THE ACQUISITIONVAND REDUCTION OF
RESPONDVING IN RATS T0 AUDITORY‘
STIMULI OF DIFFERENT INTENSITIES

Thesis for the Degree of Ph. D.
MICHIGAN STATE UNIVERSITY
EARL DONALD WALKER
1972

LIBRARY

Michigan State:
University

This is to certify that the

thesis entitled

THE ACQUISITION‘AND REDUCTION OF
RESPONDING IN RATS TO AUDITORY STIMULI
OF DIFFERENT INTENSITIES

presented by

Earl Donald Walker

has been accepted towards fulfillment
of the requirements for

Doctoral degree in Psygchology

 

M11

Major professor

Date W ell/Q 7 2

0-7639

1’ amnmc BY ‘5'

‘ "“138“? X SUNS'
BOOK BINDERY WC ::

.UBRARY swat; _ M

 

     
       

ABSTRACT

THE ACQUISITION AND REDUCTION OF RESPONDING IN RATS
TO AUDITORY STIMULI OF DIFFERENT INTENSITIES

BY

Earl Donald Walker

Auditory intensity discrimination training involv-

ing two intensities of a pure tone and a "zero" intensity
condition produced by the absence of the tone was given to
two groups of rats that differed with respect to whether
responding to the more or the less intense value of the
tone was reinforced, while the remaining tone intensity
for each group and the zero intensity condition were non-
reinforced. It was found that reinforced responding
ultimately reached higher levels in the presence of the
more intense stimulus, while the training to reduce non-
reinforced responding proceeded more quickly for the less
intense stimulus but did not progress at different rates
for the two groups during the zero stimulus condition.
The net effect was a large overall superiority in discrim-
ination learning for the group with the more intense stim-
ulus reinforced. The results by themselves and in combin-
ation with previous findings appear to offer more support
for a generalization of inhibition interpretation than

for a strict stimulus intensity dynamism position.

Earl Donald Walker

Ace/m

Robert L. Raisler
Thesis Director

Mark E. Rilling
M. Ray Denny
Gordon Wood
Committeemen

Date W‘é/771

THE ACQUISITION AND REDUCTION OF RESPONDING IN RATS
TO AUDITORY STIMULI OF DIFFERENT INTENSITIES

BY

Earl Donald walker

A THESIS

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

DOCTOR OF PHILOSOPHY

Department of Psychology

©Copyr ight by
EARL DONALD WALKER
1972

To Robert L. Raisler

ACKNOWLEDGMENTS

My sincere thanks to Dr. Rdbert L. Raisler, Dr. Mark
E. Rilling, Dr. M. Ray Denny, Dr. Gordon Wood, and Dr.
Stanley C. Ratner for their advice and support. I am
particularly grateful to Bob Raisler, my many-years dir-

ector, counselor, and friend.

iii

TABLE OF CONTENTS

Page

LIST OF FIGURES ................................ v
INTRODUCTION ................................... 1
METHOD ......................................... 2
Subjects .................................... 2
Apparatus ................................... 2
Procedure ................................... 3
RESULTS ........................................ 6
DISCUSSION ...................................... 14
REFERENCES ..................................... 20

iv

Figure

LIST OF FIGURES

Page

Percent response during the first 5 sec.
after stimulus onset to 72 db and 64 db
intensities of a 5000 Hz tone for one

group of rats given discrimination train-

ing with the louder stimulus reinforced

(72+) and another group trained with the
softer stimulus reinforced (64+). Respond-
ing to the reinforced stimulus (8+) and

the nonreinforced stimulus (8-) is represent-
ed by solid lines and broken lines, res-
pectively . . . . . . . . . . . . . . . . . . . 8

Lower portion: Percent response during
the first 5 sec. after stimulus onset

to 72 db and 64 db intensities of a 5000
Hz tone for one group of rats given
discrimination training with the louder
stimulus reinforced (72+) and another
group trained with the softer stimulus
reinforced (64+). Responding to the
reinforced stimulus (8+) and the non-
reinforced stimulus (8-) is represented
by solid lines and broken lines,respect—
ively. Data are presented for 5 equi-
distant blocks of 5 sessions,with the
first, middle, last, and two intermed-
iate blocks from the training of each
subject averaged over groups. Upper
portion: Average number of responses
per session for the above groups during
nonreinforced silent intervals that
alternated with the presentation of 5+
and 8—. Data are for the 5 blocks des-
cribed above . . . . . . . . . . . . . . . . . . 9

Figure

Page

Response speed (mean reciprocal response
latency) after the onset of a reinforced
stimulus (S+), represented by the solid
lines, and a nonreinforced zero intensity
condition (ZSC), represented by the broken
lines, for two groups of rats, one with a
72 db 5000 Hz tone as S+ (72+), and the
other with the 64 db intensity of that
same tone as S+ (64+). Five equidistant
blocks of 5 sessions are presented with
the first, middle, last, and two inter-
mediate blocks from the training of each
subject averaged over groups . . . . . . . . 12

vi

INTRODUCTION

It has generally been found that when reinforcement is
given in the presence of a stimulus (3+) but not in its
absence, responding is acquired more quickly for more intense
values of 8+ (Gray, 1965). It has not been determined, how-
ever, whether the inverse intensity relationship holds for a
nonreinforced stimulus (S-) that is added to the presence vs
absence discrimination problem, ie whether training to reduce
responding to 8- through nonreinforcement proceeds more
quickly for less intense values of S—.

The present study employed a discrimination problem
that included two intensities of a training stimulus as well
as a "zero" intensity condition produced by the absence of
training stimuli. Two groups of subjects differed with
respect to whether the greater or the lesser intensity of
the training stimulus was reinforced, while the remaining
stimulus intensity for each group and the zero intensity
condition were nonreinforced. This arrangement allowed
observation of response reduction as well as acquisition
in the presence of stimuli at different distances from the

l

nonreinforced zero intensity condition. If the above
relationships hold in the present case it is predicted
that the group with the more intense stimulus as S+

should learn the discrimination more quickly, both in
terms of response acquisition to the stimulus of greater
intensity and response reduction to the stimulus of lesser

intensity.

METHOD

Subjects

The subjects were eight experimentally naive male
albino rats of the Holzman strain 90-120 days old at the
beginning of the experiment. One subject died of respir-
atory disease midway through the study and was not replaced.
The subjects were housed in individual living cages with
food (Wayne Mouse Breeder Blox) and water available on an
ad lib basis throughout the period of experimentation.
Deprivation schedules were not used because they were found
to be unnecessary in similar previous research (Walker, 1970)
where "sucrose" pellet reinforcements were given.

Apparatus

The apparatus included an experimental space (inside

dimensions 9" x 12" x 9") painted flat black and equipped

at one end with a bar centered three inches above the

hardware cloth floor. A depression of the bar (barpress)
enabled (at the discression of the experimenter) a single
4mm x 3.3mm x 45mg. Noyes "sucrose" pellet (reinforcement)
to be automatically delivered to a foodcup located directly
below the bar. Sound stimuli were produced by a Hewlett-
Packard specification #20-222 CD wide range oscillator and
were switched on and off without audible "clicks" via photo-
cell circuitry. Signals were amplified by a McIntosh MA
5100 amplifier and presented through a 4" speaker centered
in the ceiling of the experimental space. Stimuli were
presented and data were recorded with the aid of associated
electronic programing equipment.
Procedure

The general procedure was very similar to that of
previous research by the present author (Walker, 1970).
The barpress response was conditioned by merely placing
the subjects in the experimental space for daily 30 min.
(maximum) sessions during which all barpresses were
reinforced. The first stage of this training ended with
the fiftieth barpress to occur in any two consecutive
sessions. A uniformly high rate of responding was then
conditioned by continuing the above training until 25

barpresses occured within a single 20 min. (maximum)

session. No sound stimuli were presented during original
conditioning.

Discrimination training for each subject began on the
day following the completion of original conditioning. A
discrimination trial consisted of a nonreinforced silent
interval or "zero" stimulus condition (ZSC) and a period
during which a sound stimulus was presented. A ZSC did not
end until a prescribed period of time had elapsed without
a barpress. This period was either 10, 15. 20, 25, or 30
sec. for any given ZSC, depending on a prearranged sequence
randomizing these values in blocks of five. Either a rein-
forced (S+) or a nonreinforced (3-) sound stimulus immed-
iately followed each ZSC according to a prearranged semi—
random sequence that, for blocks of fifty trials, paired
each stimulus value with each ZSC value an equal number of
times. The stimuli were, in addition, randomized in blocks
of ten (containing two successive blocks of ZSC values),
such that each value appeared exactly five times without
occuring more than three times in succession. All S-
presentations ended with the first barpress or with the
passage of five seconds from the time of stimulus onset,

wichever came first. The first barpress to occur during

S+, regardless of latency, produced a single reinforcement

and ended the S+ condition. All stimulus presentations
were immediately followed by a ZSC.

Performance was scored as correct or incorrect on the
basis of the subject's behavior during the first five sec-
onds following stimulus onset. A barpress during this
interval was scored as correct if 8+ was present and as
incorrect if S- was present. Nonoccurance of response
during the first five seconds of an S- trial was scored as
correct. All S+ trials on which response latencies were
greater than five seconds were counted as incorrect although
the S+ condition remained in effect until a barpress finally
occured. Response latencies to 8+ and following the onset
of the ZSC were recorded as totals for each session.

The training stimuli were 64 db (re: .0002 dynes/cmz)
and 72 db intensities of a 5000 Hz tone. Ambient noise
level in the absence of the tone was approximately 40 db.
All decibel values were determined and periodically checked
using the A scale of a Bruel & Kjaer sound level meter.

The subjects were randomly divided into two groups that
differed with respect to whether 3+ was the more intense
(72+) or the less intense (64+) value of the tone. The
groups were equal in size at the beginning of the study,

but one of the 64+ subjects died midway through training

and was not replaced.

Training was given on a majority of days each week.
Each subject received one 50-trial session per day until
performance averaged 80% correct or better over five
successive sessions or until a maximum of 100 sessions had

been given.

RESULTS

The average time necessary for all subjects to com-
plete original conditioning was 153 min.. There were no
statistically significant differences between groups with
respect to overall time or time to complete either stage
of this training, although two of the 72+ subjects required
some hand shaping (reinforcement of successive approxim-
ations) to meet the first stage criterion.

Acquisition of the intensity discrimination, however,
was markedly different for the two groups. The 72+ subjects
reached criterion in 46-64 sessions with a mean of 56.5,
while none of the 64+ subjects were able to reach criterion
within 100 sessions. Group discrimination performance in
terms of percent response during the first 5 sec. of 8+ and

S- is presented for blocks of 5 sessions in Figures 1 and 2.

Figure 1 shows only the first 45 sessions, after which the
72+ subjects began reaching criterion. Figure 2 represents
the entire course of training, showing group averages for

the first, middle, last, and two intermediate blocks of
sessions given each subject. In addition, the upper portion
of Figure 2 shows the average number of responses per session
during the ZSC over the same 5 equidistant blocks.

The probability of responding within the first 5 sec.
after stimulus onset was virtually the same for either group
in the presence of either stimulus at the beginning of train-
ing. The groups remained very similar in their acquisition
of responding to 5+ over the first 45 sessions. Analysis of
variance for the S+ data did not show significant differences
either between groups or for the groups-trials interaction,
although the effect of trials was highly significant (p¢:.001,
‘E = 47.25, df = 8/40). A similar analysis performed over
the data represented in Figure 2 likewise did not show an
overall significant difference between groups, but there was
a significant simple main effect of groups for the last block
of sessions (pw(.001, §'= 23.42, df = 1/25) as well as a
significant interaction of groups and trials (p<:.005, F =
6.88, df = 4/20) suggesting that S+ performance was diff-
erent for the two groups when viewed over the entire course

of training. A trend analysis subsequently showed

Figure 1. Percent response during the first 5 sec. after
stimulus onset to 72 db and 64 db intensities of a 5000 Hz
tone for one group of rats given discrimination training
with the louder stimulus reinforced (72+) and another group
trained with the softer stimulus reinforced (64+). Respond—
ing to the reinforced stimulus (3+) and the nonreinforced

stimulus (S-) is represented by solid lines and broken
lines, respectively.

 

 

PERCENT RESPONSE

 

1OO'I

 

o
‘0'".
\‘.‘
0....
72+ 0—. 72. 0-4
20- * 3'
64: 0—0 64+ Ono
10-

 

 

T I I r I I I I

1 2 3 4 5 6 7 8
BLOCKS OF 5 SESSIONS
(50 TRIALS/SESSION)

Figure l.

9

 

Figure 2. Lower portion: Percent response during the first
5 sec. after stimulus onset to 72 db and 64 db intensities
of a 5000 Hz tone for one group of rats given discrimination
training with the louder stimulus reinforced (72+) and an-
other group trained with the softer stimulus reinforced (64+).
Responding to the reinforced stimulus (5+) and the non-
reinforced stimulus (S-) is represented by solid lines and
broken lines, respectively. Data are represented for 5
equidistant blocks of 5 sessions, with the first, middle,
last, and two intermediate blocks from the training of each
subject averaged over groups. Upper portion: Average num-
ber of responses per session for the above groups during
nonreinforced silent intervals that alternated with the
presentation of 8+ and S—. Data are for 5 blocks described
above.

 

RESPISESS

PERCENT RESPONSE

 

 

   
  

100 72 ’ '
64 0 O
6 0
v””" o~.-
20 F ——o-— 29:":-
9.0 O
8 O 9 .\./
7O ‘ audio
O Q\‘
60 I ‘\ \\
I ‘\ ‘x
\\ \\
5O \\‘ .0
.s
40 9’ ‘x\
‘....
20
72 O .—. 72* ."‘".
‘ s: -
10 64+ o—o 64+ o--o
1 2 3 4 5

EQUIDISTANT BLOCKS OF 5
SESSIONS (50 TRIALS/SESSION)

Figure 2.

 

 

10

significant differences between groups both for the linear
and the quadratic components (24.05, _]:'_‘_ = 9.22, df = 1/5,
and £4.05, _1=_‘_ = 6.82, df = 1/5, respectively). These
differences in combination with the fact that group per-
formances were virtually identical over the first two
blocks of sessions suggest a significant reduction in per-
cent response to 8+ (within 5 sec.) for the 64+ group over
the latter portion of training.

The tendency to respond to S-, however, remained rel-
atively high for the 64+ group throughout training. Analysis
of variance showed significant differences in S- performance
between groups over the first 45 sessions (24.05, _F_‘_ = 7.80,
df = 1/5) and over the entire course of training (£4.01,
gj= 17.78, df = 1/5). The groups-trials interaction was
significant in both cases (p<.001, _F_ = 5.13, df = 8/40,
and 24.005, _E_‘ = 5.87, df = 4/20, respectively), and the
linear and quadratic components for the S- functions in
Figure 2 were also significantly different (34.025, 3.: =
10.93, df = 1/5, and 24.025, E = 11.39, df = 1/5, respect-
ively). All of the above contribute to the general find-
ing that S- training proceeded more quickly for the 72+
group.

In contrast to the S- condition, response reduction

11

during the ZSC was virtually identical for the two groups
(Figure 2, upper portion). The between-groups differences
and the groups-trials interactions were not statistically
significant either for the first nine blocks of sessions
or for the five blocks representing the entire course of
training. The effect of trials was highly significant in
both cases, however, (E<.001, _F_ = 46.28, df = 8/40, and
'34:.001, E = 57.25, df = 4/20, respectively), indicating
that marked response reduction took place for both groups
during the ZSC training.

Changes in ZSC performance are also illustrated in
Figure 3, which shows response speed (mean reciprocal
response latency) after the onset of the ZSC (broken
lines) and the S+ (solid lines) for the same five blocks
of sessions as in Figure 2. In keeping with the results
for number of ZSC responses, response speed to ZSC onset
decreased significantly over trials (24 .001, _F_' = 20.27,
df = 4/20) while the between-groups differences and the
groups-trials interaction were not statistically signifi-
cant.

The changes in S+ speeds shown in Figure 3 are con-
sistant with the results for 8+ presented above. As with

percent response, the groups began at virtually the same

Figure 3. ResponSe speed (mean reciprocal response lat-
ency) after the onset of a reinforced stimulus (8+), rep-
resented by the solid lines, and a nonreinforced zero
intensity condition (ZSC), represented by the broken
lines, for two groups of rats, one with a 72 db 5000 Hz
tone as S+ (72+), and the other with the 64 db intensity
of that same tone as S+ (64+). Five equidistant blocks
of 5 sessions are presented with the first, middle, last,
and two intermediate blocks from the training of each
subject averaged over groups.

 

MEAN RECIPROCAL
RESPONSE LATENCY (SEC)

.340

.3 00

.2 60

.220
.130
.140
.100
.090
.080
.070
.060
.050
.040
030
.020
.010

12

 

Xt/

’\
sag"‘°‘\\
Q..\::\flg’fl"uo
lush".."
72. 0—0 72+ 0-».
. S: zsc
64+ 0-0 64* o-uo
1 2 3 4 5

EQUIDISTANT BLOCKS OF 5
SESSIONS (50 TRIALS lSESSION)

Figure 3.

13

point and did not differ statistically overall, while there
was a significant simple main effect of groups for the last
block of sessions (24-005: E = 12.93, df = 1/25) and a sig-
nificant groups-trials interaction (24.005, _F_ = 5.49, df =
4/20) as well as a highly significant effect of trials (24:
.001, g = 12.80, df = 4/20). The linear (but not the quad-
ratic) component was also significantly different for the
two functions (E< .025, _F_‘ = 10.29, df = 1/5), supporting
the apparent downward trend that can again be seen in the
64+ data.

In summary, the prediction that the 72+ group would
learn the discrimination more quickly because of the dual
presence of a more intense stimulus for 5+ and a less
intense stimulus for S- was fully confirmed. Both the
probability and the speed of 8+ responding were ultimately
higher for the 72+ group, while the probability of respond—
ing to S- was consistantly lower for this group after the
first block of sessions. Furthermore, the fact that the
magnitude and speed of ZSC responding were nearly identical
for the two groups suggests that the 8+ and S— differences
observed were specific to the manipulation of stimulus
intensity and did not represent a generalized deterioration
in the performance of the 64+ group, whose complete failure

to achieve the discrimination had not been expected.

DISCUSSION

Hull (1949) reviewed work by several investigators
showing that asymptotic levels of conditioned responding
increased with increased intensities of the training
stimulus. To account for these findings Hull postulated
a general principle which he called stimulus intensity
dynamism (SID) denoting a positive correlation between the
intensity of a conditioned stimulus and the magnitude of the
resulting response. Perkins (1953) and Logan (1954)
independantly challenged Hull's formulation on grounds
that it was not necessary to postulate a distinct mechanism
for the SID effect. They argued that if the experimental
situation includes a zero intensity or relatively low
intensity condition during which responding is not rein-
forced (which has usually been the case in experiments
where SID has been observed), then generalization of
inhibition from the zero intensity condition could produce
the SID effect by more greatly affecting stimuli of lower
intensities.

The results for response acquisition in the present
study do not agree with previous findings of SID or with
either of the above theoretical interpretations in that

asymptotic levels of performance as shown over the first

45 sessions were virtually the same regardless of whether

15

8+ was 72 or 64 db. On the other hand, the fact that final
levels of 5+ performance were higher for the 72+ group
(Figure 2) replicates previous findings and agrees with

both theoretical interpretations. The difficulty is in
determining at what point asymptotic levels of performance
were actually reached. The significant downward trend

shown in the 64+ data over the last 3/5 of training sug-
gests that there was an erosion of 8+ performance helping

to produce the final differences observed. Although Hull's
position predicts lower asymptotic levels of 8+ performance
for the 64+ group, it does not predict a reduction of resp-
onse strength over trials. Such a reduction has not, to the
writer's knowledge, ever been previously observed, but it
can be predicted from the Perkins-Logan position by assuming
that the generalized inhibitory strength of the ZSC continued
to increase over trials after S+ responding had reached
maximal strength. This assumption cannot be directly sup-
ported by the present data since no gradients of inhibition
were measured, but the data for number of ZSC responses and
ZSC response speed indicate that response reduction during
the ZSC continued over the entire course of training, while
S+ responding clearly reached maximal levels for the 64+

group over the first 45 sessions. In addition, the fact

16

that the subjects in the present study were never food
deprived, but worked solely for the incentive value of the
"sucrose" pellets, may have made S+ responding more easily
affected by subtile influences from the ZSC. All known
previous studies in the present area have encouraged strong
S+ responding through the use of deprivation schedules or
aversive stimuli. At any rate, the downward trend in S+
performance observed for the 64+ group can be handled by
the Perkins-Logan hypothesis far more easily than by a
strict Hullian interpretation, although the lack of diff-
erences that may be seen in maximum levels of performance
remains unexplained from either standpoint.

In general, the present data partially replicate
previous findings that higher levels of responding are
produced as the intensity of the training stimulus is
increased. The major purpose of the present study, how-
ever, was to determine whether training to reduce respond-
ing may proceed more quickly as the intensity of the train-
ing stimulus is decreased. The present data are very clear
on this point. The 72+ group consistently maintained lower
levels of S- responding than the 64+ group after the first
block of sessions. The significant between-groups differ-

ences and groups-trials interactions Obtained for S-

 

17

performance reinforce the conclusion that training pro—
ceeded more quickly when S- was 64 db rather than 72 db,
while the lack of such significance for the ZSC data
suggests that the groups reduced responding equally well
when the stimulus conditions were the same.

The S- results can be interpreted in terms of either
the Hullian or the Perkins-Logan hypothesis, but the latter
position is strengthened at the expense of the former to
the extent that the ZSC training was important for the
observed differences. Generalization of inhibition from
the ZSC, once again, needs finally to be established by
generalization tests, but the importance of the ZSC training
is suggested by a comparison of the present results with
those of a previous study by Harris and Denny (1968).
These workers trained rats on a brightness discrimination
for food reward in a barpress apparatus. The brighter 0f
two white light stimuli was reinforced for half of the
subjects and was nonreinforced for the other half, in a
design similar to that of the present study. However, the
zero intensity condition (training lights out) was not
introduced until after the discrimination had been learned
to criterion. Extinction was massed during a single

presentation of the zero intensity condition, and additional

18

extinction took place immediately afterwards in the pres—
ence of the individual training stimuli. Although resis-
tance to extinction of the positive stimulus was found to
be greater for the group that had the brighter stimulus
previbusly reinforced, no statistically significant diff-
erences were found in resistance to extinction for the
previously nonreinforced stimuli. These results suggest
that a single exposure to the nonreinforced zero intensity
condition may not be sufficient to produce differential
S- performance, whereas 5- differences were reliably ob-
served in the present study, which incorporated the ZSC
throughout discrimination training. The comparison bet-
ween these two studies is admittedly loose since the stimulus
modality, levels of deprivation, and measures of response
strength (as well as other possible factors) also differed
in addition to the amount and positioning of ZSC training.
However, these studies do suggest that investigations
directly manipulating ZSC training might be helpful in det-
ermining the role that generalized inhibition may have in
the SID effect and in the reduction of responding to non-
reinforced stimuli when a ZSC is part of conditioning.

In summary, the present study has replicated the

general finding that conditioned responding is stronger

 

19

with more intense training stimuli, in that response
strength for a less intense stimulus was observed to
decrease over trials after reaching maximum levels. It
was also observed that training to reduce non-reinforced
responding proceeded more rapidly for the less intense of
two training stimuli, both of the above findings occur-
ing in the context of a discrimination problem that included
a nonreinforced "zero" intensity condition. The present
results by themselves and in combination with previous
findings (Harris and Denny, 1968) appear to offer more
support for the generalization of inhibition hypothesis of
Perkins (1953) and Logan (1954) than for Hull's (1949)

stimulus intensity dynamism position.

REFERENCES

REFERENCES

Gray, J.A. Stimulus intensity dynamism. Psychological
Bulletin, 1965, 63, 180-196.

 

Harris, L.A. and Denny, M.R. The relationship of no light
to the stimulus dimension of light intensity.
Psychonomic Science, 1968, 11, 333-334.

 

Hull, C.L. Stimulus intensity dynamism (V) and stimulus
generalization. Psychological Review, 1949, 56, 67-76.

 

Logan, F. A note on stimulus intensity dynamism (V).
Psychological Review, 1954, 61, 77—80.

 

Perkins, C.C. The relation between conditioned stimulus
intensity and response strength. Journal of Exper-
imental Psychology, 1953. 46, 225-231.

 

Walker, E.D. Changes in irrelevant cues and transfer along
the auditory intensity continuum. Unpublished Master's
Thesis, Michigan State University, 1970.

20