ABSTRACT

HEART PERIOD AND RESPIRATORY CONCOMITANTS OP
ATTENTION IN HORMALS AND RETARDATES DURING
A FIXED REACTION TIME TASK

BY
Antoinette Krupski

This study demonstrated significant differences
between retardates and nornals in reaction tine (RT)
performance and in sec-by-sec heart period (HP) activity
during preparatory intervals (PIs) of 4, 7, and l3-secs.
There were no significant group differences in respiration
frequency corresponding to the HP changes.

Normals were characterized as exhibiting signifi-
cant HP deceleration at about the time the reaction signal
occurred in all three PI conditions. Retardates, on the
other hand, were characterised as showing no significant
HP deceleration in the 4-sec and 7-sec PI conditions while
showing a significant deceleration in the l3-sec PI condi-
tion. Group differences in HP deceleration were inter-
preted as indicating an inhibition deficit, or as indi-
cating an absence of temporal conditioning in retardates.
The results were also related to Lacey's theory of atten-
tion.

Heart period changes to the onset of stimulation

also differed for groups as a function of PI. Normals
exhibited progressively larger accelerations as a function
of PI length. Retardates, on the other hand, exhibited
the same magnitude of acceleration for each PI condition.
These data were interpreted as indicating an inapprOpriate
response set, an inhibition deficit, and as reflecting a
lack of integration between environmental demands and
physiological responsiveness in retarded individuals.

HEART PERIOD AND RESPIRATORY CONCOMITANTS OF
ATTENTION IN NORMALS AND RETARDATES DURING
A FIXED REACTION TIME TASK

BY

Antoinette Krupski

A THESIS

Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY

Department of Psychology

1971

u— 7/ {/7

To my dearest and best friend, Don.

ii

ACKNOWLEDGEMENTS

I would like to express sincere thanks to Dr.
Hiram Fitzgerald who served as chairman of this disser-
tation committee for the encouragement, guidance, and
criticisms he has offered throughout my Ph. D. program.
I would also like to extend special thanks to Dr. M. Ray
Denny, chairman of my guidance committee, for the per-
sonal concern and support he has shown throughout my
training. The other committee members, Dr. William Crane,
Dr. Elaine Donelson, and Dr. Lester Hyman must also be
acknowledged for their thoughtful comments and criticisms
of this thesis.

I would like to extend sincere thanks to Fred
Beckett of the Okemos school system and David Knaggs of
St. Lawrence Community Mental Health Center for their
assistance in obtaining subjects and to Gary Connors for
his technical assistance. A final thanks goes to Steve
Porges who provided the data analysis routines and has
also been a great source of encouragement and friendship.

This research was supported by a predoctoral
fellowship from the American Association of University
Women (1970-1971), and by a NIMH predoctoral traineeship,
#HD 00158, in mental retardation (l967-l970)--both of

which were awarded to the author. It was also partially

iii

iv

supported by an NIMH Research Grant MH-18655 awarded to
Dr. Hiram Fitzgerald. Portions of this paper were pre-
sented at meetings of The Society for Psychophysiological
Research, New Orleans, 1970 and The Gatlinburg Conference

on Research and Theory in Mental Retardation, 1971.

TABLE OF

ACKNOWLEDGEMENTS. . . . . . .
LIST OF TABLES. . . . . . . .
LIST OF FIGURES . . . . . . .
LIST OF APPENDICES. . . . . .
Chapter
I. INTRODUCTION . . . . .
II. METHOD . . . . . . . .
III. RESULTS. . . . .,. . .
IV. DISCUSSION . . . . . .

REFERENCES 0 O O O O O O O O 0

CONTENTS

Page

111

vi

vii

viii

10

16

43

56

Table
l.

2.

3.

4.

S.

6.

LIST OF TABLES

Reaction Time: Mean and standard devi-
ations expressed in milliseconds for

retarded and normal groups in 4, 7, and
13-886 PI conditionse e e e e e e e e 0

Pearson product moment correlations

between total IQ, verbal IQ, performance

Wechsler IQ scores, and mean reaction

times for 4, 7, and 13-sec PI conditions

for the retarded group. . . . . . . . .

Pearson product moment correlations be-
tween mean basal heart period and mean
reaction time scores for 4, 7, and 13-
sec PI conditions for normal, retarded,
and Combined groups e e e e e e e e e e

Pearson product moment correlations
between mean HP acceleration score and
reaction time for retarded, normal, and
combined groups, and 4, 7, and l3-sec
PI conditions.. . . . . . . . . . . . .

Pearson product moment correlations be-
tween heart period deceleration differ-
ence scores I and II and mean reaction

times for 4, 7, and l3-sec PI conditions
for normal, retarded, and combined groups

Pearson product moment correlations be-

tween mean respiration frequency and mean

heart period for retarded, normal and
combined groups in 4, 7, and 13-sec PI
conditions, and for ”pre', “PI", “post“

periods and overall . . . . . . . . . . .

vi

Page

16

17

33

35

35

42

LIST OF FIGURES

Figure Page
1. A typical fixed reaction time paradigm. . 2
2. Mean heart period change as a function

of successive seconds in the 4-sec PI
condition for normal and retarded
group8...............o.. 19

3. Mean heart period change as a function
of successive seconds in the 7-sec PI
condition for normal and retarded
grOUPSe O O O O O O O O O O O O O O O O O 25

4. Mean heart period change as a function
of successive seconds in the l3-sec PI
condition for normal and retarded
groups 0 O O O O O O O O O O O O O O O O O 28

5. Mean heart period change from prestim-
ulus mean as a function of seconds past
warning light onset in 4, 7, and 13-sec
PI conditions for normal and retarded
groups.................. 31

6. Mean respiration frequency as a function

of trials in the 4-sec PI condition for
normal and retarded groups. . . . . . . . 38

vii

LIST OF APPENDICES

Appendix Page
A Instructions for reaction time task. . . 61
B Analyses of Variance summary tables. . . 64

viii

INTRODUCTION

When compared to normals, retardates typically
perform significantly slower on reaction time (RT) tasks
(Baumeister & Kellas, 1968). Several investigators have
suggested that the retardate's slower RT is due to an
inability to maintain the level of attention that is
required for a fast RT (Baumeister & Kellas, 1968; Denny,
1964). The specific purpose of the present study was to
examine heart rate (HR) changes that occur in both re-
tarded and non-retarded individuals during a RT task in an
attempt to evaluate the alleged "attention deficit“ fre-
quently attributed to retardates. Heart rate was chosen
for this assessment because recent theoretical analyses
suggest that the direction of response, that is HR accel-
eration or HR deceleration, is related to environmental
demands (Coquery & Lacey, 1966; Lacey, 1967; Lacey & Lacey,
1966; Obrist, Sutterer, & Howard, 1969a; Obrist, "ebb, &
Sutterer, 1969b). The directional HR response has the
additional advantage of being a discrete response that is
amenable to precise quantification (Brener, 1967).

The typical RT design is depicted in Figure 1. As
depicted in Figure 1, a simple RT situation typically in-
volves the successive presentation of a warning signal, a

preparatory interval (PI), and a reaction signal. For

FIGURE 1

A typical fixed reaction time paradigm.

m2:

mafiaghmuhmom maggmuucm

2.. .
J<>¢wkz_ >¢Oh<¢<au¢a

L 1

29.549. 92.2.33

 

4

example, the subject (g) is told to press a button as
quickly as he can when the green light goes off--onset of
the green light being the warning signal and offset of the
green light being the reaction signal. Reaction time
score is the elapsed time between onset of reaction signal
and occurrence of a motor response. The PI is the time
between onset of warning and reaction signals, i.e., the
time during which.§,can prepare to respond to the reaction
signal. When the PI is the same duration throughout a
block of trials, the task is called a fixed RT task. It
is generally assumed thatԤ.is attending or concentrating
during the PI if he makes a fast response.

A growing body of literature relates directional
HR changes to RT performance, and theoretically to atten-
tion. The majority of these studies have been done with
nonretarded'gs and have typically employed a fixed RT task
where the PI is 4-secs or longer (Chase, Graham, & Graham,
1968; Coquery & Lacey, 1966: Lacey & Lacey, 1966; Obrist,
‘g£_;;., 1969a, l969b; Obrist, lebb, Sutterer, & Howard,
1970; Hebb & Obrist, 1970). The resulting HR pattern
during the PI is usually triphasic in nature: sometimes a
brief deceleration is reported to occur after the warning
signal, followed by an acceleratory response, and finally
a deceleratory response in anticipation of the reaction
signal. Studies employing a classical conditioning para-
digm with a fixed CS-UCS interval greater than 4-secs in

length have yielded similar HR patterns (Hastings & Obrist,

1967; Headrick & Graham, 1969; Milson, 1969; lead a Obrist,
1964; Zeaman & Smith, 1965). Research in this area has
typically focused upon the HR deceleration which occurs
prior to the reaction signal in the fixed RT studies, and
prior to the appearance of the UCS in the classical con-
ditioning studies. This focus on HR deceleration probably
stems from the finding that magnitude of anticipatory HR
deceleration has been found to be positively correlated to
RT performance; that is, the greater the HR deceleration,
the faster the RT (Obrist 93341., 1969a, 1969b; Lacey, 1967;
Lacey & Lacey, 1966; Coquery & Lacey, 1966).

Lacey (1967) interprets the HR deceleration imme-
diately preceding a reaction signal in the fixed RT situa-
tion as a state of enhanced sensitivity to external stimu-
lation and hence, greater attention. Consistent with
Lacey's interpretation, Graham & Clifton (1966) suggest that
cardiac deceleration is a component of the orienting re-
sponse (OR) which is also presumed to reflect a state of
enhanced sensitivity. It follows from both of these ap-
proaches that a large HR deceleration occurring at about
the time a signal stimulus is to occur would result in en-
hanced sensitivity to external stimuli, enhanced attention
to environmental demands, and therefore, a faster RT.

Other investigators have suggested that anticipa-
tory HR deceleration is a temporally conditioned response
which occurs as a function of the fixed PI interval or the

fixed cs-ucs interval (Chase et al., 1968; Fitzgerald &

I) (I) Ill. In- F)

Porges, 1970; Headrick & Graham, 1969; Porges, 1970). The
temporal conditioning notion is supported by the fact that
anticipatory deceleration does not become time-locked to
the reaction signal or UCS when the PI or CS-UCS intervals
vary from trial to trial (Obrist‘g£_§l., 1969a; Porges, 1970).
Other support results from trial analyses which demonstrate
that the anticipatory HR deceleration does not occur on the
first trial, but becomes time-locked to the signal stimulus
as a function of trials (Chase‘25_gl., 1968: Hastings &
Obrist, 1967; Headrick & Graham, 1969; Porges, 1970). Chase
gngg, (1968) propose that the conditioned deceleration
which occurs in anticipation of the reaction signal corres-
ponds to a conditioned “attention” response as described by
Lacey. Consequently, their conditioning notion can be viewed
as an extension of the Lacey hypothesis.

In contrast to the Lacey and Chase‘g§_§l. contention,
Porges (1970) asserts that the observed anticipatory HR de-
celeration and RT performance are not related in a causal
fashion, but that significant correlations between the two
reflect that temporal cardiac conditioning improves over
trials as does RT. For example, Porges found that the magni-
tude of anticipatory HR deceleration increased as a function
of trials in a RT situation. Moreover, even when the magni-
tude of the decelerative HR response was greatest, it was not
significantly correlated to RT. Porges contends that if mag-
nitude of HR deceleration reflects enhanced attention magni-

tude of HR deceleration should be strongly correlated with RT.

Although anticipatory HR deceleration prior to a
reaction signal is reliably observed in a RT situation, the
specific role of this deceleration is ambiguous. Lacey
and Chase‘33_;l. relate these changes to an attentional
process whereas Porges relates them to temporal condi-
tioning which is coincidental to but not strongly correl-
ated with motor performance. (A more detailed interpre-
tation of the temporal conditioning notion will be dis-
cussed later.)

One of the purposes of the present study was to
observe the HR activity of retardates in the fixed RT
situation. It was felt that such observation might provide
some insight into the nature of the retardate's alleged
attention deficit as well as contribute to increased under-
standing of their poorer RT performance. Moreover, it was
felt that the study of anticipatory HR deceleration in
retardates would contribute to the existing theoretical
frameworks concerned with directional HR changes: theoretics
which have been based almost exclusively upon data obtained
from normal individuals.

Heart rate acceleration, which is the second “limb"
of the HR response pattern during the PI, has received less
attention than anticipatory deceleration. However there
are data which suggest it has important behavioral sig-
nificance in that it may be related to task demands.

Heart rate acceleration typically occurs after the warning

signal, sometimes preceded by a brief deceleration (Chase

_e_1_:_a_1. , 1968; Headrick 5: Graham, 1969). Coquery & Lacey
(1966) reported that HR accelerations following the warn-
ing signal in a short PI (e.g., 4-secs) were attenuated
when compared to accelerations in PIs of longer duration
(e.g., 7-secs). ChaseIg§_§;. (1968) replicated Lacey's
findings in a 4-sec PI using a button-press PT task in one
group- of S}. However, they also employed a second group
of.§s whose RT task was leg-lifting rather than button-
pressing. In this second group, significant HR acceler-
ation followed the warning signal, presumably reflecting
the more strenuous demands of the leg lift task.

A second study which relates HR acceleration to
task demands was reported by Porges (1970). He found

that HR accelerated to RT warning signals and decelergted

 

to control non-signal stimuli. Control, non-signal stimp
uli were identical to the RT warning stimuli; the only
difference between them being the instructions that §_
read. To summarize, both the Chase.g§_gl. and the Porges
studies suggest that task demands are reflected in HR
acceleration following the warning signal; the more stren-
uous the task, the more marked HR acceleration becomes.
Another purpose of the present study then, was to
examine HR changes following the warning signal in retarded
and normal individuals in order to assess retardate's
acceleration responses in PIs of both short and long
duration as well as to do comparative analyses between

retarded and normal groups.

Pu se. The general purpose of the present study
was to evaluate the “attention deficit” frequently attri-
buted to retardates. Recent psychophysiological work
done with normals suggests that this alleged deficit would
be reflected in group differences in the anticipatory HR
deceleration occurring prior to the reaction signal as
well as in HR acceleration following the warning signal.

Another interest of this study centers upon possi-
ble differences between retardates and normals in both
acceleratory and decelatory phases of the HR pattern
during PIs of different lengths. Although RTs have not
been found to interact with PI length (Baumeister &
Kellas, 1968), a time estimation study by McNutt & Melvin
(1968) suggests that group differences in different PI
conditions might exist. These investigators report that
retardates were not significantly different from normals
in their ability to estimate a 5-sec tone, but had much
greater difficulty judging the length of a l3-sec tone.
Thus, PIs of 4, 7, and 13-secs were used in the present
study. These PIs are representative of short, moderate,
and long PIs as operationally defined by previous

investigators.

METHOD

Subjects. Subjects were 12 normal males and 12-
non-institutionalized retarded males, mean age 20 and 21,
respectively. The normal.§s, from Michigan State Univer-
sity, received extra course credit for their participation,
while the retarded.§s, from the Lansing metrOpolitan area,
were paid $2.00 for their participation. Retarded males
had a lechsler mean IQ of 70 (range 55 to 82). Subjects
had no sensorimotor impairments. Neither had they received
any drugs or medicatians for at least 2 weeks prior to the
experiment.

Apparatus. The RT stimulus was a green 24 V. DC
jewel light located about 3 feet in front of.§,at eye level.
A white light, located 2 in. below the green light, served
as the rest period stimulus and became illuminated only
between blocks of trials. The presentation of these
stimuli was controlled by 2 Hunter timers. The RT appar-
atus was a microswitch which was mounted in a wooden block.
The microswitch was placed on the arm of the chair in
which'§_was seated under‘gs preferred hand. In this
position,Ԥfs forearm and heel of the hand were supported
by the arm of the chair and'gfs index finger extended above
the microswitch. Reaction time was measured in milliseconds

by a Standard electric clock. Subjects were tested in a

10

11

sound-attenuated room where the temperature was maintained
at approximately 700 F. and the ambiant noise level was
51 db.

The physiological responses, RT stimulus, and RT
responses were continuously recoreded on a 4-channel Grass
P7 polygraph at a paper speed of lOmm/sec. Heart period
(HP) and skin conductance (SC) recording sites were cleaned
with 70% ethanol prior to the application of electrodes.
Grass gold disc electrodes, filled with Grass electrode
paste, were used to record HP from EKG Lead 1. Heart
period measures were fed into a Grass Model 7P6A EKG pre-
amplifier. Zinc cup electrodes with a surface area of
3.14 sq. cm. and filled with cotton soaked in a 1% ZnSo4
solution were used to record SC. Skin conductance elec-
trodes were placed on the base of the thumb and on the
inside of the forearm about 2 in. below the elbow of'gs non-
preferred arm.

Changes in respiration were measured by a pneumo-
graph that was fitted around.§s chest with a Velcro fastener.
The pneumograph consisted of 2 wire strain gauges that
were glued to a metal arch in a half-bridge configuration.
Matching dummy resistors were used for balancing the bridge.
Any changes in chest circumference resulted in changes in
cord length of the metal arch and thus changes in the
strain gauges. Strain gauge output was fed directly into
the polygraph.

Finger vasomotor activity was measured with a

12

photoelectric plethysmograph. The pickup consisted of a
block of black phenolic plastic in which were mounted a
Clairex C1704 photocell and a General Electric 683 minia-
ture lamp powered by a 3 V. battery. The pickup was taped
flush to the tip of the finger. This arrangement held the
surface of the pickup securely against'gs fingertip with
relatively light, constant pressure, which prevented dis-
comfort to.§_and the occluding of local blood vessels. The
signal from the plethysmograph bridge was recorded DC.

Procedure. All g; were individually tested in the
developmental psychophysiology laboratory at Michigan State
University. After arriving at the lab,Ԥ,was seated in a
comfortable armchair in a soundpattenuated room. The fe-
male experimenter attached the electrodes and briefly
explained their purpose. Heart rate, respiration, finger
vasomotor, and SC were measured, but because of mechanical
difficulties SC and vasomotor data were not scorable and
will not be reported here.

After the equipment was calibrated,I§.read the in-
structions. Bach'§.was told that a green light located 3
feet in front of him at eye level would come on periodp
ically. Onset of the green light marked the beginning of
the PI; this light remained illuminated for the entire PI.
Subjects were told that their job was to press the key
(i.e., the microswitch) as quickly as possible when the
green light went off. A white light, located 2 in. below

the green light, served as a rest period stimulus and

13

became illuminated only between blocks of trials.

Bach,§,participated in a single session which con-
sisted of 3 blocks of trials and 2 rest periods. A trial
block consisted of 15 trials of either 4, 7, or l3-sec
fixed PI. Only one PI value was used per trial block.

For example, an.§ assigned to a 4-7-13-sec PI combination
received the following sequence: 15 RT trials in which
the PI was 4-secs, a 2-min. rest period, 15 RT trials in
which the P1 was 7-sec, another 2-min. rest period, and 15
RT trials in which the P1 was 13-secs. Subjects were
randomly assigned to one of the six possible combinations
of the 3 trial blocks when they appeared at the lab. The
assignment of order was counterbalanced for each group so
that there was an equal number ofigs in each order. The
inter-trial-interval varied among 10, 15, and 20-secs.

The.§‘was alone during the entire experimental
session, but could communicate with‘EDby a talk-a-phone
intercom system. Five practice trials preceded the first
trial block. If.§,did not respond or responded too fre-
quently,'§,communicated this to him during the practice
period. At the end of practice, all‘gs were told that the
real test would now begin.

W. Only the last 10 trials in
each PI condition (or, block of trials) were scored for HP
and respiration. In the counter-balanced design, there
were 6 possible orders of presentation of the 3 PI condi-

tions. Given the assignment procedure, there were Z‘gp

 

I‘lllll‘l'llltfl

14

in each of the 6 orders in each group. Although a 2-min.
rest period intervened each block of trials, the small
number of gs in each call made it difficult to statistically
determine the effect of preceding conditions on subsequent
blocks of trials. Consequently, the first 5 trials in each
PI condition were omitted from the analyses in an attempt
to avoid possible transfer effects or disruptions from
preceding conditions.

Heart period (HP) was scored by determining the
number of milliseconds between successive heart beats, or
R—R intervals. The HP reading for each l-sec interval
consisted of the number of milliseconds between R-R inter-
vals which occurred during that sec. If more than one
R—R interval was completed during any given sec, only the
first cycle was scored for that sec. The HP data was
evaluated in separate sec-by-sec analyses for each PI
condition. Seconds which were analyzed included 4-secs
prior to PI onset, all secs during the PI, and 4-secs
following PI offset.

Respiration was analyzed by counting the frequency
of initiations and terminations of inspirations during
'pre', "PI“, and ”post“ periods for each of the 3 PI
conditions. In the 4-sec PI condition, the 'pre” period
consisted of the 4-secs prior to the onset of the light,
the "PI" period consisted of the 4-secs during light pres-
entation, while the ”post" period consisted of the 4-secs

following the light. In the 7-sec PI condition, the 'pre'

15

period consisted of the 7-secs preceding the light, the “PI“
period consisted of the 7-secs during light presentation,
while the “post” period consisted of the 7-secs following
the light. The l3—sec PI condition was analyzed in 2
different ways: In the first analysis, the 'pre" period
consisted of the l3-secs preceding the light onset, the
'PI' period consisted of the 13-secs during light presen-
tation, while the "post“ period consisted of the l3-secs
following the light. In the second analysis, the l3-sec

PI was analyzed.with the “pro” period consisting of the
6-secs preceding light onset, the "PI” period was broken
down into “early PI” and "late PI" components, with ”early
PI“ consisting of the 6-secs following light onset, and
“late PI" consisting of the 6-secs preceding light offset,
while “post“ period consisted of the 6-secs following light

offset.

RESULTS

Reaction Time. An analysis of variance was per-
formed on the RT data. This analysis revealed that re-
tardates had significantly slower RT than normals in all
three PI conditions (F(l,66)=17.79, p.<:0005). The
analysis of variance summary table is presented in Appendix
B. The group x PI interaction (F(2,66)<31.0) was not sig-
nificant. These results are consistent with the majority of
studies that have examined RT in retarded and normal in-
dividuals (Baumeister & Kellas, 1968). Mean RTs and
standard deviations for each group and for each PI are

presented in Table 1. These means were calculated on the

Table 1. Reaction Time: Mean and standard deviations
expressed in milliseconds for retarded and
normal groups in 4, 7, and l3-sec PI conditions.

 

 

 

 

2:222

Normals Retardates
PI 3192 SD Mean; SD
4 293.58 76.17 408.68 240.45
7 299.73 73.34 450.65 243.00
13 324.20 83.51 465.48 214.05

 

16

17

last 10 trials of each PI condition for eachԤ,. All 10
trials were included in the calculation of the means with
the exception of trials on which false responses occurred
prior to the reaction signal. Such false responding
occurred 5 times in the retarded group and 8 times in the
normal group. In each of these cases the mean RT score of
the other trials was assigned to the false response trial
in order to maintain equal frequencies in all cells.

Correlations Between RT and IQ Scores. Pearson
Product moment correlations between mean RT score and IQ
scores were performed for the retarded group. Wechsler
combined IQ scores were available for all 12 retarded'gs;
verbal and performance scores were available for all but
one of these.§p. These correlations are presented in

Table 2. A negative correlation indicates that low IQ is

Table 2. Pearson product moment correlations between total
IQ, verbal IQ, performance Wechsler IQ scores and
mean reaction times for 4, 7 and 13-sec PI
conditions for the retarded group.

 

 

A TOTAL IQ “if-10) VERBAL IQ (Of-9) PERFORMANCE IQ (Cf-9)

 

4 -.388 -.224 -.707*

7 -.384 -.351 -.782**
13 -e347 "e336 -e750**
* p4 .02

** p( .01

18

related to a slow RT. As can be seen from the table,
significant correlations were found between performance IQ
and RT.

Heart Period. An attempt was made to evaluate the
HP data in a comprehensive analysis of variance that in-
cluded the three PI conditions. Unfortunately this type
of analysis presented several inherent difficulties: In
order to have equal cell frequencies, only the first 4-
secs or last 4-secs in each PI was examined in the com-
prehensive analysis, resulting in a deletion of a large
amount of data. Moreover, the results of this analysis
revealed a number of 3 and 4-way interactions which,
though interpretable, left more questions unanswered than
answered. Therefore, to facilitate meaningful analysis of
the data the sec-by-sec HP data was analyzed separately
for each PI condition.

The analysis of variance results for the 4-sec PI
condition is presented in Figure 2. This figure shows HP
as a function of successive secs in the 4-sec PI for each
group. Since HP is the number of milliseconds between
successive heart beats, the larger values indicate a slower
HR, or HR deceleration, and smaller values indicate accel-
eration. These values are plotted so that a downward
slope indicates deceleration, and an upward slope indicates
acceleration. Successive secs are plotted along the
abscissa; secs -4 to -1 represent the prestimulus period,

or the 4-secs preceding the warning signal. Secs 1 to 4

19

' FIGURE 2

Mean heart period change as a function of successive
seconds in the 4-sec PI condition for normal and retarded
groups.

20

 

2 3 4+I9203o4 A

 

 

s
um
an
AR 0. 1
N 1|” “A d 1
m P 0". v
m m an .r 1
HI: 8 H... e1
R 4 o a 1
m as 1
m1 1
W V; 1
d 1
G l
\
a 1
P b F b h h P P L T b p b F r h F b h n P p p 'L
O O O O O 0 O 0 o 0 O O
O 2 4 6 8 0 2 4. 6 8 O 2
7 7 7 7 7 8 8 8 8 8 9 9
Amazoowmzfizv OOEMQ Pmdmx
‘III 20-hdzm4m004 ZOERKUJMUH: 111--1'

 

 

I POST

Pl

SUCCESSIVE SECONDS

-4 -3-2 -1
PRE

l

21

represent the PI, while secs +1 to +4 follow the reaction
signal. The analysis of variance revealed significant
group differences in overall HP base levels with retardates
having significantly slower HR (or, longer R-R intervals)
than normals (F(1,22)-4.20, p<.05). In the 4-sec PI,

the HR of normal‘gp was 78 beats per minute (bpm) while
the HR of retardates was 68 bpm. The significant (F(ll,
242)-1.92, p < .038) sec-by-sec HP pattern for combined
groups during the P1 was triphasic in nature: A slight

HP deceleration followed the warning signal followed by a
small acceleration which peaked on see 3 of the PI. This
acceleration was succeeded by a deceleration which reached
its nadir on sec +1, i.e., the sec on which the reaction
signal occurred. Heart period returned to base level
during the 4 secs following the reaction signal.

There was also a significant group x sec inter-
action (F(ll, 242)-4.22, p( .0005) which is illustrated
in Figure 2. Following the warning signal, both retarded
and normal groups showed a small deceleration followed by
an acceleration. However, in normal'gs this acceleration
was followed by a deceleration while retarded‘gs diduggg
evidence this decelerative HP response.

In order to assess the statistical significance of
the accelerative and decelerative components of the HP
response during the PI, t-tests for related measures were
performed between mean basal HP and HP acceleration score,

and mean basal HP and HP deceleration scores, for each

22

group, find for each PI condition. Mean base level HP was
determined for each'§,by computing the mean HP for the 4-
secs prior to PI onset. The greatest HP acceleration
occurring within the 4-sec PI was recorded as the HP
acceleration score for each‘ga while the HP deceleration
was socred in 2 ways: The first HP deceleration score (I)
was computed by recording the greatest deceleration which
occurred during a 5-sec period which included the last 3-
secs of the PI and the first 2 secs following the reaction
signal. Thus, in this analysis, the HP deceleration score
was based on the lowest HR (or, longest R-R interval) in
any one of 5 secs surrounding the reaction signal. The
second HP deceleration score (II) was computed by recording
the HP occurring during sec +1, or the sec on which the
reaction signal occurred. Thus, in this analysis, the HP
deceleration score was based on the HP of only one sec.

In order to compensate for the problem of correlated errors
inherent in nonpindependent t-tests, the apprOpriate alpha
level was set at .01 (2-tailed test).

The t-test analysis performed on the HP accelera-
tion score in the 4-sec PI condition revealed that the
normal.§fs HP acceleration following the warning signal
was not significantly different from.HP base level (t(ll)-
2.65, p<(.05), while the retardate's HP acceleration was
significantly different from HP base level (t(ll)-4.ll,
p<£.01). In both HP deceleration analyses, the normal

group's deceleration approached the predetermined significance

23

level (I:t(ll-2.999, p (.02; II:t(ll)-2.263, p1(.05),
whereas the retarded group's HP deceleration was clearly
not statistically significant (I:t(ll)-O.836, ns; II:t(ll)=
1.502, ns). Following the reaction signal, HP acceleration
-occurred in both groups, however, in the normal group it
approached prestimulus levels, while in the retarded group
it exceeded prestimulus levels. The analysis of variance
summary table for the 4-sec PI is presented in Appendix B.
The analysis of variance of HP during the 7-sec
PI condition revealed significant group differences in
overall base levels with retardates having significantly
slower HR (or, longer R-R intervals) than normals (F(l,22)-
5.78, p1<.025). This finding is like the 4-sec PI where
retardates also had slower HRs. In the 7-sec condition,
the normal's mean HR was 78 bpm and the retardate's mean HR
was 67 bpm. There were also significant trial effects
(F(9,l98)-2.50, p (.01) which indicated that HR increased
over trials for both groups (or, R-R intervals shortened).
As in the 4-sec PI, significant sec-by-sec HP
changes were found for combined groups in the 7-sec PI con—
dition (F(l4,308)-10.97, p1<.0005). One sec after the warn-
ing signal marked the onset of HP acceleration which peaked
on sec 4 of the PI. A deceleration followed which reached
its nadir on see +1, i.e., the sec on which the reaction sig-
nal occurred. After the reaction signal, HP increased,
reaching prestimulus levels 4-secs after the reaction signal.

As in the 4-sec PI condition, a significant group

24

x second interaction was also found in the 7-sec PI con-
dition (F(14,308)a4.69, p('.0005). This interaction is
illustrated in Figure 3. As can be seen from the graph,
both normals and retardates showed a significant acceler-
ation following the warning signal (normals: t(1l)-4.82,
p( .001); retardates: t(ll)-4.18, p( .01). In the normal
group, the acceleration was followed by a dramatic de-
celeration significantly different from base level (I:
t(ll)-6.57, p{.001; II: t(ll)-6.093, p< .001) which reached
its nadir simultaneously with the onset of the reaction
signal. On the other hand, retardates showed only an
attenuated deceleration, not significantly different from
base level (I: t(ll)-l.71, ns; II: t(ll)-O.382, ns), which
reached its nadir one see before the onset of the reaction
signal and was sustained through the sec on which the
reaction signal occurred. Following the reaction signal,
as in the 4-sec condition, normals showed an acceleration
back to prestimulus level, whereas retardates showed an
acceleration which exceeded prestimulus levels. The
Analysis of variance summary table for the 7-sec PI is
presented in Appendix B.

As in previous analyses, the analysis of variance
of the l3-sec PI condition revealed significant group
differences in overall HP base levels with retardates
having significantly slower HR (or, longer R-R intervals)
than normals (F(l,22)-4.9l, p( .037). In the 13-sec PI

the mean HR of normal'gs was 78 bpm while the mean HR of

25

FIGURE 3

Mean heart period change as a function of successive
seconds in the 7-sec PI condition for normal and retarded
groups.

26

REACTION

WARNING

F
O
0
7

I

POST

 

7 SEC. P.l.

H NORMALS
°"° RETARDATES
:4
l

l

L

Av
L

V 1.

p P p P [I'm

P p p p p p P p

. p
O 0
6 8
8 8 8

ooEwm km<mr
2255.586 1|I'

r h p
O O 0
2 6 8
7 7 7

Amazoowmfiji
All 26.53433

. p
0 0 O
2 O 2
8 9 9

740?
800 -

v

2 3 4 5 6 70|+2+3+4
Pl

SUCCESSIVE SECONDS

-4 -3-2 -|
PRE

 

27

retarded'és was 69 bpm. Also consistent with previous
analyses were significant seconds effects (F(20,440)=
14.99, p<(.0005) and significant group x seconds effects
(F(20,440)-3.16, p1<.0005). Second-by—second HP changes
for combined groups, i.e., seconds effects, were character-
ized by an acceleration beginning 1 sec after the warning
signal and peaking on sec 4 of the PI. A deceleration
followed which leveled off and became sustained from sec 9
to the end of the PI. Following the reaction signal, HP
gradually increased, reaching prestimulus base levels 4-
secs after the reaction signal.

The group by second interaction is illustrated in
Figure 4. Following the warning signal, the normal group
showed a significant acceleration (t(11)=5.72, p«<.001)
which peaked on sec 3 of the PI, followed by a significant
deceleration (I: t(ll)-7.23, p< .01; II: t(11)-3.228,
p1<.01) which leveled off and was sustained from sec 9 to
sec 13 of the PI. The reaction signal was followed by a
sharp acceleration which reached the prestimulus base level
4 secs after the reaction signal. Retardates also showed
a significant acceleration after the warning signal (t(ll)-
4.83, p<:.001) which was followed by a deceleration that
attained statistical significance in analysis I (t(ll)-
5.65, p<(.001), and approached statistical significance
in analysis II (t(ll)-2.890, p.<.02). The reaction signal
was followed by a slight deceleration and then an acceler-

ation. The analysis of variance summary table for the

28

FIGURE 4

Mean heart period change as a function of successive seconds
in the 13-sec PI condition for normal and retarded groups.

I

I POST

 

9
I

)3 SEC. RI.

e—e NORMALS
0'0 RETARDATES
d

I

REACTION

29

 

 

WARNING
O

n b

P

2 3 4 5 6 7 8 9 IOII I2I3+|+2+3+4
Pl

PRE

 

-4 -3-2-1

SUCCESSIVE SECONDS

.
o
O
8

.mozoommfida ooEmn .Edwr
All 2255684 26.23385 IIIV

740 ’
60 "
780 ’

b
O
2
7

700 .

7

30

13-sec PI is presented in Appendix B.

Figure 5 depicts HP change from prestimulus mean as
a function of seconds past warning light onset for each of
the 3 PI conditions, for both groups. These are the same
data as presented in the previous 3 figures, however, in
this case they are plotted as difference scores. Heart per-
iod for each second was subtracted from the prestimulus mean
to obtain a difference score. As in the previous figures,
an upward slepe indicates HR acceleration and a downward
slope indicates deceleration. These figures clearly show
the difference between groups in both response to the warn-
ing signal and in anticipatory HP decelerations preceding
the reaction signal for each PI condition. As described
previously, the statistical significance of responses to the
warning signal and in anticipation of the reaction signal
were assessed by a t-test for related measures between
mean basal HP and acceleratkn score, and mean basal HP and
deceleration scores. Both groups showed HR acceleration in
response to the warning signal however, the magnitude of
this acceleration differed between groups as a function of
PI. Normal.§s showed a non-significant acceleration from
prestimulus base level following the 4-sec PI warning signal,
a larger and significant acceleration in the 7-sec PI, and a
progressively larger and significant acceleration in the 13-
sec PI. Retardates, on the other hand, showed about the
same magnitude of acceleration in all three PI conditions,
all of them being significantly different from prestimulus

base levels. Group differences in HP deceleration

31

FIGURE 5

Mean heart period change from prestimulus mean as a
function of seconds past warning light onset in 4, 7,
and l3-sec PI conditions for normal and retarded groups.

HEART PERIOD CHANGE FROM PRE-STIMULUS MEAN (MILLISECONDS)

4 SEC
PI

7 SEC.
P. I.

I3 SEC.
R).

32

 

40 w nachos
I 6 p
30 P ‘ I ’d’ .—-—‘ NORUALS
20 1— 1' 0,44,.” one RETARDATES
. I
10L #1!"

  
 

- .0;
-20'
-30

..o 1
-50
-60

l
l
111l114_4

I
I
__I
I 2 3 4 0|020304
‘0 'wnume ancnou

301- 0
20» '

     
 
 

IO

 

 

~50
-60

 

 

0| 0 2 039‘
“:0?“

‘-

J:

 

- IO
-20
-30
~40.
-50
-60

I

 

 

“'1

[111L111]... .lle

)2 SQ O C? COUNCMQOFQ
SUCCESSIVE SECONDS PAST
WARNING LIGHT ONSET

 

 

33

prior to the response signal were also apparent for all
three PI conditions. In each case, normal'gs showed
significant deceleration just prior to and simultaneous
with the onset of the reaction signal; for the retarded
group only the deceleration in the l3-sec PI was
significant.

Correlations between RT and HP. Pearson product
moment correlations between RT and basal HP, RT and HP
acceleration score, and RT and HP deceleration scores were
performed for each group in an attempt to assess the re-
lationship between RT performance and HP activity. The
correlations between RT and basal HP for each PI, and for

each group are presented in Table 3. Mean basal HP was

Table 3. Pearson product moment correlations between
mean basal heart period and mean reaction time
score for 4, 7, and l3-sec PI conditions for
normal, retarded, and combined groups.

 

 

g; NORMALS (dfle) RETARDATES (at-10) COMBINED GROUPS (df-ZZ)

4 .171 .218 .369#
7 .231 .298 .438*
13 .306 .221 .396##

 

# p4.10
#1 .10<p> .05
* p( .05

34

calculated, as before, as the mean HP for the 4-secs prior
to PI onset for eachig, A positive correlation in this
case means that longer R-R intervals, or slower HRs, are
related to slower RTs. As can be seen from the table,
none of the correlations for the groups are significant.
However, when the correlation is performed on the combined
groups, it is significant in the 7-sec PI and approaches
significance in the 4- and l3-sec PI. These results are
consistent with the analyses of variance results which
showed that retardates had significantly slower RTs than
normals and also significantly slower HRs (or, longer R-R
intervals) than did normals.

In the analysis relating HP acceleration following
the warning signal to RT, the acceleration score (i.e.,
the greatest acceleration following the warning signal)
was subtracted from the mean basal HP score to get an
acceleration difference score. This acceleration differ-
ence score was correlated with the mean RT score for each
PI, for each group, and for the combined groups. NOne of
these correlations reached statistical significance; they
are presented in Table 4.

In the correlational analysis relating RT to the
HP deceleration occurring at about the time the reaction
signal occurred, HP deceleration difference scores I and
II were correlated with mean RT for each PI, for each group,
and for the combined groups. The correlations between RT

and HP difference scores I and II are presented in Table 5.

35

Table 4. Pearson product moment correlations between
mean HP acceleration score and reaction time for
retarded, normal, and combined groups, and 4,
7, and l3-sec PI conditions.

A_A

 

g; NORMALS (df-lO) RETARDATES (dfalO) COMBINED GROUPS (df-22)

4 -.041 e368 e373
7 -0148 -0100 "e072
13 .518 .047 .015

 

Table 5. Pearson product moment correlations between heart
period deceleration difference scores I and II
and mean reaction times for 4, 7, and 13-sec PI
conditions for normal, retarded, and combined
groups.

 

 

I
g; NORMALS (df-lO) RETARDATES (df-lO) COMBINED GROUPS (df-22)

 

 

 

 

4 .394 .619** .613***
7 .524## .310 .492*
13 .351 —.332 .105
II
§l_NORMALS (df-IO) RETARDATES—Idf-IO) COMBINED GROUPS (df-22)
4 .534## .485 .576**
7 .523## .449 .572**
13 .754** -.208 .213

 

IIISII1 .10 < 97 .05
t

p 4.05
** p< .01

Hit p < .001

36

A positive correlation indicates that a large HP deceler-
ation is related to a fast RT. Correlations between RT
and HP deceleration difference score I revealed signifi-
cant relationships between RT and decelerations in the 4-
sec PI for the retarded and combined groups, and in the 7-
sec PI for combined groups. This means that large decel-
erations occurring in the 5-secs surrounding the reaction
signal were significantly related to faster RTs in the 4-
sec PI condition for retardates, and in the 4 and 7-sec
PI conditions for combined groups.

Correlations between RT and HP deceleration dif-
ference scores II revealed significant relationships be-
tween RT and HP decelerations in the l3-sec PI condition
for normal‘gp and in the 7 and l3-sec PI condition for
combined groups. Correlations for the normals in the 7
and 4-sec PI conditions approached significance. These
correlations mean that large decelerations occurring on
sec +1, or the sec on which the reaction signal occurred,
were significantly related to faster RTs in the l3-sec PI
condition for the normal group and in the 7 and 4-sec PI
conditions for combined groups.

Respirption Frpgpenpy. Two of the retardate's
respiration records were rendered unscorable because of
mechanical difficulties. In order to have equal cell
frequencies in the analysis of variance, lO normals were
matched with the 10 retardates on the basis of PI order.

Consequently, the respiration frequency analyses of

37

variance used only 10.§p per group.

Respiration frequency was analyzed in 4 separate
analyses of variance for 4, 7, and l3-sec PI conditions.
In the 4-sec PI analysis, frequency of initiations and
terminations of inspirations were counted during the 4-
secs prior to PI onset, the 4-secs during the PI, and the
4-secs following the PI. These periods were labeled
”pre', "PI“, and “post“. As illustrated in Figure 6, the
groups x trials interaction was the only significant findp
ing in this analysis (F(9, 162)-l.991, p< .043). Both
normal and retarded groups showed about the same level
of respiration activity until trial 7. At this point,
the normal group's respiration decreased slightly from
initial levels and became sustained for the remaining
trials. The retarded group, on the other hand, showed a
slight decrease in respiration activity on trial 8, a
sharp increase immediately after trial 8, and continued
increase across the remaining trials. The analysis of
variance summary table is presented in Appendix B.

In the 7—sec PI analysis, the “pre' period con-
sisted of the 7-secs that preceded the PI, the I'PI" period
consisted of the 7-secs during the PI, while the "post“
period consisted of the 7-secs following the reaction
signal. The analysis of variance performed on this data
only revealed significant trial effects (F(9,162)-2.046,
p < .037); respiration frequency decreased across trials

for both groups. The analysis of variance summary table

38

FIGURE 6

Mean respiration frequency as a function of trials in the
4-sec PI condition for normal and retarded groups.

RESPIRATION FREQUENCY

|.8

L7

1.6

|.5

L4

|.3

I.2

 

39

4 SEC. P.|.

e—e NORMALS
o-o RETARDATES

 

40

is presented in Appendix B.

The l3-sec PI condition was analyzed in 2 ways:
In the first analysis, the “pre" period consisted of the
13-secs preceding the light onset, the “PI“ period con-
sisted of the l3-secs of light presentation, while the
“post“ period consisted of the l3-secs following the light.
In the second analysis, the l3-sec PI was analyzed with the
'pre' period consisting of the 6-secs preceding light onset,
the “PI“ period was broken down into “early PI“ and “late
PI” components, with “early PI“ consisting of the 6-secs
following light onset, and ”late PI“ consisting of the
6-secs preceding light offset, while the “post“ period
consisted of the 6—secs following light offset. There were
no significant effects in either analysis. The analyses
of variance summary tables are presented in Appendix B.

Correlations Between Mean Respiration Fregpenpy
ppd Mean HP. In order to assess the relationship between
respiration frequency and heart period, Pearson product
moment correlations were performed between mean respira-
tion frequency and mean HP for each group, combined groups,
each period, combined periods, and each PI condition. In
the 4-sec PI condition, mean respiration frequency was
calculated for each period (i.e., ”pre”, “PI", “post')
and for each'g, Overall respiration frequency was calcu-
lated by recording the mean respiration frequency for all
three periods for each‘g, The same procedure was followed

in the 7 and l3-sec PI conditions except that in the l3-sec

41

PI mean respiration frequency was examined for 4 periods
instead of 3.

In all three PI conditions, mean HP for the 'pre'
period consisted of the mean HP during the 4-secs preceedp
ing the warning signal. In the 4 and 7 sec PI conditions
mean HP for the PI period was based on the mean respiration
frequency during the PI; for the l3-sec condition, mean HP
during the “early PI' period was based on the mean HP
during the 6-secs following the warning signal while mean-
HP during “late P1" was based upon the mean HP during the
6-secs preceeding the reaction signal. In all three PI
conditions, mean HP during the “post“ period was based on
the mean HP during the 4 secs following the reaction
signal.

Correlations between mean respiration frequency
and mean HP are presented in Table 6. A negative correla-
tion indicates that slower HR.(Or, longer R-R intervals)
are associated with fewer respiration cycles. As can be
seen from the table, significant correlations were found
for the normal group in the 7 and l3-sec PI overall means,

and for normal‘gu in the l3—sec "early PI” period.

42

Table 6. Pearson product moment correlations between
mean respiration frequency and mean HP for
retarded, normal, and combined groups in 4,
7, and l3-sec PI conditions, and for 'pre',

"PI”, “post“ periods, and overall.

 

 

4 SEC PI CONDITION

 

 

 

 

 

g§§_ g;_ gpgg. OVERALL
NORMALS (df=8) -.399 -.358 -.103 -.390#
RETARDATES (df=8) .090 .114 -.052 .060
BOTH (dfalB) .067 .046 -.000 .056
7 SEC PI CONDITION
2.3.3. 21 m _____OVERALL
NORMALS (df=8) -.493 -.627# -.521 -.612***
RETARDATES (df=8) -.472 -.232 —.378 -.367
BOTH (dfalB) -.226 -.153 -.177 -.192
19 SEC PI CONDITION
25;, ERL”PI LTE PI gppg. OVERALL
NORMALS (df=8) -.540# -.638* -.469 -.364 -.620***
RETARDATES (df=8)-.418 -.405 -.413 -.301 -.394#
BOTH (df-18) -.323 -.311 -.226 -.185 -.243
# p1(.10
* p < .05

** p < .02
*** p < .01
“u p < .001

DISCUSSION

Correlations between IQ and RT. Correlations

between Wechsler performance IQ and RT scores were signif-
icant for the retarded group. Most other studies reporting
significant relationships between intelligence and RT
performance have used MA measures rather than IQ (Bens-
berg & Cantor, 1957; Berkson, 1960; Ellis & Sloan, 1957;
Pascal, 1953). The results of the present study suggest
that performance 10 might also be a good predictor of RT
performance.

Bpse Level HP. Base level HP was significantly
different between the groups in all three PI conditions,
with retardates having slower HP, or longer R-R intervals.
This represents a discrepency with published data (Clausen
& Karrer, 1970; Holloway & Parsons, 1970; Karrer, 1966;
Wallace & Fehr, 1970); a discrepency which unfortunately
has no obvious explanation since.§s had not taken drugs
for at least 2 weeks prior to the experiment. Moreover,
no retardated'§,was institutionalized and almost all were
employed at least part-time in the Lansing community. It
also seems unlikely that the difference in base level
reflects a motivational difference as the personal obser-
vations of the experimenter suggest that if anything, re-

tarded‘gs were more highly motivated than were the normal

43

44

ga. Almost all of the retarded.§s expressed a great deal
of excitement about coming to the University campus and to
the laboratory. They also seemed very eager to “please”
the female experimenter. For example, during the practice
session, these.§s frequently asked if their RTs were "fast
enough“, and if they were "doing ok'. It appeared to the
experimenter that frequent comments such as these reflected
a high degree of motivation to perform well. One plausible
explanation for the slower HR of retardates may lie in the
sample used. Most previous studies that reported no dif-
ferences between base level HP of retardates and normals
employed retardates sampled from pepulations with 10s
that were lower than 55. Since none of the retardates in
the present sample fell below this IQ level the possibility
exists that base level varies as a function of IQ level.

Sec-by~Sec HP: Anticipatopy HP Decelerption. The
present study found significant second effects, or sec-by-
sec HP Changes for combined groups, in all three PI con-
ditions. These data will not be discussed in detail here
as they are not central to the purpose of this paper. The
main interest of the present study was to obtain informa-
tion on differences between normal and retardedlgs--effects
that are interactive in nature. Consequently, main effects
such as the seconds effects are only of peripheral interest
at this time.

Significant group differences in sec-by-sec HP

responding were found in the HP analysis. A dramatic

45

difference in HP deceleration at about the time the reac-
tion signal was to occur was found between retarded and
normal groups. In all three PI conditions, normals showed
a deceleration which reached its nadir just prior to or at
the time the reaction signal was to occur, while retardate's
HP pattern prior to the reaction signal was markedly dif-
ferent: In the 4-sec PI condition retardates showed a
complete absence of HP deceleration, while in the 7—sec PI
condition their deceleration was severely attenuated. In
the l3-sec PI condition retardates exhibited a significant
deceleration, but the pattern of this deceleration was
very irregular and markedly different in form from the
response pattern of normal.§s.

Such data would seem to demonstrate the retardate's
inability to prepare for appropriate responding regardless
of the length of the preparation time. If one accepts the
assumption that HP measures during the PI reflect an
attention process, the data clearly support the idea that
the retardate suffers from an attention deficit. Although
this notion is not a novel one in retardation research,
little real empirical evidence has been advanced to support
it. Net so incidentally then, the results of the present
study underlie the potential value of psychOphysiological
techniques for studying the capabilities of retardates.

The correlational analyses lend some support to
the notion that faster RTs are related to greater HP de-

celerations which occur prior to and during the reaction

46

signal. The HP deceleration score II, or the deceleration
which occurred on sec +1, was significantly related to RT
in the l3-sec PI condition for the normal group, and in the
7 and 4-sec PI conditions for the combined groups. Cor-
relations in the 7 and 4—sec PI conditions also approached
significance for the normal group. The specific role that
this anticipatory HP deceleration plays in RT performance
and attention has been interpreted in several different
frameworks.

Lacey (1967), for example, believes that the occur-
rence of HP deceleration facilitates the intake of informa-
tion, and hence, prepares the individual to deal with the
environment in an efficient manner. In the case of the
RT task, the enhanced attention reflected by HR decelera-
tion would enable the organism to respond more quickly.

The absence of deceleration in retardates would indicate
that retardates are less sensitive to environmental stimuli.
In the present study, a lowered sensitivity to the reaction
signal would explain the poor RT performance of the
retardate.

It is also possible to interpret the role of anti-
cipatory HP deceleration in terms of an inhibition deficit
hypothesis such as that advanced by Denny (1964). This
interpretation holds that when all task-irrelevant or com-
peting activities, including irrelevant physiological

activities, are suspended or inhibited, a fast reaction

47

time should result. This view is supported by the find—
ings of Obrist who has shown that a number of physiolog-
ical activities including muscle tension, eye movements,
respiration, as well as HR, markedly decrease simultan-
eously during a RT situation at about the time.§ is to
make a response (Obrist‘pp_pl., 1969b). The magnitude of
these decreases is directly related to RT performance.

In this framework, the retardate's primary deficit is
explained as an inability to inhibit or suspend activities
which do not contribute to good RT performance. In the
present study, the absence of HR deceleration prior to
the reaction signal can be interpreted to reflect the
retardate's deficit in the ability to suspend or inhibit
ongoing activities which interfere with the ability to
respond quickly.

A third interpretation could be based on the cardiac
temporal conditioning notion advanced by Fitzgerald & Porges
(1970), Grossman, Fitzgerald, & Porges (19700, and Porges
(1970). In this framework, the absence or attenuated
anticipatory HR deceleration in the retarded group would
reflect an absence of temporal conditioning, whereas the
dramatic decelerations which were observed in the normal
group would presumably reflect successful temporal con-
ditioning. This notion is not incompatible with Denny's
contention that retardates suffer from an inhibition

deficit. It could be that retardates are unable to make

48

use of the temporal cue because they lack the ability to
inhibit competing responses. Normal fig, on the other
hand, inhibit competing responses and simultaneously
attend to the signal value of temporally fixed stimulus
events.

Group x trial x second effects were not significant
in the present study, probably because of the deletion of
the first 5 trials. These early trials were eliminated
to avoid transfer and possibly disruption effects from
prior conditions which were different for eachug. Conse-
quently, a conditioning interpretation could not be direct-
ly assessed from the reported data. However, the possi-
bility of group differences in temporal conditioning
suggests future research which would be potentially rele-
vant to the understanding of neural mechanisms in both
retarded and normal individuals.

In a literature review on neural timing mechanisms
and conditioning, Prescott (1966) points out that use of
the classical conditioning paradigm in studying the pre-
cision of neural timing mechanisms provides a valuable
technique for research in developmental processes. He
suggests that an organism capable of inhibiting a response
until the exact moment of reinforcement would reflect a
neural system characterized by high precision and effi-
ciency. Increased error in timing, on the other hand,
would reflect a poorly integrated and biologically non-

adaptive neural system. Hence, it appears as though future

49

research designed to investigate differences in neural
timing mechanisms through the established conditioning
paradigms would have important developmental implications.
In addition to Prescott, a number of other investigators
(Brackbill & Fitzgerald, 1969; Fitzgerald & Brackbill,
1971; Fitzgerald & Porges, 1971) have made a similar case
for the use of classical conditioning to study develop-
mental phenomena including temporal.

Sec-byeagg_HP: Acceleration following the warning
Signal. Another important group difference in sec-by-sec

 

HP responding was in response to the warning signal. On-
set of the warning light is an important signal to.§ as it
indicates the beginning of a trial: this is the point at
which'g’is instructed to begin paying attention. Both
groups primarily showed HP acceleration in response to the
warning light: however the magnitude of this acceleration
differed for groups as a function of PI.

Normals responded to the warning signal differently
for each P1, with progressively greater accelerations for
the longer PIs. Retardates, on the other hand, showed the
same magnitude of acceleration for each PI. This is an
interesting difference as the warning signal onset was
identical during each PI; the only difference between PIs
was the length of time that the light remained illuminated.
These data lend support to a contention made by Baumeister
& Kellas (1968) who tentatively hypothesized that retardates

as a group have an inappropriate response set in the RT

SO

situation such that their attention is directed toward the
reception of stimuli rather than performance of the re-
sponse, in spite of instructions designed to minimize such
a set. Normals, on the other hand, maintain a response
set as long as the stimulus intensity is above some mini-
mum value. The appearance of a large acceleration at the
onset of stimulation in the retarded group would seem to
support this hypothesis.

Denny's inhibition-deficit hypothesis (Denny,
1964) would extend the Baumeister & Kellas notion by
suggesting that these data reflect the greater ability
of normals to inhibit competing responses which interfere
with a fast RT. For example, in the 4-sec PI, normals
showed a nonsignificant acceleration following the warning
signal. In this short PI a large acceleration and its
recovery would interfere or compete with the relevant
response--which is anticipatory HR deceleration. In
longer PIs, where there is adequate time for larger accel-
erations and their recovery, the normals exhibited sig-
nificantly greater accelerations. This is not the case
with the retarded‘gs who showed similar magnitude accel-
erations following the warning light in all 3 PI conditions.
In the 4-sec PI, retardates showed a significant acceler-
ation which was sustained for the entire PI. Hence, this
is a second piece of evidence supporting the idea that
retardates are unable to inhibit task-irrelevant or

competing responses.

51

Another way of looking at this early acceleration
which follows the warning signal is in terms of‘gs per-
ceiving the environmental demands of the situation. The
4, 7, and lB-sec PI vary in the amount of effort needed to
sustain attention, with the greatest amount of effort
needed to sustain attention for 13-sec and the least amount
of effort needed to sustain attention for 4—secs. If this
is the case, the small acceleration in the normal group
to the 4-sec PI warning signal would reflect the relatively
small amount of effort needed for fast RT performance.
Progressively greater accelerations in the longer PIs
would presumably reflect the higher instrumental demands
of these situations. In the case of the normal.§s, then,
onset of the warning light not only signals that they
must attend, but also signals the degree of effort needed
to attend. This interpretation lends considerable support
to the conclusions of Chase.3§;g;. (1968) and Porges (1970).

The consistent response of retardates in the present
study indicates that they are perceiving the warning light.
However, the lack of any differential responding to light
onset as a function of PI length suggests that they are
not relating onset of the light to its subsequent length
and consequent task difficulty. This apparent lack of
association between differing environmental demands and HR
change supports Holloway a Parsons (1970) notion that some
groups of retardates suffer from a disruption of the

integration between somatic and autonomic activity. They

52

prOpose that physiological responsiveness in these retard-
ates becomes dissociated from environmental demands and
that such dissociated activity reflects a source of inter-
ference with‘gs ability to attend to external signals or
to efficiently execute the appropriate response.

Sec-by-Sec HP: Overall Response Patterns. A
striking feature of the sec-by-sec HP data is that as PI

 

interval increases, the HP pattern of the retarded group
comes to more closely approximate the pattern of the normal
group. In the 4-sec PI condition, the group curves are
quite disparate, with the retardates exhibiting a progres-
sive acceleration and the normal.§s exhibiting a triphasic
pattern: deceleration followed by acceleration and a

final deceleration. In the 7-sec PI condition, there was
more overlap between the curves, but there was a sharp
divergence between them at about the time the reaction
signal occurred. At this point the normal group exhibited
a sharp deceleration while the retarded group exhibited an
attenuated deceleration. In the 13-sec PI condition, the
group curves overlapped for the first 6-secs of the PI

and then diverged, but to a lesser extent than in the
shorter PI conditions. The HP pattern of the normal group
at the point of divergence was characterized by a large,
sustained deceleration. The HP pattern of retardates at
this point was one of progressive and significant deceler-
ation, but a smaller deceleration than the normal's and one

which lacked the smooth, non-variable appearance of the

53

normal's pattern.

The progressive resemblance of the retardate's HP
pattern to that of the normal's as a function of PI can
be explained in several ways. One explanation is that
retardedigs need a longer P1 in order to recover from the
larger acceleration they exhibit and to show the subse-
quent deceleration which characterizes the normal's HP
pattern. A short PI would not provide adequate time for
recovery, but as the time interval increased, recovery
from acceleration could occur making it possible to be
followed by a deceleration. Or, in the Baumeister &
Kellas (1968) framework, a longer PI would provide the
opportunity for the retardate to transfer his attention
from the reception of stimulation to the execution of a
response.

Perhaps another alternative is that it simply takes
longer for the retardate to integrate incoming information.
In the longer PI, the longer time interval would allow the
retardedԤ,more time to integrate and.process information.

In any case, the progressive similarity in HP
patterns between the two groups as a function of longer
PIs is suggestive of future research which would poten-
tially contribute some meaningful information about re-
tardate processes. Systematic studies which explore the
conditions under which retardate HP patterns resemble
normal patterns could have important implications for

understanding the learning process in retardates. For

S4

example, it could be that Optimal learning in retardates
requires a sustained stimulus beyond some minimal time
limit. In the present study RT performance did not inter-
act with PI length, but as mentioned previously, the design
of this study was not suitable for the assessment of
learning effects. Also, a classical conditioning paradigm
might be more suitable for this kind of question. Perhaps
a study employing many trials and independent groups in
several conditions would provide a more direct evaluation
of these hypotheses.

Respiration Frequency. Analyses of variance of
the respiration frequency data did not reveal group
differences corresponding to group differences found in
the HP analyses. In the respiration frequency analyses,
only a trial x group interaction in the 4-sec PI and a
trials main effect in the 7-sec PI condition attained
statistical significance. This absence of any group or
period differences suggests that respiration frequency
measures are less sensitive to environmental demands than
are HR measures.

Correlations between mean HP and mean respiration
frequency revealed significant relationship between these
measures for normal.§s in the 7 and lB-sec PI conditions.
Two of the three significant correlations were in overall
means and not specific periods, which is similar to the
analysis of variance results.

Conclusion. The present study succeeded in

55

finding significant differences between retardated and
normal individuals in both RT performance and in sec-by-
sec HP changes--two measures which have often been asso-
ciated with the attention process. Thus, this study
provides support to the notion that retardates suffer from
an attention deficit. An attempt was made to relate the
observed differences to a broader theoretical conception
of the attention process. A number of diverse interpre-
tations were discussed, all of them being feasible ex-
planations of the reported data. Of course, the number of
these explanations and their diversity both reflects the
level of understanding that exists regarding the attention
process and also points to the need for systematic in-
vestigations for greater understanding of these processes.
Suggestions made in the discussion were aimed at providing
direction for future research in an attempt to accomplish
this goal. Implied in these suggestions is the potential
value inherent in employing retarded‘gs in such investiga-
tions. The present study demonstrates that the inclusion
of retarded individuals in the study of the attention
process provides much needed empirical data upon which

future theoretics can be based.

REFERENCES

REFERENCES

Baumeister, A. A., & Kellas, G. Reaction time and mental
retardation. In N. Ellis (Ed.), Internationgl
Review of Research in Mental Retardation, Volume
3. New York: Acad c Press, 1968. pp. 163-193.

Bensberg, G. J., & Cantor, G. N. Reaction time in mental
defectives with organic and familial etiology.
American Journal of Mental Deficiency, 1957,.gg,
5342537.

Berkson, G. An analysis of reaction times in normal and
mentally deficient young men.. III. Variation of
stimulus and response complexity. Journal of

Mental Deficiency Research, 1960,.3, 69—77.

Brackbill, Y., & Fitzgerald, H. E. Development of the
sensory analyzers during infancy. In L. P.
Lipsitt and H. W. Reese (Bds.), Advances in Child
Development and Behavior, Volume 4. New‘York:
Academic Press, 1969.

Brener, J. Heart rate. In P. H. Venables a I. Martin
(Bds.), Manual of PsychOphysiological Methods.
New York: John Wiley & Sons, 196 . pp. 103-130.
Chase, H. 6., Graham, F. M., & Graham, D. T. Components of
the heart rate response in anticipation of reaction

time and exercise tasks. Journal of E erimental
Psychology, 1968, zg, 642-348.

Clausen, J., & Karrer, R. Autonomic activity during rest
in normal and mentally deficient subjects.

American Journal of Mental Deficiency, 1970,.1§,

Coquery, J. M., & Lacey, J. I. The effect of foreperiod
duration on the components of the cardiac response
during the foreperiod of a reaction-time experiment.
Paper delivered at the Annual Meeting of the
Society for Psychophysiological Research, 1966.

Denny, M. R. Research in learning and performance. In
H. A. Stevens & R. Heber (Eds.), Mental Retardation:
A Review of Research. University of C cago Press,

1964.

S6

57

Ellis, N. R., & Sloan, W. Relationship between intelli-
gence and simple reaction time in mental defectives.

Perceptual and Motor Skills, 1957,.1, 65-67.

Fitzgerald, H. R., & Brackbill, Y. Tactile conditioning
of autonomic and somatic response in young infants.
Conditional Reflex, 1971,.g, 41-51.

Fitzgerald, H. E., & Porges, S. w. Cardiovascular effects
of paced respiration and selective attention.

Psychonomic Science, 1970,.12, 65-66.

Fitzgerald, H. E., & Porges, S. H. A.decade of infant
conditioning and learning research. Merrill-

Palmer gparterly of Behavior and Development,
1971, _l_, 79-11 .

Graham, F. R., & Clifton, R. K. Heart rate change as a
component of the orienting response. Psychological
Bulletin, 1966,.§§, 305-320.

Grossman, R., Fitzgerald, H. E., & Porges, S. W. Sex
differences in responding forearm activity during
a simple visual reaction time task. Unpublished
manuscript, Michigan State University, 1971.

Hastings, 8., & Obrist, P. A. Heart rate during condition-
ing in humans: Effect of varying the interstimulus
(CS-UCS) interval. Journal of Eyperimental
Psychology, 1967,'15, 431-442.

Headrick, M. N., & Graham, F. K. Multiple component heart
rate responses conditioned under paced respiration.

Journal of Eyperimental ngchology, 1969,.12,
486-494.

Holloway, F. A., & Parsons, 0. A. Physiological con-
comitants of reaction time performance in normal
and brain-damaged subjects. Paper delivered at
the Annual Meeting of the Society for Psycho-
physiological Research, 1970.

Karrer, R. Autonomic nervous system functions and behavior:
A review of experimental studies with mental
defectives. In N. R. Ellis (Ed.), International
Review of Research in Mental Retardation, Volume 2,
New York: Academic Press, 1966.

Lacey, J. I. Somatic response patterning and stress:
Some revisions of activation theory. In M. H. Appley

& R. Trumbull (Eds.), Psychological Stress. New
York: Appleton-Century-Crofts, l9 . pp. 14-36.

58

Lacey, B. C., & Lacey, J. I. Change in cardiac response

McNutt,

Obrist,

Obrist,

Obrist,

Pascal,

Porges,

and reaction time as a function of motivation.
Paper delivered at the Annual Meeting of the
Society for Psychophysiological Research, 1966.

T. H., & Melvin, K. B. Time estimation in normal
and retarded subjects. American Journal of Mental
Deficiency, 1968,.12, 584-587.

P. A., Sutterer, J. R., & Howard, J. L. Prepara-
tory cardiac changes: A psychobiological approach.
Paper presented at the conference on Classical
Conditioning, McMaster University, Hamilton,
Ontario, May, 1969 (a).

P. A., Webb, R. A., & Sutterer, J. R. Heart rate
and somatic changes during aversive conditioning
and a simple reaction time task. PsychOphysiology,
1969, 5“ 696-723. (b).

P. A., Webb, R., Sutterer, J. R., & Howard, J. L.
Cardiac deceleration and Reaction time: An evalua-
tion of two hypotheses. PsychOphysiology, 1970,

_§_, 695-706.

G. R. The effect of a disturbing noise on the
reaction time of mental defectives. American

Journal of Mental Deficiency, 1953,'§1, 691-699.

S. W. Respiratory and heart rate indices of
reaction time performance. Unpublished Ph. D.
Thesis, Michigan State University, 1970.

Prescott, J. W. Neural timing mechanisms, conditioning,

and the CS-UCS interval. Psychophysiology, 1966,
_2_, 125-131.

Wallace, R. K., & Fehr, F. 5. Heart rate, skin resistance,

and reaction time of mongoloid and normal children
under baseline and distraction conditions.

Psychophysiology, 1970.9, 722-731.

Webb, R. A., & Obrist, P. A. The physiological concomitants

Wilson,

of reaction time performance as a function of
preparatory interval and preparatory interval
series. PsychOphysiolggy, 1970,-g, 389-403.

R. 8. Cardiac response: Determinants of condi-

tioning. Journal of Comparative and Physiological
Psychology Monograph, 1969,.§§, No. 1, Part 2.

59

Wood, D. M., & Obrist, P. A. Effects of controlled and

Zeaman,

uncontrolled respiration of the conditioned heart

rate response in humans. Journal of Experimental
Psychology, 1964,.§§, 221-229.

D., & Smith, R. Review of some findings in human
cardiac conditioning. In W. F. Prokasy (Ed.),

Classical Conditioning: A Sypposium. New York:

Appleton-Century-Crofts, 1965. pp. 378-418.

APPENDICES

60

APPENDIX A

INSTRUCTIONS FOR REACTION TIME TASK

61

62

The pickups I have attached will record changes in
heart rate, sweating, and also changes in the blood vessels
in your finger. The elastic band around your chest records
changes in your breathing. These electrodes are very
sensitive so please try to keep movement at a minimum.

Get comfortable before the experiment begins.

This green light Q§.points to green light) will
come on periodically. Your job will be to watch this green
light ߤ,continues to point to light) and to press the
key when the light goes off. Try pressing the key now.

It is important that you press the key as rapidly as
possible when the green light disappears. Be sure not to
press the key before the light goes off. Remember not to
make any unnecessary movements and to respond as rapidly
as possible when the green light (§_points to green light)
disappears.

You will receive 2 rest periods during the experi-
ment. During these rest periods you can just relax. They
will be designated by the white light (§.points to white
light). When this white light (§.continues to point to the
white light) goes on it is a rest period and you can just
relax.

We will begin with a series of practice trials.
During these practice trials you can ask questions by
simply talking in a normal tone of voice. I will be able
to hear you in the next room and give you an answer. I

will tell you when the practice trials are over and the

63

real test is about to begin.

Remember, your job is to press this key (§,points
to key) as quickly as possible when the green light dis-
appears. IT IS IMPORTANT THAT YOU REMEMBER THIS IS A
TEST OF SPEED!

Do you have any questions?

If not, we will begin the practice trials.

APPENDIX B

ANALYSIS OF VARIANCE SUMMARY TABLES

64

65

Summary of the analysis of variance of reaction time as a
function of groups, trials, and PI conditions (4, 7, and

 

lB-secs).

_SOURCB .55. _f. 14.8. 2: 2
Groups 3317865.800 l 33l7865.800 17.789 (0.0005
PI 230030.486 2 115015.243 0.617 0.543
Group x PI 41197.858 2 20598.929 0.110 0.896
Error 12310136.183 66 186517.215

Trials 136329.939 9 15147.771 1.044 0.404
Group x Trials 68170.311 9 7574.479 0.522 0.860
PI x Trials 148320.986 18 8240.055 0.568 0.923
G x PI x T 235286.947 18 13071.497 0.901 0.578
Error 8621826.819 594 14514.860

Total 25109165.328 719

 

 

Summary of the analysis of variance of heart period as a
function of groups, trials, and seconds for the 4-sec PI

 

Total

558377.372 2879

condition.

_SOURCB 9.5 .92 3.15 5. 2
Groups 72631.378 1 72631.378 4.201 0.053
Error 380355.619 22 17288.892

Trials 836.855 9 92.984 0.791 0.625
Group x Trials 1142.743 9 126.971 1.080 0.379
Error 23276.860 198 117.560

Seconds 940.076 11 85.461 1.919 0.038
Group x Seconds 2065.359 11 187.760 4.216 <0.0005
Error 10777.440 242 44.535

Trials x Seconds 2604.566 99 26.309 0.927 0.682
G x r x 8 1908.311 99 19.276 0.679 0.993
Error 61838.165 2178 28.392

 

 

66

Summary of the analysis of variance of heart period as a
function of groups, trials, and seconds for the 7-sec PI

 

condition.

___._SOURCE .59. 9.: 11.5. .1: 2
Groups 123575.684 l 123575.684 5.778 0.025
Error 470538.271 22 21388.103

Trials 2923.360 9 324.818 2.502 0.010
Groups x Trials 1635.582 9 181.731 1.399 0.190
Error 25707.618 198 129.836

Seconds 7042.416 14 503.030 10.968 (0.0005
Groups x Seconds 3011.849 14 215.132 4.6913(0.0005
Error 14126.362 308 45.865

Trials x Seconds 3663.740 126 29.077 1.032 0.389
G x T x-S 3060.218 126 24.287 0.862 0.862
Error 78124.082 2772 28.183

Total 733409.183 3599

 

 

Summary of the analysis of variance of heart period as a
function of groups, trials, and seconds for the l3-sec PI

 

condition.

_SOURCB 5.2 .4; 11.9. .I: 2
Groups 125790.087 l 125790.087 4.914 0.037
Error 563109.346 22 25595.879

Trials 704.647 9 78.294 0.482 0.885
Group x Trials 1473.458 9 163.718 1.009 0.434
Error 32138.690 198 162.317

Seconds 11268.616 20 563.431 14.992 (0.0005
Group x Seconds 2376.700 20 118.835 3.162 (0.0005
Error 16535.846 440 37.581

Trials x Seconds 5578.840 180 30.994 1.145 0.094
G x T x S 5323.629 180 29.576 1.093 0.193
Error 107189.035 3960 27.068

Total 871488.900 5039

 

 

67

Summary of the analysis of variance of respiration frequency
as a function of groups, trials, and periods in the 4-sec PI

 

condition.

_SOURCB 9.8. 8:. 1e 1?. 2
Groups 2.042 1 2.042 1.455 0.243
Error 25.257 18 1.403

Trials 2.216 9 0.246 1.211 0.291
Groups x Trials 3.642 9 0.404 1.991 0.043
Error 32.910 162 0.203

Periods 3.000 2 1.500 2.725 0.079
Groups x Periods 0.053 2 0.027 0.048 0.953
Error 19.813 36 0.550

Trials x Periods 5.200 18 0.289 0.994 0.466
G x T x P 5.813 18 0.322 1.112 0.339
Error 94.120 324 0.290

Total 194.065 599

 

 

Summary of the analysis of variance of respiration frequency
as a function of groups, trials, and periods in the 7-sec PI

 

condition.

___SOURCB 5.5 9.5 11.9. 2'. 2
Groups 9.375 1 9.375 1.494 0.237
Error 112.950 18 6.275

Trials 4.875 9 0.542 2.046 0.037
Groups x Trials 2.875 9 0.319 1.207 0.294
Error 42.883 162 0.265

Periods 1.030 2 0.515 1.263 0.295
Groups x Periods 0.090 2 0.045 0.110 0.896
Error 14.680 36 0.408

Trials x Periods 4.770 18 0.265 0.971 0.493
G x T x P 7.710 18 0.428 1.5700 0.066
Error 88.387 324 0.273

Total 289.625 599

 

 

68

Summary of the analysis of variance of respiration frequency

 

as a function of groups, trials, and 'pre", “PI", and “post“
periods in the 13-sec PI condition.

____8<>UR<=8 .§.§. _f. 14.8. 1:. 2
‘Groups 4.335 1 4.335 0.367 0.552
Error 212.497 18 11.805

Trials 4.948 9 0.550 0.953 0.481
Groups x Trials 4.148 9 0.461 0.799 0.618
Error 93.470 162 0.577

Periods 0.663 2 0.332 0.671 0.517
Groups x Periods 0.610 2 0.305 0.617 0.545
Error 17.793 36 0.494

Trials x Periods 7.437 18 0.413 0.791 0.711
G x T x P 6.357 18 0.353 0.676 0.834
Error 169.140 324 0.522

Total 521.398 599

 

 

Summary of the analysis of variance of respiration frequency
as a function of groups, trials, and “pre', “early PI",
“late PI“, and “post" periods in the l3-sec PI condition.

 

_..___80UR<=8 _.§. 92:. .88. E. 2
Groups 4.061 1 4.061 0.954 0.342
Error 76.653 18 4.258
Trials ‘ 1.051 9 0.117 0.361 0.952
Groups x Trials 2.401 9 0.267 0.825 0.594
Error 52.373 162 0.323
Periods 0.794 3 0.265 0.718 0.545
Groups x Periods 0.704 3 0.235 0.637 0.594
Error 19.878 54 0.368
Trials x Periods 6.994 27 0.259 1.054 0.392
G x T x P 8.484 27 0.314 1.279 0.160
Error 119.398 486 0.245

Total 292.789 799

 

 

                    

HICHIGRN STRT

E UNIV. LIBRARIES
I I [III WWI“\IIIHIIHWWWI
9 2255670

13122 23110