of

binaural input.

Boothroyd (2004) assessed the benefits and limitations of a remote wireless microphone as

a hearing aid accessory for adults in laboratory and field tests (discussed later). Participants

included 8 men and 4 women (mean age 73 years) with a better ear three frequency average of

48 dB. Six of the participants had a sloping audiogram, defined ad a 30 dB difference in pure-

tone thresholds between 500 and 4000 Hz. Eleven were experienced hearing aid users. All

participants were fit with behind-the ear hearing aids from (the Free Ear from Phonic Ear),

linear, single-channel, analog aids with adjustable compression limiting. Participants also

received a Free Ear microphone-transmitter. The FM transmitter contained a compression-

limiting algorithm with a kneepoint of approximately 75 dB SPL. Gains for the hearing aid and

FM microphones were adjusted to be equal for inputs below the kneepoint of the FM transmitter.

Above this threshold, gain via the FM microphone was applied according to the parameters set

for the hearing aid microphone. Otherwise, the hearing aid was programmed to NAL targets.

Speech and noise were presented from digital stereo files via a laptop computer and amplified

single-cone loudspeakers (Roland 12C). Speech was presented as a distance of 3 feet and 0

degrees azimuth. Noise was presented from two loudspeakers at 3 feet and +60 and -60 degrees

azimuth. The two noises were desynchronized by a 200 ms delay. When testing via FM, the

remote FM microphone was placed at a distance of 6 inches from the speech loudspeaker. Root









mean square (RMS) noise level at both the listener's location and the FM microphone was 55 dB

SPL. The speech level at the FM microphone was always 15 dB greater than at the level of the

ear. Speech material included 20 isophonemic lists of 10 consonant-vowel-consonant words

recorded by a female talker from the eastern United States. Spectrally matched noise was added

and remained at a constant level. The speech was adjusted in 5dB steps from -10 dB to +20dB to

generate seven sets of stimuli. Four different carrier phrases were prepared for each set: with

and without noise. All stimuli were presented randomly via computer software. Each participant

attended three sessions. In the first session, pure-tone audiometry and tympanometry were

performed and earmold impressions were made. In the second session, a brief questionnaire was

administered to assess non-audiometric benefit (field study). Speech performance was assessed

unaided in quiet, aided with previous hearing aids in quiet, aided with experimental aids in quiet

and noise, and aided with experimental aids in noise with the FM microphone activated (no

hearing aid microphone). Counseling, demonstration and instruction in use of the FM system

was provided. Participants were given illustrated written instructions on the use of the FM

system and a diary for logging experiences with system. In the final visit (at least two weeks

later), the questionnaire was administered again. Speech performance measures with the

experimental hearing aids were conducted in quiet, noise and in noise with the FM microphone

activated. Participants were also given materials describing all commercially available BTE FM

systems at the time of the study. One subject was excluded because of unusually high noise

susceptibility. Performance intensity functions were generated for all conditions. Analysis of

variance and post hoc testing showed significant differences between aided performance in quiet

and noise at all speech levels 45-60 dB SPL (SNRs of -10 through +5) (p < .000). There was no

statistically significant difference between FM assisted performance in noise and aided









performance in quiet. When typical FM inputs were analyzed, performance at a 75 dB SPL input

at the FM microphone in a 55 dB SPL noise (SNR +20) was significantly greater than

performance with an input of 45 db SPL at the listener's location (p < .0012).

The speech perception data in Boothroyd (2004) study provides valuable information for

the ongoing development of a fitting protocol for FM used with hearing aids. It appears that the

equal gain criterion that was recommended for the hearing aid/FM mode (ASHA, 2000) may

have resulted in no perceived reduction of background noise switching from aid-only to FM-only

mode. When the person speaking via FM microphone stops talking, the gain is increased. This

would create a poor signal-to-non-simultaneous noise ratio. Reduction of FM gain may alleviate

this complaint but the FM benefit may also be reduced. Contrary to earlier investigation

(Boothroyd & Inglehardt, 1998), ideal initial settings might be closer to the equal output

criterion. The establishment of a fitting protocol for an initial FM system fit remains

challenging.

M.S. Lewis, Crandell, Valente and Horn (2004) compared speech perception in noise

performance in directional microphones and wireless FM systems. Participants included 55

adults in Gainesville, FL (n = 22, 68% male, median age of 73 years) and St. Louis, MO (n = 23,

57% male, median age 73 years) with mild-severe SNHL. The Gainesville group had

significantly higher pure-tone average thresholds than the St. Louis group. Despite this both

groups had similar word recognition performance. Subject inclusion/exclusion criteria included:

1) ear inspection via otoscopy within normal limits; 2) Normal middle ear function bilaterally

(+/- decaPascals [daPa] as indicated by tympanometry; 3) no evidence of conductive or

retrocochlear pathology as indicated by pure-tone testing and immittance measurements; 4) no

air-bone gap greater than 10 dB at any test frequency as indicated by pure-tone test results; 5)









slight (20-40 decibels Hearing Level [dB HL]) to severe (65 to 85 dB HL) high frequency or flat

sensorineural hearing loss as indicated by pure-tone test results (250-8000 Hz, including 3000

and 6000 Hz); 6) symmetrical hearing loss that does not differ by more than 15 dB at more than

one audiometric test frequency as indicated by pure-tone test results; 7) word recognition scores

of 50% or better in quiet as assessed by recorded version of Northwestern University, 6th version

(NU-6) monosyllables at the subjects Most Intelligible Level (MIL); 8) motivated to try

amplification as reported by the participant; 9) native speaker of English as reported by the

participant; 10) intact mental status as measured by the Short Portable Mental Status

Questionnaire (SPMSQ, Pfeiffer, 1975); 11) no history of chronic or terminal illness, psychiatric

disturbance, or senile dementia as reported by the participant; 12) no history of being

bedfast/chairfast as reported by the participant; 13) not home or nursing bound; 14) no history of

stroke or cerebrovascular disorder with a paresis or aphasia as reported by the participant; and

15) willing and able to give informed consent to participate in this investigation. All participants

were fit with Phonak Claro 311 digital Audio Zoom (dAZ) BTE hearing aids, select-a-vent

earmolds with size #13 or 3 mm horn tubing, Phonak MicroLink 8th version (ML8) receivers

attached directly to the hearing aid (no boot required) which can be worn in the FM only mode

(hearing aid microphone is attenuated 20 dB) or FM plus hearing aid microphone mode. The

Phonak TX3 handheld microphone transmitters were used. The TX3 allows the user to change

the microphone settings in the following mode: 1) "wide angle" which amplifies sound from all

directions equally; 2) "zoom" which provides less amplification to the signals originating from

the rear, a cardioid polar plot); and 3) superzoom" which provides less amplification to signals

originating from the sides and rear, a hypercardioid polar plot. The hearing aids were fit

according to the Desired Sensation Level (DSL) prescriptive fitting rationale. Participants wore









their hearing aid and FM systems for 30 days prior to testing. The Hearing in Noise Test (HINT,

Nilsson, Soli & Sullivan, 1994) stimuli were presented at 0 degrees azimuth located one meter

from the participant. The TX3 transmitter was positioned 0.5 meters in height at a distance of

7.5 cm from the loudspeaker. Correlated speech spectrum shaped noise was presented at 45, 145

225 and 315 degrees azimuth located one meter from the participant. Both signals were routed

through a dual channel audiometer (GSI 61). An adaptive procedure, the reception threshold for

sentences (RTS), was used with the noise level held constant at 65 dBA and the intensity level of

the sentences was varied to determine a 50% accuracy level. The RTS measures were measured

in the following conditions: 1) unaided; 2) binaural hearing aids in omnidirectional microphone

mode; 3) binaural hearing aids in directional microphone mode; 4) binaural hearing aids and one

FM receiver (ear assigned at random); and 5) binaural hearing aids with binaural FM receivers.

The FM receivers were set to the FM only mode (the hearing aid microphone was attenuated)

and the Superzoom mode on the TX3 microphone transmitter. The order of the conditions and

speech stimuli were randomized to avoid order and practice effects. Analysis of variance

(ANOVA) and least significant differences tests revealed significant differences between

conditions. Performance of hearing aids with binaural FM resulted was significantly better

thresholds than binaural hearing aids with one FM (p < .001). Thresholds for hearing aids with

monaural FM were significantly better than both hearing aid only conditions (p < .001).

Thresholds for hearing aids in directional microphones mode were significantly better than

omnidirectional microphone mode (p < .001). Thresholds for omnidirectional microphone mode

were significantly better than the unaided condition (p < .001).

In the M.S. Lewis, Crandell and Kreisman (2004) study, one goal was to replicate the

Hawkins (1984) study in adults with updated protocols that take into account the advances in









digital signal processing and directional microphones (on both hearing aids and FM

microphone/transmitters). A 3 dB binaural advantage was found in the binaural mode that was

not found in the Hawkins study. The author suggested that this advantage was probably the

result of being truly binaural (two ears, two receivers) rather than a diotic routing from one FM

receiver. A separate analysis also showed that among FM transmitter settings, thresholds for

speech performance through the FM microphone were best in the "zoom" setting and worst in

the "wide angle" mode (M.S. Lewis, Crandell & Kreisman, 2004). Perhaps the "superzoom"

setting is too directional to catch the voice of the speaker; that the acoustic focus is pointing to

the neck or chin rather than the mouth.

In sum, FM systems are an effective means of improving speech perception in those with

SNHL. Speech perception performance is best in the FM-only mode, least in EM only and with

the FM/EM, performance is somewhere in between (Fabry, 1994; Jerger et al., 1996; Boothroyd,

2004; M.S. Lewis, Crandell & Kreisman, 2004, M.S. Lewis, Valente, Horn & Crandell, 2005).

When the environmental microphone is activated, the overall advantage of the FM is

reduced by at least 3 dB due to the re-introduction of noise and reverberation. However,

listeners may prefer to have the environmental microphone activated to maintain awareness

around them. The process of achieving the ideal SNR in the FM/EM setting for optimal speech

perception via both microphones remains a challenge. In addition to the FM/EM mixing

consideration, attention should also be given to the appropriate placement and directionality

settings of the FM microphone transmitter.

Electroacoustic Performance Factors in FM Systems

In addition to evaluating the performance of the FM device, the mechanism between the

FM receiver and the hearing aid must also be considered. Electroacoustic variability has been









found in FM receivers, hearing aids, and the coupling methods used to connect the two (Hawkins

& van Tassell, 1982; Thibodeau & Saucedo, 1991; Hawkins & Schum, 1985).

Between FM systems, the volume control and frequency response have been found to vary.

The "linear" scale of the volume control on the FM receiver does not always provide linear

output (Hawkins & van Tassell, 1982; Hawkins & Schum, 1985). In addition to the nonlinearity,

different frequency responses of hearing aids were noted when coupled to FM systems in the

same manner (Hawkins & van Tassell, 1982; Hawkins & Schum, 1985; Kopun, Stelmachowicz,

Carney & Schulte, 1992)). Differences have also been found within the FM system components.

In a multi-center study, 30 units of the same FM system were analyzed and found to vary

electroacoustically in the receivers, the lapel microphone, and the neckloops (Thibodeau &

Saucedo, 1991). Neckloop couplers work by switching the hearing aid to the telecoil setting,

which allows the signal to be processed through electromagnetic energy. By nature, this method

is also susceptible to electromagnetic interference (Thibodeau & Saucedo, 1991). The strength

of the signal is dependent on distance to the telecoil located inside the hearing aid. Changes in

body position (head tilt/orientation) and physical fit (e.g. neck length) will have an effect

(Hawkins & van Tassell, 1982). Silhouette inductors are another coupling method that

minimizes the distance to telecoil by aligning between the hearing aid and the head or pinna

(portion of outer ear that "hold" BTE hearing aid), but this method is still susceptible to internal

noise (Hawkins & Schum, 1985).

The American National Standards Institute (ANSI) established a standard for

electroacoustic characteristics of hearing aids (ANSI-2003, S3.22) but not for assistive listening

devices such as FM systems. In the future, ANSI may develop standards similar to those of the

International Electrotechnical Commission (IEC) that standardized the parameters for telecoil









input (IEC 60-118-1), hand-held microphones, and wireless systems (IEC 60-118-3). Until then,

the electroacoustic variability in FM systems and their coupling methods may cause these

systems to be fit inappropriately.

D.E. Lewis, Feigin, Karasek and Stelmachowicz (1991) reviewed the three common

methods of assessing FM systems: functional gain, coupler measures and probe tube measures.

Functional gain is the difference between aided and unaided acoustic thresholds in sound field

(no headphones). The functional gain method was not recommended because it does not allow

for the measurement of input levels that are normally used with the system. Since it is a

threshold measure, the input levels are generally lower than the necessary 60 and 75 dB inputs

expected from conversational speech at the device microphone. Probe tube and coupler

measures were recommended because they can provide important electroacoustic information

such as the gain and maximum output. The authors preferred coupler measures because

harmonic distortion may also be measured.

In sum, the electroacoustic analysis of FM systems showed great variability in a number of

parameters both between and within the units and components. It is therefore, important to

control for as much of this variability as possible. The BTE-FM system would be expected to

have the most consistent response because the basic hearing aid parameters have been

standardized for quite some time. Furthermore, BTE-FM systems allow for greater flexibility in

fitting because the hearing aid has additional controls to allow for finer electroacoustic

modifications. However, it remains to be seen if the hearing aid is transparent (no change in

frequency or gain response when connected to the FM system). To ensure that the signal has not

been modified, real-ear probe microphones and/or coupler measures should be used for

verification, as recommended by D.E. Lewis et al. (1991).









Guidelines for Fitting FM Systems

Presently, there is no standard electroacoustic measurement procedure for FM systems

(Thibodeau, 2006). The American Speech-Language Hearing Association (ASHA, 2000) set

forth guidelines for fitting and monitoring personal FM systems, but does not specify

recommendations for large area FM systems. The guidelines instruct the person testing the

device to place the microphone/transmitter in the test box and deliver various input stimuli so

that the output can be measured. A large area transmitter cannot be measured with this

procedure because it does not have a microphone; it requires audio input from another audio

output source

Unless indicated, the remainder of this section refers to key points set forth in the

guidelines for fitting and verifying personal FM systems (ASHA, 2000). A similar procedure

would be followed if the attempt to verify the large area systems is made (if a reference

microphone was created for such testing).

It is assumed that the input arriving at the FM remote microphone (6-8 inches from the

mouth of the talker) is at least 15 dB greater than the input to the hearing aid at a distance of 1-2

meters from the talker, which is typically 60 to 70 dB SPL. Therefore, in the testing situation it s

recommended that the input level of the FM microphone be set 15 dB higher than the local

microphoness. The guidelines further suggest that manufacturer measurements provided with

the device should be evaluated according to the ANSI S3.22 (1987) standard, which established

parameters used in verifying hearing aid specifications. The authors on the ASHA ad hoc

committee recommended that the FM and environmental (on receiver, hearing aid, or both)

microphones be evaluated separately.









The FM microphone/transmitter may have a compression algorithm that modifies some

aspect of the signal between receiving and transmitting. It is therefore recommended that the

SSPL90 measure be made to the environmental microphone to reflect the true SSPL.

Some receivers have a volume control wheel for the FM signal and/or the environmental

signal. Care should be taken to control for these various combinations of output because

undesired effects may result. The user may also become confused as to what the appropriate

adjustment should be made so reducing the number of volume control options may facilitate ease

of use.

The general steps for adjusting gain to an FM system include: 1) measure output of the

hearing aid into a 2cc coupler for an input of 65 dB SPL at a frequency of 1000 Hz; 2) couple the

FM system with the respective method of coupling and activate the FM setting; 3) adjust the

volume on the FM system to match the output in step #1 with a 65 dB, 1000 Hz input (mark the

location on the FM volume control wheel); 4) increase the input of the test box to 80 dB SPL. If

the hearing aid output via FM increased by a step of 10 dB, the fitting is complete. If the hearing

output via FM increased beyond 10 dB, adjust the FM volume control to decrease the output to

10 dB. If the hearing aid via FM output is increased by a step that is less than 10 dB, adjust the

FM volume control to increase the output by 2 or 3 dB to provide a 7 or 8 dB advantages rather

than a 10 dB advantage. If there was no increase in output when the input changed from 65 to 80

dB SPL, you may assume that the FM transmitter has a very low compression threshold. In this

case, increase the volume control of the FM transmitter to provide a 5 dB increase of output

(giving a 5 dB advantage).

It the user chooses to wear a self-contained system (e.g. earbuds), adjust the characteristics

of the environmental microphone to match the settings of the user's personal hearing aid.









A swept tone or speech-weighted noise may be used rather than a 1000Hz pure tone. In

this case, the focus should be on the 500-2000 Hz range.

In some systems, there is an automatic reduction of gain in the local microphone when the

FM microphone is activated. This feature helps to maintain the signal-to-noise ratio benefits for

speech received via the FM microphone. Unfortunately, it also reduces the audibility of other

talkers not wearing the FM microphones. If simultaneous use of the FM and local (typically, the

hearing aid) microphones is the rule rather than the exception for a given individual, then the

adjustment of gain via the local microphone should be made with the FM channel on, but

receiving no input.

Current FM microphone transmitters and hearing aids both exhibit some type of

compression (typically when input exceeds 72-75dB). This causes difficulty in measuring and

validating the FM advantage because it affects the output measurement. Different approaches to

this dilemma have been proposed. Appendix B of the ASHA (2000) guidelines listed three

criteria for adjusting the FM gain during the simultaneous (FM and hearing aid) inputs: the equal

output criterion; the equal gain criterion; and, the +10 dB FM advantage criterion. The equal

output criterion was not recommended for simultaneous inputs because there is no SNR

advantage (this procedure can be used for 'FM only" verification). The equal gain criterion was

partially recommended because it provided an excellent SNR (+17 dB) but caution was

expressed because when no speech is present via FM, there would be an increase in noise at the

hearing aid microphone. The '+10 dB FM advantage' criterion was recommended as the basis

for adjusting gain in the FM channel because it offered a compromise between the limitations of

the two earlier criteria and a favorable SNR of +14. The FM advantage is the loudness of the

FM microphone relative to the hearing aid microphone. The SNR advantage is the difference of









SNR with and without the presence of an additional component (here, the FM system). The FM

advantage can therefore be considered the subjective correlate of the measurable SNR advantage.

The verification of simultaneous inputs remains challenging because typical clinical

equipment can only conduct sequential tests that estimate the activity at each microphone

separately. Recently, the Phonak Offset Protocol (POP), a manufacturer initiated guideline for

verifying simultaneous inputs was announced. Similar to the aforementioned criteria in the

ASHA (2000) guidelines, this procedure involves measurements taken in sequential fashion with

standard clinical equipment (Platz, 2006). However, this procedure has been compared to results

from a rigorous protocol with sophisticated equipment that allowed for Fourier analysis

(separation of simultaneous sound source contributions in a complex acoustic signal). The POP

calls for a 65 dB input to be delivered to the HA microphone with the FM receiver connected and

set to the FM+M (inside test box) and the FM transmitter off (outside test box). The output of

the HA at 750Hz, 1 kHz and 2 kHz is noted. Then, the same 65dB stimulus is delivered to the

microphone of the FM transmitter (inside test box) and the HA (in also connected and set to

FM+M position). The output of the HA at the same 3 frequencies is noted again. The first

measurement is subtracted from the second measurement. If the difference or offset between the

two measurements lies within +/- 2dB, the FM has achieved acoustical transparency. If the

offset value is greater than 2 dB, the recommendation is to reduce the FM gain by the offset

value. If the offset value is less than 2 dB, the recommendation is to increase the FM gain by the

offset value. It remains to be seen if this procedure has gained or will gain widespread

acceptance among FM audiologists. But the POP has been validated by two proofs: 1) the offset

is influenced by the parameters of the input at the HA microphone and not by signal processing









scheme); and, 2) that the offset remains constant for any FM transmitter, FM receiver and

hearing aid (Platz, 2006).

FM System Usage and Satisfaction

Although the audiometric benefit of FM systems is well established, the acceptance of this

technology among adults remains low in some studies where the participant had to pay for the

device at the end of the trial (Jerger, Chmiel, Florin, Pirozzolo and N. Wilson, 1996; Boothroyd,

2004, M.S. Lewis, Crandell, Valente & Horn, 2004) and exceptionally high where the participant

did not incur the expense (Noe, McArdle, Hnath-Chisolm et al. 2004; Hnath-Chisolm, Noe, &

McArdle, 2004). At first glance, cost appears to be the determining factor in the decision to

retain FM systems. However, other factors such as cosmetics, inconvenience (M.S. Lewis et al.

2004) and level of counseling (Boothroyd, 2004; Noe et al., 2004; Hnath-Chisolm, 2004)

complicate the findings.

While most people prefer the sound quality when the FM system was added to their

hearing aids, but the overwhelming majority of individuals preferred to wear only their own

hearing aid (Jerger et al., 1996). The author concluded that the strong preference for

conventional hearing aids in everyday life "undoubtedly reflects the fact that elderly users

usually are not willing to endure the difficulties associated with the use of remote-microphone

systems" if such systems involve a relative large transmitters and receivers with their respective

microphone and coupling wires.

Sanford and Kierkhaefer (2002) evaluated the subjective benefit of Phonak Microlink

system in 28 adults with various BTE hearing aids and degrees of hearing loss. After a 3-5 week

trial, participants were asked to rate nine listening situations with hearing aid only and hearing

aid and FM on a 4 point Likert scale: 1) poor, 2) fair, 3) good, and 4) excellent. All participants

reported that the addition of the FM improved their self-perceived hearing ability. Twenty of the









28 participants elected to purchase their FM system at the end of their trial. The participants who

purchased the FM reported greater benefit (1.2 point improvement) than those who did not (.5

point improvement) in the four most difficult listening situations. This finding may be

statistically significant, but an analysis was not reported. It was also unclear if the both groups

identified the same situations as being most difficult.

In the Boothroyd (2004) field study, all participants reported some or considerable overall

perception of benefit. The questions were designed to identify difficult listening situations and

their importance to the subject (if they experience those situations). The following situations

were described: 1) one person in quiet at a distance; 2) one person in noise at a distance; 3) in the

car; 4) one person in noise close; 5) watching TV; 5) in a meeting (including church); 6) in a

restaurant; 7) one person in quiet close; 8) listening to the radio, and 9) overall. The responses

were ordered: not used, worse, no help, some help, a lot of help. The results after at least two

weeks of FM usage showed that there were no situations in which participants felt that the FM

system made their communication worse. Benefit was reported as numbered above. Situations 1

and 2 involving distance in quiet and in noise were rated as most beneficial. A quasi-parametric

procedure in which numerical values were applied to the categorical/ordinal data: 0 for "no

use/no help", 1 for "some help" and 2 for "a lot of help". The seven beneficial listening

situations were summed. Listening to one person close by and listening to the radio, situations

where FM would not be expected to provide benefit were excluded. There was no significant

correlation between the benefit metric and pure-tone thresholds, speech perception performance

(aided in quiet and noise, FM assisted in noise) and age. Various comments were reported about

the FM microphone and typically involved sensitivity to soft sounds and occasional static

interference. One complained of localization difficulty and another was uncomfortable asking a









friend to wear the microphone. None of the 12 participants decided to purchase the hearing

system (FM integrated hearing aid and remote FM microphone). None of the participants, most

of who have worn in-the-ear hearing aids, expressed concern over the large size of the hearing

aid or the noticeable antenna. Nobody expressed concern about cost. It was not clear if the

participants were explicitly asked about these features. The author points out that some

situations were not beneficial because the listener did not experience them. They may have

altered their lifestyle because of the communicative difficulty this situation posed. The author

concluded that there is, "the need for considerable counseling, instruction and coaching,

extended over several sessions, if remote wireless microphones are to become widely accepted as

hearing aid accessories by adults with hearing loss" (p.32).

While the need for FM counseling and training seems essential, further investigation into

other aspects of the device and the behavior of the device user are needed. For example, if the

participants did not want to incur the monetary expense or believe that the device drew attention

to them, these questions should have been asked. Furthermore, the benefit data are only reported

for post-FM use. The differences between pre- and post- benefit scores would have been

informative to factor out the contribution of the non-experimental hearing aids that the

individuals were wearing prior to the experimental trial.

Noe, McArdle and Hnath-Chisolm et al. (2004) identified criteria for candidacy of FM

systems and Hnath-Chisolm, Noe, and McArdle (2004) reported the outcomes of a 7-week trial

of ear-level FM systems that were fit according to those criteria. Forty-three participants from

the Veteran Affairs Medical Center (VAMC) in Bay Pines, FL and Mountain Home, TN were

enrolled. Of the 43 enrolled, 31 have completed the protocol, 4 were still in the protocol at the

time of writing, seven withdrew from the study for various personal reasons and one withdrew









because he did not like the FM system. The inclusion criteria included: 1) at least a moderate

adult onset hearing loss with no evidence of retrocochlear pathology; 2) at least 6 months of

hearing aid experience; 3) dissatisfaction with their current hearing aids in at least one listening

environment in which an FM system would be beneficial; 4) appropriate reading and cognitive

skills to participate in the study as determined by informal clinical assessment, and 5) no known

neurological, psychiatric, or co-morbid disease. Participants were seen in five sessions over

seven weeks. In the first session, the hearing aid functioning was verified and initial outcome

measures were administered. First, goals for FM usage were established. Participants were

asked to identify situations that they most wanted to improve with the use of an FM system.

Using the Client Oriented Scale of Improvement (COSI, Dillon, James & Ginnis, 1999) the

audiologist categorized each of the open-ended situations identified into the 16 typical listening

situations. Seven situations were categorized: one or two people in quiet, one or two in noise,

conversation with a group in quiet, conversation with a group in noise, TV/radio at a normal

volume, familiar speaker on the telephone and church/meeting. The church/meeting was

identified most often as the first listening priority and often as the second (and overall) listening

priority. Although the responses were both arranged in a 5-point Likert scale, the degree of

change was termed differently between sites. Mountain Home used "degree of improvement" to

which the responses included: "worse", "slightly worse", no change", "slightly better", and

"better". Bay Pines used "final ability" to which the responses included "hardly ever (10%)",

"occasionally (25%), half the time (50%), most of the time (75%), and "almost always (95%).

The second outcome measure was 18 selected items from the Communication Profile for the

Hearing Impaired (CPHI, Demorest & Erdman, 1987; Erdman & Demorest, 1990). The items

asked the respondent to indicate how well they communicate in various situations, in which they









responded in a Likert scale from: "1 rarely, almost never", 2 occasionally, sometimes", 3

"about half the time", "4 frequently, often" and usually, almost always". The third outcome

measure administered was a group of selected items from the Marketrak surveys (Kochkin 2000,

2003) to assess hearing aid satisfaction in specific listening situations and device features.

Respondents were asked to rate "very dissatisfied" to "very satisfied" in 12 situations: one-on-

one, small groups, large groups, outdoors, concerts, worship, TV, music, leisure activities,

restaurants, cars and telephones. Another question was added to the Marketrak survey, "How

often do you find yourself embarrassed, ridiculed or rejected because you wear an FM system?".

Participants returned a week later for the second session when they were fit binaurally with the

Phonak Microlink FM system. The transmitters included the TX2 (body worn with lavaliere

microphone), TX3 (Handymic hand held unit), and TX4 (TV/phone) and were dispensed

according to communication needs. The MLx and ML8 ear-level receivers were used depending

on hearing aid compatibility (analog and digital). The FM was fit and verified with real ear

measures according to the method described by Hawkins (1984). Educational instruction was

provided in writing (with picture support) and verbal instruction on how to use the FM systems

according to their respective goals. Participants returned at 2 week intervals (sessions 3-5) to

report their perceived performance in achieving their listening goals and to receive additional

instruction as needed. At the end of the 6 week trial, mean perception of hearing ability was

significantly better ("better" or "much better") in all 7 situations identified by the COSI (p<.05).

The mean CPHI score increased one point with the addition of FM (hearing aid only = 2.44,

hearing aid with FM = 3.44, p< .00005). There was an improvement in Marketrak satisfaction

across all listening situations and a statistically significant improvement in small groups (p <

.000), worship (p< .005), T> (p < .002), restaurants (p < .02), cars ( p < .02), and telephones (p









<.000). The device feature items pertaining to cosmetics ("visibility to others" and the

embarrassment-ridicule-rejection item both showed that 90% were satisfied with these aspects.

Of the 31 participants who completed the study, all decided to keep their FM system.

Participants were also asked at the end of the trial how much they would be willing to pay for the

FM system had it not been provided free of charge. Using an anchor of hearing aid cost ("about

$4000 a pair") the participants indicated that they would be willing to pay approximately $2300

for the FM system. The high level of acceptance could be attributed to several factors: the FM

systems were provided at no personal cost to the participant; the participants received rigorous

counseling, coaching and instruction; the device has reached a level of acceptability in terms of

cosmetic appearance, ease of use or performance (M.S. Lewis, Valente, Horn and Crandell

(2005) used the complete version of the CPHI on the sample of 23 participants described in Site

II of M.S. Lewis et al. (2004). The CPHI is a 145-item self-assessment inventory that measures

hearing handicap, adjustment to hearing loss, and communication strategies. There are 25

subscales that are categorized in four areas: 1) communication performance; 2) communication

environment; 3) communication strategies; and 4) personal adjustment. Responses are arranged

in 5-point Likert ordering on a continuum of frequency of agreement/disagreement. The sample

was randomized into hearing aid only and hearing aid plus FM groups. There was crossover of

experimental groups at 3 months. The interviewer did not fit the participants and was thus

blinded from knowing group assignment. Note: The Phonak Claro 311 daZ BTE with ML8

matching receivers were used. These units have the streamlined design that was mentioned

earlier in this paper. The FM receivers were actually attached in the hearing only mode but the

FM was not activated. The CPHI was administered at baseline (hearing aid only) and once per

month during the hearing aid plus FM experimental condition. Repeated measures ANOVA









showed statistically significant differences between the hearing aid only and hearing aid plus FM

group for the following subscales: 1) importance of communication at work; 2) importance of

communication at social situations; 3) importance of communication at home; 4) problem

awareness; 5) behavior of others 6) verbal strategies and 7) stress. Communication partners

made less accommodations and participants used less verbal strategies with the utilization of FM

amplification. Despite statistical significance, all but one subscale (work) failed to reach clinical

significance because the differences did not exceed the 90% confidence interval established in

the original sample of 101 active duty military (Demorest & Erdman, 1988). The fact that many

of the adults were older (retired) made the significance in the work subscale questionable. None

of the 23 participants elected to purchase the hearing system. The difficulties that were reported

were: 1) expense 2) inconvenience (e.g. need to charge the transmitter each night); 3) cosmetic

issues (e.g. need to point the transmitter near the mouth of the talker).

It remains to be seen if the use of extensive counseling and coaching with the cosmetically

advanced devices may have an impact on the penetration of FM technology into the population

of adult hearing aid wearers. When factoring in personal expense, only one study (Sanford and

Kierkhaefer, 2002) reported a substantial uptake rate (>70%). Perhaps their clientele enjoyed a

greater socioeconomic status. While market influences are beyond the scope of this paper,

further investigation of satisfaction and benefit in multiple dimensions via subjective measures in

this population is needed.

Outcome Measures

The World Health Organization convened a group of experts known as the International

Classification of Functioning, Disability and Handicap (or ICF) to provide a unified and standard

language and framework for the description of health and health-related states (WHO, 2001).

There are two broad domains, one stated in a positively grammatical sense, "functioning" and









another that take on a potentially negative connotation, "disability". Functioning encompasses

all body functions, activities and participation. Disability encompasses all impairments, activity

limitations or participation restrictions. Impairments refer to problems with body function or

structure such as a significant deviation or loss. Activity refers to the execution of a task or

action by an individual. Participation refers to involvement in a life situation. Within the

framework, functioning and disability domains also interact with environmental (external) and

personal (internal) factors.

Validity refers to the approach of measuring whether a scale measures what it purports to

measure. In other words, do the quantifiable constructs tell you something about a concept?

Responses to the outcome measures described in the following section are the primary data

sources in the proposed study. However, audiometric data (e.g. degree and configuration of

hearing loss) demographics (e.g. age, socioeconomic status, etc) and medical history (e.g.

previous hearing aid usage etiology) information will also be important factors to consider when

interpreting the scale values. The goal in audiologic rehabilitation is to maximize the functional

abilities associated with hearing loss. Clearly that involves more than the provision of a hearing

aid to "make things louder". Audiologists must also take into account these non-audiological

factors. Vanity, for example, is a very prominent barrier to providing appropriate amplification.

If a young man is not willing to wear a BTE hearing aid because he fears being embarrassed and

stigmatized, that issue needs to be addressed. Some audiologists may take a cognitive approach

to counseling and would respond that without amplification, the concerned person might be more

embarrassed and stigmatized for a different reason; that the effects of the hearing loss are more

apparent than the sight of the hearing aid. If an elderly woman can be fit audiologically with a









completely in the canal (CIC) hearing aid, but she cannot insert and remove the device because

of limited manual dexterity, the audiologist has failed to maximize her functional ability.

There are a number of converging constructs that may lead to the valid measurement of

hearing benefit: age; years with hearing loss with and without amplification; hours of daily use;

subjective ratings of device satisfaction; severity of hearing loss; activity requirements, etc.

Investigation of seemingly generic variables may also lead to the development of operational

constructs that are sensitive detecting a change. For examples, you might not expect to see a

relationship of inter-ear differences in speech perception scores, hearing loss by gender, or inland

vs. coastal dwellings. When operational constructs are identified, the relationship to the latent

variable may be supported. In the previous examples, such operational constructs could be the

detection of lesion in corpus colosum (physiological), a history of noise exposure (occupational),

and the common report of wind noise interference to the hearing aid microphone (environmental)

respectively.

In the proposed study, it is believed that the following outcome measures validly assess the

dimensions involved with using an FM system. It is believed that the worship specific items on

these outcome measures validly assess the dimensions involved with using an FM system in a

place of worship.

Abbreviated Profile of Hearing Aid Benefit (APHAB)

The Abbreviated Profile of Hearing Aid Benefit (APHAB) is a scale that examines the

disability (or activity limitation) domain specifically. Developed by Cox and Alexander (1995),

it is a 24-item scale with four sub-scales: 1) Ease of Communication; 2) Reverberation; 3)

Background Noise; and 4) Aversiveness to Sound. The items include phrases or sentences to

which there is a 7-point Likert response scale indicating how often each situation is experienced:

always (99%), almost always (87%), generally (75%), half-the-time (50%), occasionally (25%),









seldom (12%) and never (1%). There is a response column for both unaided and aided

conditions. The difference score between unaided and aided is referred to as the benefit provided

by the hearing aid. Each subscale contains an average of the responses for 6 items with item

values corresponding to the percentage of time the situation is experience. To minimize

response bias, 6 items have been phrased in a positive communication context and 18 in a

negative communication context. The responses for all items remained the same. Prior to data

analysis, the 6 positive item values are reversed before the subscale score is calculated (APHAB

Scoring Instructions, 2006). Scores range from 1%-99%.

The APHAB was normed on 128 elderly, experienced, successful hearing aid users

(defined as wearers for greater than four hours per day) with mild-to-moderate sloping or flat

sensorineural hearing loss (Cox & Alexander, 1995). It typically takes 10 minutes or less to

complete. The APHAB was developed from the 66-item Profile of Hearing Aid Performance

(PHAP) and Profile of Hearing Aid Benefit (PHAB) (Cox & Gilmore, 1990; Cox & Rivera,

1992). Item total correlations of the original PHAB subscales were conducted in aided, unaided

and benefit scores to reduce the number of items to 6 per category. Test-retest reliability was

assessed with comparisons from the same items with a previously administered PHAP (Coz &

Gilmore, 1990) and sister measure PHAB (Cox & Rivera, 1992). The correlations for unaided

subscales ranged from .65 to .80 with the APHAB (there were no unaided measurements with

PHAP or PHAB). The correlations with the aided subscales ranged from .77 to .84 with the

PHAP and .70 to .81 with the APHAB. The correlations of benefit (difference scores) scales

ranged from .54 to .72 with the PHAB and .48to.71 with the APHAB. The lower correlations of

the benefit subscales were expected because of the combined random error with both unaided

and aided scores (Cox and Alexander, 1995). Corrected item total correlations ranged from .54









to .66, indicating homogeneity within each subscale (Cox & Alexander, 1995). Cronbach's

alpha scores ranged from 0.78 to 0.97 indicating fairly high internal consistency (Cox &

Alexander, 1995).

Glasgow Hearing Aid Benefit Profile (GHABP)

The Glasgow Hearing Aid Benefit Profile (GHABP) is an audiological outcome measure

that measures both activity limitations and disability/participation restrictions in the same scale.

The GHABP was designed to evaluate the efficacy of aural rehabilitation services for adults with

hearing loss (Gatehouse, 1999). Four of which situations include: 1) listening to the television

with other family or friends when the volume is adjusted to suit other people; 2) having a

conversation with one other person when there is no background noise; 3) carrying on a

conversation in a busy street or shop; and 4) having a conversation with several people in a

group. There are four items in which the respondent can write in the listening situations that are

specific to their needs. Questions are formulated in each of the following dimensions: 1) initial

disability (How much difficulty do you have in this situation?); 2) handicap (How much does

any difficulty in this situation worry, annoy or upset you?); 3) reported hearing aid usage (In this

situation, what proportion of the time do you wear your hearing aid?); 4) reported benefit (In this

situation, how much does your hearing aid help you?); 5) residual disability (In this situation,

with your hearing aid, how much difficulty do you now have?); and 6) satisfaction (For this

situation, how satisfied are you with your hearing aid?). For each situation, the respondent is

first asked if the event occurs in their life, to which they can reply: no or yes. If it does not

apply, all subsequent questions for that situation are omitted from analysis. The collection of

data for frequency of occurrence in these lifestyle events can serve as an indirect measure of

situation specific usage of amplification; only the pertinent environments are assessed. The

responses for each of the seven questions within each situation are arranged in Likert fashion and









include a choice "0) N/A" for instances where the item may occur only infrequently and/or is an

event that s/he does not consider to be. Five additional response choices are included that are

relevant to each scale dimension. For example, the responses for question one (initial disability)

include: 1) no difficulty; 2) only slight difficulty; 3) moderate difficulty; 4) great difficulty; 5)

cannot manage at all. The responses are coded in the same order with discrete values 1-5. Items

for each scale are average by the number of applicable responses and converted to a 0 to 100

range (by subtracting one and multiplying by 25) (GHABP Information Package, 2006).

The GHABP was normed on 293 adults (median age = 69 years) and has a block matrix

design in which a variety of listening situations were used. The GHABP was derived from the

Hearing Disability and Aid Benefit Interview (HDABI, Gatehouse, 1999), a scale that contained

14 listening situations. The method to reduce the number of items without loss of discriminatory

power was to rank difficult listening situations according to frequency of occurrence (for the

whole population) and importance of listening in the given situation (for the individual) without

duplication. For criterion analysis, the assumption was made that increased perceptions of

improved outcome of those with hearing loss should be systematically related to increasing

improvements in the audibility of the speech signal as measured by the Speech Intelligibility

Index (SII). Multiple comparison analyses of the 14 listening situations were analyzed. The four

specified 4 situations with the highest statistical power and four open-ended situations were used

in the final version of the GHABP. The reduction of situations resulted in only a minor loss of

discriminatory power.

The assumption was made that situations that were not specified would be specified by the

user according to importance. All respondents included at least one open-ended situation, 89.8%

included at least two, 80.2% included at least three and 65.9% included four situations in









addition to the pre-specified situations. Of the open-ended situations, 'listening in church' was

listed by 33.2% of the respondents, second overall only to 'listening on the telephone' cited buy

35.2% of the respondents. Examination of the order of responses to the open-ended situation

revealed that the 'listening in church' situation was ranked third as the first response (12.3%),

third as the second response (9.6%) and first as the third response, and never as the fourth

response. If the assumption is made that the order of responding to open-ended questions

reflects precedence of perceived need, then 'listening in church' might be considered one of the

top priorities for auditory rehabilitation in the normative sample.

The internal reliability of the GHABP is high (alpha > 0.7) but the authors warn that in

may be inflated because the structure of the measure identifies situations that exist only in a

listeners experience and are of relevance. Test re-test reliability assessed at three weeks show

high correlation (all greater than r = .86). The validity assessment of the GHABP could not be

assessed because there is no similar measure that has subscales that are similar to the dimensions

measured.

Glasgow Benefit Inventory (GBI)

The Glasgow Benefit Inventory (GBI) is a measure that assesses the effect of an

intervention on quality of life. The questionnaire consists of 18-items which allows for

comparison across health conditions because it is generic in nature ("hearing aid fitting" or "FM

fitting" may be substituted for "operation/intervention"). The pilot version of the GBI

consisted of 38 items that were based on interviews with patients and surgical staff. Questions

that were condition specific and difficult to understand were eliminated and the GBI was reduced

to 18 questions. Factor analysis using principal components extraction with orthogonal rotation

and yielded a three-factor solution that explained 69% of the variance. Each factor represented a

loading coefficient above .55. The three factors make up the GBI 3 subscales: a general subscale









(12 questions, numbers 1, 2, 3, 4, 5, 6, 9, 10, 14, 16, 17, and 18); a social support subscale (3

questions, numbers 7, 11 and 15) and a physical health subscale (3 questions, numbers 8, 12, 15).

The questions for the GBI are as followed: 1) Has the result of the operation or

intervention* affected the things you do? 2) Have the results of the operation/intervention* made

your overall life better or worse? 3) Since your operation or intervention*, have you felt more or

less optimistic about the future? 4. Since your operation or intervention*, do you feel more or

less embarrassed when with a group of people? 5) Since your operation or intervention*, do you

have more or less self-confidence? 6) Since your operation or intervention*, have you found it

easier or harder to deal with company? 7) Since your operation or intervention*, do you feel that

you have more or less support from your friends? 8) Have you been to your family doctor, for

any reason, more or less often, since your operation or intervention*? 9) Since your operation or

intervention*, do you feel more or less confident about job opportunities? 10) Since your

operation or intervention*, do you feel more or less self-conscious? 11) Since your

operation/intervention*, are there more or fewer people who really care about you? 12. Since you

had the operation or intervention*, do you catch colds or infections more or less often? 13. Have

you had to take more or less medicine for any reason, since your operation/intervention*? 14.

Since your operation or intervention*, do you feel better or worse about yourself? 15. Since your

operation or intervention*, do you feel that you have had more or less support from your family?

16. Since your operation/intervention*, are you more or less inconvenienced by your health*

problem? 17. Since your operation or intervention*, have you been able to participate in more or

fewer social activities? 18. Since your operation or intervention*, have you been more or less

inclined to withdraw from social situations? For each quality, responses are arranged in a

quantitative Likert fashion ranging from: 1) Much worse; 2) A little or somewhat worse; 3) No









change 4) A little or somewhat better; and 5) Much better. To control for response bias, half of

the questions are arranged from a large improvement to a large deterioration. The five ordinal

responses are coded with discrete values 1-5 and converted to a range of -100 to +100

(averaging the responses, subtracting 3 and multiplying this value by 50) (Glasgow Health Status

Questionnaire Manual, 2006).

The GBI has been used to assess the following otorhinolayngological interventions in

Scotland: middle ear surgery to improve hearing (n=181), cochlear implantation (n=184), middle

ear surgery to eradicate ear activity (n=138), rhinoplasty (n=96) and tonsillectomy (n=61)

(Robinson, Gatehouse & Browning, 1996). Each intervention had a different criterion for

defining a success varied between each intervention (e.g. tonsillectomy success meant sore throat

was better or cured).

The GBI has since been used extensively in otoloaryngological interventions since its

inception, including: quality of life studies (Morzaria, Westerberg, & Anzarut, 2003); middle ear

implantation (Sterkers et al., 2003); bone-anchored hearing aids (McLarnon, Davison, &

Johnson, 2004); Hol Bosman, Snik, Mylanus, & Cremers, 2005; cochlear implantation

(Lassaletta, Castro, Bastarrica,de Sarria, & Gavilan, 2005); tumor surgery (Eikelboom, Eager, &

Atlas, 2005; Myrseth et al., 2005; Subramaniam, Eikelboom, Eager, & Atlas, 2005); septum

surgery (Konstantinidis et al., 2003; Konstantinidis, Triaridis S, Triaridis A, Karagiannidis &

Kontzoglou, 2005; Uppal, Mistry, Nadig, Back, & Coatesworth, 2005); sinus surgery (Mehanna,

Mills, Kelly, & McGarry, 2002; Salhab Matai, & Salam, 2004) and surgery to treat snoring

(Uppal, Nadig, Jones, Nicolaides, Coatesworth, 2004).

Hearing Handicap Inventories (HHI)

There are two versions of the Hearing Handicap Inventory (HHI), one for the Elderly

(HHIE) (Ventry & Weinstein, 1982) and another for adults (HHIA) (Newman, Weinstein,









Jacobson, & Hug, 1990). Both are 25-item measures designed to assess the perceived emotional

and social/situational consequences of activity limitations resulting from hearing loss which are

not apparent from the audiogram. Perceived handicap is an important measure because it

emerges as a construct that is separate than more ojective measures such as hearing sensitivity,

speech perception and central auditory processing (Jerger & Chmiel, 1997). They have high

internal consistency and high test-retest reliability. They are relatively easy to administer, score

and interpret. For each item the responses are: yes (4 points), sometimes (2 points), and no (0

points). Therefore, scores range from 0 (no handicap) to 100 (significant perceived handicap).

The HHIA was slightly modified from the HHIE to assess adults less than 65 years of age

(Newman, Weinstein, Jacobson, & Hug, 1990). Three items were changed in the HHIA: one

emotional question regarding occupation, one social question regarding occupation, and one

social question regarding leisure activities. Both scales take 10 minutes or less to complete.

Initial development of the HHIE items was generated by questions from five audiologists

in various clinical settings with experience working with older adults. They intended to identify

situational difficulties in their patients and probed the effect the difficulty had on limiting or

reducing the involvement of the individual in these situations (Ventry & Weinstein, 1982). A

total of 42 items comprised the original version. Preliminary data collection on 42 participants

(> 65 years) generated reliability coefficients were .93 for the emotional subscale, .83 for the

social/situational subscale and .24 for the sensitivity subscale. A few items in each group were

eliminated due to low inter-item and item total correlations and those with a high rate of "not

applicable" responses. Due to a low reliability coefficient, all but two items on the sensitivity

subscale were eliminated. Multiple regression analysis revealed two items from the sensitivity

subscale that contributed significantly toward the overall variance and were incorporated into the









final version: one social item and one emotional item. The final version of the HHIE comprised

a 25-item scale with a reliability coefficient of .95 including a 13-item emotional and 12-item

social/situational subscales with reliability coefficients of .93 and .88 respectively.

A subsequent study on mode of administration showed that reliability was maintained in

face-to-face (r = .96) and pencil-and-paper (r = 0.84) formats (Demorest and Walden, 1984).

Demorest and Walden conceded that if the HHIE was to be used as an index of change due to an

intervention, it should have low standard error. Test-retest reliability at 6 weeks was evaluated

in a sample of 27 elderly adults with hearing loss and 95% confidence intervals were calculated

to determine the necessary change in HHIE scores for such an attribution to occur (Weinstein,

Spitzer & Ventry, 1986). The face-to-face and paper-and-pencil versions resulted in confidence

intervals (and standard errors) of 18.7% (SE = 6.6) and 36% (SE = 13) respectively. Therefore,

the required difference in HHIE score to attribute a true change with 95% confidence using the

written administration is approximately double of the verbal administration. Audiometric data

were not reported in the Weinstein et al, 1986) article, but the authors caution that there may

have been lesser variability in the verbal administration because of yea or nea saying; the

respondent may have offered more consistent responses because the question was misheard. In

such a situation, it is plausible that the respondent elected repetition instead of requesting

clarification prior to responding. The recently cited critical difference for paper-and-pencil

administrations is 19.2% (Weinstein, personal communication 2006) indicates that these

constructs leading to the measurement of handicap have long term stability (only a .5%

difference since 1986).

The HHIE has shown significant differences in hearing aid outcomes at 3 weeks (Malinoff

& Weinstein, 1989) and at one year post-fitting (Newman & Weinstein, 1988).









Twenty-two of the 25 questions between the HHIE and the HHIA are identical. Three

questions have been modified. In the HHIE, the following questions were targeted for change:

1) Does a hearing problem cause you to feel "stupid" or "dumb"? 2) "Do you have difficulty

hearing when someone speaks in a whisper? and 3) "Does a hearing problem cause you to attend

church services less often that you would like?. Instead, the HHIA omitted the above questions

and included the following: 1) Does a hearing problem cause you to feel frustrated when talking

to coworkers, clients, or customers? 2) Does a hearing problem cause you difficulty in the

movies or theater? and, 3) Does a hearing problem cause you difficulty hearing/understanding

coworkers, clients or customers?.

The balance of 13 emotional and 12 social questions remains the same between versions.

The HHIA has high internal reliability with a Cronbach's alpha value of .93 for the entire

measure, and .88 and .85 for the emotional and social subscales respectively (Newman,

Weinstein, Jacobson, & Hug, 1990a). The HHIA has high test-retest reliability with Pearson

product-moment correlations of .97 for the entire measure and .93 and .95 for the emotional and

social subscales respectively (Newman, Weinstein, Jacobson, & Hug, 1990b).

The combination of HHIE and HHIA items will be called the Hearing Handicap Inventory

(HHI), a 28-item collection that administered to all adults and analyzed according to their age

(above and below age 65) and compared to their respective version (HHIE or HHIA).

Psychosocial Impact of Assistive Technology Scale (PIADS)

The Psychosocial Impact of Assistive Technology Scale (PIADS) was developed as an

outcome measure to investigate the reasons for abandonment of assistive devices (Day & Jutai,

1996; Day, Jutai, Woolrich & Strong, 2001). The 26-item measure was developed on two

premises the authors believed were central to retaining an assistive device: 1) that the assistive

device (AD) improves function, and, 2) that the AD must function to improve the quality of life









of its adopter. The authors make the distinction between these two effects because although a

user may be satisfied with the AD, they may be unhappy overall with the AD for other reasons

(e.g. they are embarrassed when using the device).

Device abandonment presents a monetary concern from a service delivery standpoint.

Identification of recipients who retain and use their devices may be critical for appropriate

spending from the perspectives of public and private service providers (Phillips & Zhao, 1993).

The assumption had been that adoption of a working AD, not necessarily an appropriate AD,

contributed to an improvement in quality of life. Vash (1983) noted that personality, motivation,

perceived and desired roles, and the amount of effort expended in using ADs compared to the

rewards experienced in using them, are psychosocial factors that appear to affect satisfaction.

There existed no widely known mechanism to operationalize or measure the constructs

toward AD satisfaction that were centered on sustaining an improvement in quality of life (Day

& Jutai, 1996). Through focus groups of people with disabilities, constructs were devised and

tested on a sample of eyeglass and contact lens wearers. Item analysis reduced the number of

items based on extreme distribution and redundancy. To account for the potential for a negative

impact on an individual's quality of life, the scale was revised to reflect a bi-directional Likert

type scale ranging from -3 (decreases) to +3 (increases) points. There are three subscales that

comprise the overall PIADS score: the competence subscale (which reflects items such as

productivity, usefulness, performance and independence); adaptability (which reflects items such

as ability to participate, willingness to take chances and eagerness to try new things); and the self

-esteem subscale (which reflects items such as emotional health and happiness). Each subscale

derives a score from -3 to +3. To control for response bias, one item in each scale was reversed

and these items are adjusted before averaging scores.









The directions state: "Each word or phrase below describes how using an assistive device

may affect a user. Some might seem unusual, but it is important that you answer every one of

the 26-items. So, for each word or phrase, put an "X" in the appropriate box to show how you

are affected by using the (device name)". The items include the following words/phrases: 1)

competence 2) happiness; 3) independence; 4) adequacy; 5) confusion; 6) efficiency; 7) self-

esteem; 8) productivity; 9) security; 10) frustration; 11) usefulness; 12) self-confidence; 13)

expertise; 14) skillfulness, 15) well-being, 16) capability; 17) quality of life; 18) performance;

19) sense of power; 20) sense of control; 21) embarrassment; 22) willingness to take chances;

23) ability to participate; 24) eagerness to try new things; 25) ability to adapt to the activities of

daily living; 26) ability to take advantages of opportunities. The PIADS can be completed in 5-

10 minutes.

The PIADS has shown high internal consistency with a Cronbach's alpha values of .95 for

the total scale. The reliability of each subscale was also high as alpha coefficients were .92, .88,

and .87 for competence, adaptability, and self-esteem respectively (PIADS manual version 4.21,

2006; Scherer, 1996; Day, Jutai & Campbell, 2002; Demers, Monette, Descent, Jutai & Wolfson,

2002. These values indicate that the three constructs mentioned above represent a large

proportion of the latent variable of interest, psychosocial impact. The PIADS has shown high

test-retest reliability (Scherer, 1996; Jutai & Campbell, 2002; Demer, Monette, Descent, Jutai &

Wolfson, 2002) demonstrating temporal stability within the measure. The PIADS is also

sensitive to device stability in that the effect of the AD intervention shows long-term

repeatability (Scherer, 1996; Jutai & Day, 2000; Day, Jutai & Campbell, 2002). The PIADS has

been used as a generic outcome measure for hearing aid trials (Jutai & Saunders, 2003; Jutai &

Saunders 2003; Jerger, 2004; Saunders & Jutai, 2004).









Concurrent validity of the PIADS was assessed on a sample of 157 eyeglass wearers with

the Pleasure, Arousal and Dominance (PAD) (Mehrabian & Russell, 1974), a measure used to

evaluate emotional responses to a diversity of conditions. The Pearson correlation coefficients

were significant (p < .05) for the pleasure (r = .46-.59)and dominance (r = .21 to .34) subscales,

but not the arousal subscale (r = .06-.17). These correlations indicate that the PIADS and the

PAD may theoretically predict the relationship of one another. Construct validity was evaluated

using principal components analysis. From a sample of 307 subjects with eyewear devices, the

three sub-scales account for 61.1% of the variance (Day & Jutai, 1996).

Spiritual Well-Being Scale (SWBS)

The Spiritual Well-Being Scale (Paloutzian & Ellison, 1982; Ellison, 1983) was developed

as a global psychological measure of one's perception of spiritual well-being. The authors

borrowed terminology from Moberg and Brusek (1978) and developed a scale based on two

dimensions: a vertical dimension that refers to one's sense of well-being in relationship to God

and a horizontal dimension that refers to one's perception of the purpose of life and life

satisfaction apart from any specifically religious preference. The two sub-scales were based on

these concepts: the former dimension was conceptualized as religiosity spiritual well-being and

the latter dimension as existentiality. Collectively, the total score is intended to reflect the

construct of spirituality.

An independent review contended that the SWBS is not a measure of spiritual health or

spiritual maturity (Boivin, Kirby, Underwood & Silva, 1999). The SWBS is a psychological,

rather than theological measure that is nonsectarian and can be used in a variety of religious,

health and research contexts (Boivin et al., 1999). The 20-item scale consists of evenly divided

items that are believed to measure religious and existential dimensions. The items in the religious

subscale mention "God" in the item. The items in the existential subscale do not mention "God",









instead asking about life purpose, satisfaction and relations with other people and situations.

Each Item is rated in a 6-point Likert scale ranging from: SA "strongly agree"; MA "moderately

agree"; A "agree"; D "disagree", MD "moderately disagree" and SD "strongly disagree".

Approximately half of the items are worded in a reverse direction so that disagreement with the

item reflects higher well-being. Point values range from 1-6 with the higher number

representing higher well-being. The range is 10-60 points per subscale and 20-120 total. It takes

about 10-15 to complete.

The SWBS was normed on 206 college student and additional studies employing over 500

of various age, gender and geographical representations. Subsequent investigations have used

the SWBS on participants with various physical and mental disorders. The internal reliability,

based on data from over 900 participants, ranged from .89-.94 for the total score, .82-.94 for the

religious subscale and .78 to .86 for the existential subscale (Boivin et al, 1999). Test-retest

reliability coefficients for four different samples with 1, 4, 6 and 10 weeks between

administrations ranged from .82-.99 for the total SWBS, .88-.99 for the religious subscale and

.73-.98 for the existential subscale (Boivin et al., 1999). The vertical/religious and

horizontal/existential approach to conceptualizing the construct of spirituality appears to be

valid. An initial factor analysis revealed that all religious items loaded on the religious factor

and most existential items loaded on the existential factor (Ellison, 1983). The scores on the

SWBS have been positively correlated with measures of physical well-being such as ideal body

weight and perceived health and negatively correlated with pain (frequency, intensity and

duration) (Granstrom, 1987).









CHAPTER 3
METHODS

Research Question

The goal of this study was to determine if a six month trial period with a large area

frequency modulation (FM) system had an impact on individuals with hearing impairment who

desired to hear better in their place of worship by using an FM device. Specifically, the impact,

if any, was assessed in self-reported measures of: 1) auditory functioning (disability/activity

limitation and handicap/participation restriction); 2) health related quality of life; 3) psychosocial

impact to assistive devices, and 4) spiritual well being. It is hypothesized that usage of an FM

system in addition to current amplification with hearing aids will improve each of dimensions

that are measured with their respective metrics. In this experiment, it is expected that only items

related to the place of worship will exhibit the increase in score. Therefore, the specific

hypothesis for this study is that the FM users in this study will show significant increases in their

scores pertaining to worship related items. Worship related items include: items 18 and 21 of the

Abbreviated Profile of Hearing Aid Benefit (APHAB), the customized listening situations

(discussed later) on the Glasgow Hearing Aid Benefit Profile (GHABP); item S11 on the

Hearing Handicap Inventory for the Elderly (HHIE); and, all items on the Spiritual Well-Being

Scale (SWBS). The remainder of items in the previously listed measures, all items on the

Glasgow Benefit Inventory (health-related quality of life) and Psychosocial Impact of Assistive

Devices (PIADS) measures are expected to show increases in mean scores, but are not expected

to reach significance as they are more generalized, global measures.

Pilot Investigation

To estimate current usage and potential need of assistive listening devices (ALDs), a self-

addressed, stamped envelope was mailed with a letter and questionnaire to the 227 places of









worship in the Gainesville area listed in the yellowpages.com (accessed 9/1/05) website with

complete, unduplicated addresses. The letter briefly explained about ALDs and their function.

The questionnaire portion of the mailing asked about the representative's general knowledge and

provision of ALDs, the estimated physical dimensions of place of worship, and a description of

the listening experiences of the members including how many were reported difficulty. The

questionnaire was anonymous. Of the 227 letters sent, 27 were returned from the postmaster as

undeliverable. Of the 200 mailings, 43 were returned for a response rate of 21.5%. One

questionnaire was omitted from the analysis because most questions did not contain responses.

The results of the pilot investigation are summarized in Table 3-1. In the 42 completed

questionnaires, 40 of 42 (95.2%) respondents reported the provision of an amplified public

address (PA) system during their services. Sixteen of the 40 (38.1%) places of worship with a

PA system also reported having assistive listening devices (ALDs). Of the 16 places of worship

with ALDS: 11 (68.8%) described the ALD as an FM system; 3 (18.8%) did not know or specify

the type of ALD; and 2 (12.5%) described the ALD as an infrared system.

Among all 42 respondents: 33 (78.8%) reported awareness of ALDs; 7 (16.8%) reported

no awareness of ALDs; and 2 (4.8%) reported being "somewhat aware". The two respondents

that did not have a public address system also reported being unaware of ALDs.

When asked if there is a need for ALDs in their house of worship: 24 representatives

(57.l1%) indicated that there is a need, 9 (21.4%) responded that there is not a need; and 8 (19%)

responded that there might be a need. The majority of places of worship surveyed reported

awareness/limited awareness (83.3%) and reported need/potential need for ALDs (76.1%).

However, only 38.1% reported having ALDs. This finding suggests that places of worship are

knowingly not providing the listening accommodations for all of their congregants.









When the respondent was asked if congregants with normal hearing experience difficulty

hearing the presenter during the service: 28 (66.7%) believed that they do not experience

difficulty; 6 (14.3%) believed that they experience difficulty "sometimes"; 4 (9.5%) believed

they experience difficulty; and 4 (9.5%) were "not sure" if they experience difficulty. When

asked if congregants with hearing impairment experience difficulty hearing during the service:

16 (38.1%) believed that they experience difficulty; 9 (21.4%) believed they experience

difficulty "sometimes"; 9 (21.4%) believed they did not experience difficulty; and 8 (19%) were

"not sure" if they experience difficulty. Collapsed into a binary question, the respondents

believed that 23.8% of congregants with normal hearing and 59.5% of congregants with hearing

impairment have experienced difficulty hearing the service.

Two of the questions were aimed toward estimation of the physical space of the venue.

When the respondents were asked to estimate the physical volume inside the main area of the

place of worship: 16 (38%) described the space as "medium (20,000-100,000 cubic feet)"; 12

(28.6%) described the space as "small (less than 20,000 cubic feet)"; 7 (16.7%) could not

estimate the volume; and 6 (14.3%) described the space as "large (more than 100,000 cubic

feet)". When asked to estimate the seating capacity in the main area of the place of worship: 20

(47.6%) estimated less than 200 people; 12 (28.6%) estimated between 200-399 people; 8 (19%)

estimated between 400-799 people; and 2 (4.8%) estimated 800 or more people. The majority of

the places of worship in Gainesville were estimated to have physical volumes of less than

100,000 cubic feet (66.7%) and less than 400 people (76.1%).

Correlational analyses were conducted by Pearson chi-square tests. Responses were

collapsed into binary responses ("yes" and "maybe" treated as "yes"; and "no" and not sure"

treated as "no"). Results of the comparisons are shown in Table 3-1. Four of the 10









relationships were significantly correlated (c = .05). Respondents that reported a need for ALDs

were also likely to report that: congregants with normal hearing were perceived to have difficulty

during the service; congregants with hearing loss were perceived to have difficulty during the

service; and the estimated seating capacity was high in their place or worship. Respondents that

reported greater awareness of ALDs were also likely to report that congregants with hearing loss

had difficulty during the services. It appears that when the respondents suspect a need for ALDs

that they are aware of their existence and that the congregants with different hearing abilities

have listening difficulty during the service, particularly when congregating in a larger venue. It

is unclear if the congregants were perceived to have residual difficulty when using the ALD

because the questionnaire did not probe this dimension.

Main Investigation

Participants were recruited through convenience sampling of various places of worship in

Gainesville, FL, a mid-central Florida city of approximately 100,000 (U.S. Census, 2006).

Larger facilities were targeted first because they had a greater pool of potential participants and a

greater likelihood of adverse acoustical characteristics. An attempt was made to contact places

of worship of different faiths and denominations. The Christian and Jewish faiths were

represented in the sample. The Islamic faith could not be represented because there was no such

place of worship listed in the yellowpages.com search. Each place of worship was contacted via

telephone and a representative was identified. The questionnaire data from the previous mailing

was not used to identify any potential participants/venues because it was anonymous.

Of the 14 place of worship (PW) contacted, 9 were interested in learning more about the

study and 5 declined to participate. Although the representatives were not asked for reasons for

declining participation, 3 offered explanations. Two representatives believed that their









congregations were preoccupied with atypical activity: one congregation was moving to a

different location and the other was participating in a series of retreats. The third representative

reported, "we are managing the [FM system] fine. We have a device for people with hearing

aids and for people without hearing aids so we don't need your help". Of the two PW that

declined without reason, one representative simply stated that they were not interested and the

other representative did not return phone calls.

A slide presentation was created on Microsoft Powerpoint and burned to digital video disc

(DVD) to disseminate information about the research study. The DVD was delivered to each

interested representative and they were asked to review its contents.

Interviews with interested representatives were arranged to answer questions, clarify the

purpose of the study and arrive at mutual agreement on the terms of the participation. Nine

interviews were conducted. One PW declined to participate two weeks after the interview. The

representative explained that he polled the congregation informally and determined that there

was no interest in participating.

The remaining 8 PW agreed to participate. One PW had only one potential participant and

he decided to attend another PW instead, one that had already been enrolled in the study. The

participant conveyed that his decision was not made on the basis of the study but that, "it is

something we [he and his wife] had been meaning to do anyway so we could join our daughter

and her family [at the other PW]".

A total of 7 PW were enrolled in this study. The participants by place worship are

summarized in Table 3-2. Installations of Listen Technologies (LT-800) large area FM

transmitters were arranged at the convenience of the representative. The procedure for installing

the device is similar to installing a home electronic component (e.g. DVD player) in that there









was only a couple of necessary connections: one for the audio input and another for a power

supply. The audio input possibilities included: an XLR microphone cable, a 1/4 inch stereo

connection, or RCA connections. The principal investigator installed four large area transmitters

and the remaining three were installed by either the music minister or house manager of the

respective PW. Six of the installations used XLR inputs and 1 used a 14 inch audio input.

The 7 PW that were enrolled in the study included 3 Methodist churches, 2 Catholic

churches, 1 Lutheran church and 1 Synagogue. The PW that chose not to participate included 3

Baptist churches, 2 non-denominational Christian churches, 1 Catholic church, and 1 Episcopal

church.

All participating PW agreed to allow the CD presentation to be shared with interested

members of the congregation during an information session. An announcement was made in the

weekly bulletin and an on-site location was reserved. One PW included the information session

in their newsletter (Methodist A) and another PW (Methodist B) invited the principal

investigator to make announcements during two Sunday services, one week prior to the

information session. Five places of worship arranged weekday evening meeting times. One

place (Lutheran) reserved time during the day because it facilitated transportation for the

primarily senior population. One place (Methodist B) conducted the informational session on

Sunday following their most heavily attended service. Meeting attendance varied from one

person (Catholic B) to approximately 60 people (Synagogue).

The information session included the same Microsoft Powerpoint slides as the

informational CD, projected on screen or white wall. The principal investigator narrated the

presentation in person to facilitate speech reading and to offer clarifications to several questions.









The session outlined that each participant may use an FM system for a 6 month trial

without any monetary obligation. Those who were current hearing aid wearers at the time of the

information session and still perceived difficulty listening to the services in their place of

worship were eligible to participate. A sign-up sheet was made available and the potential

participants were contacted via their preferred means (phone or email) to schedule a clinic

appointment.

The clinic appointments were scheduled for two hours. The average length of each

appointment was 2 hours and 15 minutes. At the beginning of the clinic visit, the participant

read and signed the Informed Consent form approved by the Institutional Review Board at the

University of Florida. A copy of the Informed Consent form was provided to each participant.

An audiological evaluation was performed including: otoscopy, immittance audiometry

and pure-tone air/bone conductance audiometry and speech audiometry. These measures reflect

functional aspects of outer ear, middle ear, inner ear and central processing respectively (but not

exclusively). A medical case history form was also administered.

Video otoscopy (Panasonic video camera GRK 5162, Panasonic power supply #521248,

Sony Trinitron monitor) was performed to rule out excessive cerumen and to ensure the integrity

of the ear canal and tympanic membranes. None of the participants displayed symptoms of

soreness or signs excessive cerumen, inflammation/redness, or perforation of the tympanic

membrane. One participant was referred to an otolaryngologist for follow-up because of the

presence of glue-like substance in ear canal of one ear.

Immittance audiometry was performed on a Grason Stadler Instruments (GSI) Tympstar

immittance bridge. A computer monitor was attached to TympStar for ease of viewing.

Immitance testing included ipsilateral (tone presented and measured in same ear) acoustic









reflexes at .5, 1, 2, and 4 kilohertz (kHz). If acoustic reflexes were present at .5 kHz or 1 kHz,

acoustic decay was measured at the same frequencies. Acoustic decay is the reduction in

response magnitude within 10 seconds from a pure tone stimulus presented at 10dB SL (decibels

Sensation Level re: acoustic reflex threshold). The presence of decay indicates a positive

finding, and would be grounds for exclusion from participating in this study. None of the

enrolled participants presented abnormal immittance results.

Pure tone audiometry was performed on a GSI-61 dual channel audiometer. Thresholds

were defined as the lowest level at which the participant responded to the presence of a tone 50%

of the time. Thresholds were obtained at .25, .5, 1, 2, 4 and 8 kHz. Inter-octave thresholds at

.75, 1.5, 3 and 6 kHz were obtained when there was an inter-ear difference of 15 dB or greater at

adjacent octave frequencies. Narrow band masking noise was routed to the contralateral ear via

the second channel of the audiometer when indicated; when crossover auditory perception was

possible.

Speech recognition threshold (SRT) was defined as the lowest level that speech is

understood half of the time and was measured using spondaic words (bi-syllabic words with

equal stress on each syllable, e.g. toothbrush). Speech perception performance was measured as

a percent correct score on the Northwestern University auditory test #6 (NU-6) monosyllabic

word lists. The 50-item word lists were spoken by a female talker on a compact disc recording

produced by the Department of Veteran's Affairs in Mountain Home, TN (version 2.0, lists 17-

24). Fifty words were presented at the most comfortable level (MCL) for each condition. The

conditions included: monaural left ear in quiet; monaural right ear in quiet; binaural in quiet; and

binaural in noise (pink/speech noise at + 10dB signal-to-noise ratio (SNR). Pure-tone and









speech audiometry was performed in a single walled sound booth. Stimuli were routed through

Etymotic Research (ER3A) insert earphones or TDH-50P headphones.

A routine professional calibration was performed March 8th, 2006. The technician

recommended usage of headphones because the calibration ensured that the impedence of the

transducer matched that of the audiometer. Participants evaluated prior to noon on March 8,

2006 used insert earphones and those evaluated afternoon on March 8, 2006 used headphones.

Insert earphones were initially used instead of headphones because they are more hygienic,

facilitate ease of testing because need for masking is reduced, and avoids erroneous thresholds

due to ear canal collapse. To ensure that the insert earphones thresholds were accurate, two

individuals (one with normal hearing and one with a moderate to profound sensorineural hearing

loss) were tested with both inserts and headphones in the research sound booth and in the regular

clinic sound booth. There were no differences in the thresholds obtained; all thresholds were

within test-retest reliability (+/- 5 decibels).

FM receiver selection

Following the audiologic testing, the information from the on-site session was reviewed

with all 32 participants. All options for coupling FM receivers to hearing aids were discussed

and differentiated. The participant was provided a paper handout that listed the advantages and

disadvantages of each. The decision matrix included factors such as signal quality, cosmetics

and financial considerations for using the devices) during and after the trial (Appendix A).

None of the participants chose Direct Audio Input (DAI).

The participants were reminded (from informational session) that they could purchase, at

the end of the 6-month study, any new Phonak FM product available at a significantly reduced

cost. One hearing aid model, the Phonak iLink with completely integrated FM, was also

included as an option for purchase. This particular hearing aid model was made available









because the FM receiver could not be separated from the hearing aid because it was an internal

component. Because this unit was primarily a hearing aid with FM as a feature, payment was

required at the beginning of the study. However, the standard trial period that allowed the user

to return the hearing aid-FM system within 45-days for a full refund (including reimbursement

for the purchased earmolds) was extended for the duration of the study. The price quotes for all

FM products, including the iLink hearing aid were given only when the participant asked.

Of the 29 participants who completed the study, 23 chose to use an FM system and to

participate in the experimental group and 6 chose not to use an FM system and to serve in the

comparison group. Of the 23 participants in the experimental group, 12 chose ear level receivers

(5 ML8S, 2 iLink hearing aids, 2 ML9S, 2 MLxS, 1 ML4S) and 11 chose belt level receivers (5

earbuds, 3 neckloops, 3 headphones). The selection of receivers is summarized in Table 3-2.

The three participants who chose the iLink hearing aid-receiver were fit by a licensed,

certified audiologist. The fittings were conducted at the University of Florida Speech and

Hearing Clinic in Dauer Hall. The fitting fee for each of these participants was paid using the

research grant, funded by Phonak Hearing Systems. One participant was initially fitted with the

iLink but during her hearing aid trial upgraded to a more sophisticated Phonak Eleva behind-the-

ear (BTE) hearing aid with ML9S receiver. This participant was the only monaural fit of the

sample.

Participants were counseled on how to change between their microphone mode settings. A

hearing aid wearer that uses an FM system in conjunction with their hearing aid can choose to

have their hearing aid microphone activated, their remote FM microphone activated and both

hearing aid and FM microphones activated simultaneously. Suggestions were given on when it

might be preferred to change the active microphone settings based on the type of listening









situation (Appendix B). A hands-on demonstration was given and the participant was asked to

repeat the steps to demonstrate understanding.

The participants were counseled on the appropriate use of the FM system. They were

instructed to wear the FM system each week during their worship servicess. Traditional users

were given one AA battery and were instructed to replace the battery every 6 weeks. Wireless

users were reminded that the receiver will draw more power from the hearing aid battery. Both

groups were advised to bring replacement batteries with them to their weekly services.

Enrollment by place of worship

The procedure for advertising the study, enrollment of participants, and selection of

receivers are summarized in Table 3-2.

At the Lutheran church, 8 people expressed interest at the session on 2/1/06 and 7 enrolled

(5 experimental, 2 comparison). Of the 5 enrolled participants in the experimental group, 3

chose a Microvox belt level receiver (2 with headphones and 1 with neckloop) and 2 chose ear

level receivers (ML8S and ML9S). Two participants in the experimental group at Lutheran

church dropped out of the study. One participant, who wore ML9S ear level receivers, dropped

out because he experienced a serious, unrelated injury and could not attend services. The other

individual, who wore the belt level receiver with headphones, dropped out because he lost

interest in participating since enrollment.

At Catholic Church A, 7 people expressed interest at the information sessions on 2/5/06

and 2 of them were enrolled (1 experimental, 1 comparison). The participant in the experimental

group chose a Microvox belt level receiver with earbuds.

At Catholic church B, 1 person expressed interest at the session on 2/7/06 and was enrolled

(experimental) but participant decided to attend Catholic church C instead. Catholic Church B









was then dropped from the potential list of study sites. This person was omitted from the final

data set because they did not return the 12 or 24 week questionnaire booklet.

At the synagogue, 4 people expressed interest at the session on 2/8/06 and 2 of them were

enrolled (one experimental, one comparison). The participant in the experimental group chose a

Microvox belt level receiver with a neckloop. The participant in the comparison group initially

chose and was fit with MLxS ear level receivers but within the first week, reported too much

anxiety about losing them and chose to participate in the comparison group instead.

At Methodist church A, 5 expressed interest 2/9/06 and 4 were enrolled. All 4 participated

in the experimental group: 2 wore ear level receivers (ML4S and ML8S) and two chose

Microvox belt level receivers with earbuds. The two participants with the ear level receivers

were experienced FM users: one for 2 years and the other for 20 years. However, neither

experienced user used their FM system during their worship service until this study.

At Methodist church B, 13 expressed interest at the session on 2/26/06 and 10 were

enrolled. Of the 10 enrolled participants, 5 chose ear level receivers (2 MLxS, 2 iLinks, and 1

ML9S) and 5 chose belt level receivers (2 earbuds, 2 headphones, 1 neckloop). The participant

with the ML9S receivers was fit at the VA medical center in Gainesville during the enrollment

period. He was also fit with a Smartlink microphone/transmitter. Two participants were fit

binaurally with Phonak iLink hearing aids. One of the iLink users was also fitted with an

Easylink microphone/transmitter.

At Catholic church C, 5 people expressed interest at the session on 2/27/06. A total of 6

participants were enrolled (including one participant from former site Catholic B). Five

participants elected to be in the experimental group and one elected to serve in the comparison

group. Of those enrolled in the experimental group, 3 chose ML8S ear level receivers, one chose









a Microvox belt level receiver with earbuds and one withdrew from the study before choosing an

FM receiver. The participant who withdrew did not feel comfortable answering the questions on

the Spiritual Well Being Scale and cited this reason for declining further participation. The

participant in the comparison group decided not to use FM because he developed a strategy to

compensate for his hearing difficulty. Because he served as an usher, he was able to position

himself near the loudspeaker in the crying room at various times during the service. This

member of the comparison group may therefore not be representative as a suitable comparison

unit as he no longer experienced listening difficulty.

At Methodist church C, 4 expressed interest at the session on 3/1 and 2 were enrolled in

the experimental group. One participant chose to be fit monaurally with an iLink monaural

hearing aid but later switched to Phonak Eleva BTE hearing aid with a ML9S receiver and a

Smartlink microphone/transmitter. The other participant chose a belt level receiver with

neckloop.

Hearing aid adjustments

Current hearing aid settings were downloaded with the HI-PRO hearing aid hardware

device and Noah software program and saved with a code (no participant identifiers). If hearing

aid adjustments were desired, they were made according to patient preference. Five participants

requested changes to their hearing aid programs. Frequency response changes included: a 4 dB

increase in high frequency gain; a 2 dB decrease in low frequency gain; and an increase in the

compression threshold (kneepoint) for quiet sounds. Program changes included: activation of the

telecoil (as T-only) for one participant and activation of the hearing aid microphone in addition

to the telecoil (T+M) for another participant. The final program settings were saved to the Noah

database with and the hearing instrument.









The three within-program hearing aid adjustments were verified with real-ear acoustic

measurements. Using the Audioscan Verifit, the gain and output was be measured in the main

program of the hearing aid at soft (55 dB SPL), average (65 dB SPL) and loud (70 dB SPL)

inputs of speech weighted noise. This procedure did not assess the gain or output of the FM

receiver through the large area transmitter. There are currently no guidelines for evaluating such

a system (Thibodeau, 2006 personal communication).

Quality control protocol for verifying large area FM

In effort to validate the electroacoustic variability of a sample of FM receivers used in this

study, a quality control protocol of a sample of the equipment was developed. Using the same

procedure described in American Speech Language Hearing Association (ASHA, 2000), a

sample of 3 Phonak Microvox receivers, 3 Monacor MD-300 headphones and 3 Radio Shack

earbuds were randomly selected. Each of nine receiver/coupler combinations were verified at

55, 70, 75 dB sound pressure levels (SPLs) of speech-shaped noise from the Audioscan Verifit

real-ear loudspeaker at 2 receiver volume settings (corresponding to dial positions 3 and 6). A

Phonak Smartlink microphone transmitter was placed 10 inches from the Verifit loudspeaker.

The headphones and earbuds were positioned on the Knowles Electroacoustic Manikin for

hearing Aid Research (KEMAR) which was positioned 3 feet from the Verifit loudspeaker. The

average output data for the left and right ears have been collapsed. The real-ear outputs at the

level of the ear canal for earbuds at volume settings 3 and 6 are presented in Figures 3-5 and

Figure 3-6 respectively. The real-ear outputs at the level of the ear canal for headphones at

volume settings 3 and 6 are presented in Figures 3-7 and Figure 3-8 respectively. In all four

figures, the output lines are not symmetrical which indicates variable operation of a compression

algorithm across the frequency range. More importantly, none of the outputs at any frequency

exceeded 90 dB SPL in any of the four conditions, which verifies that the sound pressure level in









the ear canal is not excessive. Despite this, users of FM systems in this study were instructed to

adjust their volumes to a comfortable level.

Questionnaire booklets

A number of relevant self-report outcome measures were compiled into a Microsoft Word

document with 14-point size text and Times New Roman font. All measures were printed and

tape-bound into booklets. All measures were arranged in a Likert response format that quantifies

or orders the attribute being measured. All responses were arranged vertically in the original yet

unnumbered order. The questionnaire booklet for the FM group is included in Appendix C. All

administrations were written self-reports from each participant. During the initial administration,

the investigator was present and provided assistance with completing the forms as needed. Each

participant typically had one question or request for clarification. The most common request for

clarification occurred during the administration of the Abbreviated Profile of Hearing Aid

Benefit (APHAB) when respondents were asked to answer each question in the situation where

they are not wearing hearing aids. The respondents reported stated that they are not in these

situations without their hearing aids. They were instructed to imagine that they were in that

given situation today without their hearing aids and to response hypothetically. The glossary for

the items in the Psychosocial Impact of Assistive Devices Scale (PIADS) was also provided

during the initial administration to clarify any meaning of definitions used in the measure, as

recommended by Pauloutzian and Ellison (1982). Participants were instructed to respond to all

items.

Subsequent administrations were also written self-report measures that were nearly

identical to the initial administration. The only difference to the repeated measures was the type

of amplification. In subsequent administrations, the amplification pertained to the hearing aid

and FM system combined. It was identified by a teal booklet cover. If the participant was part









of the comparison group, s/he received another copy of the initial administration (beige booklet)

that pertained to amplification that was limited to hearing aid only.

The subsequent administrations were mailed to arrive on Monday following the 12th and

24th weekend service. They were asked to complete and return in the postage-paid return

envelope prior to their next weekend service.

All clinic administrations referred to outcomes experienced by the participant when using

the hearing aid alone. For the comparison group, the mailed versions referred to outcomes

experienced by the participant when using the hearing aid alone. For the FM group, the mailed

versions referred to outcomes experienced by the participant when using the FM in addition to

their hearing aid.

Six participants in the sample used a personal FM system in addition to the large area FM

system installed at their place of worship: 3 were fit prior to study enrollment (two with 2 years

experience and one with 20 years experience); and 3 were fit during the study (two at baseline

and one at approximately 12 weeks). Therefore, 6 participants had additional exposure to the

FM signal in addition usage in the place of worship. Three of the 6 personal had cumulative FM

exposure prior to the study. Additional analyses were performed on the 'personal FM'

participants (discussed later).

The outcome measure booklet contained the following measures, in order : the

Abbreviated Profile of Hearing Aid Benefit (APHAB; Cox & Alexander, 1995); the Glasgow

Hearing Aid Benefit Profile (GHABP) (Gatehouse, 1999); the Glasgow Benefit Inventory (GBI,

Robinson, Gatehouse & Browning, 1996); the Hearing Handicap Inventories (Ventry &

Weinstein, 1982; Newman, Weinstein, Jacobson, & Hug, 1990); the Psychosocial Impact of

Assistive Devices Scale (PIADS) (Day & Jutai, 1996); and the Spiritual Well-Being Scale









(SWBS) (Paloutzian & Ellison, 1982). Where applicable, the terms "hearing aid only" and

"hearing aid with FM" were used for their respective versions. The "hearing aid with FM" or

experimental version of the outcome measure booklet is included in Appendix C.

Adaptations to outcomes measures

The original version of the Glasgow Hearing Aid Benefit Profile (GHABP) was

administered at baseline. A revised version of the GHABP reflecting the addition of FM was

administered at 12 and 24 weeks. For both versions, a total of 9 listening situations were

described: items 1-4 were pre-specified (previously mentioned) and items 5-9 were created

specifically to address listening in the place of worship. Recall that in the normative sample,

'listening in church' was listed by 33.2% of the respondents, second overall only to 'listening on

the telephone' cited buy 35.2% of the respondents. It could be assumed that if the number of

pre-specified items in the GHABP measure were expanded, an item pertaining to the place of

worship would likely be included. Because listening in this situation is commonly reported as

difficult and because FM exposure in this investigation is limited to the place of worship, five

items have been specified. Item 5 describes, "listening to the main presenter in a house of

worship (e.g. pastor, priest, rabbi, etc.)". Item 6 describes, 'listening to other presenters in a

house of worship (readers, guest speakers, people who make announcements, etc.). Item 7

describes, "listening to music in a house of worship (understanding the lyrics of people

singing)." Item 8 describes, "having a one-on-one conversation in a house of worship before or

after the service (many people speaking simultaneously)." Item 9 describes, "having a one-on-

one conversation in a house of worship during the service (while one presenter is speaking)."

For all listening situations, the following domains were investigated: handicap, reported usage,

reported benefit, derived benefit (residual disability minus initial disability) and satisfaction. The

Abbreviated Profile of Hearing Aid Benefit (APHAB) was administered at baseline. For each









item on this administration, the participant was asked to respond as if they were unaided and

aided with only their hearing aids. A revised version of the APHAB reflecting the addition of

FM was administered at 12 and 24 weeks. For each item on this administration, the participant

was asked to respond as if they were aided only with their hearing aids and also aided with their

hearing and FM system. Benefit scores were derived by subtracting the aided score from the

unaided score. Benefit scores in the positive direction showed the listening difficulty was

experienced less often and scores in the negative direction showed that listening difficulty was

expressed more often.

In the baseline administration of the APHAB, the 'unaided' response column asked the

participants (who were experienced hearing aid users) to state hypothetically the amount of time

they would experience difficulty in the given situation without their hearing aids, and the 'aided'

response column asked the participants to state how much difficulty they experienced in the

given situation with their hearing aids until the time of administration. In the subsequent

administrations, 'unaided' reflected amplification with hearing aid only and 'aided' reflected

amplification with hearing aid and FM system.

The 28 items from the Hearing Handicap Inventory for Adults (HHIA) and Hearing

Handicap Inventory for the Elderly (HHIE) were combined and administered at baseline, 12

weeks and 24 weeks. The 25-item HHIA and the 3 items that differed on the 25-item HHIE

version were combined to form the HHI section. However, the HHIA and HHIE were analyzed

separately as their original 25-item scales.

For the baseline administration, the intervention in the Glasgow Hearing Benefit Inventory

(GBI) was "since your hearing aid fitting". The intervention in the GBI at 12 and 24 weeks was

"since your FM system fitting".









For the Psychosocial Impact of Assistive Devices Scale (PIADS), the assistive device was

specified as an FM system in the directions.

Compensation/Other Benefit

To enhance participant compliance, the participants were compensated monetarily ($50) at

the completion of the study. The place of worship was permitted to keep the transmitter if there

was a perceived need. All places of worship were expected to keep their FM transmitters

because at least one participant from each venue expressed interest in ongoing use. Participants

were sent a letter prior to the completion of the study that gave the purchase price of the

receivers) if they decided to keep after the trial. Collection of FM items is ongoing so a

summary of purchases and returns cannot be completed at the time of this writing.

Analyses

Data were analyzed with the Statistical Package for the Behavioral Sciences (SPSS)

statistical analysis software program. For the Stage 1 pilot questionnaire data, Pearson chi-

square analyses were run between items for each respondent. The remainder of the analyses was

conducted for the Stage 2 main investigation. For each subscale of each outcome measure a

within-subjects, a repeated measures Friedman analysis of variance (ANOVA) was performed on

all 23 participants in the FM group to determine if there are any differences between scores at

pre- (baseline), mid- (12 weeks) and post-trial (24 weeks) administration at c = .05 significance

level. If significant differences were found, Wilcoxon rank sum tests were performed on the

pairwise comparisons to determine if the significance remained at a specific time interval.

Between-subject comparisons with the Mann Whitney U test were performed to determine

if there were differences between FM group (n = 23) and the comparison group (n = 6) on: pure

tone audiometry results, speech audiometry results, items 5-9 of the GHABP, the PIADS









measure and the three original religious items (Items 18 and 21 on APHAB and Item S 11 on the

HHIE) to determine if there were any significant differences at the a =.05 level:

Between subject comparisons with the Mann Whitney U test were performed for items 5-9

on the GHABP, the PIADS measure and the three original religious items (Items 18 and 21 on

APHAB and Item S11 on the HHIE) to determine if there were any significant differences at the

a =.05 level: 12 users with ear level receivers and 11 users with belt level receivers; 12 poor

performers (<20%) and 11 fair performers (>/= 20%) on word recognition performance in noise

at + 10dB SNR; and the 17 users that used FM only in the place of worship and the 6 users that

used personal FM systems.

Hypotheses

It was hypothesized that a 24 week trial of an FM system used in the place of worship will

have a significant effect on: 1) disability (lesser); 2) handicap (lesser); 3) quality of life (greater);

4) psychosocial impact (more positive); and, 5) spiritual well-being (greater); as measured with

their respective metrics in items pertaining to the place of worship listening environment.

It was also hypothesized that FM users with ear level receivers will have significantly

greater (more positive) psychosocial scores than FM users with belt level receivers as reflected in

the PIADS measure.

It was also hypothesized that the 24 week trial with an FM system used in a place of

worship will have no difference on worship specific outcome measures between: the poor and

fair performers of speech perception in noise; and between FM users that were exposed to the

FM signal only in the place of worship and FM users who were exposed to the FM signal in

more than the worship setting.











Table 3-1. Pilot investigation data summary: pearson inter-item chi-square values.
Seats Need Awareness Normal Hearing
hearing impaired
Seats

Need ** 9.79

Awareness 2.64 2.65

Normal 8.02 ** 12.89 4.93
hearing
Hearing 6.31 *** 31.45 ** 9.82 14.86
impaired


Seats
Need
Awareness
Normal Hearing

Hearing Impaired

*
**


Informant estimate of seating capacity.
Informant estimate of the need for Assistive Listening Devices (ALDs)
Informant estimate of awareness of Assistive listening devices (ALDs)
Informant estimate of perception of listening difficulty among congregants
with normal hearing
Informed estimate of perception of listening difficulty among congregants
with hearing impairment
p < 0.10
p < 0.05
p < 0.001









Table 3-2. Main investigation enrollment summary: participant selection of FM receivers and
transmitters.
Ear Level Personal
Receivers Belt Level FM Users
Date Interested Enrolled (type- Receivers Selected (Transmitter
(2006) Denomination Participants Participants quantity) (coupling-quantity) model)
2/1 Lutheran 8 7 2 3 0
(ML8S-1) ** (headphones-2)
(ML9S-
1) (earbuds-1)
2/5 Catholic A 7 2 0 1 0
(earbuds-1)
2/7 O Catholic B 1 0 0 0 0

2/8 Synagogue 4 2 0 1 0
(neckloop-1)
2/9 Methodist A 5 4 2 2 2
(ML4S-1) (earbuds-2) (Smartlink)
(ML8S-1) (HandyMic)
2/26 Methodist B 13 10 5 5 2
(iLink-2) (headphones-2) (Smartlink)
(MLxS-2) (earbuds-2) (Easylink)
(ML9S-1) (neckloop-1)
2/27 Catholic C 5 ***6 3 1 2
(Smartlink-
(ML8S-3) 00 (earbuds-1) 2)
3/1 Methodist C 4 2 1 1 0
(ML9S-1) (neckloop-1)
TOTAL 47 33 13 14 6
* This participant dropped out due to non-study related injury.
** One participant dropped out due to lack of interest.
*** One participant dropped out during clinic visit prior to FM receiver selection. Also,
another participant from Catholic B joined Catholic C.
0 Participation site eliminated due to lack of enrolled attendees. No FM installed.
00 Omission due to non-compliance (no 12 or 24 week data).









Table 3-3. Individual participant data for the 23 participants in the FM group: amplification
history and audiometrics.
Word Recognition


Participant
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
Mean
SE


Age
79
84
93
71
93
61
74
91
81
74
71
82
86
74
80
88
76
78
66
75
56
77
63
77.18
10.06


Yrs
HA
Use
11
23
26
5
4
39
NR
1
20
6
8
NR
21
9
8
10
21
4
31
2
13
36
1
14.24
2.54


Age
1st
HA
68
61
67
66
89
22
NR
90
61
68
63
NR
65
65
72
78
55
74
35
73
43
41
62
62.25
3.65


PTA
52
27
40
27
52
88
65
38
63
48
32
52
47
35
50
37
47
38
58
22
58
38
15
44.67
3.35


Left Ear
54
CNT
32
48
50
0
8
80
16
28
70
76
58
76
80
56
42
48
34
74
18
88
82
50.09
5.97


Right Ean
76
86
58
70
70
18
6
66
0
36
34
0
60
58
84
64
46
12
58
100
78
84
98
54.86
6.40


Binaural
88
92
58
70
78
18
16
70
38
40
76
70
60
74
86
76
44
72
48
96
68
88
96
66.17
6.02


Binaural
+10
SNR
32
24
0
20
16
0
0
10
0
8
32
30
22
12
22
24
4
26
14
22
26
8
40
17.04
2.46


Age
Yrs HA Use
Age 1st HA
PTA
Word Recognition
Left
Right
Binaural
Binaural +10 SNR


NR
CNT


In years
Years of hearing Aid use
Age at first hearing aid
Pure tone averageof thresholds at 500, 1000 and 2000 Hz
Percent of correctly repeated words
Left ear only in quiet at most comfortable level
Right ear only in quiet at most comfortable level
Both ears in quiet at most comfortable level
Both ears in noise at most comfortable level with competing pink noise at
10 decibels hearing level (dB HL) below speech level
No response
Could not test









Table 3-4. Individual participant data for 6 participants in the comparison group: amplification
history and audiometrics.
Word Recognition


Age
1st
HA
57
71
NR
43
66
73
62.00
17.69


PTA Left Ear
35 32
48 48
62 72
33 88
38 18
63 82
46.50 56.67
8.48 4.88


Right Ean
60
74
98
32
28
82
62.33
5.47


Binaural
44
92
96
82
46
82
73.67
7.88


Binaural
+10
SNR
18
24
42
22
0
12
19.67
3.46


Age
Yrs HA Use
Age 1st HA
PTA
Word Recognition
Left
Right
Binaural
Binaural +10 SNR

NR
CNT


In years
Years of hearing Aid use
Age at first hearing aid
Pure tone averageof thresholds at 500, 1000 and 2000 Hz
Percent of correctly repeated words
Left ear only in quiet at most comfortable level
Right ear only in quiet at most comfortable level
Both ears in quiet at most comfortable level
Both ears in noise at most comfortable level with competing pink noise at
10 decibels hearing level (dB HL) below speech level
No response
Could not test


Participant
1
2
3
4
5
6
Mean
SE


Yrs
HA
Use
25
6
NR
10
5
3
9.80
3.29


Age
81
77
81
53
71
76
73.17
22.52











Frequency


1000


(Hz)
2000


4000


8000


Figure 3-1. Left ear pure tone air thresholds (dB HL) by frequency (Hz) for the 23 participants
in the FM group: means (+/- 2 standard errors). indicates no response at 4000 Hz
for 1 participant. ** indicates no response at 8000 Hz for 6 participants.













500


Frequency (Hz)

1000 2000


4000


8000


Figure 3-2. Right ear pure tone air thresholds (dB HL) by frequency (Hz) for the 23 participants
in the FM group: means (+/- 2 standard errors). indicates no response at 2000 Hz
for 1 participant. ** indicates no response at 4000 Hz for 1 participant. *** indicates
no response at 8000 Hz for 5 participants.


****


90

100

110

120











Frequency (Hz)

1000 2000


4000


8000


U-
10
20
30
40
5 50
S60-
70
80
90
100
110
120

Figure 3-3. Left ear pure tone air thresholds (dB HL) by frequency (Hz) for the 6 participants in
the comparison group: means (+/- 2 standard errors).













500


Frequency (Hz)

1000 2000


4000


8000


U I I
10
20 -
30
40

50
60
70
80
90 _
100
110
120


Figure 3-4. Right ear pure tone air thresholds (dB HL) by frequency (Hz) for the 6 participants
in the comparison group: means (+/- 2 standard errors). indicates no response no
response at 8000 Hz for 2 participants.











100
90



370
60
50 -
h 40-
0
30 -30

1 0 --------------------------


LE RE BIN BIN+10 SNR
Condition

Figure 3-5. Word recognition performance in percent correct scores on Northwestern University
word list #6 (NU-6) for the 23 participants in the FM group. LE) Left Ear in Quiet at
Most Comfortable Level (MCL). RE) Right Ear in Quiet at MCL; Binaural in Quiet
at MCL. BIN) Binaural at MCL. BIN +10 SNR) Binaural with Pink Masking Noise at
10 decibels below Binaural MCL for a +10 dB Signal-to-Noise Ratio.






















BIN


Condition
Figure 3-6. Word recognition performance in percent correct Scores on Northwestern University
word list #6 (NU-6) for the 6 participants in the comparison group. LE) Left Ear in
Quiet at Most Comfortable Level (MCL). RE) Right Ear in Quiet at MCL; Binaural
in Quiet at MCL. BIN) Binaural at MCL. BIN +10 SNR) Binaural with Pink Masking
Noise at 10 decibels below Binaural MCL for a +10 dB Signal-to-Noise Ratio.


- I
























-*-Soft (55dB)
-U- Average (70dB)
Loud (75dB)
MPO (85 dB)


250 500 750 1000 1500
Frequency (Hz)


2000 3000 4000 6000


Figure 3-7. Quality control measures for electroacoustic verification: average acoustic output (dB
SPL) by frequency (Hz) of earbuds at volume 3. The outputs correspond to 4 input
stimuli at the same frequency (Hz): soft speech weighted noise (55 dB SPL), average
speech weighted noise (70 dB SPL), loud speech weighted noise (75 dB SPL) and
Maximum Peak Ouput (MPO) narrowband signals. The averages represent 18
conditions: 3 Microvox Receivers (1, 2, 3); 3 Transducer Pairs (A, B, C); and 2 Ears
(Left and Right).


M 60


50


40
0

'3






















-*- Soft (55dB)
-U- Average (70dB)
Loud (75dB)
MPO (85 dB)


250 500 750 1000 1500
Frequency (Hz)


2000 3000 4000 6000


Figure 3-8. Quality control measures for electroacoustic verification: average acoustic output
(dB SPL) by frequency (Hz) of earbuds at volume 6. The outputs correspond to 4
input stimuli at the same frequency (Hz): soft speech weighted noise (55 dB SPL),
average speech weighted noise (70 dB SPL), loud speech weighted noise (75 dB SPL)
and Maximum Peak Ouput (MPO) narrowband signals. The averages represent 18
conditions: 3 Microvox Receivers (1, 2, 3); 3 Transducer Pairs (A, B, C); and 2 Ears
(Left and Right).















90

80


I 70

60
-*- Soft (55dB)
S5- Average (70dB)
50
Loud (75dB)
MPO (85 dB)
40

30

20


10


0
250 500 750 1000 1500 2000 3000 4000 6000
Frequency (Hz)

Figure 3-9. Quality control measures for electroacoustic verification: average acoustic output of
(dB SPL) by frequency (Hz) of headphones at volume 3. The outputs correspond to 4
input stimuli at the same frequency (Hz): soft speech weighted noise (55 dB SPL),
average speech weighted noise (70 dB SPL), loud speech weighted noise (75 dB SPL)
and Maximum Peak Ouput (MPO) narrowband signals. The averages represent 18
conditions: 3 Microvox Receivers (1, 2, 3); 3 Transducer Pairs (X, Y, Z); and 2 Ears
(Left and Right).
















90

80


S70


60
S-- Soft (55dB)
-U- Average (70dB)
50
Loud (75dB)
SMPO (85 dB)
S40

30

20


10

0
250 500 750 1000 1500 2000 3000 4000 6000
Frequency (Hz)

Figure 3-10. Quality control measures for electroacoustic verification, average acoustic output of
(dB SPL) by frequency (Hz) of headphones at volume 6. The outputs correspond to 4
input stimuli at the same frequency (Hz): soft speech weighted noise (55 dB SPL),
average speech weighted noise (70 dB SPL), loud speech weighted noise (75 dB SPL)
and Maximum Peak Ouput (MPO) narrowband signals. The averages represent 18
conditions: 3 Microvox receivers (1, 2, 3); 3 transducer Pairs (X, Y, Z); and 2 ears
(Left and Right).