EPA-600/1-77-038
June 1977
Environmental Health Effects Research Serie<
      BEHAVIORAL  AND  PHYSIOLOGICAL  CORRELATES
                    OF VARYING NOISE ENVIRONMENTS
                                    Office of Health anil Ecological Effects
                                      Office of Research and Development
                                     U.S. Environmental Protection Agency
                                             Washington, D.C. 20468

-------
                RESEARCH REPORTING SERIES

Research reports of the Office of Research and Development, U.S. Environmental
Protection Agency, have been grouped into nine series. These nine broad cate-
gories were established to facilitate further development and application of en-
vironmental technology.  Elimination of traditional grouping  was consciously
planned to foster technology transfer and a maximum interface in related fields.
The nine series are:

      1.  Environmental  Health Effects Research
     2.  Environmental  Protection Technology
     3.  Ecological Research
     4.  Environmental  Monitoring
     5.  Socioeconomic Environmental Studies
     6.  Scientific and Technical Assessment Reports (STAR)
     7.  Interagency Energy-Environment Research and Development
     8.  "Special" Reports
     9.  Miscellaneous Reports
This report has been assigned to the ENVIRONMENTAL HEALTH EFFECTS RE-
SEARCH series. This series describes projects and studies relating to the toler-
ances of man  for unhealthful substances or conditions. This work is generally
assessed from a medical viewpoint, including physiological or psychological
studies. In addition to toxicology and other medical specialities, study areas in-
clude biomedical instrumentation and health research techniques utilizing ani-
mals — but always with  intended application to human health measures.
This document is available to the public through the National Technical Informa-
tion Service, Springfield, Virginia 22161.

-------
                                              EPA-600/1-77-038
                                              June 1977
   BEHAVIORAL AND PHYSIOLOGICAL CORRELATES OF

           VARYING NOISE ENVIRONMENTS
                       by

                Lawrence F. Sharp
                 John F. Swiney
                Mickey R. Dansby
                Stephen C. Hyatt
                Dale E. Schimmel
Department of Behavioral Sciences and Leadership
United States Air Force Academy, Colorado  80840
            Contract No. IAG-D4-0537
                 Project Officer1

                 George R. Simon
             Health Effects Division
     Office of Health and Ecological Effects
              Washington, DC  20460
     OFFICE OF HEALTH AND ECOLOGICAL EFFECTS
       OFFICE OF RESEARCH AND DEVELOPMENT
      U.S. ENVIRONMENTAL PROTECTION AGENCY
              WASHINGTON, DC  20460

-------
                              DISCLAIMER
This report has been reviewed by the Office of Health and Ecological Effects,
U.S. Environmental Protection Agency, and approved for publication.  Approval
does not signify that the contents necessarily reflect the views and policies
of the U.S. Environmental Protection Agency, nor does mention of trade names
or commercial products constitute endorsement or recommendations for use.

-------
                                 ABSTRACT

Eighty male college juniors and seniors were dichotomized into either High
or Low Anxiety groups.  Each subject experienced a household noise profile
under a quiet (50 dBA), intermittent (84 dBA) and continuous (84 dBA)
noise condition, while performing either an easy or difficult pursuit
tracking task.  Heart rate, electromyographic potentials, and tracking
error responses were evaluated.  Results indicated significant (P<.01) main
effects for task difficulty and noise condition and significant (P<.01)
interaction effects for task difficulty, noise condition and anxiety level
(as measured by the IPAT Self Analysis Form) of subjects.  The significant
noise effect occurred for the difficult task condition during the second
tracking period (which includes transfer of training effects) indicating
that factors such as task difficulty, direction of task transfer effects,
duration of noise exposure as well as anxiety level of subjects appear to
be important variables affecting human psychomotor performance in noise
environments below 85 dBA.  These findings appear to be consistent with
previous research which suggests that task difficulty is the variable deter-
mining the direction of stress (noise) effects on psychomotor performance and
the nature of the interaction between stress and anxiety level.  The present
findings are therefore seen as supporting the concepts of the response
interference hypothesis and the inverted-U function between stress and
performance.

-------
                                CONTENTS

                                                             Page

Abstract                                                     in
List of Figures                                               vi
List of Tables                                              viii
Acknowledgements                                              ix
Sections
I      Conclusions                                            1
II     Recommendations                                        4
III    Introduction                                           5
IV     Method                                                 8
V      Results                                                20
VI     Discussion                                             48
VII    References                                             61
VIII   Glossary                                               65
IX     Appendices                                             67

-------
                             FIGURES
No
1    Sten Scores of Experimental Sample and National College     9
     Sampl e
2    1/3-Octave Band Spectral Analysis of Noise Profile          13
3    Schematic Diagram of EMG Signal Conditioner Circuitry       15
4    Low Anxiety Track Error - Data 1                            22
5    High Anxiety Track Error - Data 1                           23
6    High Anxiety Track Error - Data 2                           26
7    Low Anxiety Track Error - Data 2                            27
8    Easy Tracking Task - Data 1                                 35
9    Difficult Tracking Task - Data 1                            35
10   Easy Tracking Task - Data 2                                 36
11   Difficult Tracking Task - Data 2                            36
                                 VI

-------
                                 TABLES
No.
1    Experimental Design                                      10
2    Stroop Color Word Chart Arrangement                      17
3    Analysis of Variance for the Electromyographic Potential  20
     Dependent Variable - First Task
4    Analysis of Variance for the Electromyographic           21
     Potential Dependent Variable - Second Task
5    Analysis of Variance for Track Error - First Task        24
6    Track Error Cell Means and Standard Deviations -         24
     First Task
7    Track Error Row Means for the First Task - High Anxiety  25
     Group
8    Track Error Row Means for the First Task - Low Anxiety   28
     Group
9    Analysis of Variance for Track Error - Second Task       29
10   Track Error Cell Means and Standard Deviations -         30
     Second Task
11   Track Error Row Means for the Second Task - High Anxiety 31
     Group-Task Effect
12   Track Error Row Means for the Second Task - Low Anxiety  31
     Group-Task Effect
13   Track Error Column Means for Second Task - High Anxiety  32
     Group-Noise Effect
14   Track Error Column Means for Second Tracking Task - Low  33
     Anxiety Group-Noise Effect
15   Analysis of Variance for Stroop Color Word Responses     37
16   Stroop Color Word Response Times by Cell Mean            38
17   Analysis of Variance for Heart Rate Responses - Data 1 - 38
     First Task
                               vii

-------
                       TABLES  (continued)
No.                                                          Page
18   Analysis of Variance for Heart Rate Responses - Data 2-39
     Second Task
19   Heart Rate Mean Response - First Task                    40
20   Heart Rate Mean Response - Second Task                   41
21   Analysis of Variance for the Restricted Regression       42
     Model for Tracking Error - Data 1
22   Regression Summary Table for the Restricted Model for    43
     Tracking Error - Data 1
23   Analysis of Variance for the Full Regression Model for   43
     Tracking Error (Data 1)
24   Regression Summary Table for the Full Model for Tracking 43
     Error - Data 1
25   Analysis of Variance for the Restricted Regression Model 44
     for Stroop Response Time
26   Analysis of Variance for the Restricted Regression Model 45
     for EM6 (Data 1)
27   Regression Summary Table for the Restricted Regression   46
     Model for EMG (Data 2)
28   Analysis of Variance for the Full Regression Model for   46
     EMG - Data 2
29   Regression Summary Table for the Full Regression Model   46
     for EMG - Data 2
                              vm

-------
                            ACKNOWLEDGMENTS

This research was funded by the United States Environmental  Protection
Agency, Health Effects Division, under EPA-IAG-D4-0537 and EPA-IAG-D5-0537.

The views expressed in this paper represent those of the authors and do
not necessarily reflect the views of the United States Air Force or the
Department of Defense.

In conduct of this research the authors wish to acknowledge the adminis-
trative and technical assistance of Mrs. Berthe Giroux, Mrs. Kathy Rockefeller,
Dr. Karl Kryter, Valentin W. Tirman, Stephen Stitch, Bryon Russell and
Mark 0'Green.
                                     ix

-------
                               SECTION I
                              CONCLUSIONS

    The experimental results indicate that the relationship between physio-
logical and motor skill variables when noise is used in combination with
other variables is extremely complex.  In this experiment, there were no
significant physiological response differences in electromyographic potential
or heart rate changes as a function of either noise profile, anxiety con-
dition, or task difficulty.  Previous studies of physiological  response to
noise stress show similarly conflicting findings.  There appears to be an
overlap of anxiety and stress in the physiological area and in  the awareness
of pressure and tension with the result that the degree of familiarity of
the stimulus, and the tendency of individual differences to cancel each
other out must be taken into consideration before attempting to make signifi-
cant practical predictions concerning the effects of moderate noise levels
on human physiological response.
    Anticipation of noxious stimulations also appears to be an  important
factor in predicting physiological stress reactions to noxious  stimulation.
Research (Speilburger, 1972) shows that most of the autonomic stress reaction
may take place prior to the administration of the noxious stimulation.  With
time to appraise the situation, subjects are able to develop self-assuring
coping responses which can lead to lowered physiological activation during
the experimental period.  This factor may have contributed to the lack of
practical significance between baseline and experimental condition physio-
logical arousal levels in the present experiment.  In summary,  human
physiological response to moderate environmental noise is a complexly con-
trolled system and may not be a reliable indicator of arousal level and
stress at noise levels below 85 dBA.
    The average performance of each subject on the Stroop Chart reading task
improved as a result of the experimental conditions reflecting  the arousing
but not debilitating effect of the experiment.  Previous noise  research which
dealt with high (115 dBA) sound pressure levels have shown decrements in

-------
Stroop Chart reading times.  Apparently, 84 dBA noise levels are not
debilitative on this task.  Instead, the data reflect that merely being in
a controlled environment, or attending, independent of noise condition,
to a demanding task could have produced the improvement in reading times.
    In terms of psychomotor performance, task difficulty, anxiety level,
and noise condition all interacted to.determine tracking error with the
result that the poorest tracking performance occurred when high anxiety
subjects were required to switch from the easy to the difficult task during
exposure to intermittent noise.  Overall, the best tracking performance oc-
curred when high anxiety subjects performed the easy tracking task, regardless
of noise condition and task transfer direction.  Tracking performance on the
difficult task, however, improved significantly when low anxiety subjects
were exposed to noise and also when high anxiety subjects were not exposed
to noise.  Apparently, moderate intensity (84 dBA) household noise serves  as
a stressor for high anxiety subjects and as a facilitator for low anxiety  sub-
jects performing a difficult psychomotor task.
    The interaction between noise condition and anxiety level occurred only
during the second five minute tracking/noise exposure period (data 2) when
the task was difficult, which suggests that duration of stress (noise and
tracking) as well as task transfer effects are important variables in moderate
level (below 85 dBA) noise research.
    These findings appear to be consistent with previous research which sug-
gests that task difficulty is the variable determining the direction of stress
(noise) effects on psychomotor performance and the nature of the interaction
between stress and anxiety level.  The present findings are therefore seen as
supporting the concepts of the response interference hypothesis and the inverted-
U function between stress and performance.  Anxiety then, as a personality
variable, when predicting the effects of moderate noise on psychomotor per-
formance, should be evaluated as a probable determinant of moderate noise  ef-
fects on human behavior.
    In summary, moderate intensity, (84 dBA) household noise, appears to act as
a stressor for high anxiety subjects performing a difficult psychomotor task
and particularly for those who experience noise as:  1) primary overstimulation,

-------
i.e., feelings of being overwhelmed and bombarded with stimuli;  2)  or who
lack an appropriate course of action to resolve undirected arousal, anxiety
and frustration as a result of unavoidable noise exposure; 3)  a  violation
of expectancies; 4) or where noise serves to interrupt cognitive pro-
cesses resulting in the inability to carry out a cognitive plan  when
experiencing interruption or environmental disorganization due to noise ex-
posure.  Consequently, it may be that one of the primary effects of moderate
environmental noise is its interrupting, disorganizing quality,  which would
be particularly debilitating to those subjects who already experience sub-
stantial internal arousal as the result of elevated trait anxiety levels.
In conclusion, except for that segment of the population that could be
characterized as possessing low trait anxiety and who actually appear to
profit (at least temporarily) from moderate noise stimulation on difficult
psychomotor tasks, or for high and low anxiety subjects on easy psychomotor
tasks, by adding stimulation from the environment (noise) that violates or
precludes the development of expectancies (intermittent noise),  it can then
be expected that decrements in psychomotor performance will probably occur
on difficult psychomotor tasks, and as part of the "cost" to the person,
frustration can also be expected to occur with its special relevance for
maladaptive behavior and the normal stresses of everyday living that now
appear to plague our highly industrialized, urban society.

-------
                               SECTION II
                            RECOMMENDATIONS

    The results of this study suggest that future research investigating
the physiological effects of stress due to broadband noise below 85 dBA
should consider employing measures other than average heart rate and
electromyographic potentials from which to predict human physiological
response.  Additionally, attention should be focused on the degree to
which subjects have achieved physiological baseline prior to rioise
exposure.
    The results also suggest that the internal arousal level of subjects
(anxiety) is an important variable although a more heterogeneous subject
population may have resulted in an even larger effect due to the anxiety
variable.  In this regard, factors such as task difficulty, direction
of task transfer effects and duration of noise exposure appear to be
important variables requiring careful consideration when predicting the
interaction of previous arousal level (anxiety) and stress (moderate
noise) on psychomotor performance.
    Due to subject unavailability, sex differences were not evaluated in
the present study.  Future research should attempt to determine if such
differences exist, and the applicability of the present findings should
be experimentally expanded to include a cross section of the general popu-
lation performing psychomotor tasks that are representative of the work
activities experienced by that population.
    Finally, in practical terms, even though the independent variables in
the present study resulted in significant mean differences in psychomotor
performance, only a small proportion (19%) of the total variance in
tracking performance has been accounted for.  This would indicate that
individual differences as well as other factors have not been fully ex-
plored and, therefore, substantially limit the applicability of these
findings to similar subject populations.

-------
                              SECTION III
                              INTRODUCTION

The 1973 International Congress on noise as a public health problem,
especially the psychological consequences, began with a rather somber
summary.  Gulian (1973) noted that most noise research since the 1950s is
generally controversial and no firm conclusions can be drawn.  Gulian
attributes the disparate research results to the extraordinary complexity
of factors which, intenvene between noise as the independent variable and
the dependent measures.  Even a cursory examination of the literature
during the past five years dramatically substantiates this complexity and
the paucity of variables involved.  For example, some researchers have
found no-ise induced: facilitation of learning and cognitive performance
(Fechter 1972) while others observe a decrement (Renshaw, 1973).  Others
have found sex (Kumar, 1969 and Elliott, 1971) age (Mathur, 1972) and
social class (Anderson, 1973) differences in response to noise.  To further
complicate interpretation of the findings, Harcum and Mont-i (1973) found
that subjects will "cooperate" with the experimenter on noise disturbance
ratings unless this factor is controlled.  Although much research seems
to confirm the absence of the main effects for noise alone, some research
is beginning to emerge which shows the interactive nature of noise as an
independent variable.  For example, Harris and Schoenberger (1970) demon-
stratedvthat the detrimental effect of noise is additive to that of
vibration when both are presented simultaneously.  When combined with noise,
the additional stressor of shock (using rodents) (Campbell, 1968) and
neomycin (Jauhiainen, Kohonen' and Jauhiainen, 1972) produce a synergistic
effect.  Indicative of the^problems associated with these research findings
are Grether's (1972).research which, failed.to demonstrate combined effects
of heat, noise and vibration and Kryter's (1970) caution when using rodents
and rabbits as subjects in. noise research.;

The data on threshold shift (temporary or permanent) and hearing loss is
certainly well founded, especially for extreme noise environments, and

-------
with appropiate protective apparatus, damage can be attenuated or avoided.
There seems to be a lack of appreciation however, for the effect of
moderate noise environments and for the effects these environments have
on non-auditory or non-physiological responses.  Recent research by
Bull (1973), Edsell (1973) and Glass and Singer (1973) provide good evi-
dence that even "low" (84 dB) noise environments result in important
changes in socially relevant behavior; e.g., tolerance for ambiguity
decreases in a noise environment (Bull, 1973); perception of others
assumes negative dimensions (Edsell, 1973); and frustration tolerance
decreases (Glass and Singer, 1973).  These effects obviously represent
the psychological cost the organism pays for exposure to unavoidable
environmental noise.  In fact, behavioral responses to noise and behavior-
al differences between subjects may be among the most important indicants
of noise effects and a major source of variation in the various dependent
measures assessed.

Some recent studies of noise and personality have focused on introversion
or extroversion as contributors to psychomotor performance differences
under noise conditions.   These studies have been generated by Eysenck's
theory of personality and cortical  arousal.  In general, extroverts were
found to display greater decrements in psychomotor performance while
experiencing noise stimulation than were introverts.  Di Scipio (1971)
showed that white noise facilitates psychomotor response for an optimal
period of time, after which decrements were observed.  This effect was
heightened for extroverts.  Even though extroverts are more prone to
noise distraction, Elliott (1971) showed that they will tolerate greater
intensities of white noise than will introverts. Results of other studies
on noise and social behavior are diverse.  Edsell (1973), indicates that
subjects in a game situation perceived other players as more disagreeable,
disorganized, and threatening under noise as opposed to no-noise conditions,
Jansen & Hoffman (1971)  demonstrated that increasing the loudness of a
noise stimulus augmented subjective annoyance, with neurotic personality
tendencies contributing to this effect.  Angrier speakers were found to
use more high frequency elements in their speech (Mason 1969).

-------
Stephens (1970) showed that test anxiety scores correlated positively with
the slope of a loudness judgement function.

The above considerations and the data to be presented in this paper readily
attest to the fact that the main effects of noise, especially moderate
levels (60-90 dB) are elusive, depend to some extent on the psychological
structure of the recipient, and, potentially, can be confounded with a
seemingly endless array of other factors.

The research reported in the paper was conceived and conducted to:  (1)
specifically assess a noise profile to which a large proportion of both
urban and suburban dwellers are exposed on a daily basis;  (2) examine
these effects on a relatively homogenous population with respect to sex,
age, physical fitness, intellectual ability, psychological structure, and
environmental stress; and (3) provide more adequate control in terms of
research design, of individual differences which could potentially contri-
bute to between group differences in noise responses.

-------
                              SECTION IV
                                METHOD

SUBJECTS

Eighty male junior and senior Air Force Academy cadets provided data
for this study.  Subjects were volunteers, solicited from upper
division Behavioral Science and Life Science classes.  No inducements
were offered.  Eighty subjects initially agreed to participate and
completed "Informed Consent" certificates in accordance with HEW standards.

Participating subjects were administered an anxiety scale to assess their
relative levels of state-trait anxiety.  On the basis of this measure,
subjects were dichotomized as above (High Anxiety) or below (Low Anxiety)
the group median score.  Each subject was then randomly assigned to one
of the four task sequence groups, which resulted initially in eight cells
of ten subjects each. .Figure 1 displays the frequency distribution of
anxiety measure scores of the experimental sample versus national, male
                                                             2
college norms.  These distributions differed significantly (x.   =25.81,
df_ = 9,-£ < .005) and the group mean,-STEN scores were significantly
different (t_ = -10.88, df = 40, £ < .005).

DESIGN

A 2 x 3 x 4 factoral design was employed.  Table 1 shows the variables and
levels involved.

Forty subjects served under each of the two anxiety conditions.  Within
each anxiety condition, ten subjects served under each of the four task
sequence conditions.  Each cell of ten subjects experienced the three
noise conditions in a counterbalanced, repeated measures sequence.
                                    8

-------
  25
  23
  21
1U
   19
2  17
g
UJ
UJ
0.
15
13
11
 9
 7
 5
 3
 1
                                                     NATIONAL NORMS
                                                     EXPERIMENTAL SAMPLE
                           456
                            STENSCORES
                                                         10
    FIGURE!
             STEN SCORES OF EXPERIMENTAL SAMPLE AND NATIONAL
             COLLEGE SAMPLE.

-------
             TABLE-  1.
                         EXPERIMENTAL DESIGN


EASY TO
EASY11
EASY TO
DIFFICULT
DIFFICULT
TO EASY
DIFFICULT
TO
DIFFICULT
HI ANXIETY
Qa
SI
THRU
S10
S11
THRU
S20
S21
THRU
S30
S31
THRU
S40
IN








CN
•^
^*
>_. .
•^


>_
>_
••
LO ANXIETY
Q
S41
THRU
S50
S51
THRU
S60
S61
THRU
S70
S71
THRU
S80
IN








CN


>^

>_



aO  =  QUIET CONDITION (50 dBA MASK)
IN  =  INTERMITTENT PROFILE
CN
CONTINUOUS PROFILE
bREFERS TO DIFFICULTY LEVEL OF TWO, SUCCESSIVE, 5 MINUTE TRACKING TASKS, ACCOMPLISHED UNDER
EACH NOISE CONDITION.
                                     10

-------
APPARATUS

The experimental sessions were conducted in the Behavioral  Sciences
Laboratory at the United States Air Force Academy.  A controlled acoustic
environment was provided through the use of ventilated, Industrial
Acoustics (AIC) audiometric examination booths.  Each chamber was
equipped with a Hewlett Packard 1205A dual trace oscilloscope which
displayed a randomly moving, horizontal "target" line as well as the
subject's "controlled" line.  A 60 Hz sign-wave was superimposed on the
controlled line to aid in subject differentiation between the two lines.
Total system inputs and outputs are shown by block diagram in appendicies
1 and 2 respectively.
       T
A Weston 1242 digital multimeter was located on top of each oscilloscope,
adjusted to display (-10 to +10) volts, and represented real-time in-
tegrated tracking error, which provided immediate feedback to subjects
during the training portion of each experimental session.

Subjects sat facing the oscilloscope at an approximate viewing distance
of 33 cm.  The oscilloscope was placed at eye level to minimize parallex
distortion.  A Measurement Systems Model 542, 2 axis, gimballed joystick
was installed at the end of the subject's right arm rest.  Coil springs,
set at .45 kg maximum deflection force, were used to return the handle to
center.  Maximum possible stick deflection was 28° from center in each
direction.

The experimental variable, noise, was introduced via a high-fidelity
(AR2-ax) bookshelf loudspeaker located immediately above and facing the
seated subject.  Fifty dBA of background acoustical masking was provided
continuously through a "Sound Shield" random noise generator placed
adjacent to the loudspeaker.  A 2-way intercom station was installed in
each chamber as well as the necessary EKG and EMG leads and electrodes.

The audio input was generated by combining signals from a Hewlett-Packard
(HP) 8057A Precision Noise Generator set for "pink" noise, and a HP 3722A

                                   11

-------
Noise Generator with a selective sequence length of N = 15, clock period
of 100 ms and gaussian noise bandwidth of .5 Hz with the variable output
set for "binary".  The generated signal was magnetically recorded on a
Crown Model 824SX 4-track tape recorder, the output of which was then
1/3-octave band shaped in real-time by a Bruel  and Kjaer (B&K) Model
125 1/3-octave Graphic Frequency Response Equalizer and amplified by a
Crown IC-150 preamplifier and DC-300A laboratory amplifier.  The audio
profile chosen represented 84 dBA of typical suburban household noises
and was generated by magnetically recording a central heating system,
television program, and a canister type vacuum cleaner at operator ear
level in a carpeted and draped living room.  A calibrated (+_ 1 dB 20 -
30 KHz) Crown 824SX tape recorder was used with input provided by a B&K
2619 Microphone Preamplifier, B&K 4133 1/2" Condenser Microphone with a
UA 0386 Nose Cone, a B&K 2804 Power Supply and calibrated for absolute
Sound Pressure Level (SPL) by a B&K 4220 Pistonphone.  Overall SPL was
measured by a General Radio (GR) 1558-BP Octave Band Noise Analyzer for
both dBA and dBC values.  Calibration was performed prior to measurements
with a GR Type 1562 Sound Level Calibrator.  The resultant magnetic tape
was then analyzed using a HP 8064A Real-Time Audio Spectrum Analyzer, the
output of which was connected to a HP 7004B X-Y Recorder and automatically
plotted on HP 08064-9010 dB scaled graph paper.  Primary spectral energy
content centered between 200 and 300 Hz with a peak at 400 Hz (see Figure
2).

The final audio profile was verified by real-time 1/3 octave band analysis
in the sound chamber at subject ear level.  Microphones, equalizers, and
amplifiers, etc. were the same items used in the original recording pro-
cess.  Final A-weighted SPL settings were done with a test subject in
place and the chamber door closed.

The random tracking signal input was generated by the HP 3722A low fre-
quency random noise generator with a selective sequence length of N = 4,095,
clock period of 333 ms and gaussian noise band width of .15 Hz.
                                    12

-------
  dB
-10
-15
-20
-25
-30
-35
                                                                LEGEND

                                                                MICROPHONE:   B&K4133
                                                                RECORDER:    CROWN SX-824
                                                                CENTRAL HEATING-VACUUM-T.V.
                                                                 (78dBA-80 dBC)
      50  63 80 100  125 160  200250 315 400 500 800 1.K 1.25K1.6K 2.K 2.5K3.15K4K 5K 6.3K 8K

                                     FREQUENCY-Hz

    FIGURE 2.  1/3-OCTAVE BAND SPECTRAL ANALYSIS OF NOISE PROFILE
10K

-------
Subject tracking output and tracking signal input were fed into an EIA
TR-20 analog computer which was pre-programmed to present zero order
(position control) or first order (rate or velocity) control feedback
to the subject and provide an integrated error voltage to drive the
digital multimeter display.  The analog computer also provided a real-
time absolute error voltage output which was magnetically recorded on
an Ampex FR-1300 14 channel FM instrumentation recorder.

The EMG (Electromyographic) potentials recorded off the frontal is muscle
group, were filtered, rectified and magnetically recorded on the FM
recorder.  The rectifying/filtering circuit was unique and not readily
available in the literature.  Figure 3 shows the circuit used.

The electrocardiograph (EKG) signal  was processed through a Gould/Brush
Model 4307 Biomedical-Tachometer Coupler and magnetically recorded.

A Datametrics SP-425 time code generator was used to differentiate the
various experimental conditions and to generate a time track on the FM
recorder.  The subsequent analog data were processed and digitally analyzed.
                                    14

-------
                              RECTIFIER
                              (FULL WAVE)
RC FILTER NETWORK
              IN
OUTPUT
FROM
UNIVERSAL
PREAMPLIFIER
                                                                    OUT
                                                                      (D.C. COMPONENT)
FIGURE 3.   SCHEMATIC DIAGRAM OF EMG SIGNAL CONDITIONER CIRCUITRY.

-------
PROCEDURES

Prior to each day's subject runs, complete minimum and maximum calibration
voltages were magnetically recorded.   A cardiac simulator was used to
calibrate heart rate.

After reading and signing an informed consent form, each subject was
administered the Institute for Personality Ability Testing "Anxiety
Scale Questionnaire", (self-analysis  form) a brief, non-stressful,
clinically validated questionnaire for appraising free anxiety level.

Results of the test were used to divide the subject pool into statisti-
cally significantly different high and low anxiety groups.

Upon entering the experimental room,  eye dominance was first assessed,
then each subject was randomly administered one of two versions of a
standard Stroop Chart "Pretest".  Table 2 shows each version of the
Stroop Color Word chart.  The word in parentheses indicates the color of
the associated word.

Time required to successfully complete the chart was then recorded. If
the subject made an error, he was instructed to correct the error and
continue as quickly as possible.

Normally, two subjects were run simultaneously since two instrumented
sound chambers were available.  Subjects were instructed to remove their
upper body clothing.  EKG and EMG electrodes were attached and then the
subject seated in the chamber with feet flat on the floor, arm on arm rest,
hand on tracking handle, facing the oscilloscope.  Subjects were told to
relax, and cautioned against extraneous movements of body, head, jaws and
random eye movements.  Use of the 2-way intercom was explained (it did
not require subject manipulation), duration of the experiment was given,
and subjects informed that all further instructions would be issued over
the speaker located overhead, and that any subsequent questions would be
                                   16

-------
              Table 2.   STROOP  COLOR WORD  CHART ARRANGEMENT9
Stroop A
1 -Yellow
(orange)
3-Green
(blue)
6-Orange
(yellow)
9-Yellow
(green)
11 -Red)
(yellow)
14-Yellow
(orange)
16-Green
(blue)
18-Green
(yellow)
21 -Green
(orange)
2-Black
(green)
4-Blue
(red)
7-Red
(green)
10-Black
(orange)
12 -Green
(blue)
15-Blue
(red)
17-Orange
(blue)
19-Red
(blue)
22-Blue
(red)

5-Blue
(orange)
8-Blue
(orange)

13-Blue
(red)


20-Red
(green)

Stroop B
1 -Green
(blue)
4-Blue
(red)
6-Red
(blue)
9-Green
(orange)
12-Orange
(blue)
14-Black
(orange)
16-Blue
(orange)
19-Orange
(yel 1 ow)
21 -Green
(blue)
2-Red
(green)
5-Yellow
(orange)
7-Black
(green)
3-Blue
(yellow)

8-Blue
(red)
10-Red 11 -Blue
(yellow) (red)
1 3-Green
(blue)
1 5-Yellow
(green)
1 7-Red
(green)
20-Blue
(red)
22-Yellow
(orange)


18-Green
(yellow)


alt is important to note that both charts were identical  except for  sequence.
 The same number of words appeared on each chart, and they appeared  in  the
 same colors, but in different orders.  For example, the  red-colored word
 BLUE appeared four times on each chart:   On chart A it  was  the 4th, 13th,
 15th and 22nd word presented, and on Chart B the 4th, 8th, llth and 20th
 word presented.         '
                                    17

-------
answered via the intercom.

The chamber door was then closed, and after 2-3 minutes,  one minute of
physiological data was magnetically recorded (Baseline 1) under the quiet
(50 dBA mask) condition.  If subjects were scheduled to be exposed to
either the continuous or intermittent noise condition, then a second one
minute (Baseline 2) of physiological data was recorded with the appropri-
ate noise input.

Two minutes of taped instructions immediately followed.  Instructions
included the use of the tracking handle, oscilloscope, digital  multimeter
and the tracking task was explained in detail.   A one minute period of
questions and answers followed.

Subjects were then exposed to two, 5 minute experimental  sessions with a
one minute rest period between sessions.  These sessions, Data 1  and Data 2
respectively, consisted of exposing the subject to one of three noise condi-
tions; quiet (50 dBA generated by the Sound Shield white  noise generator),
(2) continuous noise (84 dBA), or (3) intermittent noise  (84 dBA) of an
identical audio spectrum but randomly cycled, both on and off between one
and nine seconds with a total "on" duration of 50% of the five minute
experimental session.  The identical noise condition used for Data 1 was
repeated for Data 2.  Each subject was scheduled on a daily basis for three
separate experimental sessions,  insuring exposure to all  three noise
conditions and minimizing noise adaptation and fatigue effects.

During each experimental session (Data 1 and Data 2), subjects were re-
quired to perform one of two tracking tasks.  The task was labeled "easy"
(position feedback) or "difficult" (rate or velocity feedback), and the
serial presentation of the task was counter balanced (Easy, Easy;  Easy-
Difficult;  Difficult-Easy;  and Difficult-Difficult) and remained un-
changed during the three experimental sessions for a given subject.

Subjects practiced the tracking task for four minutes with the integrate
error feedback display on.  During the fifth minute, the error display was
                                    18

-------
remotely turned off, and the FM recorder turned on generating the Data 1
and Data 2 information file.  Heart rate, EMS level, and absolute tracking
error voltages were recorded at this time.  The audio signal  and time
code generator signal were also recorded.

At the completion of Data 2, all recording equipment was turned off, and
each subject was immediately administered the Stroop posttest.

To alleviate potential anxiety associated with being exposed to noise and
instrumentated with EKG and EMG electrodes, recorded popular music was
played prior to and immediately after each experimental session.
                                   19

-------
                               SECTION V
                                RESULTS
ANALYSIS STRATEGY
Multiple analysis of variance for a design involving two between and one,
within subject factors was used to assess the significance of the main and
interaction effects of the independent variables.  This procedure is de-
scribed in Myers (1972) and provides for analysis of a repeated measure-
ments factor.  In this research, each subject served under each of the
three noise conditions, hence repeated measurements were made on each sub-
ject, for all dependent variables under a quiet, intermittent, and continu-
ous noise environment.

ANALYSIS OF ELECTROMY06RAPHIC CHANGES
For either Data 1 (First Tracking Task) or Data 2 (Second Tracking Task)
none of the independent variables produced significant changes in general
muscular tension as evidence by changes in electromyographic potentials
taken from the Frontal is Group.  Table 3 shows the analysis of variance
for these data for the first tracking task (Data 1).

 Table 3.  ANALYSIS OF VARIANCE FOR THE ELECTROMYOGRAPHIC POTENTIAL
           DEPENDENT VARIABLE - FIRST TASK
Source of variation
Anxiety
Task difficulty
Anx x tsk dif
Error
Noise condition
Anx x noise cond
Task dif x noise cond
Anx x tsk dif x noise
condition
Total
DF
1
3
3
72
2
2
6
144
239
SS
18.2803
33.9461
44.6793
564.5801
5.4246
39.6191
46.5791
1132.7677
1940.0469
MS
18.2803
11.3154
14.8931
7.6200
2.7123
19.8096
7.7632
7.8664

F
2.3989 N.S.
1.4894 N.S.
1.9544 N.S.

.3447 N.S.
2.5182 N.S.
.9868 N.S.


                                    20

-------
Table 4 shows the analysis of variance for EMG for Data 2 (Second Tracking
Task).
Table 4.  ANALYSIS OF VARIANCE FOR THE ELECTROMYOGRAPHIC POTENTIAL
          DEPENDENT VARIABLE - SECOND TASK
Source of variation
Anxiety
Task difficulty
Anx x task dif
Error
Noise condition
Anx x noise cond
Tsk dif x noise cond
Anx x tsk dif x noise
condition
Error
Total
DF
1
3
3
72
2
2
6
6
144
239
SS
4.9713
39.9109
18.6738
426.0356
14.9123
24.8510
69.3777
54.8425
878.6613
1526.3565
MS
4.9713
13.3036
6.2246
5.9171
7.4562
12.4255
11.5629
9.1404
6.1018

F
.8401 N.S.
2.2483 N.S.
1.0519 N.S.

1.2219 N.S.
2.0363 N.S.
1.8949 N.S.
1.4979 N.S.


ANALYSIS OF TRACKING ERROR RESPONSES

For both the first (Data 1) and second (Data 2) tracking tasks, task diffi-
culty significantly influenced .tracking error performance.  Table 5 shows
the analysis of variance for the first tracking task.

Figures 4 and 5, show, for both anxiety groups, that tracking error
responses approximately double (10% to 20%) when the task was the more
difficult rate or velocity tracking.

-------
ro
ro
        100
         90
         80
      Z 70
      CO
      <
         60
cc
O
CC
cc
UJ


o  50
      uj 40

      D
         30
         20
         10
                                                                Q= QUIET

                                                                I = INTERMITTENT

                                                                C = CONTINUOUS
                               EE
                                            ED
DE
DD
                                         TASK TRANSFER GROUPS

            FIGURE 4: LOW ANXIETY TRACK ERROR - DATA 1

-------
ro
CO
        100
DC

g
DC
UJ


O
<
QC
         80
         70
         60
50
         40
      O 30
      CO
      CO
         20
         10
                             QIC
                              EE
                                         ED
                                         TASK TRANSFER GROUPS

           FIGURE 5: HIGH ANXIETY TRACK ERROR - DATA 1
                                                            Q= QUIET
                                                            I = INTERMITTENT
                                                            C = CONTINUOUS
                                                               DE
DD

-------
Table 5.  ANALYSIS OF VARIANCE FOR TRACK ERROR - FIRST TASK
Source of variation
Anxiety
Task difficulty
Anx x task dif
Error
Noise condition
Anx x noise cond
Tsk dif x noise cond
Anx x tsk dif x noise
condition
Error
Total
Table 6 shows the cell
tracking error.
Table 6. TRACK ERROR
DF
1
3
3
72
2
2
6
6
144
239
means and
CELL MEANS
SS
3.9106
5349.6581 1
295.6206
3443.9221
94.9963
3.3320
82.4734
730.5761
5158.1347
15469.1981
MS
3.9106
783.2194
98.5402
47.8322
47.4982
1.6660
13.7456
121.7627
35.8203

standard deviations for
AND STANDARD
DEVIATIONS
F
.0817 N.S.
37.2807 p <.01
2.0601 N.S.

1.3260 N.S.
.0465 N.S.
.3837 N.S.
3.3992 p <.01


the first task
- FIRST TASK
                         Hi anxiety
Q
EE
ED
DE
DD
X"
X 12.65
s.d.10.00
X" 6.69
s.d. 4.97
123.04
s.d. 15. 47
X" 20.34
s.d. 15. 56
15.68
IN
In. 76
s.d. 9.40
X" 7.65
s.d. 6.10
121.16
s.d. 15. 08
I 24.25
s.d. 17. 61
16.18
CN
X" 12.93
s.d. 9.96
X" 9.47
s.d. 6.02
X" 33.29
s.d. 13. 60
X" 25.08
s.d. 17. 48
20.19
X
12.45
7.94
25.83
23.19

                         Lo anxiety
EE
ED
DE
DD
X"
X 12.01
s.d. 9.38
X" 10.25
s.d. 8.04
120.12
s.d. 15. 01
118.45
s.d. 11. 68
15.21
X 12.25
s.d. 9.78
X" 10.95
s.d. 9.05
124.73
s.d. 17. 62
X" 17.65
s.d. 11. 94
16.40
X" 22.76
s.d. 16. 35
JTlO.31
s.d. 7.77
124.64
s.d. 16. 51
I 15.34
s.d. 11. 34
18.26
24
15.67
10.50
23.16
17.15


-------
Neuman-Kuhls analysis of row means for both anxiety groups showed that,
within each anxiety group, the difficult tracking task resulted in signifi-
cantly higher error scores than the easy tracking task.  Table 7 shows the
Neuman-Kuhls analysis for the High Anxiety Group.
Table 7.
TRACK ERROR ROW MEANS FOR THE FIRST TASK - HIGH ANXIETY GROUP
                    Noise Condition
Task
E
E
D
D
Q
12.65
6.69
~ 23.04
20.34
IN
11.76
7.65
21.16 ,
24.15
CN
12.93
9.47
33.29
25.08
7
12.45
7.94
25.83
23.19

T2
Tl
T4
T3
Ti
           4.51(2.10)*
15.28(2.49)*
10.77(2.10)*
17.89(2.73)*
13.38(2.49)*
 2.61(2.10)*
  q2 = 2.10;  q3 = 2.49;  q4 = 2.73,  £ <.05
                          /
Table 8 shows the row means and Neuman-Kuhls analysis for the low anxiety
group.

In the high anxiety group, both groups which performed the easy task first
had significantly smaller track errors than either of the difficult task
groups.  Subject differences are apparently operating however, in that the
two groups performing the easy task differed significantly.  Additionally,
the two difficult groups differed significantly from each other.  Similar
                                    25

-------
ro
       100


        90


        80


     2 70
     oc
     o
     £60
     LU

     Q 50
     £  40
     D
     I
        30
        20
        1.0
                            QIC
                              EE
ED
                                         TASK TRANSFER GROUPS
           FIGURE 6: HIGH ANXIETY TRACK ERROR - DATA 2
                                                                       Q= QUIET
                                                                       I = INTERMITTENT
                                                                       C = CONTINUOUS
DE
DD

-------
ro
       100  _
        90
        80
      Z 70

      EC
      O
      £60
< 50
DC

LU 40

D

§30
CQ


  20



  10
                              EE
                                            ED
                                        TASK TRANSFER GROUPS

            FIGURE 7: LOW ANXIETY TRACK ERROR - DATA 2
                                                              Q= QUIET
                                                              I = INTERMITTENT
                                                              C = CONTINUOUS
DE
DD

-------
results hold for the low anxiety groups, with the exception that one of the
easy task groups (T«) did not have significantly poorer performance than
one of the difficult task groups (T3).
  Table 8.  TRACK ERROR ROW MEANS FOR THE FIRST TASK- LOW ANXIETY GROUP
                                Noise Condition
Task
E
E
0
D
Q
12.01
10.25
20.12
18.45
IN
12.25
10.95
24.73
17.65
CN
22.76
10.31
24.64
15.34
X
15.67
10.50
23.16
17.15

T2
Tl
T4
T3
                                                            '4
                      5.17(2.10)*
6.65(2.49)*
   1.48(2.10)
12.66(2.73)*
7.49(2.49)*
6.01(2.10)*
 q2 = 2.10;  q3 = 2.49;  q,= 2.73,  £ <.05
Table 9 shows the analysis of variance for the second tracking task.
(Data 2).  These data now include the effects of transfer of training.

Again, significant main effects are present for task difficulty.   Addition-
ally, these data now show significant main effects for noise condition.

Figures 6 and 7 show, that for both high and low anxiety groups,  the  second
task (Data 2) tracking error responses increased when the transfer was  from
an easy to a difficult tracking task.
                                    28

-------
      Table 9.   ANALYSIS OF VARIANCE FOR TRACK ERROR - SECOND  TASK
Source of Variation
Anxiety
Task difficulty
Anx x task dif
Error
Noise cond
Anx x noise cond
Task dif x noise cond
Anx x task dif x noise
condition
Error
Total
DF
1
3
3
72
2
2
6
6
144
239
SS
158.9432
7806.7954
261.7668
7035.2170
244.7820
56.8519
507.0714
194.5361
3186.5429
19222.5994
MS
158.9432
2602.2651
87.2556
97.7113
122.3910
28.4259
84.5129
32.4227
22.1287

F
1.6266 N.S.
26.6321 p_<.01
.8929 N.S.

5.5308 p_<.01
1.2845 N.S.
3.8191 p_ <.01
1.4651 N.S.


The Data 2, analysis of variance indicated significant main effects for
noise condition on tracking error, as well as for task difficulty.

Table 10 shows the cell means  and standard deviations for the second task
(Data 2) tracking error responses.

Neuman-Kuhls analysis of row means for both anxiety groups showed that
within each anxiety group, the difficult tracking task resulted in signifi-
cantly higher error scores than the easy tracking task.  Table 11 shows the
row means and Neuman Kuhls analysis for the high anxiety group.

Table 12 shows the row means and Neuman-Kuhls analysis for the low anxiety
group (Data 2).

In the high anxiety group (second task), the difficult task (T4) preceded by
a difficult task resulted in significantly higher track error scores than
                                    29

-------
either the difficult task (T~) preceded by an easy task or an easy task (T-j)
preceded by an easy task.  This finding also applied for the difficult task
(T2) followed by an easy task.  Regardless of prior (first) task, as long as
the second task was difficult, error scores were significantly higher than
when the second task was easy.  For the low anxiety group, the difficult
task (T.) preceded by an easy task resulted in significantly higher track
error scores than either the difficult task (To) preceded by a difficult
task or the easy task (T2) preceded by an easy task or the easy task following
a difficult task (T^).

Table 10.  TRACK ERROR CELL MEANS AND STANDARD DEVIATIONS - SECOND TASK
                              Hi anxiety
Q
EE
ED
DE
DD
X"
X 9.99
s.d. 8.68
115.43
s.d. 12. 07
X" 11.43
s.d. 9.36
JT22.12
s.d. 16. 52
14.74
IN
X" 9.78
s.d. 8.56
140.29
s.d. 8.92
X" 12.96
s.d. 9.45
X" 27.54
s.d. 17. 18
22.64
CN
X" 11.06
s.d. 9.39
JM3.27
s.d. 10. 24
X" 12.44
s.d. 8.06
JT35.97
s.d. 14. 48
18.19
X
10.28
23.00
12.28
28.54

                              Lo anxiety
EE
ED
DE
DD
X"
X 9.99
s.d. 8.57
JT37.79
s.d. 12. 42
X" 11.59
s.d. 9.83
X" 15.59
s.d. 11. 54
18.74
X 11.11
s.d. 9.24
X"27.74
s.d. 17. 82
X" 16.86
s.d. 12. 29
X" 14.49
s.d. 11. 12
17.55
X 20.25
s.d. 14. 06
JT21.75
s.d. 15. 71
JM1.20
s.d. 9.16
I 18.78
s.d. 11. 89
18.00
13.78
29.09
13.22
16.29

                                     30

-------
Table 11.  TRACK ERROR ROW MEANS FOR THE SECOND TASK -  HIGH ANXIETY GROUP-
	TASK EFFECT	

                             Noise condition
Task
E
D
E
D
Q
9.99
15.43
11.43
22.16
IN
9.78
40.29
12.96
27.54
CN
11.06
13.27
12.44
35.97
X
10.28
23.00
12.28
28.54

Tl
T3
T2
T4
Tl
T2
T3
T4
Table 12.


q2 = 5.11;
TRACK ERROR
TASK EFFECT
2.0(5.11)

q3 = 6.05; q4 =
ROW MEANS FOR THE
12.72(6.05)*
10.72(5.11)*

6.62 £ <.05
SECOND TASK -
18.26(6.62)*
16.26(6.05)*
5.54(5.11)*
LOW ANXIETY GROUP-
                             Noise Condition
Task
E
D
E
D
Q
9.99
37.79
11.59
15.59
IN
11.11
27.74
16.86
14.49
CN
20.25
21.75
11.20
18.78
X
13.78
29.09
13.22
16.29

T2
T4
Tl
T3
'1



'2 '3
56(5 11 } ? f)7(fi 05)
	 o c-ifti TH

'4
15 87(6 12)*
15 31(6 05)*
12.80(5.11)*
               ~ 5.11; q~ = 6.05; q. = 6.62  £<
.05
                                    31

-------
Tables 13 and 14 show the noise group cell  and column means for each anxiety
group, along with the Neuman-Kuhls analysis.
Table 13.  TRACK ERROR COLUMN MEANS FOR SECOND TASK - HIGH ANXIETY GROUP -
           NOISE EFFECT

Task
E
D
E
D
X"

Tl
T2
T3
q2 = 2.10;
Noise
Q
9.99
15.43
11.43
22.16
14.74
Tl
Tl


q3 = 2.53, £<
condition
IN
9.78
40.29
12.96
27.54
22.64
T3
T2
3.45(2.10)*

.05

CN
11.06
13.27
12.44
35.97
18.19
T2
T3
7.90(2.53)*
4.45(2.10)*


Neuman-Kuhls analysis of column means revealed that only the high anxiety
group (second task) displayed significant decrements in tracking  accuracy
as a result of noise condition.  In the high anxiety group,  the IN condition
    *
produced the greatest decrement, followed by the CN condition.   In the
low anxiety group, the Q condition produced the largest mean decrement in
tracking error, however, this result was clearly not significant.
                                    32

-------
Table 14.  TRACK ERROR COLUMN MEANS FOR SECOND TRACKING TASK   LOW ANXIETY
	GROUP - NOISE EFFECT	

                          Noise condition
Task
E
D
E
D
JT
Tl
Tl
T3
Q IN CN
9.99 11.11 20.25
37.79 27.74 21.75
11.59 16.86 11.20
15.59 14.49 18.78
18.74 17.55 18.00
T3 Tl T2
T2 T3
.45(2.10) 1.19(2.53)
	 .74(2.10)
    q2 = 2.10;  q3 = 2.529,  £ <.05
                                    33

-------
Figures 8 and 9 show, for the easy and difficult task respectively,
the significant main effect of task difficulty and the significant inter-
                                                     *
action effect of anxiety, task difficulty and noise condition on tracking
error for the first task (Data 1).  Figures 10 and 11 (which include
transfer of training effects) show, for the easy and difficult task
respectively, the significant main effects of task difficulty and noise
condition, as well as the significant task difficulty by noise condition
interaction effect on tracking error for the second task (Data 2).  By
inspection, the significant main effect of noise condition occurred  during
the difficult tracking task (Figure 11).

ANALYSIS OF STROOP RESPONSES
                         7 '     '                                >
Table 15 displays the analysis of variance for the Stroop color-word
responses.  There was a significant main effect for noise condition.

Table 16 shows the cell and column means for the Stroop color-word responses.
These data are collapsed over anxiety groups.

In the Neuman-Kuhls analysis, these data were collapsed across anxiety groups
with the resulting column means:
                   Q  =  1.24 sees.
                  IN  =  1.33 sees.
                  CN  =   .87 sees.
                                     34

-------
O
a:
UJ
a,
ffl
  34 r-
  30
  28
  26
  24
  22
  20
  18
  16
  14
  12
  10
   8
g
uj
O
                                 HIGH ANXIETY	
                                 LOW ANXIETY	
               QUIET
                          INTERMITTENT   CONTINUOUS
                          NOISE CONDITION
     FIGURES: EASY TRACKING TASK - DATA 1
   34
   32
   30
   28
   26
   24
   22
   20
   18
   12
   10
<  8
                                 I
               QUIET
                           INTERMITTENT    CONTINUOUS
                          NOIES CONDITION

     FIGURE 9: DIFFICULT TRACKING TASK - DATA 1
                           35

-------
   34 _
   32
   30
   28
cc
g  26
£  24
X  22
^  20
£E  18
uj  16
3  14
   12
   10
    8
      HIGH ANXIETY
      LOW ANXIETY	
                                                 I
                QUIET
INTERMITTENT
CONTINUOUS
                      NOISE CONDITION
   FIGURE 10. EASY TRACKING TASK-DATA 2
   34
S5  32
Z  30
cc  28
g  26
£  24
*  22
<  20
cc  1R
I-  1°
^  16
3  14

8  1n -
<  10
    8
                                                _L
                QUIET      INTERMITTENT

                       NOISE CONDITION

   FIGURE 11. DIFFICULT TRACKING TASK - DATA 2
                CONTINUOUS
                          36

-------
   Table 15.   ANALYSIS OF VARIANCE  FOR STROOP  COLOR WORD  RESPONSES
Source of variation
Anxiety
Task difficulty
Anx X tsk dif
Error
Noise cond
Anx X noise cond
Tsk dif x noise cond
Anx x tsk dif x noise
condition
Error
Total
DF
1
3
3
72
2
2
6
6
144
239
SS
11.2509
3.7769
6.2205
285.8412
30.5040
6.2864
51.6820
43.3612
624.0214
1054.7181
MS
11.2509
1.2590
2.0735
3.9700
15.2520
3.1432
8.6137
7.2269
4.3334

F
2.8339 (N.S.)
.3171 (N.S.)
.5209 (N.S.)

3.5196*
.7253 (N.S.)
1.9877
1.6677 (N.S.)


  £  < .05

In every cell, subjects read the color-words faster, after being exposed
to the experimental condition.  These ranged from a low of .11  seconds
faster to a maximum of 2.79 seconds faster.

ANALYSIS OF HEART RATE'RESPONSES

Heart rates changed significantly as a function of task difficulty and noise
condition (Data 1- First Task) and Task Difficulty alone (Data  2- Second task),
Tables 17 and 18 show the analyses of variance for both first and second
tasks (Data 1 and Data 2) respectively.
                                     37

-------
          Table 16.  STROOP COLOR-WORD RESPONSE TIMES BY CELL MEAN
High Anxiety Low anxiety
^*k Q IN CN Q IN
EE 1.37 1.71 .24 1.15 1.10
ED 1.38 2.62 .67 .80 .28
DE .70 .95 2.22 .71 1.24
DD 2.79 1.68 .96 1.07 1.10
X" 1.56 1.74 1.02 .93 .93
T,-Q T0-IN T,-CN
1 2 3
T] — .37 .46
T2 — .09 -
T3
CN X
.11 .94
1.02 1.12
1.15 .83
.64 1.37
.73





        q2 = .64     q3 = .76    £< .05 (N.S.)




Table 17.  ANALYSIS OF VARIANCE FOR HEART RATE RESPONSES-DATA 1-FIRST TASK
Source of variation
Anxiety
Task difficulty
Anx x tsk dif
Error
Noise cond
Anx x noise cond
Tsk dif x noise cond
Anx x tsk dif x noise
condition
Error
Total
DF
1
3
3
72
2
2
6
6
144
239
SS
11.2873
1004.5748
73.0760
7871.6596
223.3829
123.3906
154.6267
156.1343
5263.0169
14813.7616
MS
11.2873
334.8583
24.3587
109.3286
111.6915
61.6953
25.7711
26.0224
36.5487

F
.1032 (N.S.)
3.0628*
.2228 (N.S.)

3.0559*
1.6880 (N.S.)
.7051 (N.S.)
.7119 (N.S.)


*p <.05
                                    38

-------
TABLE 18.  ANALYSIS OF VARIANCE FOR HEART RATE RESPONSES-DATA 2-SECOND TASK
Source of variation
Anxiety
Task difficulty
Anx x tsk dif
Error
Noise cond
Anx x noise cond
Tsk dif x noise cond
Anx x tsk dif x noise
cond
Error
Total
DF
1
3
3
72
2
2
6
6
144
239
SS
91.1439
1141.0576
261.6626
8754.8715
187.8425
81.9342
169.4013
108.1718
5069.4002
15875.5860
MS
91.1439
380.3525
87.2209
121.5924
93.9212
40.9671
28.2335
18.0286
35.2041

F
.7495 (N.S.)
3.1280*
.7173

2.6679 (N.S.)
1.1637 (N.S.)
.8019 (N.S.)
.5121 (N.S.)


*p <.05
Table 19 shows the cell, column and row means for the Data 1 (First Task)
heart rate responses.  Since there was no main or interaction effects for
the anxiety measure, these data are collapsed across the anxiety variable.

Neuman-Kuhls1 analysis of column means showed that although the F test was
significant, the differences between noise condition, heart rate means was
not large enough to produce a significant multiple comparison.   Row mean
analysis indicated that mean heart rate for one of the Difficult Task
groups (T,) was significantly lower (57.54 BPM) than any of the other task
groups.  This again apparently reflects the operation of uncontrolled
individual differences.  There is no a priori reason to expect T^ versus
TO differences since both groups performed the same task.
                                    39

-------
 TABLE 19.  HEART RATE MEAN RESPONSE - FIRST TASK


                    Noise condition
Task Q
E 59.53
E 75.14
D 63.52
D 56.24
IN
63.17
62.17
67.25
60.11
CN
71.38
63.10
68.58
56.29
X
64.69
66.80
66.45
57.54

(T2)

-------
Table 20 shows the heart rate means for the second tracking task.   Since
there were no main or interaction effects for either anxiety or noise
condition, these data are collapsed across both variables.

       TABLE 20.  HEART RATE MEAN RESPONSE - SECOND TASK
E
X 63.84
T2
Tl
Tl -
T2
T3
T4
D
70.01
T4
T2
4.70
	


E D
68.41 59.14
T3 Tl
T3 T4
9.27* 10.87*
4.57 6.17
1.6
— — — —
         q2 = 7.00;  q3 = 8.42;  q4 = 9.26;  £ <.05

The T, group mean heart rate was significantly lower than either the Tg
or T, group.  These data indicate significantly lowered heart rates for
a difficult task group preceeded by a difficult task or an .easy task.
Again, regardless of prior task, as long as the second task was difficult,
heart rates tended to be lower.

REGRESSION ANALYSES

Since track error, heart rate, and Stroop chart reading times reflected
the effects of either task difficulty or noise conditions, it was deemed
appropriate to assess the contribution of these independent variables to
the dependent variables in terms of the proportion of variance accounted
for.
                                   41

-------
ANALYSIS STRATEGY

Multiple stepwise regression analysis was used to evaluate the contri-
bution of the independent variables to the dependent variables.  The
major independent variables were:  Anxiety Level, Noise Condition, and
Tracking Task Difficulty.  These variables were manipulated either by
selection and assignment of subjects (Anxiety level) or directly as a
function of the experimental conditions (Noise Condition and Tracking
Task Difficulty).  Two additional variables, Eye Dominance and Time,of Day,
known to be correlated with motor skills and tracking task performance
were controlled statistically by adding them to a second regression analy-
ses along with the three major independent variables.  The restricted
model includes only the major independent variables.  The full model in-
cludes the former plus those variables whose effects are to be controlled
statistically.  An analysis of regression comparison of the two models
yields the same results as the traditional analysis of covariance (Roscoe,
                                        (
1975).  The coefficients of determination for each model are compared
using an £ test.  If the F_ is significant, the covariates contribute a
significant proportion of variance to the dependent variable.  Using this
technique six regression analyses were performed for Data 1 (First Tracking
Task); a full and a restricted model for each of the three dependent variables,
TRACK ERROR DATA 1 (RESTRICTED MODEL)
Table 21 reflects the summary data for the restricted model regression
analysis.  The three major independent variables account for a signifi-
cant (£= 18.4727, df_= 3,236, £<.01) but relatively small (R2 = .19)
proportion of the Tracking Error variance.  Since Data 1 represents the
first of each subjects two tracking tasks, the contribution of task
                                     p
difficulty alone to tracking error (R  =  .19) represents the difference
between the easy and the difficult tracking task.
Table 21.  ANALYSIS OF VARIANCE FOR THE RESTRICTED REGRESSION MODEL FOR
           TRACKING ERROR - (DATA 1)

R
R2
S.E.

.43598
.19008
12.02125
SV
Regression
Residuals
Total
DF
3
236
239
SS
5290.70
22543.62
27834.32
MS
1763.56
95.52

F
18.4627*

                                     42

-------
Table 22.  REGRESSION SUMMARY TABLE FOR THE RESTRICTED MODEL FOR TRACKING
           ERROR - (DATA 1)
Variable
Task difficulty
Noise condition
Anxiety
Constant
R
.40506
.43585
.43598

R2
.16407
.18997
.19008

R2 Change
.16407
.02589
.00011

r
.40506
.16754
.06731

b
4.06
2.61
.06
.37
TRACK ERROR-DATA 1 (FULL MODEL)

Table 23 shows the regression analysis of variance for tracking error for
the full regression model which includes time of day and eye dominance.

Table 23.  ANALYSIS OF VARIANCE FOR THE FULL REGRESSION MODEL FOR TRACKING
           ERROR - (DATA 1)

R
R2
S.E.

.45537
.20736
11.93057
SV
Regression
Residuals
Total
DF
4
235
239
SS
5771.86
22062.46
27834.32
MS
1442.96
93.88

F
15.3703*


Table 24 shows the summary data for the full model.

Table 24.  REGRESSION SUMMARY TABLE FOR THE FULL MODEL FOR TRACKING ERROR
           (DATA 1)
Variable9
Task difficulty
Noise condition
Time
Eye dominance
Constant
R
.40506
.43585 '
.45213
.45537

R2
.16407
.18997
.20442
.20736

2
R Change
.16407
.02589
.01445
.00295

r
.40506
.16754
-.12766
-.10133

b
4.52090
2.59801
-.83190
-.99007
5.45925
aThe contribution of anxiety to residual reduction was less than the
"F to enter" the equation, therefore anxiety does not appear.
                                     43

-------
As described, the restricted and full regression models were compared
using an F_ test.  The full model was not significantly different from the
restricted model.  (£ = .8537 d_f = 2,238) indicating that neither eye
dominance nor time of day contributes significantly to the reduction of
residual variance.

EMG - DATA 1 - RESTRICTED AND FULL MODELS

The independent variables in either the restricted or full model did not
contribute significantly to electromyographic changes.  The restricted
model multiple R was .14960 which was not significant (£ = 1.8024, df_ =
3,236, 2<.05).  In the full model, the multiple R was .16824 which was
not significant (£ = 1.7112,  df= 4,235, £ <.05)

HEART RATE- DATA 1 - RESTRICTED AND FULL MODELS

Although task difficulty and noise condition produced significant changes
in heart rate means as evidenced by the analysis of variance, the inde-
pendent variables in both the full and restricted models were not signi-
ficantly related to heart rate changes.  For the restricted model, the
multiple R with heart rate was .08489 (£ = .8602, df = 2,237- £< .05).
In the full model, the multiple R was .14953 (£ = 1.799, df = 3,236, £<.05.)

STROOP RESPONSES - RESTRICTED MODEL
The independent variables in the restricted model did not account for a
significant proportion of the variance in reading time.  Table 25 shows
the regression analysis of variance.
Table 25.  ANALYSIS OF VARIANCE FOR THE RESTRICTED REGRESSION MODEL FOR
           STROOP RESPONSE TIME

R =
R2 =
S.E.

.13171
.02301
= 2.43239
SV
Regression
Residuals
Total
DF
3
211
214
SS
45.67
1276.24
1321.91
MS
15.2243-
6.048

F
2.517(N.S.)


The multiple R (.13171) was not significant (£ = 2.68, df = 3,211, £ <.05)
                                     44

-------
DATA 2 - REGRESSION ANALYSIS
The second tracking task of each experimental session comprised the Data 2
information.  Independent variables in the restricted model  were not signi-
ficantly related to tracking error nor did inclusion of eye dominance or
time of day improve the relationship (£_ = .6762, df = 2,237 p_ > .05,
F_ = 1.1117, df = 5,234 £ > .05, respectively).  Multiple analysis of variance
of these same data however, reflected significant main effects for task se-
quence on tracking error and noise condition on tracking error.  It is quite
evident from these data that a significant difference in tracking error was
present for the groups differing in task difficulty of the second tasks.
The student of analysis of variance vis-a-vis regression analysis will recog-
nize the apparent conflict between regression analysis and analysis of
variance across tracking tasks.  For Data 1, subjects were relatively consis-
tent in the tracking error performance across noise conditions as evidenced
by the significant multiple R and group means were different across task
difficulty.  For Data 2, although, groups differed on their mean tracking
performance as a function of task difficulty as evidenced by the signifi-
cant main effects for task sequence, there was obvious inconsistency in indi-
vidual performance across noise conditions as indicated by the absence of a
significant multiple R.
EMG - DATA 2 - RESTRICTED AND FULL MODELS
Variables in the restricted model were significantly related to EMG changes
(£ = 4.5935, df_ = 3,236, p_ < .01).  Table 26 shows the analysis of variance
for regression and Table 27 displays the summary data.
Table 26.  ANALYSIS OF VARIANCE FOR THE RESTRICTED REGRESSION MODEL FOR EMG-
	(DATA 2)	
	SV	 DF	SS	MS	F	
R     .23816   Regression     3       64.09       21.36       4.5935
R2    .05672   Residual      236    1065.89        4.51
S.E. 2.66571
                                     45

-------
Table 27.  REGRESSION SUMMARY TABLE FOR THE RESTRICTED REGRESSION MODEL FOR
           EMG - (DATA 2)
Variable
Task difficulty
Noise condition
Anxiety
Constant
R
.19092
.23513
.23816

R2
.03645
.05529
.05672'

2
R change
.03645
.01884
.00143

r
.19092
-.13385
.03646

b
.46393
-.45992
-.05028
37.72333
The variables involved in the full model were significantly related to £MG
(F = 3.0905, df = 5,234, £ <.01).


Table 28.  ANALYSIS OF VARIANCE FOR THE FULL REGRESSION MODEL FOR EMG-DATA 2
                   SV
DF
SS
MS
R = .24102     Regression    5       65.642     13.1284    3.0905*

R2 = .05809    Residuals    234    1064.348      4.548

S.E.= 2.68170  Total
Table 29.  REGRESSION SUMMARY TABLE FOR THE FULL REGRESSION MODEL FOR EMG
           DATA 2
Variable Multiple R
Task difficulty
Noise
Anxiety
Eye dominance
Time
.19092
.23513
.23816
.24080
.24102
R Square
.03645
.05529
.05672
.05798
.05809
Simple R
.13385
.13385
.03646
.02583
.00258
Beta
.19779
.13387
.04014
.03562
.01040
                                    46

-------
The multiple R (.24102) was significant (F_ = 3.09, df = 5,234, p_ < .01
and the variables now account for 5.8% of criterion variance composed
of 5.67% in the restricted model.  Comparison of the two models indicates
that addition of time of day and eye dominance does not account for a
significantly increased proportion of criterion variance. (f_ = 1.7618,
df = 2,234; £ >.05)

HEART RATE - DATA 2 - RESTRICTED AND FULL MODELS.

The variables in both the restricted and full models were not significantly
related to the heart rate criterion.  For the restricted model, R^= .07332;
£= .4251, df = 3,236, £ >.05 (N.S.).  For the full model, R = .19283;  F_ =
2.26, df = 5,234, £ >.05 (N.S.).
                                    47

-------
                              SECTION VI
                              DISCUSSION
Data analysis indicate the complexity of the relationships between
physiological and motor skills variables when noise is used in combination
with other variables.  There were no significant mean differences between.
groups in electromyographic potential changes as a function of either
noise profile, anxiety condition or task difficulty.   Tracking error
responses reflected significant changes as a function of some but not
all the variables.  For example, in both tracking tasks, errors increased
when the task was the more difficult rate plus velocity tracking.  This
is not an altogether unexpected result.  This increase in error was
evident, independent of the anxiety and noise conditions.  Within the
anxiety groups however, there were some anomalous results.  In the high
anxiety subjects, the two groups which performed the  easy task first,
differed significantly but not greatly from each other, as did the two
groups performing the difficult task first.  This is  clearly an indication
of subject differences operating.  Similar significant, but slight dif-
ferences exist in the 40 low anxiety subjects.  For the first tracking task,
there were no main effects for either noise condition or anxiety level of
subjects.

Tracking error data from the second task, which includes transfer of
training effects, again shows significant main effects of task difficulty
and also noise condition (high anxiety only).  Again, track errors in-
creased for the difficult task.  In the high anxiety  group, regardless of
prior (first) task, as long as the second task was difficult, error scores
were significantly higher than when the second task was easy.

For the second task, both intermittent (IN) and continuous (CN) noise
profiles resulted in significant decrements in track  error performance
compared to a quiet (Q) condition.  In high anxiety subjects, IN produced
                                    48

-------
the greatest decrement; however,  within the low anxiety group,  the  IN
profile did not produce differential  effects compared to CN  or  Q.   High
anxiety subjects are apparently more  susceptible to the effects of  dif-
fering noise profiles than are low anxiety subjects.

The poorest tracking performance occurred when high anxiety  subjects were
required to switch from the easy to the difficult task during exposure to
intermittent noise.  The best tracking performance was demonstrated by
high anxiety subjects also, who performed the less difficult task  under
the quiet (no noise) condition.  These results are similar to those
found by Goodstein, Speilburger, Williams, and Dahlstrom, 1955 (Speil-
burger, 1966) where performance of the high anxiety subjects was inferior
to that of the low anxiety subjects for the more difficult tasks and
superior on the less difficult tasks.  The results, according to the
authors, would be predicted from Drive Theory.  An extension of Drive
Theory might be called the "response interference hypothesis", which
states that task-irrelevant responses, which in some situations may
interfere with efficient performance, are more easily elicited in  high
than in low anxiety subjects (Spence, 1956; Taylor, 1956; Taylor,  1959).
According to Child (1954), "high anxiety subjects tend to react emotionally
to many experimental situations, even those in which stress  stimulation
is not explicity employed" (Speilburger, 1966).

Previous investigations concerning stress effects on complex task perfor-
mance between anxiety groups have "demonstrated all varieties of relation-
ships, suggesting that these stress conditions are complex in their effects
and interact with a number of variables to determine performance"  (Farber,
1955; Lazarus, Deese, & Osier, 1952;  Speilburger, 1966).  One relatively
consistent finding has been that with certain complex speed  tasks,  "the
performance of high anxiety subjects will decline earlier on the stress
continuum than low anxiety subjects,  and at any given point, be more pro-
nounced".  I. G. Sarason and Palola (1960) found that with simpler speed
                                    49

-------
tasks, the high anxiety subjects improved their performance under stress
(Speilburger, 1966).  This finding is consistent with the results of the
present experiment in which high anxiety subjects performed the more simple
tracking task with lower average error scores than the low anxiety subjects
regardless of noise condition.  Additionally, I. G.  Sarason and Palola
(1960) suggest that "task difficulty was the variable determining the
direction of the effect of stress on performance and hence the nature of
the interaction between stress and anxiety level" (Speilburger, 1966).
This relationship was also evident in the present experiment since there
was a significant interaction between stress, (noise condition) anxiety
and task difficulty.  By visual  inspection, the significant anxiety,
stress (noise condition), and task difficulty interaction only occurred
during the more difficult tracking task.

STROOP RESPONSES

The average performance of subjects in each cell improved as a result of
the experimental conditions.   Stroop Chart reading times  were lower after
the experimental session regardless of the independent variable combina-
tion.  Intermittent noise produced the largest significant reduction in
reading time.  However, the quiet condition produced a significantly
larger reduction than did the continuous conditions.  It  is not clear
that the noise profile alone, or its mode of presentation are directly
responsible for the reduction.  It is conceivable that merely being in
a controlled environment, or attending to a demanding tracking,task
produced the reading time reductions.  It is therefore inappropriate,
as a result of these data, to infer noise facilitated performance.  The
arousal value of attending to a demanding task, independent ..of noise
stimulation, is indicated by the data in Table 16.  When  both the first
and second tasks were the more difficult rate or velocity tracking, the
largest facilitation of Stroop Chart reading occurred (2.79 - cell;
1.37 - group average).
                                   50

-------
HEART RATE RESPONSES

Heart rates changed by less than one full  beat per minute as a func-
tion of the differing noise environments.   This result has no partic-
ular practical significance.  Across task difficulty groups there was
a significant decrease in heart rate as a function of performing the
more difficult rate plus velocity tracking in one group performing
this task.  This result is seen as spurious in that a similar group
of subjects performing the same task did not show a like decrease
(Table 19).

Results of previous studies of heart rate changes following the pre-
sentation of a noxious stimulus (loud tone) show similarly conflict-
ing findings.  Epstein (Speilburger, 1972) concludes that "the find-
ings on heart rate are, to say the least, surprising."  Epstein found
marked accelerative and decelerative reactions in different individuals
which cancel each other out.  "It is apparent that heart rate is a
complexly controlled system, and that strong stimulation does not
always produce acceleration."  Similar results are also reported by
Eason et. al. (1964).  Interestingly, a possible explanation for the
lack of significant heart rate changes as a function of differing
noise environments, particularly the intermittent noise condition
may coincide with Epstein's conclusion that "it is only after the
stimulus is familiar within the experimental context, such as after
it has been presented a number of times in the count-up, that its
presentation by surprise is as apt to produce decelerative as
accelerative reactions."  Epstein further states that a "possible
explanation of the differential response to a familiar and unfamiliar
strong stimulus presented by surprise is that in the former case some
degree of habituation has already taken place, and the stimulus, there-
fore, is less threatening, and less apt to evoke orienting reflexes, with
corresponding heart-rate deceleration.  An unfamiliar strong stimulus, on
the other hand, is apt to evoke defensive reflexes, with corresponding
heart-rate acceleration."
                                   51

-------
Finally, in considering the inconsistencies of human physiological
response to stress, Cattel & Nesselroade (Speilburger, 1972)  state  that
there is an overlap of anxiety and stress in the physiological  area,
and in the awareness of pressure and tension and "it is evident that
a person can be stressing himself most when he is calm, concentrated,
and successfully working hard."

Because subjects were run at different times on experimental  days and
because of the different visual orientation of the subjects to  the
display depending on experimental booth, these two factors were
statistically controlled by adding them to a regression equation with
the major independent variables.  In effect, these former factors were
considered as covariates.  Coefficients of determination for  each
regression model were compared by an F Test.  Addition of the time  of
day and eye dominance variables did not significantly reduce  residual
variance for any of the dependent measures for either the first or  second
task.  Further, when considering only the major independent variables
(anxiety,, noise condition and task difficulty), significant but relatively
small proportions of variance were accounted for.  For example, in  Data 1,
the major independent variables accounted for only 19% of the variance
in track error, 2% of the variance in Stroop response time, 19% of  the
variance in EMG changes and 6% of the variance in heart rate.  In practical
terms therefore, the independent variables, though they may have resulted
in significant mean differences, do not contribute much to our  under-
standing of the differences.  Instead, the data suggest that  no simple
empirical or theoretical statement about the influence of noxious stimu-
lation (noise) can be made and that in predicting its effect  on performance,
account must be taken of such variables as the nature of the  task materials,
the nature and direction of task transfer effects, the manner in which the
noxious stimulation or its anticipation is introduced into the  situation,
the instructions subjects are given about its significance, the trait
anxiety level of the subjects, and the number and intensity of  the  noxious
stimulation.  Previously, such variables had not been systematically in-
vestigated and the interactions that exist among them have been little
understood, if at all.
                                    52

-------
ANXIETY AS A MODERATOR OF NOISE EFFECTS ON HUMAN PSYCHOMOTOR PERFORMANCE

Previous findings by Elbiet and DiScipio (1971), that extroverts
were found to display greater decrements in psychomotor performance
while experiencing noise stimulation than introverts, were not fully
supported by the results of this study.  Instead, it appears that
the psychomotor performance of those assigned to the high anxiety
group was effected to a greater extent by the noise environment
than was the performance of the low anxiety subjects.  The relation-
ship between A-STAIT, A-TRAIT anxiety and introversion is described
in the MMPI Handbook (Dahlstrom et. al, 1973) High "0" scale males
(introverts), display many of the personality traits of high anxiety
subjects with marked insecurities and worries.  Anxiety, as measured
by the IPAT STAI-A scale, simiarly involves feelings of tension,
nervousness, worry and apprehension with high scores reflecting
states of intense apprehension and fearfulness approaching panic
and low scores reflecting feelings of calmness and serenity.

The finding that level of anxiety is a variable that differentially
effects psychomotor performance (track error) is both interesting
and significant in terms of the possible consequences regarding the
effects of moderate noise on human behavior.  Watson (1930) postu-
lated only two innate fear stimuli:  loss of support and loud noises.
It would seem inappropriate to classify 84dBA as a loud noise.  The
question that then arises is, what is  it about moderate noise as a
stimulus that interacts with and degrades the psychomotor performance
of subjects who demonstrate an elevated (but not pathological) level
of anxiety?  That is, what possible mechanisms are operating that
would allow us to understand the relationship between anxiety, noise,
and psychomotor performance.  We believe that fright (as an explanation)
is a class of fear-related emotions which are relatively stimulus
bound, moreover, fright reactions may differ among themselves depending
                                   53

-------
on-their objects (Speilburger, 1972).  It does not appear likely
that there is, in this experiment, the necessary appropriate stimuli
to warrant the development of the emotion of fear.  Substantial
efforts were made to reduce fear related stimuli (i.e., pleasant
background music was played prior to the start of the experiment,
instructions were low keyed, relaxed and designed to reduce the  impact
of upper body disrobing and EKG electrode application, and fellow
students normally assisted during the experiment.

Lazarus & Averill (Speilburger, 1972) state that "there can be little
doubt that theorizing with respect to anxiety is still in the
elementary stages, somewhat like the concept of air in 18th century
chemistry."  However, there is a significant body of literature  on
the subject of the causes and consequences of anxiety (Speilburger,
1972) which classifies anxiety into three general categories in
terms of etiology.  They are (1) Primary Overstimulation  (2) Response
Unavailability and  (3) Cognitive Incongruity.

According to the neobehavioristic learning theorists, pain is the
unconditioned stimulus for fear and anxiety.  Organisms have an
upper limit to stimulation.  Overstimulation is associated with
feelings of being overwhelmed and bombarded with stimuli, corresponding
to the statement, "Stop it, I can't stand it anymore."  Epstein
(Speilburger, 1972) states that "organisms are energy systems that
are responsive to energy inputs and must maintain their levels of
excitation, no less than their other internal states, within
homeostatic limits in order to survive.  Small increments in arousal
cause the individual to attend to his environment and to register the
stimulus associated with the increment.  Large increments in arousal
cause a reduction in receptivity to and registration of stimulation
and are experienced as unpleasant.  It appears that arousal is con-
trolled through inhibition, which is intimately associated with  the
                                  54

-------
establishment of expectances.  Through the process of inhibition,
stimuli such as moderately loud noises that were initially attended
to only because of their energetic properties become registered
or "learned1, and thereafter can be responded to in terms of their
cue properties."  It may be that for low anxiety subjects, the
introduction of moderate noise as an environmental stimulus serves
as a general facilitator of performance because it causes the
organism to attend to its environment and raises its general activation
level.  Evidence for this can be seen in the consistent decrease in
the time needed to successfully respond to the post-experimental
STROOP color-word chart.  In every cell, subjects read the color-
words faster after being exposed to the experimental condition.
This finding is contrary to other studies involving high sound pressure
levels (Sommer & Harris, 1972) and indicates the arousing effect of
the experimental condition alone.  Several studies (Scott, 1966; Malmo,
1957) show that performance increases as a function of activation  level
up to a point and then declines as general arousal is further increased.
This concept would be in agreement with Di Scipio's (1971) finding that
white noise facilitates psychomotor response for an optimal period of
time, after which decrements were observed.  However, with subjects dis-
playing higher than normal anxiety with concomitant internal feelings of
tension, nervousness, worry and apprehension, further increases in external
sources of stimulation, such as moderate noise, may well be debilitating
to some degree due to the requirement that the organism cope with
heightened levels of general arousal, particularly since "increases in
arousal are produced by any stimulation, internal or external" (Epstein, 1967)
                           i
According to Lazarus (1966), anxiety is viewed as a state in which the
individual experiences diffuse arousal but is unable to direct that
arousal into purposive action.  Moreover, "it is the arousal and the
defenses against it, and not the anxiety, and the defenses against it,
                                     55

-------
that are responsible for the primary symptoms of the behavior dis-
orders.  In this respect, it has been established that experimental
neurosis can be evoked by conditions that are arousing, but not
frightening, such as difficult discriminations and the disconfirmation
of established expectancies."

The notion that anxiety is more noxious to an organism than fear appears
to have adaptive value.  Epstein notes that "one of the most common
distinctions made between anxiety and fear is that in fear the source
of the threat is known and in anxiety it is unknown."  Epstein further
states that fear is an avoidance motive.  If there were no restraints,
internal or external, fear would support the action of flight.  As
mentioned previously, anxiety can be defined as unresolved fear, or
alternately, as a state of undirected arousal following the perception
of threat.  Given a crisis, it is important that the organism rapidly
assess the situation and take immediate action.  In the course of
evolution, man, as an animal, has also developed the ability to, when
danger is perceived, generate a heightened state of arousal that
provides nonspecific preparation for flight or fight.  "Thus, it is
adaptive for the state of diffuse arousal  to be an acutely unpleasant
one, and for it to become more so in time, thereby providing the animal
(man) with a powerful incentive to resolve indecision and to select
a course of action."

Thus, it would appear that noise need not be so loud as to generate
fear to be arousing or anxiety provoking.   Diffuse arousal, resulting
from moderately noxious stimulation (noise) coupled with heightened
states of trait anxiety should, according to Epstein, be perceived by
the organism as acutely Unpleasant, thereby providing the input to
select a course of action.  However, in the case of environmental
noise, particularity intermittent, uncontrollable noise, there may be
no appropriate course of action available to the organism, resulting in
unresolved and undirected arousal, anxiety and frustration.
                                  56

-------
Epstein (Speilburger, 1972) concludes by noting that the relationship
between frustration and anxiety is evident.  "Frustration has been
accorded special significance in psychological  theorizing because
of its relevance for maladaptive behavior and the normal stresses of
everyday living.  The consequences of heightened states of diffuse
arousal have been observed to include restlessness and tension,
aggression, apathy, withdrawal, disorganized behavior, regression
and escape."  All of which appear to be common symptoms of the
"urban din" we must live in.

An additional mechanism by which environmental  noise may interact with
the second basic source of anxiety can be described as Response Un-
availability.  In this experiment, as in real life, the subject
could not control the noise.  It was just there.  Mandler (1961)
states that "in a state of arousal, the organism who has no behavior
available to him, who continues to seek situationally or cognitively
appropriate behavior is 'helpless' and also may consider himself, in
terms of the common language, as being in a state of anxiety."
Mandler further states that "any such arousal - peripheral or central,
environmentally or behaviorally induced - will  lead to a feeling of
helplessness when no cognitive or behavioral sequence appropriate
to the situation is available, or when no substitute sequence or
escape from the field provides a means of terminating the state
of arousal."

This raises the question of what would happen if escape from a source
of threat (noxious stimulation) were blocked.  Studies by Seligman and
his colleagues (Speilburger, 1972) indicate that "a state of helplessness
produced by unavoidable noxious stimulation is extremely debilitating
and tends to be self-maintaining."  This factor may also apply in this
experiment, particularly for those subjects who displayed higher than
normal states of anxiety and who were locked in a sealed acoustic chamber
and subjected to unpleasant stimulation over which they had no control.
                                   57

-------
A third basic parameter that affects arousal  level is Cognitive
Incongruity which is determined by the consistency of one's expecta-
tions and the ability to establish an adequate cognitive model  of events
(Epstein, 1967).  In the present experiment,  subjects were not  able
to predict the onset or duration of the intermittent noise.  (Event
Certainty, Temporal Uncertainty).  There is considerable evidence in the
research literature indicating that in a normal subject population the
violation of expectancies beyond a certain point induces anxiety.  "It
would appear that an accurate expectation with regard to the occurrence
of noxious stimulation is generally sought" (Epstein, 1967; Speilburger,
1972).  In the case of intermittent noise, the violation of expectancies
is clearly evident, since the experimental results indicate that high
anxiety subjects were more affected by the intermittent noise condition
than by the continuous noise condition, whereas the low anxiety subjects
were not significantly affected by noise conditions at all.  Again,
demonstrating the need to specify anxiety level when predicting the
effects of moderate noise on human behavior.

Interestingly, the concept of Cognitive Incongruity may also be a
factor in the apparently anomalous finding that low anxiety subjects
(2nd task) displayed the greatest decrement in track error performance
when transferring from the easy to difficult task under the quiet (no
noise) condition.  Under both anxiety conditions, the transfer  from the
easy to the difficult task resulted in the greatest track error per-
formance decrement.  This finding appears consistent with the concept
of Cognitive Incongruity in that expectations concerning the difficulty
of the tracking task were most severely violated under this condition.
A possible explanation for the low anxiety subject's poor performance
under these conditions (quiet) may be linked to overall activation
levels, i.e., insufficient internal and external stimulation to perform
well under these conditions.  This would be consistent with Scott's
(1966) findings concerning activation level and performance previously
mentioned.  Several other studies have also suggested an inverted U-shaped
                                   58

-------
function as describing the relationship between motivation and perfor-
mance (Malmo, 1957, 1958, & 1959) generally by appealing to the notion
that "difficult" tasks elicit more emotionality than "easy" ones,
particularly in high anxiety subjects.  Such studies have generally con-
firmed the notion that "subjects who were low in both anxiety score and
experimentally manipulated stress and those who were high in both were
poorer in performance than members of the other two groups" (Speilburger,
1966).

•Normally, anticipation is of great importance for research on anxiety
(and stress in general), due to the emerging cognitive appraisal by
the person of the significance of the event.  Breznitz (Speilburger,
1972) speaks of the "incubation of threat", observing in his research
and that the longer subjects had to wait, the greater the stress, as
measured by heart rate, just prior to shock.  Breznitz found that most
of the stress reaction (as measured autonomically) took place during
the. anticipatory period, with little further increment during the noxious.
stimulus period itself.  Furthermore, a dissertation by Folkins (1970)
showed that degree of stress varied significantly as a function of antici-
pation as time increases from 5-30 seconds, with a further rise up to 1
minute, with a drop in disturbance from 3-5 minutes - though it rose
slightly at 20 minutes.  In the present experiment, baseline autonomic
data was taken during the 2nd minute period, followed by 2 minutes of
instructions, with the onset of the noise following during the 4 to 9
minute period, thereby elevating the baseline response and depressing
the experimental condition response.  It appears, according to Folkins,
that with 3-5 minutes to appraise the situation, "subjects are better able
to develop self-assuring coping responses, and therefore display less
stress."  These findings may well have contributed to the lack of
practical significance between baseline and experimental condition phys-
iological arousal levels in the present experiment.
                                   59

-------
Finally, interruption of cognitive processes appears to generate a
state of heightened arousal  (Mandler, 1961).  Dr.  Gordon Davis,
University of California Medical  School  at Davis,  provides a
"clinical" description of interruption (due to noise)  in its purest
form (Speilburger, 1972).  The case study of a 9 year  old boy is de-
scribed - when noise or activity  going on at home  reaches a certain
point - an explosion of behavior  results - "he may go  out and ride
his bicycle to get away from it," apparently as the result of a
deficient ability to carry out a  cognitive plan when experiencing
interruption or environmental  disorganization.  It may be that one
of the primary effects of moderate environmental noise is its
interrupting, disorganizing quality, which would be particularly
debilitating to those subjects who already experience  substantial
internal arousal as the result of elevated trait anxiety levels.

In sum, "human beings are motivated to structure their world and to
find ways of dealing with it largely because of the characteristics
of their anxiety system.  At low  levels  of anxiety, the process  is
a constructive one, leading to expanded  awareness  and  increasing
control of nature.  At high levels, it produces defensive retrenchment,
including delusional interpretations of events (any explanation  is
better than none), and compulsive rituals for dealing  with them
(any action is better than none)" (Epstein, 1967;  Speilburger, 1972).
Thus, for organisms that already  possess heightened internal states
of arousal (high trait anxiety),  by adding stimulation from the  environ-
ment (noise) that violates or precludes  the development of expectancies,
(intermittent noise) it can then  be expected that  decrements in  psycho-
motor performance should result,  and as  part of the "cost" to the
organism, frustration can also be expected to occur with all of  the
negative ramifications that now appear to plague our highly industrial-
ized, urban society.
                                   60

-------
                               SECTION  VII

                               REFERENCES


 1.  Anderson, P.  A., "Social  Class,  Noise  and  Performance," Dissertation
     Abstracts International.  33,  No.  6-B,  pp 2784  (1973).

 2.  Bull, A.  J.,  "Effects of  Noise and Intolerance of Ambiguity Upon
     Attraction for Similar and Dissimilar  Others," Journal of Social Psychology,
     88, No.  1, pp 151-152 (1973).

 3.  Campbell, B.  A., "Interaction of Aversive  Stimuli:   Summation or Inhi-
     bition?," Journal  of Experimental  Psychology,  78, No.  2, pp 181-190
     (1968).

 4.  Child, I. L., "Personality,"  Annual  Review of  Psychology, No. 5, pp 149-
     170 (1954).

 5.  Dahlstrom, W. G.,  Welsh,  G. S.,  and  Dahlstrom, L. E.,  "On MMPI Handbook,
     Volume I," University of  Minnesota Press  (1973).

 6.  Deese, J., Lazarus, R. S., and Keenan, 0., "Anxiety, Anxiety Reduction
     and Stress in Learning,"  Journal  of  Experimental Psychology, 46, pp 55-60
     (1953).

 7.  DiScipio, W.  J.   "Psychomotor Performance  as a Function of White Noise
     and Personality Variables," Perceptual and Motor Skills, 33, No. 1,
     pp 82 (1971).

 8.  Dyer, F.  N.,  "The  Stroop  Phenomenon  and It's Use in  the Study of Perceptual,
     Cognitive, and Response Processes,"  Memory and Cognition, 1, No. 2, pp
     106-120 (1973).

 9.  Eason, R. G., Harter, M.  R.,  and Storm, W. F., "Activation and Behavior.
     I:  Relationship Between-Physiological 'Indicants' of  Activation and Per-
     formance  During Investigation of Nonsense  Syllables  Using Differing
     Induced Tension Conditions,"  Perceptual, Motor Skills. No. 19. pp 95-
     110 (1964).

10.  Edsell, R. D., "The Effect of Noise  on Some Interrelationships Between
     Social Interaction and Anxiety," (Doctoral dissertation, Drexel University)
     Ann Arbor, Mich:  University Microfilms. No. 734040  (1973).

11.  Elliott,  C. D., "Noise Tolerance and Extraversion in Children," British
     Journal of Psychology, 62, No. 3, pp 375-380  (1971).
                                     61

-------
12.  English, H. B., and English, A.  V.,  "A Comprehensive Dictionary of Psycho-
     logical Terms, David McKay Co.,  New  York (1958).

13.  Epstein, S., "Toward a Unified Theory of Anxiety,"  In B.  A.  Mather (Ed.),
     Progress in Experimental  Personality Research,  Academic  Press,  New York,
     Vol 4, pp 1-89 (1967).

14.  Epstein, S., "The Nature of Anxiety  with Emphasis  Upon It's  Relationship*to
     Expectancy," In C. D. Spiel burger (Ed.), Anxiety,  Current Trends in Theory
     and Research Vol  II, Academic Press, New York,  pp  291-337 (1972).

15.  Farber, I.  E., "The Role of Motivation In Verbal Learning and  Performance,"
     Psychological Bulletin, No. 52,  pp 311-327 (1955).

16.  Fechter, J. V., "The Effects of  Noise on Human  Learning," (Doctoral  dis-
     sertation,  University of South Dakota) Ann Arobr,  Mich:   University Micro-
     films, No.  72-32728 (1972).

17.  Folkins, C. H., "Temporal  Factors and the Cognitive Mediators  of Stress
     Reactions," Journal of Personality and Social Psychology, No.  14,  pp 173-
     184 (1970).

18.  Glass, D. C., and Singer,  J. E., "Experimental  Studies of Uncontrollable
     and Predictable Noise," Representative Research in  Social  Psychology,  4,
     No. 1, pp 165-183 (1973).

19.  Goodstein,  L. D., Spielburger, C. D., Williams, J.  E., and Dahlstrom,  W. G.,
     "The Effects of Serial Position  and  Design Difficulty on  Recall  of the
     Bender-Gestalt Test Designs," Journal of Consultive Psychology,  No.  19,
     pp 230-234  (1959).

20.  Grether, W. F., "Two Experiments on  the Effects of  Combined  Heat,  Noise,
     and Vibration Stress," Aerospace Medical Research  Laboratory,  Wright-
     Patterson AFB, Ohio. No.  AMRL-TR-71-113 (1972).

21.  Gulian, E., "The Factor 'Difficulty  of the Task1 and It's Influence Upon
     Performance Level Under Noise Conditions," Revista  de Psihologie,  18,  No.
     3, pp 323-330 (1972).

22.  Gulian, E., "Psychological  Consequences of Exposure to Noise,  Facts and
     Explanations," Proceedings  of the International Congress  on  Noise as a
     Health Problem, U. S. Environmental  Protection  Agency, Washington, D.  C.,
     No. 550/9-73-008.

23.  Harcum, E.  R. and Monti,  P. M.,  "Cognitions and Placebos  in  Behavioral
     Research in Ambient Noise," Perception & Motor  Skills, 37, No.  1,  pp 75-
     79 (1973).
                                     62

-------
24.  Harris, C. S., and Schoenberger,  R.  W.,  "Combined  Effects  of Noise  and
     Vibration on Psychomotor Performance,"  US.AF AMRL Technical  Report,  No.  70-
     14, pp 24 (1970).

25.  Jansen, G., and Hoffman, H., "Noise-Induced Changes  in  Fine Motoricity
     and Noise Induced Feelings of Annoyance Which Depend Upon  Definite  Person-
     ality Dimensions," Army Foreign Science and Technology  Center,  Charlottes-
     ville, VA., No. FSTC-HT-23-236-71 (1971).

26.  Jauhiainen, T., Kohonen, A., and Jauhiainen, M.,  "Combined Effect of  Noise
     and Neomycin in the Cochlea," Acta Oto-Laryngologica, 73,  No.  5, pp 387-
     390 (1972).

27.  Kryter, K. D., "The Effects of Noise on Man," Academic  Press,  New York  (1970).

28.  Kumar, P., -and Mathur, C. N., "Sex and Noise Destractibility,"  Indian Journal
     of Applied Psychology, No. 6, pp 13-14 (1969).

29.  Lazarus, R. S., Psychological Stress and the Coping  Response,  Vol 3,  McGraw-
     Hill, New York, pp 96-102 (1966).

30.  Lazarus, R. S., and Averill, J. R., "Emotion and  Cognition:  With Special
     Reference to Anxiety," In C. D. Spielburger (Ed.), Anxiety, Current Trends
     in Theory and Research, Vol II, Academic Press, New  York,  pp 242-283  (1972).

31.  Lazarus, R. S., Deese, J., and Osier, S. F., "The  Effects  of Psychological
     Stress Upon Performance," Psychological Bulletin,  No. 49,  pp 293-317  (1952).

32.  Malhur, C. N., "Age as a Factor in Noise Distractibility," Manas, 19, No. 1,
     pp 31-33 (1972).

33.  Mai mo, R. B., "Anxiety and Behavioral Arousal," Psychological  Review, No. 64,
     pp 276-287 (1957).

34.  Malmo, R. B., "Measurement of Drive:  An Unsolved  Problem  in Psychology,"
     In M. R. Jones (Ed.), Nebraska Symposium on Motivation, Lincoln, Nebraska,
     University of Nebraska Press, pp 229-265 (1958).

35.  Malmo, R. B., "Activation:  A Neuropsychological  Dimension," Psychological
     Review, No. 66, pp 367-386, (1959).

36.  Mandler, G., "Helplessness:  Theory and Research  In  Anxiety,"  In C. D.  Spiel-
     burger (Ed.), Anxiety. Current Trends in Theory and  Research,  Vol II, Academic
     Press, New York, pp 360-378 (1972).

37.  Mason, R. K., "The Influence of Noise on Emotional States," Journal of  Psycho-
     mat ic Research, 13, No. 3, pp 275-282 (1969).
                                      63

-------
38.  Myers, 0. L., Fundamentals of Experimental  Design.   Allyn  and  Bacon,  Boston
     (1966).

39.  Renshaw, F- M., "The Combined Effects of Heat and Noise in Work Performance,"
     Dissertation Abstracts International, 33, No. 8-B,  pp 3699 (1973).

40.  Roscoe, J. T., Fundamental Research Statistics for  the Behavioral  Sciences,
     Holt, Rinehart and Winston, Inc.  (1969).

41.  Sarason, I. G., and Palola, E.  G.,  "The Relationship of Test and General
     Anxiety, Difficulty of Task, and  Experimental Instructions to  Performance,"
     Journal of Experimental Psychology, No. 59, pp 185-191  (1960).

42.  Scott, W. E., JR., "Activation  Theory and Task Design," Organizational  Be-
     havior and Human Performance, No.  1, pp 3-30 (1966).

43.  Spielburger, C. D., Anxiety and Behavior, Academic  Press,  New  York (1966).

44.  Spielburger, C. D., Anxiety, Current Trends In Theory and  Research, Vol  II,
     Academic Press, New York (1972).

45.  Spence, J. T., and Spence, K. W.,  "The Motivational  Components  of Manifest
     Anxiety:  Drive and Drive Stimuli," In C. D.  Spielburger (Ed.), Anxiety and
     Behavior, Academic Press, New York  (1966).

46.  Spence, K. W., Behavior Theory  and  Conditioning,  Yale University Press,  New
     Haven, Conn. (1956).

47.  Stephans, S. D., "Personality and  the Slope of Loudness Function," Quarterly
     Journal of Experimental Psychology, 22, No. 1, pp 9-13 (1970).

48.  Taylor, J. A., "Drive Theory and Manifest Anxiety,"  Psychological  Bulletin,
     No. 53, pp 303-320 (1956).

49.  Taylor, J. A., "Manifest Anxiety,  Response Interference and Repression,"
     Experimental Foundations of Clinical Psychology Symposium, University
     of Virginia Medical School, April  (1959).

50.  Watson, J. B., Behaviorism (2nd ed.), Norton & Company, New York (1930).
                                     64

-------
                             SECTION VIII

                               GLOSSARY


Activation Level - The preparation or the tendency toward action.  The
level of activation of a whole system is the degree of tension.

Anxiety - A feeling of threat, especially of a fearsome threat, without the
person's being able to say what he thinks threatens.

Anxiety, Free - A chronic state of anxiety which attaches to almost any
situation or activity of the individual.

Anxiety, Trait - A chronic state of anxiety which remains unattached and
constant over all situations and activity of the individual.

A-STATE - State anxiety may be conceptualized as a transitory emotional
state or condition that varies in intensity and fluctuates over time.

A-TRAIT - Trait Anxiety refers to relatively stable individual differences
in anxiety proveness.

Eye Dominance - A tendency to fixate objects with one eye rather than with
both and to depend primarily upon the impressions of that one eye, though
the nbn-preferred eye is not blind.

Extravert - A person who tends strongly to the attitude of extraversion.
Extraversion has three aspects to include outward oriented interests, ease
of social adjustment and open behavior.

Homeostasis - The maintenance of consistency of relationships or equilibrium
in the bodily processes, whether physical or psychological.  Any departure
from the equilibrium sets in motion activities that tend to restore it.

Inhibition - A mental state in which the range and amount of behavior is cur-
tailed, beginning or continuing a course of action is difficult, and there
is a peculiar hesitancy as if restrained.

Introvert - A person who tentis strongly to the attitude of introversion.  In-
troversion has three aspects to include inward oriented interests, difficulty
of social adjustment and secretive behavior.

IPAT - Institute for Personality Assessment and Testing.

Neurosis - A mental disorder ill-defined in character but milder than psychosis,
Neurosis are usually characterized as disfunction within the individual, as
opposed to between the individual and his environment.
                                     65

-------
STAI-A - State-Trait Anxiety Inventory - Scale A - consists of twenty
statements that ask people to describe how they generally feel.

Stroop Chart - A collection of stimulus words describing colors  which are
printed in a visual color other than the word.  This maximizes the inter-
ference between the written and reproduction colors.
                                  66

-------
                          SECTION IX
                          APPENDICES
                                                         Page
A.  System Inputs - Block Diagram                         68
B.  System Outputs - Block Diagram                        69
                               67

-------
                    INPUTS
       BOOTH 1
 BOOTH 2
oc
*
85
                  	i~r
     Q O
     TR-20
  (ANALOG COMPUTER)
       I
                         REAL TIME SHAPED NOISE
                INTEGRATED
                ERROR
                                     POWER AMP
     RANDOM
     SIGNAL
      GEN.
   I
FREQ. SHAPER
   I
                                        TAPE
                       68

-------
   EMG
 DC
Mv
 SIGNAL
 INT/REC
 CIRCUIT
   I
UNIVERSAL
 PRE-AMP
              OUTPUTS
              EKG
 TRACKING
 VOLTAGE
             BIO-TACK
           DC VOLTAGE
 chir2    ch, 1
   TR-20
(ANALOG COMPUTER)
                        ABSOLUTE ERROR
                            TAPE
                             REC.
                                  TCG
                     NOISE  CONDITION  	|
               chl 3      chl 4     chll4
         FM RECORDER
                   69

-------
                                   TECHNICAL REPORT DATA
                            fflease read /nLtntcti
-------