Table Of Contenttm goiE of maotisuis of kkward in m rnxxm m
inSfmmMkh FBBFOBIKUrCB ism *
Donald E*e" Swisher
A dissertation submitted In partial fulfillm ent of the
requirements for the degree of Doctor of Philosophy
in the Department of Psychology in the
Graduate College of the State
University of Iowa
August 1951
ProQuest Number: 10598631
All rights reserved
INFORMATION TO ALL USERS
The quality of this reproduction is dependent upon the quality of the copy submitted.
In the unlikely event that the author did not send a complete manuscript
and there are missing pages, these will be noted. Also, if material had to be removed,
a note will indicate the deletion.
uest
ProQuest 10598631
Published by ProQuest LLC (2017). Copyright of the Dissertation is held by the Author.
All rights reserved.
This work is protected against unauthorized copying under Title 17, United States Code
Microform Edition © ProQuest LLC.
ProQuest LLC.
789 East Eisenhower Parkway
P.O. Box 1346
Ann Arbor, Ml 48106 - 1346
ACKHOW hEDGMEOTS
fhe w riter wishes to express his indebted-
ness to Professor Kenneth W. Spence fo r his gen*
erous assistance through a il phases of this re
search and in the preparation of the manuscriptj
to Professor E# G# Gross for the generous loan of
experimental animals from the Pharmacology vivar
ium; and to Ire# Elisabeth MeG* Swisher for tech
nical assistance in the research#
11
TABLE OF GOmmtB
Page
Introduction ........................................... I
Experimental Method ............. 10
Apparatus . * . ................................. 10
Subjects ............. 11
Procedure ........................................ 12
Results .................... 19
Treatment of the Bata • • ........... 19
In itia l T rain in g ............. ....................... 20
Training Under the Changed Reward
Conditions .................... 25
Analysis of the Post-*shift Data for
the Hooded Strain .................. 32
Analysis of the P ost-shift Data for
the Albino Strain ........... 56
Analysis of the Pooled Post-switch
Bata ............................ 40
Discussion of Results ........................ 45
The In itia l Training ........................... 45
Training Under the Changed Reward
Conditions ....................... 65
Summary and Conclusions 60
References ..........*............................................ 62
Appendix A ............................ ................. 63
Appendix B ............... .. 65
ill
TABLE OF TABLES
M S> £SSS
I Summary fable of the Two-dimensional
Factorial Analysis of Variance Performed
on the Response Speed Measures at the
Final T rial of In itia l Training . ♦ ...............*.. 24
IX Summary Table of the Analysis of Vari-
anoe on Trends of the four Switch Groups
of Hooded Animals, the 2.5 and 0*05
Hooded Control Croups .......... 54
XXI Summary Table of the Analysis of Vari
ance on the Trends of the Four Switch
Croups of Albino Animals, the 2*5 and
0*05 Albino Control Croups . . . . . . . ........................ 37
IV Summary Table of the Analysis of Vari
ance on the Trends of the Four Switch
Groups, the 2*5 and 0.05 Control Croups
for Both Strains of Animals .......... 41
V Summary Table of the Simple Analysis of
Variance Performed on the Lumped Scores
for the Final Four Trials of Switched
Training for Four Switch Groups, Strains
Pooled .............. *........... 44
iv
TABLE OF FIGURES
Figure Fftge
1 Mean response speed as a function of
tria ls for a ll animals under two con
ditions of magnitude of reward in
in itia l tra in in g . ............ 21
2 Mean response speed as a function of
tria ls for two strains of animals un
der two conditions of magnitude of
reward In In itia l training *********•*****•.* 25
3 Mean speed of response over blocks of
two tria ls for five group® undergoing
varying degrees of change in reward and
for two control groups remaining at lev
el® of reward experienced In in itia l
training «*»«*«•...................• *. . * •.......... 26
4 Tops Mean speed of response over block®
of two tria ls for the groups shifted to
*026 gau reward, and for the 2*5 gnu
control groups# Bottoms Mean speed of
response over bloeEa of" two tria ls for
the group® shifted to *060 gm* reward,
and for the *050 gnu control groups . • * * 30
5 Topi Mean speed of response over blocks
of two tria ls for the groups shifted to
*100 gm* reward, and for the 2*6 gm*
control groups* Bottoms Mean speed of
response over bloc Iks of two tria ls for
the groups shifted to *200 gm* reward,
and for the 2*3 gm. control group® *****......... 51
V
I
Chapter I
im m m onon
In the Principles of Behavior (4) Hull assume#
that varying the magnitude of reinforcement in an instru
mental learning situation would affect the maximum (Mg) to
which the hahit being learned would grow* This assumption,
taken together with the assumption to the effect that hab
i t strength, once acquired, is not subject to dissipation,
led to the deduction that, other things being equal, once
learning had reached or nearly reached maximal strength
for the particular experimental conditions being used, a
reduction in reward should lead to no decrement in
performance*
This theoretical structure, in common with a
great deal in the Principles, was a wbeet guess” based on
the admittedly inadequate experimental data available at
the time. Subsequently there have appeared several stud
ies centering on the effect of change in reward conditions
during learning, and on the influence of different reward
conditions from thb beginning of learning. Sfotable among
these were studies by Grespl (1, 3) and Seaman (IS) which
dealt with both these variables and which gave results
2
sharply at variance with the conclusions drawn from Hullfs
V
theoretical position with regard to the effect of change
in reward conditions* Using an instrumental running situ
ation, both experimenters found that although d ifferen tial
magnitudes of reward did lead to different asymptotic la
tency values, a reduction in magnitude of reward led to an
almost immediate decrement in speed of performance* In ad
dition it was found that this decrement was so considerable
as to render the post-change asymptotes actually higher
(i.e ., longer latencies) than those for control groups
which had experienced only the smaller rewards throughout
the course of learning* To this phenomenon Crespi gave
the name ^depression effect,0 and spoke of i t as the re
sult of changes in motivation accompanying the change in
reward as a side-effect*
In an attempt to rectify his systematization In
the direction of accounting for the Zeaman and Crespi data,
Hull (5) has recently proposed several modifications of
his postulate system* According to Hull* a new position,
magnitude of reward Is assumed to affect the value of a
new construct (K) which is conceived as functioning much
as did motivational factors In the earlier system of the
Principles. Thus K becomes one of a set of Intervening
f5
variables that m ultiply existent habit strengths (&%) to
determine excitatory potential (&%) * te ls effectively
makes the acquisition of habit or cognitive factors inde
pendent of magnitude of reward, while retaining the depen
dence of performance measures on this variable*
$here Is serious reason, however, to question
both the systematic wisdom of this modification and the
adequacy of the data on the basis of which i t was adopted*
Systematically i t has the fau lt of implying that K is
uniquely determined by magnitude of reward, and that there
fore any performance change accompanying change in reward
should take place after only one experience with the new
reward* Although the data reported by Zeaman and Crespi
lend some support to the belief that such changes do take
place quite rapidly, the typical curve of experimental ex
tinction, representing the effects of change to aero reward,
suggests that the change is far more gradual than either
the Hull hypothesis or the Zeaman and Crespi data would
Indicate*
It should be noted as well that both the studies
of Zeaman and Crespi utilized learning situations involving
the traversing of a runway. Ihe Instrumental activity of
locomotion or approach as such must be presumed, in the
4
mature ra t, to have high habit loadings with a great many
stimuli* Consequently runway learning Is at best a pro-
seas of giving greater relative strength to one of the few
highest members of the habit family heirarchy. there la
no a p rio ri reason for assuming that the results from such
runway studies ean then be safely generalized to the ease
where in itia l habit strength Is low*
It is probable that such Improvement In perform
ance as does appear in the course of eighteen to twenty
tria ls in the runway situation does not reflect substantial
change in the strength of the tendency to approach the food
receptacle so much as It reflects extinction of exploratory
tendencies and adaptation to the fear-evoking properties
of the experimental surround* Farber (3) has demonstrated
that this la st can be experimentally a very important vari
able, and that I t is markedly affected by eating in the
presence of the fear-evoking stimuli* Consequently It
should be expected that as more opportunity to eat is af
forded with successive tria ls , an improvement in perform
ance should be noted* It should also follow that, inasmuch
as a larger reward involve® more eating In the presence of
the fear-evoking stim uli, animals receiving larger rewards
should perform better than animals receiving small rewards*