5J4:

3,“ 41. at. 4 («.qu
-é‘ug‘lt‘x- I'l‘ué i ‘ i";

'-§R4$:L

:1

"in ”L3 3 .=‘.\.;.i."’.z='“

 

- ._ us Q‘!
' -"'~1 4' .
Q“ N 4 i‘.‘ 1‘44: ix 4- a‘ an“

an 4 . n1 .
rmzis 93:; ,: fin:- Elana 333 :21 3:2. 3.
.3 rig;- 3'47. r‘ H: a! "“4 Qa‘u :g; 5' In‘: a .
~L-fS-éwl'é-i‘c‘élk-L‘; ’. . ‘u :‘V‘LV‘ ‘ @351".

.... ‘. r .

3 , -g . ‘ 4 . 4:?

um 2.". .“LQfi' I‘d! VP 9954

rapt
.994»

$5

thSit’:

 

LIBRARY

Michigan State
University

 

This is to certify that the

thesis entitled
Lu- ELI IIxGT ZIRUUQEI STII. UIUS 3 iL'L‘JI ION:

03E TRIAL A DAY
presented bg

RUTH HCII‘A WELIS

has been accepted towards fulfillment
of the requirements for

Eih. Do degree in PSYChOlOQ‘y

%&%/

Mala professor

Date October 19, 1956

 

 

 

 

ABSTRACT

Recently learning theorists have been attempting to determine the
effects of the numerous factors which are involved in alternation behavior.
The present thesis was concerned with several of these factors. The pri-
mary interest was in an attempt to determine whether or not the stimulus
satiation developed in a specific situation decreases to zero in less than
24 hours. This was tested. The effects of maze cues during the early
trials of the learning period, before the formation of a habit, were com-
pared to the effects of the same cues after'a turning-response habit had
been acquired. Lastly, a comparison was made of the reversal tendency of
satiated subjects with that of food-deprived animals.

The apparatus was a standard T-maze; black cardboard inserts
were used to line one alley when differential maze cues were required.

Forty-seven experimentally naive male rats were used in the
study. Group 1, tan animals, received the following training; (1) 52 days,
one trial a day, 2Lle ratio, (2) 6 days, 3 trials a day with a fifteen_
minute inter-trial interval, 2L:IR ratio, (3) 26 days, free trials, fol-
lowed by a three-day delay period, (4) 1 day, free trial, with three sub-
jects satiated and two subjects food-deprived, (5) 3 days, free trials,
maze rotated 180'. During the entire experiment, the left alley was black
and the right alley was natural wood finish. Group 2, eleven animals, re-
ceived the folloming training; (1) 24 days, 2R:1L ratio, left alley -
black, and right alley - natural wood finish, (2) 3 days, free trials, (3)
33 days, 2L:lR ratio, (4) 12 days, 2141B ratio, left alley - natural wood
finish and right alley - black, (5) 12 days, free trials, followed by a

three-day delay period, (6) 1 day, free trial, with four animals satiated

and three animals food—deprived, (7) 3 days, free trials, maze rotated
180°. Group 3, five animals, received the following training; (1) 19
days, 2L:1R ratio, left and right alleys - black, (2) 14 days, left alley -
natural wood finish and right alley - black, (3) 16 days, Forced L, Forced
I” Free-choice sequence, (4) 8 days, maze rotated 180°. Group 4, five
animals, received the following training; (1) 19 days, 2R:1L ratio, left
and right alleys - black, (2) 14 days, left alley - natural wood finish
and right alley - black, (3) 16 days, Forced L, Free-choice, Free-choice
sequence, (4) 8 days, maze rotated 180‘. All trials for Groups 2, 3, and
4 were on a one-trial-a-day basis. Group 5, seven animals, received the
following training; (1) 31 days, 2Lle ratio, left alley - black and right
alley - natural wood finish, (2) 10 days, free trials, one a day, (3) 3
days, free trials, five a day with twenty-minute, inter-trial intervals.
Group 6, nine animals, received the following training; (1) 31 days, 2RzlL
ratio, left alley - black and right alley - natural wood finish, (2) 10
days, free trials, one a day, (3) 3 days, free trials, five a day with
twentyeminute, inter-trial intervals.

The following results were obtained. The acquisition of a habit
did occur with all six groups even though the inter-trial interval was 24
hours. Groups 1, 2, 5 and 6, with which differential intraemaze cues were
used, showed a tendency to respond to the less frequently experienced
stimulus complex. Groups 3 and 4, with which non-discriminable maze cues
were used, learned to go to the left alley, presumably on the basis of
extra-maze cues. The running response which was learned was extremely re-
sistant both to reversal when the maze stimuli were changed and to extinc-
tion when the trials were all free—choice. The tendency for satiated
animals to show more reversals than food-deprived animals was discussed.

There was an indication that extra-maze cues have some eliciting value

after a running habit has been developed as shown by the effects of maze
rotation after 88 days of running trials.
The limitations of various theoretical approaches to the problem

of alternation phenomena were discussed.

LEARNINU'THROUGH STIMULUS SATIATION:

ONE TRIAL A DAY

By

RUTH HOLM WELLS

A THESIS

Submitted to the School of Advanced Graduate Studies of Michigan
State'University of.Agriculture and Applied Science
in partial fulfillment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY

Department of PsycholOgy

1956

_,/- 7 - 5' 5’?

L132;

ACE‘IOiWT

The author wishes to express her gratitude for the
indispensable assistance and advice in formulating the experi-
mental problem and carrying out the research to Dr. M. Roy

Denny, under whose supervision this study was done.

TABLE OF CONTENTS

Page
LISP OF TABLES iii
LIST OF FIGURES iv
INTRODUCTION 1
STATm’EENT OF THE FROBIJM 15
EXPERIMENTAL PROCEDURE
Apparatus 14
Subjects 16
Procedure 1'7
RESULTS 22
DISCUSSION 33
SUMMARY 39

BIBLI OGRAHIY I 42

111

LIST OF TABLES

Page
1. Experimental design of study 20

INTRODUCTION

Tolman (26), in 1925, discussed a specific phenomenon of rat
behavi or in the following manner:

"A simple T-maze was used, and it was arranged so that

the animal could get back to the food box in identical

fashion whether he chose the left or the right angle of

the T. Either route met with success . . . . (There was)

a very pronounced tendency toward continuous and regular

alternation - left, right, left, right, or right, left,

right, left . . . It appeared, in short, that even where
either side was equally "sati sfactory" there was in our

rats a positive tendency left over toward variation of

response . . . a positive tendency in and of itself.”

This positive tendency, now referred to as spontaneous alter-
nation, has been investigated extensively in recent years, inasnuch as
the behavioral phenomena related to it present a difficult set of data
for any existing theory of behavior to explain.

Attempts to explain alternating behavior within a Hullian
framework have been based on the concept of inhibition. Hull's (12)
Postulate 8 includes three ideas relevant to spontaneous alternation
patterns: (1) A primary negative drive, Ir, results from the occurrence
of a response and lessens the probability of the repetition of the re-
sponse, (2) The amount of Ir which develops is an increasing function
of the amount of work or effort involved in making the response, and
(3) Ir dissipates according to a decay function of time. Heathers (11)
has reported experimental evidence relating the amount of alternation
in a T-mazs situation to the length of the inter-trial interval which
seemed to support the Bullion postulate. Using 15, 30, 60, and 120-

second intervals between five massed trials, he found the percentage

of alternation on a free trial ranged from 85.6 with the lS-second
interval down to 65.6 with the lZO-second interval. 4A test trial fif-
teen minutes after the massed trials yielded 50.3% alternation and on
the first trial on the following day, (24phour delay) 50.4% alternation.
Thus, his data showed no evidence for any reaction decrement after a
fifteeneminute period.

Zeaman and House (32) made three deductions from Hull's postu-
late which were specifically related to the spontaneous alternation
problem and then attempted to test these ideas directly. They stated,
FAlternation tendency will be (1) a negatively accelerated decreasing
function of the time between responses; (2) a positively accelerated
increasing function of the work involved in the execution of the re-
sponse; and (3) a simple linear increasing function of the number of
evocations.‘ Thus, if the amount of reactive inhibition were to be
built up by increasing the number of responses in a short period of
time, the percentage of alternation responses should increase propor-
tionately. In.the case of a Tamaze experiment then, the greater the
number of times a response to one side of the maze is forced in massed
trials, the greater should be the tendency for the alternate response
to occur on a test trial. Supposedly, this rise in the number of al-
ternating responses demonstrates the growth of inhibition. Zeaman and
House also intended to demonstrate the decay of inhibition.by showing
a fall in the curve relating the percentage of alternation to the time
interval between the massed, forced trials and the subsequent test
trial. 'Using varying numbers of forced trials and varying delay periods
between the massed trials and the test trial, Zeaman and House obtained
the following results. The percentage of rats alternating increased

linearly with increasing numbers of forced trials. The alternation

percentage decreased as a negatively accelerated function of the length
of delay period between the massed trials and the test trial. The per-
centage of alternation decreased gradually from 100% after no delay
betwaen the massed trials and the test trial to 71% at 12 hours delay
and to 51% at 24 hours. This contrasts markedly with the conclusion
drawn by Heathers to the effect that the tendency to alternate drops to
50% after a fifteen-minute period.

Berlyne (l) , in a similar approach to the problem, has de-
veloped two postulates and three corollaries based primarily on Hull's
Postulate 8. Berlyne's Postulate I states, "When a novel stimulus
affects an organi sm's receptors, there will occur a drive-stimulus-
producing response." This response Berlyne called curiosity. Postu-
late 11 states, "As a curiosity-arousing stimulus continues to affect
an organism's receptors, curiosity will diminish.“ The corollaries
include the following ideas; (1) "An organism will learn to respond to
a curiosity-arousing stimulus with activity which will increase stimu-
lation by it”, (2) after a time, the activity, labeled exploration by
Berlyne, will cease, and (3) I'After a further lapse of time, if the
curiosity-arousing stimulus is again affecting the organi en's receptors,
there will be a second spell of exploration, but this time there will
be less exploration than during the first spell; such spells of ex-
ploration will recur at intervals, until exploration of the stimulus in
question ceases altogether.” Berlyne's experimental results were in
line with the predictions following from his corollaries; the rats
spent more time exploring a novel stimulus than they did exploring
stimuli which had been explored previously and they spent less time

exploring stimuli the second time the stimuli were encountered. This

type of explanation can be used in dealing with an alternation problem;

the stimuli of the most recently visited T-maze arm will have less
effect in arousing the drive-stimulus-producing response, curiosity,
than the other arm and hence, the animal alternates.

Solomon (25) reasoned that if the concept of reactive inhibi-
tion is a valid explanation of the occurrence of alternation behavior,
the tendency to alternate should be greater when more work is involved
in the reactions. He tested this hypothesis experimentally by inclin-
ing the goal arms of a T-maze at a 16° angle from the horizontal level
so that greater effort would be expended by the animals in getting to
the goal boxes. The alternation percentage increased from '74 to 90,

a significant change. However, another attempt to increase the amount
of effort in attaining the goal which involved strapping weights to the
backs of the rats did not lead to increased alterration.

Data from numerous other studies have not been adequately
explained with Hull's concepts. Dennis (4) has found that in a
multiple-unit maze, animals did not alternate on a right-left basis
with individual maze units but rather they alternated their total
pattern, for instance, from left, left, left, right turns to right,
right, right, left turns. This is in direct contradiction to the re-
sults to be expected if the alternation tendency were a function of a
response-produced decrement. Dennis also found that increasing the
length of the final section of the haze from O to 38 feet resulted in
no change in the amount of spontaneous alternation, in spite of the in-
crease in inhibitory effects which one would hypothesize from Postu-

late 8. Wingfield (31) has shown that human subjects show more alter-

nation in a situation which calls for a choice of lights when the lights
are different in color than when the lights are the same.

Glanzer (8) has presented in a review article the major find-

ings in the area of spontaneous alternation up until early 1952. He
has pointed out the above-mentioned inadequacies of the response-theory
type of explanation and has formulated a postulate and corollaries in
terms of the stimulus aspect of the situation. The first part of the
postulate states, ”Each moment an organism perceives a stimulus-object
or stimulus-objects, A, there develops a quantity of stimulus satiation
to A. Glanzer posits that the amount of stimulus satiation developed
over a period of time is an increasing linear function of time, that
satiation to A will be generalized to other stimulus—objects B, and
that stimulus satiation reduces the organism's tendency to make any
response to A. Glanzer ms experimentally tested twa deductions from
his postulate and has contrasted the results obtained to those which
would have been predicted from a response-type theory. The first hy-
pothesis states that exchanging the stimulus-objects of a two-alterna-
tive maze betwaen trials will result in a repetition rather than an
alternation of turns. This was tested by using a cross-shaped maze
with twa starting boxes. 0n half of the eight experimental days, the
26 animals were started from one starting box for the first trial and
were started from the Opposite box for the second trial which followed
immediately. The animals alternated alleys (stimuli) rather than re-
sponses, whereas any response-type theory would predict opposite results.
The second hypothesis states that increasing the time interval between
trials will have varying effects on spontaneous alternation depending
on the place of the delay. Glanzer found that a delay of ten minutes
in the alley which had been chosen on the first trial resulted in the
greatest amount of spontaneous alternation as compared to an equal de-
lay period in the starting box or at the choi ce-point. A response-type

theory would regard the amount of alternation as a function of the time

interval and not as a function of the stimulus complex during the delay
interval. The verification of these two hypotheses is accepted by
Glanzer as evidence for the accuracy of explaining alternation phenomma
in terms of the effects which exposure to stimuli has on the organian.

Montgomery (17, 18, 19, 20) has reported several studies and
an original theoretical developnent on the problem of spontaneous al-
ternation. In one study the relationships between increased work and
increased inter-trial intervals and the tendency to alternate were in-
vestigated. Montgomery varied the work by placing weights on the arms
of a bar-pressing apparatus and found that increasing the weights did
not result in an increase in the alternation of turn-responses. How-
ever, inasnuch as Hull posits reactive inhibition as being highly spe-
cific to a response and in this situation the relevant response was a
turning-response, not a bar-pressing response, the date cannot be con-
sidered a test of Hull's concept. The data did confirm earlier studies
in showing that with increasing inter-trial intervals, there is a cor-
responding decrease in the tendency to alternate.

From research designed to study the relationship befleen ex-
ploratory behavior am the general activity level, several. important
findings have been reported by Montgomery (21, 22, 23). Animals de-
prived of food or water showed significantly fewer exploring responses
than satiated animals showed. Animals deprived of their normal amount
of activity by constant enclosure in a cage which was too small for
any activity showed no more exploratory behavior in a test period in a

maze than a control group which was allmed constant access to an

activity wheel. From these data, Montgomery concludes that exploratory
behavior is independent of the general activity level. With considerable

evidence that neither Hull's concepts nor any other formulated theory

 

 

III.|iIIIi[I‘I|l-|Ill|ll‘lll

could explain the phenomena of exploratory behavior and spontaneous al-
ternation, Montgomery has posited a unique drive, exploratory drive. He
regards the drive as a primary drive which is relatively independent of
the other primary drives, although it shows some tendency to decrease
in strength.when another primary drive is heightened, as in the case of
food or water deprivation. His specific hypothesis states, "(a) that

a novel stimulus evokes in an organism an exploratory drive Which moti-
vates exploratory behavior, and (b) that strength of exploratory drive,
as measured by amount of exploratory behavior, decreases with.time of
continuous exposure to a given stimulus situation and recovers during a
period of non-exposure.” From.this theoretical orientation, it was
hypothesized that any decrement in exploratory behavior from the ex?-
posure to a stimulus situation would generalize to other situations and
that the amount of the generalized decrement would decrease as the simi-
larity of the stimulus situations decreased. These hypotheses were
tested and confirmed and on the basis of these results, Montgomery has
stressed the importance of external stimuli as a factor in exploratory
behavior as contrasted to the importance of the response.

Walker and his coworkers (28, 29, 30) have studied the func-
tional relationships among several of the variables operative in spon-
taneous alternation behavior. One study yielded results contradictory,
in part, to the studies of Glanzer and Montgomery in regard to the in-
fluence of the stimulus and the response in determining alternation.
Walker, like Glanzer, used a cross-shaped T-maze but found half of his
animals alternating the response and the other half of the animals al-
ternating the stimulus, although Glanzer's study showed significantly
more animals alternating the alley (stimulus). When the maze arms Were

inclined, however, his data substantiated the other two studies; that

is, his animals alternated the stimulus complext In a study using dis-
criminable maze cues, white and black maze covers, Estes and Schoeffler
(7) randomized the turning responses during the ten massed trials, and
found alternation on the basis of the maze stimuli significant at the
.01 level'while alternation of the turning response, right or left, was
52% or chance.

In.a study utilizing the cross-shaped maze and adding the
feature of rotation of the maze in order to change the extra-maze stim.
uli, a comparison was made of the strength of place (extra-maze cues),
stimulus (maze cues), and response as determiners of alternation.
Walker (30) found that the stimulus had significantly more influence
than place, place had significantly more influence than reaponse, and
that the response, in this case, was not operative in increasing or de-
creasing alternation. The general conclusion drawn from this study was
that the tendency to alternate would be at a minimum when the stimuli
in the two alternatives were the same and would increase as the stimuli
became increasingly discriminable. Estes and Schoeffler (7) ran an al-
ternation study with a maze designed in such a way that for an animal
to alternate at the choice point, it would be necessary to make the
same turning response that had been made moments before. Using ten
massed trials and then a test trial, it was found that 100% alternation
resulted; that is, all animals ran to the new stimulus complex even
though this involved repeating the response. In.a maze rotation experi-
ment (90' rotation) in which the specific response could be alternated
but the place, extraqmaze stimuli, could not be alternated, Estes and
Schoeffler found sqt alternation or chance behavior. In.contrast,
Rothkopf and Zeaman (24) rotated a maze 180’ between the forced trials

and the test trial and found 69% ”response alternation".

In another study, Walker (27) has found that animals which
were rewarded in the goal box of a standard T-maze showed significantly
more alternation behavior than animals that were not rewarded. He
states, on the basis of these data, "the effect of reward is to increase
the tendency to alternate and by implication to increase the size of the
react]. on decrement".

In a study designed to determine whether or not exposure to a
stimulus outside of the maze situation would influence the choice of
that stimulus on a test trial, Walker (29) obtained negative results.
Exposing the animals to both the stimulus color and the place by putting
them in the goal box for a 45-second period did not result in any tend-
ency to avoid the color on a test trial. At this point, Walker hy-
pothesized that animals had to make a choice in the stimulus situation
before any reaction decrement or stimulus satiation would take place.
His next study (14) was designed to test this idea. 0n the first trial,
animals were placed in a T-maze which had glass doors at the choice
point so that the alleys could not be entered but they could be observed.
Both alloys were the same color, either black or white. 0n the seconi
trial, one of the alleys was changed in color, the glass doors were re-
moved, and the animals were allowed to make a choice, run to either alley.
The results indicated a tendency to run to the alley with the changed
stimulus color, significant at the .01 hvel. Dember (3) hypothesized
that these results did not indicate necessarily the effects of stimulus
satiation as differentiated from a tendency to explore novel stimuli.

In an experimental setup similar to that used by Kivy, Dember placed

his animals in the straight alley of a T-maze with glass doors at the
choice point but with one alley black and the other alley white.

On Trial 2, both alleys were the same color, the glass doors were re-

10

moved, and the animals were allowed to respond. The subjects respond-
ed to the alley which had been changed in color, significant at the
.001 level of significance. The stimulus satiation postulate does not
seem to be adequate for explaining these data.

Two recent experiments dealing with the intensity of the af-
ferent feedback from.a response and with.the discriminability of the
afferent feedback of different responses as they affect alternation
have been reported by Walker (28). The study in regard to intensity or
the amount of effort involved in making a response yielded results
similar to studies by Solomon and Montgomery; there was no increase in
alternation. ‘However, when the maze was built in such a way that the
responses to be made were quite different, the tendency to alternate
increased significantly. Walker states that these data cannot be ex.
plained in terms of any response-type theory.

De Valois (6) studied alternation.behavior in a maze with
several equally long routes to the goal. Groups of rats running under
four different conditions, 6 hours of thirst, 22 hours of thirst, a
light shock, and a strong shock, were divided so that half of the animals
‘were shifted to the other condition of the same type half way through
the experiment. iHis results indicate that the "low motivation” subjects
were more variable than the ”high motivation” subjects, increasing the
strength of motivation leads to a decreased amount of variability, and
that decreasing the strength of motivation leads to an increased amount
of variability in the case of subjects originally run on.moderately
Strong.motivation (strong thirst) but not significantly in the case of
Subjects originally run on extremely high motivation (strong shock).

In an attanpt to determine whether a massed series of forced

trials increases the tendency to avoid a specific maze location or to

ll

avoid recently visited locations, Estes and Schoeffler ('7) ran another
experiment. After ten massed forced trials to one side of a T-maze,
animals were given ten massed test trials. The tendency to avoid the
specific T-maze alley to whi ch the animals had been forced to run drop-
ped to a chance level on the third test trial while the tendency to al-
ternate alleys increased to and remained at approximately the 65%
level. This seems to indicate that the animals tend to avoid the spe-
cific maze location and that the general tendency to alternate alleys
is not affected.

Several unpublished studies by Danny and his students (13,
15, 16) at Michigan State University have dealt primarily with two
problems related to alternation behavior. One of these problems was
to compare the eliciting value of two stimulus complexes when.one com-
plex has been experienced twice as frequently as the other. For in-
stance, in a T-maze situation if some forced trials are used so that
animals run to the left alley twice and to the right alley once in each
block of three runs, all runs being rewarded, will the free trials in-
dicate a tendency to turn left or right? The second main problem with
which these studies have been concerned is the length of time during
which the reaction decrement is operative. Matus and Lange's study
(16) used two trials a day with a thirty-minute interval. Two groups
of animals were used, one group on a L L R ratio and the other on a
R R L ratio, to control for direction preference. By the melfth day,
80% of the animals Were running to the alley which had been experienced

less frequently. In a similar study, Marti ndale (15) found that after
animals had formed a running habit to the less frequently visited alley

and the schedule of forced trials was reversed, the animals' running

habit also was reversed and the animals again were running to the stimu-

lus complex less frequently experienced. Julian (15) used three trials
a day with a two-hour inter-trial interval and obtained similar results.
Bis animals also selected the less frequently experienced alley on free
trials. After Julian's group was showing consistent behavior, lOQ$
going left on free trials on days five through nine, the schedule was
reversed from.R.R I.to L L R and with nine more days, the percentage of
left turns had fallen to 50. Since Julian's animals developed rapidly
the tendency to turn to the less frequently experienced alley even with
a two-hour interval, research for the present thesis was planned with a
much longer interval in an attempt to determine the limit to the para-

meter.

 

 

 

l.1'lll.ll ’1')"

STATEMENT OF THE PROBLEM

The present research was designed primarily to determine
whether or not the effects of stimulus satiation are present after
a 24-hour period. If a 24-hour period does not allow for complete
dissipation of the satiation effects from a single event, trials
administered on a one-a-day schedule should result in a habit of
running to the less frequently visited side of a T-maze with dis-
similar arms.

The same T-maze, but with similar arms, also was used in
order to determine the role of the differential intra-maze stimuli
in the developnent of an alley preference.

Two other variables, food satiation and extra-maze stimuli,

also were manipulated experimentally.

APPARATUS

The apparatus used in this experiment was a standard Tamaze
as shown diagrammatically in.Figure 1. It consisted of a starting box,
a combination stem and choice-point section, 14" long, two alleys, 20"
long, and two goal boxes. All of the sections were 6" deep and 4?
wide. The entire maze was natural wood and was covered with hardware
cloth. Black cardboard was used to line the stem and one of the two
alleys when the experimental design called for differential stimuli.
Guillotine-type doors were placed at the exit of the starting box, at
the choice point, and at the entrance to the goal boxes, and were used

to prevent retracing.

15

 

 

 

 

.ensfiun. no 95393 omega .H one»:
son
mourns
"""" J <
988 33328 - a .o .o .m .a
m n n m
, . u
afim flashes“ “ .593 n Home

.

H _ n n

n a e m

 

 

 

SUBJECTS

The subjects used in the present study were 54 experimentally
naive male rats from the colony maintained by the department of
psychology of Michigan State University. The animals, 27 albino rats
and 27 hooded rats, ranged in age from 120 to 150 days at the beginning

of the experiment.

PRO CEDURE

All animals were handled and then fed in individual feeding
cages for ten days prior to the beginning of the experiment. During
this period, they were fed 8 grams of Purina Dog Chow every 24 hours,
water being available at all times. At the end of the ten-day period,
the animals were randomly divided into four groups, except for the con-
trol of the proportion (50-50) of hooded and albino animals in each
group and the total number for each group.

The basic design for all four groups involved specifying the
proportion of runs to each maze alley. ‘I'wo ratios were used, L L R
and R R L. For animals on a L L R ratio, each three-day period includ-
ed two trials of running to the left alley and one trial of running to
the right alley and for animals on a R R L ratio the reverse was true.
During each three-day period, the first day's trial was a free-choice
trial; both maze alleys were unblocked and the animals could go left
or right. The second day's trial was forced or free depending upon the
direction of the first day's run. For an animal on a L L R ratio, if
the first run were to the right alley, the second and third trials were
forced left runs. However, if the first run were to the left alley,
1=he second trial was again a free-choice trial, and only the third trial
Was forced. Sliding panels located at the entrance to each alley were
used on forced trials; the animal could enter only one of the alleys.
A pellet of food was in the goal box of each alley on each trial so that

all runs were rewarded.

Group 1 (N-twelve animals) was given 52 days of training on a

L L R ratio, with the left alley of the maze lined with black card-
board and the right alley unlined, or natural wood finish. On Days
53 through 58, the subjects were given three trials a day at fifteen.
minute intervals. Following the training period, 26 extinction trials
were run, one free-choice trial a day. In an attempt to determine the
influence of satiation and extra-maze cues on the learned response, the
following procedure was utilized. A three—day delay period was intro-
duced and three of the five remaining animals (with five animals the
response had been extinguished) were placed on an ad libitum feeding
schedule while the other two were continued on their regular feeding
schedule of 8 grams a day. On Day 88, these animals (three-satiated,
two-food-deprived) ran one trial. On Day 89, the maze was rotated 180',
reversing the extra-maze cues, and on Days 89, 90, and 91, the subjects
were given one free-choice trial a day. Two animals were discarded
from Group 1 on Day 2 when they failed to leave the starting box for
the second consecutive day. A 10-minute criterion period was used.
Group 2 (N-twelve animals) was given 24 days of training on
a R R L sequence, with the left alley of the maze lined with black card-
board and the right alley unlined. The animals then were given three
free-choice trials, after which, the training sequence was reversed
from R R L to L L R and was continued for 33 days. Then, the black
cardboard was moved so that the left alley was natural wood and the
right alley was lined with black and training trials were continued
for 12 additional days. An extinction period of 12 days followed. .
rinally, a three-day delay period was used, during which four subjects
were placed on an ad libitum feeding schedule while three animals
continued on the deprivation feeding schedule. One free-choice trial

was run on Day 88. The maze was rotated 180° and free-choice trials

19

were run on Days 89, 90, and 91. One animal was discarded from Group
2 on Day 2 when it failed to leave the starting box for the second con-
secutive day. .A 10-minute criterion period was used.

Groups 3 (N;five animals) and 4 (N-five animals) were given
19 training trials with both maze alleys lined with black cardboard,
Group 3 on a 2Lle ratio and Group 4 on a 2R:1L ratio. At the end of
the 19 days, the cardboard was removed from the left alley, leaving
the left alley a natural wood finish and the right alley black. Ac.
quisition trials were continued for 14 days. The procedure was changed
at this point so that Group 3 was forced to run to the left alley on
the first two trials of each block of three trials and Group 4 was
forced to run to the left on the first trial only. After the subjects
had run 16 trials on this schedule, the maze was rotated 180' and
eight final trials were run with the extraamaze cues reversed.

A partial replication was done with Groups 5 (N—ten animals)
and 6 (N-ten animals) in an attempt to eliminate the influence of the
unknown extra-maze stimulus operative with Groups 1 and 2. Groups 5
and 6 were given 51 training trials with the left alley lined with
black cardboard and the right alley a natural wood finish, Group 5 on a
2L:1R sequence and Group 6 on a 2R:IL ratio. The extinction period con-
sisted of ten days of one-a-day free-choice trials followed by three
days of five massed trials, with an inter-trial interval of approxi-
.mately twenty minutes. Three animals from Group 5 and one animal from
Group 6 were discarded on Day 2 when they failed to leave the starting
box for the second consecutive day. .A 10-minute criterion period was

usede

Group 1 N - lO
Acquisition:

hti nc ti on:

Group 2 N - ll
Acquisition:

Extinction:

Group3 N-5
Acquisition:

Group 4 N - 5
Acquisition:

52 days

26

19

14

(I)

19

14

16

20

TABLE I

DESICH‘I

1 trial a day, L L R, L - black, R - natural
3 trials a day, 15-minute interval

1 trial a day, free-choice trials

delay period - no trials

free-choice trial, 3 animals satiated

2 animals food-deprived
free-choice trials, maze rotated 180’

1 trial a day, R R L, L - black, R - natural
free-choice trials

L L R schedule

L L R, L - natural, R - black

free-choice trials

delay period - no trials

free-choice trial, 4 animals satiated

3 animals food-deprived
free-choice trials, maze rotated 180°

ltrialaday,LLR, LandR-black
R - black, L - natural
force L, force L, free-choice

force L, force L, free-choice
maze rotated 180°

1 trial a day, R R L, L and R - black
R - black, L - natural
force L, free-choice, free-choice

force L, free-choice, free-choice
maze rotated 180’

21

Group5 N-7

Acquisition: 31 days 1 trial a day, L L R, L - black, R - natural
Extinction: 10 1 trial a day, free-choice trials
3 5 trials a day, 20-minute inter-trial interval,

free-choice trials

Group 6 N - 9

Acquisition: 31 1 trial a day, R R L, L - black, R - natural
Extinction: 10 1 trial a day, free-choice trials
3 5 trials a day, 20-minute inter-trial interval,

free-choice trials

RESULTS

The experimental results have been analyzed in terms of the
responses made on the test trials, the first trial of each block of
three days. The number of left-turning responses made by each group
was used as the basis for statistically comparing performance trends
during acquisition and extinction.

Acqgisition Data

Figure 2 shows the percentage of animals in Group 1 running to
the right alley of the T-maze on free trials. This group was trained on
a 2L:1R ratio and after 52 days showed a consistent pattern with five
animals going to the right alley on each free trial and five animals
going to the left. During the six days, Days 53-58, when the subjects
were given three trials a day at fifteen-minute intervals, there were
no reversals in choice of alley. Each of the ten subjects ran all six
free trials in the same direction that it was running at the end of the
52 days of training with one trial a day.

On the 2R:1L ratio, 100% of the animals in Group 2 ran left on
five of the eight free trials during training, as can be seen in Figure
3. Free trials on three consecutive days, Days 25-27, yielded 100, 91,
and 83% of the subjects turning left. The training sequence was reversed
from.2R:lL to 2L:1R for 33 days and the percentage of animals making
left turns on any specific trial varied from.75 to 100. In.cther words,
the group showed no significant tendency to reverse; alternating re-
sponses did not occur at this stage of learning. On the thirteenth day

of this reversal period, a sudden drop in temperature resulted in the

28-3 :8 .8383". one .84 :8

ensuwusea mousse Henna one» gone so noesoaeeu mansnsauaomgu no «season .a shaman .

mm when

muse on as as as «a as 3 3 me 3 3 a... on. x..- .C. 2. as. an u~ 3 3 2 a
he: as I l2. 2 R «a at Me 3 3 Pk «at... C. t. on an .n .3 2 2 a.

 

l

 

| 'l' I'll

 

 

.III'I'II' ‘((1“‘1

 

 

 

a nsouc

 

OH

cu

on

on

93.

UUIDL tufita JO guessed

24

death of three animals in this group. .All three of the animals had
been running consistently to the left alley.

In a comparison of the mean number of left-turning responses
during learning made by the animals in Group 1 (M = 5.3, S.D. = 2.53)
and.Group 2 CM = 7.91, S.D. 8 .668) a.p value of 3.01 was obtained,
significant at the .02 level of confidence.

Figure 4 shows the percentage of animals in Group 3 trained on
the 2L:1R ratio making right turns and Figure 5 shows the percentage of
animals in Group 4 trained on the 2R:1L ratio making left turns. Both
groups were trained for 21 days with the entire maze lined with black
cardboard. On the test trial at the end of the period, 80% of Group 4
and 100% of Group 3 ran to the left alley of the maze. With identical
maze stimuli, these animals acquired the same response habit, namely,
running to the left alley. With the cardboard removed from the left
alley so that the maze stimuli were discriminable (natural wood finish
in the left alley and black cardboard in the right), twelve additional
trials showed no significant change in the number of animals running
left on free trials.

Figure 6 shows the Percentage of animals in Groups 5 and 6
making running responses to the less frequently visited alley of the
maze, right-turning responses for Group 5, left-turning responses for
Group 6. Both groups were trained for 31 days with the left alley
lined with black cardboard and the right alley a natural wood finish.
At the end of acquisition trials, Group’5 showed a tendency to make
right-turning responses on free trials, 71% on.Days 28 and 31, and
Group 6 showed the opposite tendency, 70% left turns on Day 28 and 89%
on Day 31.

In a comparison of the mean number of left-turning responses

Percent of Left Turns

100

90

80

70

60

50

30

20

25

WWW—

 

 

 

 

3/ 73

77
7" 91 9V

I 7 13 11 2f 27 33 3? yr‘ r1 r7 ‘3 ‘9 73 7r 77
q 10 [6 22. 2‘ 1" 36 V1 ‘lf rt! 60 u 71 74/ 7‘ 7:

Day.

rigure 3. Percent of left-turning responses on each test trial
during learning, Days 1-72, and extinction, Days 73-84.

26

made during the total acquisition period by the animals in Group 5
(M = 4.3, S.D. = 2.81) and Group 6 (M = 7.0, S.D. = 1.82) at value
of 2.05 was obtained which approaches the 5% level (2.145).
Extinction Data

In this study, extinction does not refer to the omission of
food but to the omission of the forcing procedure. During the extinction
period for Group 1, an extinction criterion of two reversals during a
period of three days was used. During the first 26 extinction trials,
the five animals which were turning left consistently at the end of
acquisition trials showed no reversals, each animal turned left on 26
consecutive days. Of the five animals which had been turning right at
the end of learning trials, four extinguished, i.e., ran twice to the
left alley within a three-day period and the fifth animal showed two
reversals but not on successive trials. After each animal extinguished,
no trials were given for a period of seven days. This was followed by
one free-choice trial; all four animals turned left on this test trial.

During the ten oneA-day extinction trials for Groups 5 and 6,
only one animal, a right-turning animal in Group 5, varied its running
pattern; it turned left on.Days 9 and 10. Day 31, the last day of
vauisition trials was a free trial and was followed by ten days of ex-
tinction or free trials. Massed extinction trials began on Day 42.
Including the trial on Day 31, the trials on Days 32 through 41 and the
first free trial of Day 42, there were twelve consecutive one-a-day
extinction trials. The total number of left-turning reaponses made by
each group during this twelve-day period was used to compare the per-
formance of the groups as it existed at the end of acquisition and
during spaced extinction. A _1_:_ value of 2.624 was obtained, significant

at the 2% level (Group 5, M a 3.7 left turns, S.D. = 5.29; Group 6,

Percent of Right Turns

100

90

80

70

60

50

20

10

- Group 3

 

Figure 4. Percent of right-turning responses on
each test trial.

28

M ='10.7 left turns, S.D. = 3.77).

It seems clear from these data that the subjects in the
differential alley maze did show learning on a one-trial-a-day schedule.
The response which was learned was running to the less frequently ex-
perienced alley, the R of the 2L:IR ratio (Groups 1 and 5) or the L of
the 2RzlL ratio (Groups 2 and 6) even when the left-turning tendency
due to an unknown extra-maze stimulus had been eliminated (Group 6).
All animals were exposed to both alleys throughout the learning trials
since they were forced on the third trial of each block to one alley
or the other, depending on which alley was entered on the test trial.
At least four possible response patterns could have developed on a one-
trial-a-day schedule in a T-maze with discriminable cues. The animals
could have shown no learning or random alley selection, they could have
alternated turning responses, they could have learned to turn in the
direction that had been most rewarding or, as they actually did, they
could have learned to run to the less frequently visited maze arm.

When the intraqmaze cues were reduced by lining both alleys with black
cardboard, Groups 3 and 4, there was no significant difference between
the response patterns of the two groups and thus no evidence of a de-
veloped preference for the less frequently visited side.
Subsidiary:Findings

Several subsidiary findings have been noted from the data.

The first question is that of whether or not original position prefer-
ences account for the turning habit acquired later in the experiment.
That is, did these animals ”learn? to go to the alley to which they re-
sponded on the first trial? 0n.Day1, three of the ten animals in.Group
1, three of the eleven in Group 2, three out of five in Group 3, three

out of five intsroup 4, four out of seven in Group 5, and two out of nine

'Eg 9H-‘ “0 “i-“nfl-“

Percent of Left Turns

30

20

10

29

r

 

 

Group 4

 

I 7 13 I, 22 2? 3r :9 W W 97 to 3'3 5‘6,
9 In I6 23‘ 31 3‘ 37 92 w‘ 49 .r/ 14/ :7

Day!

rigire 5. Percent of left-turning reeponeee on each test trial.

in Group 6 made left-turning responses. There is no indication from these
totals that the habit acquired was the same response as that made on Day
1, inasmuch as Groups 2, 3, 4, and 6 later acquired a left-turning re-
sponse habit while Group 5 acquired a right-turning response habit.
.Another factor apparent in the data of all six groups is the
strength of the habit once it has been.acquired. The extinction trials
can be regarded as a demonstration of fixation, considering how seldom
the animals responded to the non-habitual alley. With Group 2, for
instance, after the initial acquisition period of 24 days (2R:1L),
during which the group learned to make left-turning responses, there was
no change in response pattern during the following 33 days on the op-
posite schedule (2141B). During a subsequent twelve days of training
when the right alley was lined with black cardboard and the left alley
was natural wood finish, the percentage of animals turning left varied
from 75 to 100, the same as during earlier acquisition trials, Days 4-
60. On the first free trial after the colors of the alleys‘were re~
versed, the median running time went from the usual 3 seconds to 7.5
seconds. Hewever, the animals still ran to the left alley; possibly,
the habit was maintained due to the effects of the extraqmaze stimuli.
At the start of extinction trials, seven animals in.Group 2 were running
consistently to the left alley of the maze and one animal was running
consistently to the right alley. From.Day 73 to Day 77, extinction
trials, the percentage of left-turning responses varied from.75 to 100.
Then, from.Day 78 to Day.84, extinction trials, the seven animals which
had been turning left consistently at the end of the learning period
made only left-turning responses and the one animal which had been run-
ning to the right alley during learning trials made only right-turning

I‘OBPOIISBSe

31

$19.. :8 .3383". as. .34 :8 .2253 9.35 new» «.3 5..
no head. venue? ”avenues.“ need 23 ea eeeqoneeu «o useeuem .o earn

 

 

 

 

 

.58
at a) «a \k e) on us an 3.3 an an .«a R oak... an ex ex 2 es s r .

.1 o
. 3

. on

fi. 8

o in a 38.3

. 3

l'llllllllll'llllllle llll.el 8
i 8

> e

'/ .
P ”(IO)” P 4 . 8
. 8

tone pester; innenbea; not 0: cum: :0 sneezed

Massed trials were used with Groups 5 and 6 after the ten one-
a-day extinction trials to determine whether or not massing would have
any.more effect on the fixated habit of these animals than did the one-
a-day trials. During the massed extinction trials, five a day with 20-
minute inter-trial intervals, the animals in Group 5 made 13 reversal
responses out of a total of 105 responses or 12% while Group 6 made 11
such responses out of a total of 135, or 8%. Even this massing of trials
resulted in very little extinction; the fixation of the learned response
is apparent.

Some evidence regarding the effect of satiation on the running
habit has been tabulated. On Day 88, after three animals from Grmup l
and four animals from Group 2 had been satiated, one free-choice trial
was given. Two of the seven satiated animals ran to the alley opposite
to their learned response, one animal with a left-turning habit ran to
the right alley and one animal with a right-turning habit ran to the
left alley. None of the five control animals, food-deprived, reversed
directions.

0n.Days 89, 90, and 91, after the maze had been rotated 180°
so that the extraqmaze cues were reversed, six out of twenty-one re-
sponses or 28.6% by the satiated animals were reversals while four of
fifteen responses or 26.6% by the control animals were reversals. The

following table summarizes these data.

Group 1 Group 2
Reversals - Total responses
Satiated Control Satiated Control
Day 88 1-3 0.2 1-4 0-3
Maze Rotated
Day 89 1-3 1-2 3-4 2-3
Day 90 1-3 0.2 1-4 0-3
Day 91 0-3 192 0-4 0-3

The number of subjects included in this part of the study is

too small for the data to be anything more than suggestive.

DISCUSSION

These data seem to indicate that learning does occur with a
time interval of 24 hours between trials when there are discriminable
intra-maze cues. In approximately 30 days the animals in all six
experimental groups acquired strong habits either of running to the

left alley or of running to the right alley on each free-choice trial.

 

Previous theorizing in this area (11,27) would have led to no such i
prediction, on the basis of stimulus satiation or response factors,
since a 12-hour interval was considered to be the longest interval
over which a response was influenced by a preceding stimulus or response
event. Hullian theorizing would have led to the prediction that in the
event any learning took place in a situation involving a 24-hour delay,
the response which would be learned would be that of running to the
alley where the reward had been obtained the more frequently, the L of
the 2L:lR ratio or the R of the 2R:lL ratio. The animals in this ex-
periment did not learn to go to the more frequently rewarded alley nor
did they simply alternate their responses. An alternating habit would
have shown up during extinction, daily free-choice trials, in the form
of alternate right and left turns. As has been pointed out earlier,
the running habit, once acquired, was resistant to extinction in.most
cases.

Walker's study (30) in.which he analyzed the influence of the
three variables, indicated that intraqmaze cues were the most important,
extra-maze cues were of next importance, and the previous response was

of little, if any, importance in determining alternation behavior.

‘Using a cross-shaped maze, Walker found more animals making a response

which involved different maze stimuli than making a response which in-
volved different extra-maze stimuli or a different turning response, as
such. In the present study, Groups 1 and 2 which had discriminable
maze cues, learned different responses than Groups 3 and 4 which had
non-discriminable maze cues. 'Most of the animals, 8 out of 10, in
Groups 3 and 4 developed a habit of running to the left alley on free-
choice trials. Inasmuch as the lighting was constant and an auditory
making stimulus, a fan, was used, the discriminable extra-maze cues

to which these animals were responding are not known. However, the point
is, the extra-maze stimulus pattern was identical for all four groups
and Groups 1 and 2 with discriminable maze cues present learned on the
basis of the intraqmaze cues rather than on the basis of the extra-
maze cues. In.the absence of obvious intraqmaze cues, as was the case
with Groups 3 and 4, the extraqmaze cues apparently became the elicit-
ing agent in the situation. (Another fact which seems to indicate the.
importance of extra-maze stimuli in the present study is present in the
extinction data. The only animals which showed a tendency to extinguish
were those in Group 1 which had acquired a habit of running to the right
alley. The eliciting strength of the extraqmaze cues, apparently
located to the left of the maze, was sufficient to change the response
pattern of the right-turning subjects. It also seems to have been the
case that cues, intraqmaze or extraqmaze, play a more important role
early in the learning sequence than they do later. When Group 2 was
switched from a stimulus situation involving a black left alley and a
natural wood right alley to the opposite pattern on the 63rd day of
learning, no significant change in running habits occurred. Likewise,
With Groups 3 and 4 which were changed from.a stimulus situation in-

volving a black left and a black right alley to a black left alley and

 

35

a natural wood right alley on the 22nd day, no significant change in
response took place. A1though.Martindale (15) and Julian (13) found
that their animals reversed their running responses when the stimulus
pattern was reversed, the present study did not find this. Martindale
used an inter-trial interval of fifteen.minutes and Julian used an
inter-trial interval of two hours contrasted to the interval of 24 hours
used in the present study. It may be that after a response becomes
established on a schedule using a long inter-trial interval, drastic
changes in the stimulus complex are necessary to elicit a new response.
The maze rotation on Day 89 resulted in 28.6% reversal responses by the
satiated subjects and 26.6% reversal responses by the food-deprived
subjects on the following three days' trials. ,Maze rotation, i.e.,
change in extra-maze cues, on Day 89 resulted in more reversals than
changing the intra-maze cues on Day 63 with Groups 1 and 2.

The data of De Valois (6) indicate that low motivation, 6
hours of water deprivation or a light shock, leads to more variability
of response in a multiple maze than does high motivation, 22 hours of
water deprivation or a strong shock. The results on Day 88 in the
present study would seem to lend some support to this idea although the
number of animals was too small at that stage for being more than sug-
gestive. However, two of the seven satiated subjects did reverse their
turning response while none of the five control animals did. Inasmuch
as these animals had developed a habit of approximately three months'
duration, it is interesting that any of them.reversed their response,
and that the two which reversed were the satiated animals, low.motivation.

No theoretical analysis of the various research results re-
lated to alternation phenomena has been sufficiently comprehensive.

The attempts by various investigators to account for the research find-

 

ings invariably have left some contradictory evidence unexplained.
Hull's concept of a response-produced inhibitory or negative drive
state, Ir, includes the ideas that alternation of responses should in-
crease as a function of the increasing effort of the response and that
II. and the tendency to alternate should dissipate within a short time.
Experiments designed to test the first of these ideas, namely, that
alternation of response should increase as a function of increasing
effort of response, have yielded contradictory results. Solomon's (25)
animals increased alternation responses when inclined planes were used
as the method of increasing the effort of the response but they did
not increase alternation responses when weights tied to their backs
were used. Dennis (4) and Walker (28) have used increased effort as
the experimental variable and have found no increase in the tendency to
alternate responses. The work of Glanzer (8), Estes (7), and Walker (30)
'which found animals alternating intraemaze stimuli rather than the
specific response would not be predicted from the D, postulate.
Glanzer noted the theoretical weaknesses of Hull's concept,
Ir, and formed a new postulate regarding stimulus satiation, a state
which supposedly increases as an organism perceives a stimulus or
stimulus complex. Glanzer's concept can account for his own research
data, the results from which indicated that animals tend to alternate
stimuli rather than responses. However, the concept does not explain
Dember's data (3) which showed animals responding to a change in the
stimulus situation even though the stimulus involved was that to which
the subjects had been exposed, satiated. Neither does it explain
Walker's experiment (29) in which the passive perception of a stimulus
outside the maze.or in the goal box of the maze if the animal is

placed there instead of running there did not lead to an increase in

 

37

alternating responses. Glanzer's postulate and its corollaries do not
lead to the prediction of differential amounts of alternation result-
ing from inactive perception or active responding, running.

Montgomery (22) has attempted to clarify the situation by
naming the tendency of an organism to respond to new stimulus complexes,
a primary drive, the exploratory drive. Those drives with which
psychologists have been working in the past, e.g., hunger, thirst,
etc., have been shown to be lacking any precise definition or any
explanatory value (2). Therefore, adding one more specific drive seems
to be something less than beneficial to theory development.

Estes (7) states that alternation behavior may represent
transfer of training which has taken place during pretraining or in an
incidental manner in the laboratory environment, the home cages of the
animals. For instance, if there were a negative correlation between
the probability of receiving reinforcement in a given location and
the recency of visiting that location, alternation or responding to
new stimuli would be predicted. In.one study, Grossberg and Sprinzen
(10) reinforced some rats when they repeated an approach response to
a specific stimulus complex and reinforced other rats only when they
varied their responses. When tested in a Y-maze situation, the latter
group of animals alternated maze arms significantly more than the form-
er group.

The learning theory currently being developed by Danny and
Adelman (5) states in the Stimulus Satiation Postulate, "Thus, for all
practical purposes, learning occurs if and only if a response is elicit-
ed in a consistent manner." In the present experiment, when discrimi-
nable maze cues were used, it would seem to be the case that the stimulus

complex of the less frequently visited alley had higher eliciting value

than did the other alley and hence elicited a response in a consistent
manner. Apparently, this holds true only during the first few days,
approximately 30, while the turning habit is being acquired since after
that time few alternating responses occurred with changes of stimuli

or with extinction trials. One factor which undoubtedly contributed

to the relative absence of alternating responses late in the course

of the experiment was the long period of food deprivation of the animals.

Montgomery found less exploratory behavior by food-deprived subjects

 

than by satiated subjects and De Valois found fewer alternations in

the case of highly motivated animals than in the case of rats with low
motivation. Hence, an increase in the number of hours of food-depriva-
tion would seem to lead to less alternation or to a greater tendency

to fixate. The animals in the present study were kept on short rations
for approximately two—three months and this may have increased the

tendency to form fairly rigid response patterns.

SUMMARY

The present thesis was concerned with several of the factors
related to alternation behavior. The primary interest was in an at—
tempt to determine whether or not the stimulus satiation developed in
a specific situation decreases to zero in less than 24 hours. This F
was tested. A study also was made of the eliciting values of maze and !l

extraqmaze cues by the use of differential maze stimuli. The effects

 

of maze cues during the early trials of the learning period, before the
formation of a habit, were compared to the effects of the same cues
after a turning-response habit had been acquired. Lastly, a comparison
was made of the reversal tendency of satiated subjects with that of
food-deprived animals.

Forty-seven experimentally naive male rats were used in the
study. Group 1, tan animals, received the following training; (1)
52 days, one trial a day, 2L:IR ratio, (2) 6 days, 3 trials a day with
a fifteen-minute inter-trial interval, 21413 ratio, (3) 26 days, free
trials, followed by a three-day delay period, (4) 1 day, free trial,
wdth three subjects satiated and two subjects food-deprived, (5) 3 days,
free trials, maze rotated 180°. During the entire experiment, the left
alley was black and the right alley was natural wood finish. Group 2,
eleven animals, received the following training; (1) 24 days, 2RzlL
ratio, left alley .. black and right alley — natural wood finish, (2)
3 days, free trials, (3) 33 days, 2L:lR ratio, (4) 12 days, 2L:1R
ratio, left alley - natural wood finish and right alley - black, (5)
12 days, free trials, followed by a three-day delay period, (6) 1 day,

free trial, with four animals satiated and three animals food-deprived,

 

(7) 3 days, free trials, maze rotated 180°. Group 3, five animals,
received the following training; (1) 19 days, 2L:1R ratio, left and
right alleys - black, (2) 14 days, left alley - natural wood finish and
right alley - black, (3) 16 days, Forced L, Forced L, Free-choice se-
quence, (4) 8 days, maze rotated 180’. Group 4, five animals, receiv-
ed the following training; (1) 19 days, 2R:1L ratio, left and right

alleys - black, (2) 14 days, left alley - natural wood finish and right

4:7

alley a black, (3) 16 days, Forced L, Free-choice, Free-choice sequence, I

(4) 8 days, maze rotated 180°. All trials for Groups 2, 3, and 4 were

 

on a one-trial—a-day basis. Group 5, seven animals, received the
following training; (1) 31 days, 2L:1R ratio, left alley - black and
right alley - natural wood finish, (2) 10 days, free trials, one a day,
(3) 3 days, free trials, five a day with twentyeminute, inter-trial
intervals. Group 6, nine animals, received the following training;

(1) 31 days, 2R:1L sequence, left alley - black and right alley -
natural wood finish, (2) 10 days, free trials, one a day, (3) 3 days,
free trials, five a day with twentyeminute, inter-trial intervals.

The following results were obtained. The acquisition of a
habit did occur with all six groups even though the inter-trial
interval was 24 hours. Groups 1, 2, 5 and 6, with which differential
intraemaze cues were used, showed a tendency to respond to the less
frequently experienced stimulus complex. Groups 3 and 4, with which
non-discriminable maze cues were used, learned to go to the left alley,
presumably on the basis of extraemaze cues. The running response which
was learned was extremely resistant both to reversal when the maze
stimuli were changed and to extinction when the trials were all free-
choice. The tendency for satiated animals to show more reversals than

food-deprived animals was discussed. There was an indication that

41

extraemaze cues have some eliciting value after a running habit has
been developed as shown by the effects of maze rotation after 88 days
of running trials.

The limitations of various theoretical approaches to the

problem of alternation phenomena were discussed.

 

2.

3.

4.

5.

6.

13.

BIBLIOGRAPHY

Berlyne, D. E. Novelty and curiosity as determinants of exploratory
behavior. Brit. J. Psychol., 1950, 41, 68-80.

Behan, R. A. The Quantification of Drive: III. Privation of food
and water. Unpublished Ph.D. Dissertation, Michigan State Uni-
versity, 1954.

Dember, W. N. Response by the rat to environmental change. L
compdhysiol. Psychol., 1956, 49, 93-95.

Dennis, W. Spontaneous alternation in rats as an indicator of the
persistence of stimulus effects. J. comp. Psychol., 1959, 28,
305-512.

Denny, M. R., and Adelman, H. M. Elicitation Theory: I. An
analysis of two typical learning situations. Psychol. Rev.,
1955, 62, 290-296.

De Valois, R. L. The relation of different levels and kinds of
motivation to variability of behavior. J. exp. Psychol., 1954,
47, 392-398.

Estes, W. K., and Schoeffler, W. 8. Analysis of variables in-
fluencing alternation after forced trials. J. comp. physiol.
Psychol., 1955, 48, 557-562.

Glanzer, M. Stimulus satiation: An explanation of spontaneous
alternation and related phenomena. Psych. Rev., 1953, 60, 257-268.

Glanzer, M. The role of stimulus satiation in spontaneous alter-
nation. J’. exp. Psychol., 1953, 45, 387-393.

Grossberg, J ., and Sprinzen, H. Unpublished research, University
of Indiana, 1954.

Heathers, G. L. The avoidance of repetition of a maze reaction in
the rat as a function of the time interval between trials. _J_'_._
Psychol., 1940, 10, 559-580.

Hull, 6. L. Principles of Behavior. New York: Appleton-Century.
Crofts, 1945.

Julian, J. Unpublished research, Michigan State University, 1955.

 

14.

15.

16.

17.

18.

19 .

20.

21.

22.

23.

24.

25.

26.

2'7.

28.

29.

Kivy, P. N., Earl, R. W., and Walker, E. L. Stimulus context and
satiation. J. comp. phjfiiol. Psychol., 1956, 49, 90-92.

Martindale, R. Unpublished research, Michigan State University, 1954.

hiatus, 3., and Lange, E. Unpublished research, Michigan State
University, 1955.

Montgomery, K. C. Exploratory behavior as a function of similarity
of stimulus situations. J. comp. physiol. Psychol., 1955, 46,
129-153.

Montgomery, K. C. Spontaneous alternation as a function of time
between trials and amount of work. J. exp. Psychol., 1951, 42,
82-95.

Montgomery, K. C. Exploratory behavior and its relation to spon-
taneous alternation in a series of maze exposures. J. comp.
physiol. Psychol., 1952, 45, 50-57.

Montgomery, K. C. The relation between exploratory behavior and
spontaneous alternation in the white rat. J. comp. physiol.
Psychol., 1951, 44, 582-289.

Montgomery, K. C. The effect of the hunger and thirst drives upon
exploratory behavior. J. comp. physiol. Psychol., 1955, 44, 515-519.

Montgomery, K. C. A test of two explanations of Spontaneous alter-
nation. J. comp. physiol. ngchol” 1952, 45, 287-294.

Montgomery, K. G. The effect of activity deprivation upon explora-
tory behavior. J. comp. physiol. Psychol., 1955, 44, 458-441.

Rothkopf, E. Z., and Zeaman, D. Some stimulus controls of alter-
nation behavior. J. Psychol., 1952, 54, 255-255.

 

Solomon, R. L. The influence of work on behavior. Psychol. Bull.,
1948, 45, 1-40.

Tolman, E. 0. Purpose and cognition: The determiners of animal
learning. Psychol. Rev., 1925, 52, 285-297.

Walker, E. L. Reward, Time, and Variability. Mimeographei paper,
1955.

Walker, E. L., Dember, W. N., Earl, R. W., Fawl, C. L., and Karoly,
A. J. Choice alternation: III. Response intensity vs. response
discriminability. J. comp. physiol. Psychol., 1955, 48, 80-85.

Walker, E. L., Dember, W. N., Earl, R. W., Fliege, S. E., and
Karoly, A. J. Choice alternation: II. Exposure to stimulus or
stimulus and place without choice. J. comp. physiol. Psychol.,
1955, 48, 24-29.

 

44

50. Walker, E. L., Dember, W. N., Earl, R. W., and Karoly, A. J.
Choice alternation: I. Stimulus vs. place vs. response. J. comp.
phypiol. Psychol., 1955, 48, 19-24.

 

51. Wingfield, R. 0. Some factors influencing spontaneous alternation
in human subjects. J. comp. Psychol., 1945, 55, 257-244.

52. Zeaman, D., and House, B. J. The growth and decay of reactive
inhibition as measured by alternation behavior. J. exp. Psychol.,
1951, 41, 177-186.

 

{1'
c" {315511 5:1
4 f\ h‘ I‘ ‘9 '£.- l
. , «9,, a
‘ L‘J“

Demco-293

 

Date Due

 

NW!

UBR
8 3065

V.
-II
S
R
E
W

l AN STATE UN

JIHIWIHH

mmnn

3 1293 0317

WWW

IHIIH