Group Title: contribution of nonassociative factors to cebus monkey eyelid conditioning performance
Title: The contribution of nonassociative factors to cebus monkey eyelid conditioning performance
CITATION PDF VIEWER THUMBNAILS PAGE IMAGE ZOOMABLE
Full Citation
STANDARD VIEW MARC VIEW
Permanent Link: http://ufdc.ufl.edu/UF00097790/00001
 Material Information
Title: The contribution of nonassociative factors to cebus monkey eyelid conditioning performance
Physical Description: viii, 83 leaves : illus. ; 28 cm.
Language: English
Creator: Cook, William Andrew, 1942-
Publication Date: 1968
Copyright Date: 1968
 Subjects
Subject: Eyelid conditioning   ( lcsh )
Monkeys   ( lcsh )
Psychology thesis Ph. D   ( lcsh )
Dissertations, Academic -- Psychology -- UF   ( lcsh )
Genre: bibliography   ( marcgt )
non-fiction   ( marcgt )
 Notes
Thesis: Thesis - University of Florida.
Bibliography: Bibliography: leaves 77-83.
Additional Physical Form: Also available on World Wide Web
General Note: Manuscript copy.
General Note: Vita.
 Record Information
Bibliographic ID: UF00097790
Volume ID: VID00001
Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
Resource Identifier: alephbibnum - 000559206
oclc - 13447206
notis - ACY4654

Downloads

This item has the following downloads:

contributionofno00cookrich ( PDF )


Full Text






THE CONTRIBUTION OF NONASSOCIATIVE
FACTORS TO CEBUS MONKEY EYELID
CONDITIONING PERFORMANCE










By
WILLIAM ANDREW COOK


A DISSERTATION PRESENTED TO THE GRADUATE COUNCIL OF
THE UNIVERSITY OF FLORIDA
IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE
DEGREE OF DOCTOR OF PHILOSOPHY










UNIVERSITY OF FLORIDA
1968


































TO LYTN AND KONTY












ACKNO'LBDC L I `71-F


I am especially indebted to Dr. F. S. PennyDacker

who, in addition to being a conscientious employer and

chairman, has been a good friend. Without his unwaivering

support and patient guidance this project would not have

reached completion.

I want to sincerely thank Dr. Bradford N. Bunnell,

Dr. C. Michael Levy, Dr. William Mendenhall, and Dr. Wise

B. Webb for the special efforts they have made to insure that

I have acquired in my graduate career the skills and knowledge

necessary to make a useful contribution to the profession of

psychology, and for their assistance in the development of

this dissertation.

I would like to acknowledge the staff and facilities

of the University of Florida computing center for their

significant role in analysis of this study. I am especially

grateful to Edwin Bradley for his assistance and for the use

of his regression program, UP STAT 10.

I would like to acknowledge the invaluable assistance

given so cheerfully by my wife, Lynn, and by Gale Lee who so

ably typed and retyped the various phases of this study.

This research was jointly suonorted by TUnited States

Public Health Service Grants MH-08S87 and -H-06379.










TABLE OF CONTENTS



ACKNOWLEDGMENTS ........................................ 1i i

LIST OF TABLES ......................................... v

LIST OF FIGURES ....................................... vi

INTRODUCTION ...........................................

METHOD ..................................... .......... 12

RESULTS ........... .......... ........ ... .............. 24

DISCUSSIO ............................................. 59

SUMMARY ................................................ 69

APPFE' IX A ........................................... 71

APPENDIX B ............................................. 73

APPENDIX C ............................ ........ ....... 75

OREFERENCES ...... ............................... ....... 77

BIOGRAPHICAL SKETCH ................... ............... ...












LIST OF TABLES


Table Page

1 Overall Experimental Design .................. 16

2 Factorial Paradigms for Stage 2 ................ 17











LIST OF FIGURES


Figure Page

1 Monkey Restraining Chair and Headholder ........ 13

2 Off-trial Response Rate as a Function of
Blocks of 25 Trials in Stage 1 ................. 26

3 Off-trial Response Rate Corrected for
Regression on Initial Blink Rate as a
Function of Blocks of 25 Trials for the
CS-UCS Factorial Experiment in Stage 2 ......... 28

4 Off-trial Response Rate Corrected for
Regression on Initial Blink Rate as a
Function of Blocks of 25 Trials for the
CS/UCS Factorial Experiment in Stage 2 ......... 29

5 Mean Uncorrected On-trial Response Rate as
a Function of Blocks of 25 Trials in Stage 2 ... 30

6 On-trial Response Rate Corrected for
Regression on Off-trial Response Rate as
a Function of Blocks of 25 Trials for
the CS-UCS Factorial Experiment in
Stage 2 ........................................ 31

7 On-trial Response Rate Corrected for
Regression on Off-trial Response Rate as a
Function of Blocks of 25 Trials for the
CS/UCS Factorial Experiment in Stage 2 ......... 33

8 Off-trial Response Rate Corrected for
Regression on Initial Blink Rate as a
Function of Blocks of 25 Trials for the
CS-UCS Factorial Experiment in Stage 3
with Stage 2 Treatments Continued .............. 37

9 Off-trial Response Rate Corrected for
Regression on Initial Blink Rate as a
Function of Blocks of 25 Trials for the
CS-UCS Factorial Exoeriment in Stoae 3
for Ss Shifted to CS-UCS ....................... 3






LIST OF FIGURES (CONT.)


Figure Page

10 Off-trial Response Rate Corrected for
Regression on Initial Blink Rate as a
Function of Blocks of 25 Trials for the
CS/UCS Factorial Experiment in Stage 3
with Stage 2 Treatments Continued .............. 39

11 Off-trial Response Rate Corrected for
Regression on Initial Blink Rate as a
Function of Blocks of 25 Trials for the
CS/UCS Factorial Experiment in Stage 3
for Ss Shifted to CS-UCS ....................... 40

12 On-trial Response Rate Corrected for
Regression on Off-trial Response Rate as
a Function of Blocks of 25 Trials for the
CS-UCS Factorial Experiment in Stage 3 with
Stage 2 Treatments Continued .................... 41

13 On-trial Response Rate Corrected for
Regression on Off-trial Response Rate
,as a Function of Blocks of 25 Trials
for the CS-UCS Factorial Experiment in
Stage 3 for Ss Shifted to CS-UCS ............... 42

14 On-trial Response Rate Corrected for
Regression on Off-trial Response Rate as a
Function of Blocks of 25 Trials for the
CS/UCS Factorial Experiment in Stage 3
with Stage 2 Treatments Continued .............. 43

15 On-trial Response Rate Corrected for
Regression on Off-trial Response Rate as a
Function of Blocks of 25 Trials for the
CS/UCS Factorial Experiment in Stage 3 for
Ss Shifted to CS-UCS ........................... 44

16 Off-trial Response Rate Corrected for
Regression on Initial Blink Rate as a.
Function of Blocks of 25 Trials for the
CS-UCS Factorial Experiment in Stage 4
with Stage 2 Treatments Continued in
Stage 3 ...............**** *....... ............ 49

17 Off-trial Response Rate Corrected for
Regression on Initial Blink Rate as a
Function of Blocks of 25 Trials for the
CS-UCS Factorial Experiment in Stage 4 for
Ss Shifted to CS-UCS in Stage 3 ................ 50


vii





LIST OF FIGURES count. )


Figure Pare

18 Off-trial Response Rate Corrected for
Regression on Initial Blink Rate as a Function
on Blocks of 25 Trials for the CS/UCS Factorial
Experiment in Stage 4 with Stage 2 Treatments
Continued in Stage 3 .......................... 51

19 Off-trial Response Rate Corrected for
Regression on Initial Blink Rate as a
Function of Blocks of 25 Trials for the CS/UCS
Factorial Experiment in Stage 4 for Ss Shifted
to CS-UCS in Stage 3......... ............... 52

20 On-trial Resnonse Rate Corrected for
Regression on Off-trial Res-onse Rate as
a Function on Blocks of 25 Trials for
the CS-UCS Factorial Ex eriment in
Stage Li with Stage 2 Treatments Continued
in Stage 3 .................................... 53

21 On-trial Response Rate Corrected for
Regression on Off-trial Response Rate
as a Function of Blocks of 25 Trials
for the CS-UCS "actorial T::-orinent in
Stage 4 for Ss Shifted to CS-UCS in
Stage 3 ....................................... 5L

22 On-trial Response Rate Corrected for
Regression on Off-trial Response Rate as a
Function of Blocks of 25 Trials for the
CS/UCS Factorial Experiment in Stage 4 with
Stage 2 Treatments Continued in Stage 3 ....... 55

23 On-trial Response Rate Corrected for Re-
gression on Off-trial Response Rate as a
Function of Blocks of 25 Trials for the
CS/UCS Factorial Exoeriment in Stage 4 for
Ss Shifted to CS-UCS in Stage 3 ............... 56


viii












INTRODUCTION


The traditional definition of classical conditioning

is in terms of the classic experiment of Pavlov--

". . the intake of food (UCS) by the animal
takes place simultaneously with the action of
a neutral stimulus which has been hitherto in
no way related to food, the neutral stimulus
(CS) readily acquires the property of eliciting
the same reaction in the animal as would food
itself." (Pavlov, 1927, p. 26)

However, there are some serious difficulties with defining

the supposedly associative process, classical conditioning,

in such simple terms. These difficulties are of three

basic types. For most of the responses used in conditioning

studies there are few, if any, absolutely neutral stimuli.

Secondly, in addition to true conditioning, certain non-

associative processes are active in the conditioning

situation which also result in an increased rate of responding.

Thirdly, in most types of classical conditioning, the situation

is further complicated by the possibility of the spontaneous

occurrence of the "conditioned" response. We shall briefly

consider the past research on each of these contaminating

factors.

Orienting Responses

Novel stimuli tend to elicit an orienting response;

and not infrequently, the response which is being conditioned

is one of the components of the generalized orienting response.

1




2

Hence, the "neutral" stimulus may elicit the very response

being conditioned. This phenomenon has been found in con-

ditioning of the eyelid (Hilgard, 1934), galvanic skin

response (GSR) (Prokasy & Ebel, 1967), salivation (Kotaki

& Miyata, 1958; Pavlov, 1927), and cardiac reflexes (Zeaman

& Smith, 1954). The orienting response component (OR) can

sometimes be distinguished from the true conditioned

response (CR) by the form or latency of the response. Such

ORs have been isolated for eyelid conditioning (Hilgard,

1934) and for GSR conditioning where a relatively long CS-UCS

interval is used (Prokasy & Ebel, 1967; Stewart, Stern,

Winokur, & Fredman, 1961).

In most other cases it has been assumed that the

novelty of the CS is decreased over the first few condition-

ing trials or over a short series of adaptation (CS alone)

trials. Thus, by reducing the novelty of the CS, the likeli-

hood of the occurrence of the OR is assumed to be sufficiently

reduced as to eliminate the OR as a contaminating factor.

However, Korn and Welch (1962) have shown that, at least in

the case of the GSR, the effect of adaptation trials is only

partially effective and is also unstable. In the last half

of 25 light-only trials these investigators found a residual

response rate of 50o. Although Kimmel and Goldstein (1967)

found a residual effect of GSR adaptation trials from one

session to another, their data still showed considerable

recovery between sessions. This latter finding suggests

that studies using more than one session would require





3

adaptation prior to each session--the effects of such a

procedure would not then be clearly distinguishable from

partial reinforcement. Furthermore, it has been shown ,.

Grant, Hake, and Schneider (1948) for the eyelid response and

Carlton and Vogel (1967) for the conditioned emotional res-

ponse (CER), that habituation to the CS seriously impairs

conditioning. The results of Grant et al. (1948) cannot be

attributed to the removal of the ORs since they compared only

responses of latencies longer than that of an OR.

It should be pointed out here that the eyeblink OR

to a tone has not been demarcated for conditioning in the

cebus monkey. It is possible that the response, if it

exists, may differ either in latency or in form from that

found in human eyeblink conditioning.

Pseudoconditioning and Sensitization

In addition to the difficulties mentioned above in

regard to adaptation, several studies have shown an increase

in both frequency and amplitude of the OR with the introduc-

tion of the UCS. This increment has sometimes been referred

to as pseudoconditioning (Martin, 1962) or as sensitization

(Wendt, 1933). The term pseudoconditioning will be used here

in a broader context meaning all response increments which

are not solely "a function of the repetition of the condi-

tioned and unconditioned stimuli in Drecise relationship"

(Hilgard & Marquis, 1940, p. 42). Hull (1934) used the term

alpha conditioning to designate the increase in size or

frequency of the unconditioned response (UCR) to the CS or




4

the reappearance of an habituated OR resulting from the

presentation of the CS and UCS together. Several positive

instances of alpha conditioning have been reported for human

eyeblink conditioning (Bernstein, 1934; Grant & Adams, 170L:

Weber & Wendt, 1942; Wendt, 1933) and for GSR conditionin-

in humans (Gale & Stern, 1967; Stewart et al. 1961; Ziry,

Stern, & Fjeld, 1966). -Ho;%ever, there have been some

failures to find alpha conditioning in human eyeblink

conditioning (Grant, 1943a; Eilgard & Fiel, 1937; cAllister,

1953a; Weber & Wendt, 1942).

Sensitization may occur without the pairing of the

CS and UCS. It may occur after presentations of the UCS

alone or after unpaired CS and UCS presentations (CS/UCS)

(Grant & Meyer, 1941; Grether, 1937; Grosberg, 1962; arlow,

1939; T-arris, 1941a, 1941b; Martin, 1962; Sears, 193-;

Shipley, 1933; Wickens & 'ickens, 19L2; Zimny et al., 1966).

The frequency of such occurrences of this tyoe of pseudo-

conditioning has resulted in the almost standard use of a

control group which is presented both the CS and the UCS

but unpaired. There are some serious theoretical difficul-

ties with the use of this as a control, however.

Zimny et al. (1966) and Prokasy, Fall, and -awcett

(1962) have shown very different performance in extinction

for Ss receiving the CS and UCS unpaired and Ss receiving

the UCS alone. In both studies the latter group responded

at a significantly higher rate. Zimny et al. explained the

difference in terms of the Sokolovian hypothesis which





5
stresses the importance of the significance of a stimulus

which is to serve as a CS; according to this hypothesis

repeated presentations of the CS leads to a decrease in the

significance of the stimulus. The conditioning group (CS-

UCS) in both studies had a higher response probability than

the CS/UCS and UCS alone control groups during extinction.

This could be merely because the response tendency is en-

hanced more by the conditioning process than by sensitization;

or it could be that when the CS and UCS occur together, the

CS fails to lose its significance (Jouvet, 1960).

Rescorla (1967) makes the point that the CS may even

serve as a sort of safety signal in the typical paradigm in

which the CS and the UCS are presented in an unpaired

sequence. This occurs, according to Rescorla, because the

CS is only presented in the absence of the UCS. His remedy

for this situation is to present the CS and the UCS in a

truly random fashion with two independent programs.

There are yet other factors that may result in sen-

sitization of responses to the CS. Grant and his co-

workers (Grant, 1943a, 1943b; Grant, 1945; Grant & Norris,

1946; Grant, Norris, & Boissard, 1947; Grant & Norris, 1947)

found a long latency UCR to a visual CS which was sensitized

by dark adapting S:s eyes. The response, named the beta

response, was very similar in form and latency to the CS.

Grant and Norris (1947) were able to differentiate between

the latencies of the beta response and the CR by comparirn

latency distributions of Ss receiving conditioning trials in




6

either light or dark. Studies since the findings of Grant

and his co-workers have eliminated the problem of beta re-

sponding by using light adaptation rather than dark adapta-

tion. Many studies prior to that time, however, were

undoubtedly seriously contaminated by beta responses.

Spontaneous Responding

For several of the response modalities used in

classical conditioning there have been observed spontaneous

occurrences of the responses. Spontaneous responses are

particularly common in GSR conditioning (Korn & Welch, 1962)

and in salivary conditioning (Kotaki & 1Hiyata, 1958). Ost

and Lauer (1965) found spontaneous responding of such

magnitude in their canine salivary conditioning that they

analyzed their results in terms of difference scores (Inter-

stimulus salivation rate Pre-trial salivation rate).

Perhaps the limiting case of contamination by spontaneous

responding of conditioning phenomena is found in hecrt-rate

conditioning where the task is that of modification of the

rate of a continuous sequence of responses (Elack, 1965).

Spontaneous blink rate (SBR) has been the object of

attention in articles dealing with a wide variety of exoreri-

mental settings. Among these aro article deall""'- ;ih

visual-motor tasks (Drew, 1951), ease of reading (Luckiesh,

1946, 1947), illumination (Bitterman, 1948; Ponder & Kennedy,

1928), humidity (Ponder & Kennedy, 1928), effort extended in

mental work (Bitterman & Solway, 1946), ease of seeing

(Luckiesh & Moss, 1942), manifest anxiety (Doehring, 1957),





7

introversion-extraversion (Martin, 1958), neuroticism ( ranks,

1963; Martin, 1958), muscle tension (Meyer, 1953), alcohol

consumption (Franks, 1964), visual attention (Kennard S

Glaser, 1964), and instrumental conditioning (Doehring &

Ferster, 1962).

Few studies, however, have analyzed the role of

spontaneous blinking in the eyelid conditioning laboratory.

Primarily, the corrections for SBR have been indirect in

nature: Kimble (1947) used CS-only presentations as catch

trials during acquisition; McAllister (1953b) used subtraction

of the percentage of responses on the first block of trials

from the succeeding blocks; Prokasy, Ebel, and Thompson (1063),

and also Foneau (1958), counted as CRs only those responses

which overlaoped the UCS onset; Hanche and Grant (1960)

corrected for S3R by fitting a straight line to their inter-

stimulus interval function and using the deviation from the

line as the corrected measure of conditioning.

Recently, SBR itself has been measured in several

eyelid conditioning studies. Levey and Martin (1967) measured

SER prior to conditioning. It has been measured throughout

the inter-trial interval (a-ttson ?- notre, 1964; Moore &

yewman, 1964). It has also been measured during a 10 sec.

period immediately prior to the CS onset (Liokin & A'oore,

1966). In another study the number of blinks every 60 sec.

irrespective of the Dresence of a trial was counted and

STR calculated (Deaux & Deaux, 1967). The S?7 measure for

Deaux and Deaux (1967) is correlated a -riori with their




0

measure of conditioning since the former measure includes

the latter. Also, the pre-trial asure of Likin and :oore

(1966) may not be wholly independent of the coditicned re-

sponse frequency measure because of the refractory period of

the blink apparatus.

Cook (1966) measured S3R for 60 sec. before and after

each block of 25 trials in ar eyelid conditioning study

with ccbus monkeys. Despite the crudeness of the S3E

measures and the fact t S 3 fluctutes widely fro- moment

to moment (Drew, 1951: Ponder & ::7 :-"- 1928), he still got

a correlation bet"we- n inter-tr il af on-triLa respor i-

of 0.90 for accuisLtion and C.95 for extinction. Both of

these correlations were hihly significant. The high cor-

relations between SBR and frequency of Cs found by Cook

suggest that any effect wich the ex:eri mental treatments

have on S3R will be of pra.nount i-portance in determining

the frequency of CRs. Spance and Deoux (1966) and Deaux

and Deaux (1967) both found an increase in 3S2 in hur 2a Ss

which were changed from CS-UCS to CS/UCS. Deaux and Deau

(1967) also noted a significant difference in SIR during

acquisition between the four highest conditioned res.oners,

An experiment is needed which will allow the CR to

be defined, if conditioning is actually taking place, in tho

attempts to condition the blink reaction of the c ..

monkey. Hil'ard and Iara.ic (1936) in their .'-.--nt v.o

condition the rhecus monk'y rcp rted data .whiUMhwere

unlike that found in the I'tui by Cook (105f a ----7.





9

and Cook (1967) in the cebus and squirrel monkey. Instead of

the initial maximal level of responding and large within-S

variation found by the latter authors, Hilgard and Marquis

reported a smooth sigmoid curve for individual Ss beginning

at an almost zero level of responding and increasing gradually

to a level of better than 90% over several days.

The low response rate found in the monkey eyelid con-

ditioning studies of Cook (1966) and Pennypacker and Cook

(1967) suggests that the CS may actually have some inhibitory

effect which counteracts the effects of conditioning. Pos-

sible support for this hypothesis is seen in the finding

of Cook that for some blocks of trials SBR was higher than

the on-trial response rate. Mourant (1965) also found

evidence for an inhibitory effect of the CS in squirrel

monkey eyelid conditioning in that a loud CS produced con-

ditioning significantly inferior to that of a soft CS. All

three of these studies, however, had an additional result

which suggests that rather than a simple CS effect their

results reflected the interaction of the CS and the UCS. All

three studies showed an increase in response rate for some

Ss in the initial trials of extinction. This phenomenon,

labeled inhibition of reinforcement by Hovland (1936), is

not uncommon in human eyelid conditioning (Prokasy, 1958).

Hence, for a truly definitive study it is essential

that an assessment of the effect of the presence of the CS

on blink probability be provided. A measurement of the

degree to which this effect changes across time and with




10

unpaired presentation of the UCS is also needed. Yot only

does SBR appear to require investigation as a potent

variable, but the degree to which SBR interacts with all

of the above-mentioned effects seems to warrant consideration

as well. The rate of blinking can be measured either during the

CS presentations or during a comparable blank trial--on-trial

response rate (On-trial RR)--or it can be measured during

the inter-trial interval--off-trial response rate (Off-trial

RR). The comparisons outlined above require both On-trial and

Off-trial RR measurements for Ss receiving no CS, the CS

alone, and the CS and UCS--both paired (CS-UCS) and unaaired

(CS/UCS). To completely assess the interaction of the CS and

the UCS it is also necessary to include Ss which receive the

UCS only. Jensen (1961) has pointed out that the use of a

2x2 factorial design with CS presence and absence orthogonal

to UCS presence and absence would be an efficient means of

assessing the interaction of the CS and UCS on response rate.

The presence of such an interaction, depending on whether the

CS and UCS are paired or unpaired, can be taken as evidence

for associative conditioning or interactive (pseudo) condi-

tioning, respectively.

Additional important information can be derived by

following these different types of stimulation with condition-

ing trials (CS-UCS). This procedure would provide within-

S examination of the effects of the various other procedures

with conditioning. A measure of the effects of adaptation

to different stimuli on subsequent eyelid conditioning perfor-

mance would also be provided. If, for example, the CS is





11

actually a source of response inhibition in the cebus monkey

eyelid conditioning situation, then the effect of prior CS-

only presentations on conditioning would not be expected to

be the typical diminution in CR rate (Grant et al., 1948;

Prokasy et al., 1962). Instead, the CS adaptation would be

expected to decrease the inhibitory effect and thereby enhance

the CR rate.

Presentations of the UCS alone have generally been

found to impede subsequent conditioning (Brimer & Kamin,

1963; MacDonald, 1946; Taylor, 1956). However, instances of

sensitization caused by presentation of the UCS are not

uncommon (Brimer & Kamin, 1963; Prokasy, et al., 1962).

Prokasy et al. (1962), however, found the increment in re-

sponse rate due to sensitization to be short-lived as measured

by their extinction trials.











METHOD


Subjects.--The Ss were 20 male and 10 female cebus

monkeys (Cebus albifrans). They were jungle-born and ac-

quired from Tarpon Springs Zoo at Tarpon Springs, Florida

a few weeks prior to their use in the experiment.

Apparatus.--The restraining chair, programming equip-

ment, and recording method have been described previously

(Pennypacker, King, Achenbach, & Roberts, 1966). The chair

was modified by the addition of the headholder and bucket

seat seen in Figure 1. In addition to moving the microtorque

potentiometer used in recording S's eyeblink, the method of

linking the potentiometer arm to S's eyelid was changed

slightly. In the present study a piece of .01 cm. thick

copper wire was attached to the potentiometer arm by clipping

it in a small coil spring soldered to the end of the poten-

tiometer arm. The other end of the wire was bent into a

.33 cm. loop and wrapped with a piece of plastic tape. A

triangular piece of plastic tape 4 mm. high by 2 mm. wide

was stuck at its apex to the end of the wire and at its base

to the monkey's eyelid.

The experimental chamber was a Model 401-A I.A.C.

soundproof room. A 60-watt light in a conical reflector was

situated 30 cm. above and 35 cm. behind S's head to provide

12


























































Fig. 1. Monkey Restraining Chair and Headholder. a)
Full View of Chair, b) Close-up View of the Seat, c) Close-up
View of the Headholder.





14

illumination of the chamber. In addition to the programming

and recording equipment described by Pennypacker et al. (1966),

a Schmitt trigger was added to the eyeblink recording circuit.

Ey means of this trigger all closures 1.5 mm. beyond S's

baseline were automatically scored. A two-channel Texas

Instruments' printing counter and a half-second timer were

used to record the approximate duration of each inter-trial

interval and the number of blinks during that interval as

scored by the Schmitt trigger. This method provided a

simple and fairly accurate count of S's inter-trial response

rate. Latency and form measurements of the intra-trial

responses were determined by analysis of the Grass polygraph

records which were obtained as described by Pennypacker et al.

(1966).

The CS was a 1,000 Hz, 65 db (re .0002 dynes/cm2) tone

produced by a Hewlitt Packard oscillator. The tone was

delivered through two 10cm. sneakers--one mounted on each

side of the chair--14 cm. from S's ears for a duration of

1,575 msec. The background noise level provided by a ven-

tilation fan was 55 db (re .0002 dynes/cm2). The UCS was a

2.5 psi puff delivered for a duration of 100 msec. through

the system described by Pennypacker et al. (1966) except that

the source of air was a 2,400 psi cylinder of compressed air.

The inter-trial interval ranged from 26.7 to 31.7

sec. and averaged 30.0 sec. except for those Ss in the

CS/UCS condition in Stages 2 and 3. The range and mean

inter-trial interval for those Ss in the CS/UCS condition





15

were halved to equate session length for all Ss. All

inter-event intervals and durations were controlled by

Grason-Stadler programming apparatus.

Design.--The design for the entire experiment is

diagrammed in Table 1. The experiment was divided into four

consecutive stages for purposes of explication and also in

terms of the theoretical problems involved. Across all

stages the three groups of 10 Ss were treated as three ran-

domized blocks. The ten female Ss were all assigned to the

second replication. The sex effect was confounded with the

group effect.

Stage 1 was solely concerned with the changes for SBR

over time--both within days and across days. Seventy-five

consecutive measures of SBR were taken from each S for three

consecutive days. The basal SBR measured in Stage 1 was then

available to be used as a covariate to correct for individual

differences in SBR in successive stages of the experiment.

Stage 2 was primarily designed to assess the effect of

the CS and the UCS, separately as well as jointly, on SBR.

Stage 2 thus permitted the assessment of conditioning, pseudo-

conditioning, the effect of the presence of only the CS on

blink probability, and the interaction of these factors

with SBR. As is shown in Table 2, two different 2x2

factorial paradigms were constructed with CS presence and

UCS presence as factors. The CS and UCS can be presented

together either in a paired or in an unpaired paradigm.

The degree of positive interaction would reflect conditioning









TABLE 1-Overall Experimental Design


S y 1 2 3 4 5 6 7 8 9 10 11

1 STiO. STIU.
NO STIC.
2 CS-UCS
3 CS OICY --cs-OLY
4 CS-UCS CS
5 Ucs 0...Y
5 NO UCS ONLY o --S-O"'^Y
6 CS-UCS H.Y
7 STiFULATION CS/US '- S/UCS
8 riXED ECS-UCS
9 CS-UCS CS-UCS
10 PD CS-RC _______
11 0 STS:.N.
12 CS-UCS
13 CS C.:'Y
14 > CS-"cS -- CS
15 0 S L o UCS C: .Y
iKS ONLY
16 STIULATICS-US OLY
STIF.ULATION
17 CS/UCS CS/UCS
18 faiSXED CS-UCS
19 CS-'CS CS-UCS
20 PAPSEP D CS-UCS
21 O- SI STII.
225 CO S ------------------
22 CS-UCS
23 CSCS OLYLY
24 CS-UCS CS
25 No UCS CILY
U2S OLY O --
26 o CS-UCS ONLY
27 ST8IMULATION cS/UCS ,, CS/UCS
28 ,IXED CS-UCS
29 CS-!CS CS-UCS
30 PARED (S-UCS
A 1 2 2' 3 4













TABLE 2

Factorial Paradigms for Stage 2





CS


Absent Present

no CS
stimulation only

UCS CS-UCS
only (Daired)

a.) Paired Factorial Design






CS

Absent Present

no CS
stimulation only

UCS CS/UCS
only (unpaired)

b.) Unpaired Factorial Design


UCS


UCS


Absent


Present


Absent


Present





18

in the first case and pseudoconditioning in the latter case.

Stage 2 included repeated measures on SBR and also on the

On-trial RR (response rate during CS or, for groups not

receiving CS, during a comparable blank trial). There were

trials administered each day for two consecutive days. Ad-

ditional on-trial response measures were provided at the end

of the second day of Stage 2 when Ss were presented 15
1
CS-only trials.

Stage 3 was designed to permit a comparison of the

effects on "conditioned" response rate of the five different

treatments applied in Stage 2. The treatments applied in

Stage 2 continued to be applied to half of the Ss which

previously received these treatments. This procedure pro-

vided four control conditions: no stimulation; CS only;

UCS only; and CS/UCS. Both Ss in each replication which

received CS-UCS trials in Stage 2 continued to do so. The

two 2x2 factorials in Stage 2 hence became 2x2x2 factorials

in Stage 3. Again repeated measures were employed both

across days and within days.

Stage 4, the extinction phase, was designed to detect

the actual effects on conditioning in Stage 3 caused by prior

treatment in Stage 2, with the control Ss in Stage 3 still

serving as controls of various types. Modification of SBR

by the current treatment was controlled for in Stage 4 since

all Ss received the CS alone. The same two 2x2x2 factorial


1Because of an experimenter error the procedure
actually included these test trials only for Replications
1 and 3.




19

models were used in the conceptualization and analysis of

Stage 4 that had been used in Stage 3.

In conjunction with the development of the experimental

design a linear model was constructed for each stage of the

experiment which contained, in addition to the parameters for

CS presence, UCS presence, and their interactions, parameters

for the two time variables represented by trials within days

(Trials) and days (Days) and the interaction of the time var-

iables with the treatment variables (see Appendices). The

model also included a covariate--either initial blink rate

or concurrent blink rate. The time variable was separated into

Trials and Days because of the rather typical finding that

carry-over from the last block of trials on one day to the

first block of trials on the next is incomplete. The co-

variate was included because it provides a means of separating

the contributions to response rate of the different processes.

For example, the difference in SBR between two monkeys due

to pre-experimental differences in general activity level can

be isolated from that difference in SBR produced by giving the

two Ss different treatments in Stage 2 by using the mean SBR

in the first stage, where all Ss are treated identically, as

the covariate in the second stage. To assess changes in

responding which resulted in the bowing of the curves, either

within days or across days, rather than simply linear trends,

a quadratic parameter for Trials was included, and where

there were three days, a quadratic for Days. Furthermore,

since Trials was expected to change from day to day as asymp-

totic response levels were approached, parameters were





20

included which provided for the interaction of Trials by

Days. Interaction parameters were added for the time

variables interacting with the treatment variables since the

changes across time of the effects of the treatments were

expected to differ depending upon which treatments were

involved. On the basis of prior experimentation the assumn-

tion was made that the within-S and between-S variances were

not sufficiently different to warrant distinct parameters

in the model.

The different questions which this study was designed

to answer can consequently be related to tests of coefficients

or linear combinations of coefficients in a linear model

fitted to the data. The empirical and statistical internre-

tations of some of the questions are given below for illus-

trative purposes:

1. Do UCS presentations facilitate Off-trial RR?

a. Do Ss receiving the UCS (alone or with CS, paired

or unpaired) demonstrate higher Off-trial RR

than Ss not receiving the UCS after correcting

for initial differences in SBR?

b. Is the magnitude of the coefficient for the UCS

parameter significantly greater than zero in the

linear model with Off-trial RR as the response

and with the mean SBR for each S, as measured

in Stage 1, as the covariate? This question

would apply for models fitted to either Stage 2

or Stage 3 data in either of the factorial

experiments.





21

2. Is there evidence of pseudoconditioning?

a. Is the On-trial RR for the CS/UCS factorial

experiment greater for Ss in the CS/UCS

condition than can be accounted for by adding

the effect of the CS and the UCS alone after

correction for possible effects of treatments

on Off-trial RR?

b. Is the coefficient for the CSxUCS interaction

parameter significantly greater than zero for the

model fitted to the On-trial RR in Stage 2 of

the CS/UCS factorial experiment where the Off-

trial RR of Stage 2 is the covariate?

Procedure.--Throughout the course of the experiment

all Ss were housed in individual cages with ad lib water.

They were given their daily ration of Purina monkey chow

and fresh fruit each morning.

The order in which the Ss in a given replication were

used on a given day was randomly determined. The assignment

of Ss to a given treatment was also randomized. The 10 Ss

within a replication were assigned to one of the ten different

treatments depicted in Table 1. This outline determined which

treatment each S received in Stage 2.

At the beginning of each daily session, S was.placed

in the chair and the recording equipment attached. Prior

to SER measurement each S received three daily chair adapta-

tion sessions 40 min. in length. Three days of SBR recording

were next obtained for each S. The 75 measures obtained each




22

day were made during the inter-trial interval of 76 blarnk

trials (average inter-trial interval was 30 sec.), using the

Schmitt trigger and printing counter.

In an attempt to obtain greater independence of the

Off-trial and On-trial RRs the off-trial measure was stopped

2 sec. prior to the CS or blank trial presentation. In an

attempt to minimize SBR distortion by immediate consequences

of the UCS presentation the off-trial response measure was

not begun until 1.20 sec. after the offset of the UCS or

1.30 sec. after the CS or blank trial offset.

On the day following the third SBR measurement session

Stage 2 began. The Ss received 75 trials of the condition

to which they were assigned (150 trials for the CS/UCS

condition) on that day and also on the following day. In

Stage 2 one pair of Ss in each replication received either

no stimulation (NS), the CS only (CS), the UCS only (UCS),

the CS and UCS paired (CS-UCS), or the CS and UCS unpaired

(CS/UCS). The Ss receiving the CS/UCS condition actually

had 150 trials per day--75 presentations of the CS mixed

with 75 presentations of the UCS. The order of presentation

was CS, CS, UCS, CS, CS, UCS, CS, UCS, UCS, CS, UCS, UCS,

etc. A fixed sequence was used rather than the independent

presentation of the CS and the UCS suggested by Rescorla

(1967) because of the difficulty of measuring the Off-trial

RR while using such a procedure.

After the prescribed number of trials on the second

day of Stage 2 fifteen CS-only test trials were given to





23

Ss in the first and third replications. The interval

separating the treatment trials and the test trials was of

the usual inter-trial interval length.

The three days of Stage 3 immediately followed Stage

2. In Stage 3 one member of each pair of Ss receiving the

same treatment in Stage 2 continued to receive that treat-

ment while the other S received paired CS-UCS trials. Each

S continued to receive 75 trials per day (150 trials for Ss

in the CS/UCS condition).

The three days of Stage 4 immediately followed those

of Stage 3. All Ss received 75 CS-only trials per day for

each of the three days. The inter-trial interval remained

an average of 30 sec.










RESULTS


Response Definition


Two different measures of blink rate were used in this

study. The measures differed with respect to the lengths of

their measurement periods, the time of measurement, and the

method of recording.

Off-trial RR.--Off-trial RR was the measure of S3R

taken during the inter-trial interval. The mean length of

this measurement period was 26.8 sec. The Off-trial RR was

recorded by means of the Schmitt trigger and printing

counter, the Schmitt trigger being activated by a lid closure

greater than 1.5 mm. beyond S's base line. During the course

of the experiment occasional difficulty was encountered in

maintaining a zero input to the Schmitt trigger because of the

changes in S's lid position. Thus, the most probable error

was that of counting as full blinks miniature blinks accom-

panied by an elevated base line. On other occasions, however,

the base line shifted in the other direction and responses of

1.5 mm. amplitude were not scored.

On-trial RR.--On-trial RR was the 1,600 msec. measure

of the blink rate taken from the onset of the CS to a point

30 msec. after UCS onset or during a blank trial of comparable

length. For Ss receiving the CS the On-trial RR was always




25

measured during the CS presentations, and for Ss receiving

the UCS it was always measured during the 1,600 msec. im-

mediately preceding the UCS onset. The On-trial RR was

measured by visual inspection of the Grass polygraph record.

The criterion for an On-trial response was that the response

be at least 1.5 mm. over the concurrent base line. In cases

where there were more than one response per trial, S's lid

had to return to the base line before the onset of the next

blink for the latter blink to be scored as an On-trial

response.


Stage 1


Effect of time in chair on SBR.--As measured by the

Off-trial RR the effect of time in the chair on SBR is shown

in Figure 2. An anticipated trend in SBR across trials

within days did not materialize nor was there a change in

SBR across days. The coefficients of the linear and quadratic

parameters for Trials and the linear and quadratic para-

meters for Days were all nonsignificantly different from zero

according to t tests performed on a simple linear model

fitted to Off-trial RR.2 An analysis of variance further

showed that the overall effects of Trials and Days were non-

significant.

Effect of between-S differences on SBR.--There was no

significant difference in overall Off-trial RR due to a


2For all significance tests to be reported,cc= .05.
It is recognized that because of the large number of tests
performed the power of the tests is reduced and replication
will be necessary before any conclusions can be made other
than those of a very tentative nature.
















Q)
bC


CD












r1




cO
....O




CO
\ o
S cit

























C







it
-^ h 0

.^-"^ "AC l




27

difference in S's sex. There were large individual differences

in Off-trial RR. There differences were statistically re-

liable, F (27, 232) = 18.000.


Stage 2


This stage of the experiment was of a dual nature.

It not only provided a measure of the effects of the experi-

mental stimuli on Off-trial RR (Figs. 3 & 4) but also provided

a measure of conditioning based on comparisons with control

groups--with the effect of the treatments on Off-trial RR

removed by covariance techniques (Figs. 5 & 6).

Assumption of single error component in model.--A

comparison of the within-S and between-S variance for both

the On-trial RR and Off-trial RR showed no difference between

the two. The F ratios obtained by dividing the within-S

variance into the between-S variance were nonsignificant.

Effect of UCS presentations on SBR.--The anticipated

increase in Off-trial RR due to puff presentations was found.

The US, CS/UCS, and CS-UCS conditions produced higher rates of

responding than the NS and CS conditions (Figs. 3 & 4). The

coefficient of the UCS parameter in the linear model (see

Appendix A) fitted to the Stage 2 Off-trial RR with the over-

all SBR for each S in Stage 1 as the covariate was significant

according to a t test for both the CS-UCS factorial, t (117)

6.653, and the CS/UCS factorial experiments, t (117) = 7.053.

Habituation of the UCS, a decrease in the facilitatory

effect of the UCS across trials and/or across days, is not






































F-


--'


I'
*1


3!V8 v',t~ 91~r

~JO1 C~3D~R6O) J.RVB


~------%140





C20<,r^ 1C '"120
.C 'j <;-- l,.. ^


0

r
C\.,













C CC


H E-





fO7

0
C)L



0-)








0




C~
C)C

fcr

0U








HO
C
C-
C
C' 0
' : 5
^ <












0 &


C-C-.













4-)
CC'







0 H
S<*
















0 .0 71
cqo



b C)
QHE










0-4
CO
o




-4
. ; 0
o I I C,
. o







. 0 0I
0 0
b \e D
























-4
Sq 0
o. O. .. H-


















o
\.0 0




.1 o s c



HO
1 Ct
o ;





Crt





'-----------------------------------WO !--l
<" D3










Co
r-












































I /
I



71


N" /


N
N.
>1


< <


o LI, 0 0
.s C., Co r *-C r-


K




E---
I __





LzE"
M-^


(H






0











Co
0













c,
tra






















oN





CO
-.
O

C




























0























--
liP

1e
Ci


TJ


Lra L.,
LJ-B *Q


I I I I 1 _


C3 LJ
-- C3l


~i"ii`d 35~0e3S~U




































./ Fl


\. /
\ ti
/


I I

L8U r,~rr: 0~~ld:~ Li "r:~:'~



z:~r.


0












SF-.
C6







Ol
C-

c w





cr
-_
SCE















OO
0 0
co


















00
.-4
0 E












OC)
SC)



O







O










I
CO







*O
--4





* n
00



4-1


0
Cl


0-




32

seen in Figures 3 and 4. The linear model coefficients for

the interaction of the UCS with Trials, linearly or cuadra-

tically, and with Days were all near zero.

Effect of CS presentations on SBR.--There was found

to be no reliable difference in Off-trial RR between those

Ss receiving the CS and those not. There was, however, an

interaction between the CS and UCS in the CS-UCS factorial

experiment (Fig. 3) which was not found in the CS/UCS

factorial experiment (Fig. 4). The facilitation of Off-

trial RR by the UCS was markedly attenuated by preceding

the UCS with the CS. The different results for the CS-UCS

and CS/UCS factorial experiments appear to be due to the

difference in the temporal relationship of the CS and UCS.

The linear model coefficient for the CSxUCS interaction

parameter was significantly negative for the CS-UCS factorial

experiment, t (117) = -2.351.

To determine whether or not responding was inhibited

during the CS presentation attention must be turned to the

On-trial RR data summarized in Figures 5, 6, and 7. After

removal of the effects of differences in Off-trial RR there

remained little difference in responding under the various

conditions (Figs. 6 & 7). Comparison of the NS and UCS

conditions with the CS and CS-UCS or CS/UCS conditions re-

vealed no inhibitory effect of the CS.

Pseudoconditioning.--Pseudoconditionin, would be

evidenced by higher On-trial RR for the CS/UCS condition

than predicted by adding the effects of the CS- and UCS-

alone presentations. The CSxUCS interaction, or the lack of










C\J

O C)
o 0b


Ct C

4-)





P.
to C





ct


F-4
\ 00




-4o
m c
mc6














C-4
SI C
// io II
.. \ I 0 0

i"" .




< o* co to

\o
2 "






0 0




10 I-
SI r
*. O/\
























o*




b0
'"3 -., .. .. S i .:V -, C)

Co-


Co

C)
c*tL





34

additivity of the CS and UCS effects, was negligible according

to a t test of the interaction parameter in the CS/UCS fac-

torial linear model (Fig. 7).

Conditioning.--The curves in Figure 6 show that the

CS-UCS condition produced no divergence from the simple

summation in the corrected On-trial RR of the CS and UCS

effects. If conditioning had been a slow process it might

have appeared only as an increase across trials or days

instead of an overall effect since Stage 2 was only two days

long. Analysis of the data also failed to reveal any gradual

deviation from additivity of the CS and UCS for the CS-UCS

factorial experiment. The linear model coefficients for the

CSxUCSxTrials-linear, CSxUCSxTrials-quadratic, and CSxUCSx

Days interactions were all nonsignificant.

One could assume that conditioning would be reflected

in Off-trial RR as well as in On-trial RR through generalized

fear conditioning or some such mechanism. Then it could be

argued that using Off-trial RR as the covariate camouflaged

the measure of conditioning. Considering the On-trial RR of

the CS-UCS factorial experiment with no covariate makes it

possible to determine to what extent the covariate analysis

may have camouflaged conditioning. Dropping the covariate

adjustment did not result in a significant increment in re-

sponse rate for the CS-UCS condition relative to the response

of the control conditions (Fig. 5) as might have been expected

from an inspection of the Off-trial RR curves (Fig. 3). The

coefficients for the CSxUCS interaction and for the hi-her-





35

order interactions involving the CS, UCS, Trials, and/or Days

parameters were all near zero.

Relationship between initial SBR and subsequent Off-

trial RR.--The covariate used with Off-trial RR in Stage 2

was the mean SBR measure for each S in Stage 1. It accounted

for a significant proportion of the variance in Off-trial

RR. The results of t tests of the covariate coefficients

in the models were both significant, t (117) = 6.585 for the

CS-UCS factorial experiment and t (117) = 3.857 for the

CS/UCS factorial experiment.

Relationship between On-trial and Off-trial RR.--For

both the CS-UCS and CS/UCS factorial experiments a major

Portion of the variance in On-trial RR could be accounted for

in terms of variance in the covariate, Off-trial RR. The

coefficients of the covariate parameter in the linear models

fitted to the On-trial RR of Stage 2 were significantly

greater than zero for both the CS-UCS analysis, t (117)

12.396, and for the CS/UCS analysis, t (117) = 15.446.


Stage 2 Test Trials


For the CS-only trials given at the end of Day 5 the

mean On-trial RR was .042, .250, .188, .406, and .344 while

the mean Off-trial RR was .204, .340, .241, .408, and .267

for the TS, CS, UCS, CS-UCS, and CS/UCS treatments, respective-

ly The linear model used to analyze the data includes the

first seven components and the error component of the model

in Appendix A.





36

Effect of prior CS presentations.--The analysis of the

test trial means indicates some inhibitory properties of the

CS. The Ss which had previously experienced the CS responded

more during the test trial after correction for regression on

Off-trial RR, than Ss who had not received the CS previously.

The CS parameter coefficient was significant only for the

CS/UCS factorial analysis, t (10) = 2.331.

Pseudoconditioning.--There was no evidence of a

CSxUCS interaction in the CS/UCS factorial experiment. There

was, however, strong evidence for sensitization. In both the

CS-UCS and the CS/UCS data Ss with a prior history of UCS

presentations had a higher On-trial RR after correction for

regression on Off-trial RR than Ss who had not received the

UCS, t (10) = 2.870 and t (10) = 2.859, respectively

Conditioning.--There was no evidence of a CSxUCS inter-

action in the CS-UCS analysis of the test trial data.


Stage 3


For half of the Ss (Non-shift Ss) Stage 3 was a

continuation of Stage 2 (Figs. 8, 10, 12, & 1L). By com-

bining Stage 2 and 3 for the Non-shift Ss a five-day parallel

to Stage 2 is provided but with half the number of Ss. The

other half of the Ss (Shift Ss) received the CS-UCS condition

in Stage 3 (Figs. 9, 11, 13, & 15).

Effect of UCS presentations on SBR.--As in Stage 2

there was a noticeable difference in SBR in Stage 3 between

the Non-shift Ss receiving the UCS and those not receiving







































I \\
*a -i ^1\


1a


I I
a a


Si






I I

/







a I
I
--------






['o i:o;5smE ^


rC-







C 4
cO


0C

H c




ct







C

0I O
rco





c0 *.












o C)
0
HO
OP
-1
cf-












I 0


0 0 CO

SO O
r-l









ao t








CC m
O -P





cz
!-!



[Cti


0 '..-. '


I.JY I B 'N i1 Y 'A13


Ec~a




E~a


Wm,


\e


'

































-1. < ~ -





/
I


.1


~\ ~\. <*
N/ ~
\ \
1< ~







~ jl\


S|c





S C --- N






80O G(DFl03) 2'.2 BS:COES8 "V'n'0


F-




S:



- -


--

--


0) 0

C)O











CO




cr

t0 0



C



0


CO


C-,



CI O



0'0








0 C
c^-i

IO c



















c o

C ^c
zo


C/
o u








C2
CCr



cr
.0











c ^J





































0
0
\
o


I.
I/


.I .


1

/)
*0I

0


0


0
/ *
I.<


C U? C= o CU
C, C*^ C4 -- =


U)





.HD
r 0

rH




05H C
S

Ha







80 p
C.




r)




Ocr

c6C)

0






oo



SC
a 0)
oC)









0 c?-



cbca

c o


CO
r a


0sr~

U -P,
Q) ?-


L"


j-


t I I . . .


<"-


C U-
^3- f?


t


rR-~
h"~~ ''?
II*~


HUI :'"19 17i.l


~04 Q~a~,`:3~ 31~8 arI:~ir"'3 %V8~9-3;10






40





CI-
o..o
cr




CO
cO
.u










W 0

















4-) 0
O













cJ
< ^ <; F ", ,-




0 \
tl 0
o .~ O


S<" C CO
Vo

< \ < \ < O (r )






0 U
S*U
\ 2 0















U'3 G3 31 I m
-.0 4 e











SOC!

O-d
oco
,oo \ Ou



ct 0O



+-O CO



O





X cP








































. >


i vI- o


0 4-0









aC







*C)
C x


0 r-





Qdo




0








0
bo.



0 c-
" u






p 0)








-I O



Cl)









o
CO
0 0









*NM 0
r--,
CO0
P-P4






C .D C
C )










CC e/
r-l C


"~l^








!U-t

e>^~


'._ C ',


1


3]' SlSOdSu 18VILA-1,0 NO


















C 18




bc








0-1


**. I
C**. C



;--4
C












CO C



S* ,

0 c
1 0
I* i















U\ 1 0
ICID












I -.





















co


C-
1 V5 :


--- I-- c- '

























0 CO



06
0)- r















cF-c






































I .




0\/



b .*





\ i / "<


0- o.o-O -o0


0 'I *4


Uo as NiJUO Evg 3-so0ds Ii M -SSW


0 )

0 pC

CI

C)








0



c04 0
OH




P -i
to,




C10
Ob-















0 rz
WO
03











C .


























0 ?U)
Q 0Z













oU CO











a,4
ocv
t o
?i 0



o o6

















--OP
v^ 4> ,-





fr 0C
K) ^












cM
0 OC



Co-










..ol






O .

cD
So I C
-1* I s-oK










C) x.
0

S.











04
-< ', <-

















0 .
OO




















C,-m
c- i, I










I Ott)




v co ':


o 0
w C0


O)
31Vy3SN~S3^ ~iajaO M N08S]^9H 0^

HO^ I]Dr 3 3Afl S^OG'S^ IS~l-O ^




oj CQ





45

the UCS (Figs. 8 & 10). For the Shift Ss, who were all re-

ceiving the same treatment, there was no difference between

Ss which had previously received the UCS and those who had

not (Figs. 9 & 11). This effect of UCS presence resulted in

a significantly non-zero linear combination of coefficients

for both the CS-UCS factorial experiment, t (141) = 2.253,

and the CS/UCS factorial experiment, t (141) = 2.811, where

the linear combination represented the difference between

those Non-shift Ss receiving the UCS and those not.

Effect of CS presentations on SBR.--Inspection of

Stage 3 Off-trial RR (Figs. 8 through 11) revealed no evidence

of an inhibitory effect of the CS on Off-trial RR. Inspec-

tion of the data did reveal a depressing effect of the CS

on the UCS facilitation of Off-trial RR for those Non-shift

Ss receiving the CS-UCS treatment (Fig. :3). The linear

combination of coefficients for the CSxUCS interaction for

the Non-shift condition was significantly less than zero,

t (141) = -2.577.

The On-trial RR in Stage 3 revealed a significant

difference between Non-shift Ss receiving the CS and those

not receiving it but no difference between Shift Ss who had

previously experienced the CS and those who had not (Figs.

12 through 15). For the 6 Ss who received the CS/UCS treat-

ment in Stage 2 a within-S comparison was available of On-

trial RR as measured during the CS and also during a blank

trial preceding the UCS presentation. Inspection of the

difference between these two measures revealed no effect of





46

the CS on responding. Both of these findings can be construed

as evidence against an inhibitory effect of the CS on con-

current responding.

Conditioning and oseudoconditioning.--Stage 3 produced

no evidence for either conditioning or pseudoconditioning

(Figs. 12 through 15). The CS-UCS and the CS/UCS conditions

did not produce higher corrected On-trial RR than did the

NS, CS, or UCS conditions either for the Shift or the Non-

Shift Ss. The conditioning paradigm did result in some inter-

esting increases in Off-trial RR. Of the 6 Ss which had

received conditioning in Stage 2, five showed an increase

in responding from the first to the third block of trials on

Day 6. All 6 Ss increased in Off-trial RR from the mean of

Day 6 to the mean of Day 7 (Figs. 8 & 9). The probabilities

that these events occurred by chance are less than .06 and

.03, respectively (Walsh test, Siegel, 1956, p. 83). For

those Ss receiving conditioning in Stage 2, Day 8 was the

third day of conditioning. The Shift Ss demonstrated a

similar increase on Day 8 which was also their third day of

conditioning (Fig. 11). Of the 12 Ss in the NS--CS-UCS,

CS--CS-UCS, UCS--CS-UCS, and CS/UCS--CS-UCS conditions

eleven increased in Off-trial RR from the first to the last

block of trials on Day 8. The probability of these increases

occurring by chance would be less than .01 (Walsh test,

Siegel, 1956, p. 83). The joint probability that an increase

in Off-trial RR on the third day of conditioning for both

these groups of Ss resulted from chance alone would, indeed,

be quite small.





47

Relationship between initial SBR and Off-trial RR in

Stage 3.--The covariate employed in the analysis of Off-trial

RR in Stage 3 was the mean SBR for each S in Stage 1. That

covariate accounted for a significant proportion of the

variance in Off-trial RR for both the CS-UCS analysis t (141) =

8.744, and the CS/UCS analysis, t (141) = 5.406.

Relationship between On-trial and Off-trial RR.--In

Stage 3 the Off-trial RR accounted for 59.4% of the variance

in the On-trial RR. The effect of Off-trial RR was significant

for the CS-UCS analysis, t (141) = 11.581, and for the CS/UCS

analysis, t (141) = 12.9L4.

Stages 2 and 3 combined.--The combined Non-shift data

do not provide any new insights into the data but, instead,

reiterated the findings of Stages 2 and 3 when considered

singly.

Off-trial RR data.--The UCS produced a significant

increment in Off-trial RR, t (139) = 7.618 and t (139) =

4.452 for the CS-UCS and CS/UCS factorial experiments, re-

spectively. This facilitatory effect was reduced by im-

mediately preceding the UCS with the CS. Such attenuation of

the UCS effect on Off-trial RR resulted in a significantly

large coefficient for the CSxUCS interaction parameter in

the CS-UCS factorial experiment, t (139) = -2.888.

On-trial RR data.--The combined On-trial RR data

revealed no evidence of an inhibitory effect of the CS,

conditioning, or pseudoconditioning (Figs. 12 & 14). As in

the analysis of Stages 2 and 3 alone, there was found to be a





48

substantial contribution to the variance in On-trial RR from

Off-trial RR. The covariate coefficient was significantly

greater than zero for both the CS-UCS, t (139) = 13.997, and

CS/UCS factorial experiments, t (139) = 19.223.


Stage 4


All Ss received the same treatment in Stage 4--only

CS-alone presentations. The differences between the Ss re-

ferred to differences in the stimuli previously experienced

in Stages 2 and 3.

Relationship of SBR to Off-trial RR in extinction.--The

covariate, mean SBR in Stage 1, accounted for 34.4 of the

total variance in Off-trial RR in Stage 4. The amount of the

variance accounted for in terms of the covariate was signifi-

cant both for the CS-UCS analysis, t (141) = 10.244, and the

CS/UCS analysis, t (141) 8.578.

Relationship between Off-trial and On-trial RR.--The

On-trial RR, after removal of the effect of regression on

Off-trial RR, is depicted in Figures 20, 21, 22, and 23. The

coefficient of the covariate parameter of the linear model

was significant for the CS-UCS analysis, t (141) = 13.209,

and for the CS/UCS analysis, t (141) = 14.758. The covariate

accounted for 61.6% of the variance in On-trial RR.

Carry-over from Stages 2 and 3.--Both Off-trial RR

(Figs. 16 through 19) and On-trial RR (Figs. 20 through 23)

in Stage 4 revealed no differences due to whether or not

Ss had previously experienced the CS. Only in the CS/UCS











,c








*Hc
COf
D-



--4











C ,
0t
Cr-

















Oct
CO





n C
C6 4 .




























0t






0 p
C)
1. ( r..











0 0
S ,- o

























-'p
0, ,





C O
bO 0C















0)
60
h3 (o 6
g^ Io
^n ^
?- 1 r (U
*'- 0
0 U -P
t^
^i 4- ora
^ cC ^
o ca o
^a O k
ca ( ri5

S0} 4
b O C
O.?-l
OP'm





^-i -
?- o -
*P ^
Ie 0



0 p'

(d 3


" .


Cr
.1


1 1
LD O
-


HYu 1m1m IVSIMI NO No ISSRSI9


Ma G3X o) 31va IsMods3l IvaI-jEo


-A-


*


c




m












m



N


"~:'~

ck~
















T--------t


Q



9P'
f




















0p
OC)
Cdb-





















0 S0
0 pU
-PO
a ) .



















0
a0a
5















E .,.
I
.
0'
1 W NIf -,







.o





















CO-
K,-/ o *




C
U 05
sd -A-M V -O O- Kl



















*1-<











c4-)






Hci








0 x
.0
V3-










Dll








0o ;-4
CO





0. bc








4-!
Hc






o a 0












0-0- .-A











* 0 -0 N
0 0
I0 0,D












.o0 C




*OO



Oc
Flo-


-- -(D-
S-00
-^i "3 or E ^
.g 4- or
^^ cC CM
^ 0 r; a
-i w h
<3 1 <1 ^ e s. 6 / *CO CO

\ ;/* ^






































N
N .
N
.. \


*. \
. 0
0
i
0


A


- N
N
7.
/


.314,
0 N
.* o \







31 10I I O OISS 9









0A GODIUM 31iV 3 "SOdS3J 15M1-:


*P 0




CO


rO:







-4
El


-I 0





C"



C14
CO







cca
to












03 C
4 0











0
-- OCt,
,41 4H


















0o
CO


-P




















0 ,- )
Z CO
c)



















oM4
44











0o0 to
CC








S ' 0 (T;
^ -1 CO
ea ^^~
- racdco
c: T-
0 ^ *












O

O4-)



*to







* O H2
-4

* IC






bc,


C r





o ?
E-












*. {I
S 0













4-)0
" I 0 0






11) M 0
C "-
,-- O 'r-
CCo











"" --0 U 3 L 0-
31VN/ II fI ON r

























bo no
O~ 13 \ I-O 0 bO

bO c
S1-3













C.) .,
c tf











..--F;
cf-drC





































d. ~ ..







I.


4-Q -Z


I .


4.
\ /" .








I I



31V0 3SDO )SI3 %1m l-ijO MNO O0SSS13?393


t1O-J G3DS_^O 31VtJ 3 .'IS O

O-bO
0 --
OCo
OU)




0 C



(LI

v





0

O)
N c





0 .. r
CC4
























O 0
O G




P cU
O
4-3 ( M
0 f-l
- ~' (U r
OPT-
r 05i





- ~ P, 0 4


)
0


S


0 0







4-)


to!
r- 0T
CMOE(


I












mO-
O -)



CO








0
* *H

*C
S ol o
..o



0 E
p- 0



0**
.. A-"' -. Oc1






/ __

S4-) 4-



0 ". o






N N









*H Zo
ro o 0





S.) S *
4-H -
^{3 '' N -o oP



<,-- S








N ,



~ 4 .4- 0
4s. cS a i1"5a b?- *
^ 0 ~~cr;^
F T-
<}o~~~~t <. ^ wc u^
*^o ''* w N ^
<\ s-^C4 ^0^ *** ^^ r, 0 cdr1

%!ton 'i'l '<3. >a ~ P-0 ?






































C ~ *?

4/


31VU 3SNMOdS3U 1VMPI-1'O NO NOISS13933


N1OJ G3D3R O)1 O 3mIv ImSodSg iVi-NO


* I


-a


U-_
0




' 0
-a


C)
Ou
0.
PCO













O
to













0
?E-























C/)
0 c
O3













0 0




bO -4


-qco 9o
-N O
















l(D
CC ..
ci1
O CO
COU

















c6
0 ,

uO
-P T-l




o < r
P- -





-b~l C






-I -


`-d


<^^





57

analysis of the Off-trial RR was there a significant difference

due to prior experience with the UCS, t (141) = 2.834. The

Ss which had received the UCS in Stage 2, either alone or

unpaired with the CS, responded at a higher rate during

the Stage 4 inter-trial interval than Ss who had not received

the UCS in Stage 2 (Figs. 18 & 19). There were no significant

interactions indicating that either the conditioning paradigm

or the pseudoconditioning paradigm had affected extinction

performance.


Other Behavioral Measures


Latency of responses.--Preliminary analysis of the

latency distributions of the On-trial responses with Chi

Square contingency tests revealed no differences due to

treatments and no trends through the stages of the experiment.

On the basis of these results more detailed analyses were not

conducted.

Struggling and emotionality.--A record was kept of

the behavior of each S during his removal from his home cage,

transportation to the experimental room, and placement in the

restraining chair. No differences were observed between Ss

receiving the various treatments or between the various

stages of the experiment. There was little evidence of

habituation to handling or placement in the chair. The Ss

showed a small decrease in fighting the chair during the

first part of the three days of chair adaptation. The re-

duction in struggling did not continue throughout the study,





58
however. Subjects were struggling as much on Day 11 as

they were on Day 1, the day following chair adaptation.











DISCUSSION


The major finding of this study was that cebus monkey

eyelid conditioning performance was significantly influenced

by the Off-trial RR which was in turn a function of the type

of stimulation the monkey received. This finding is of

primary methodological importance for future classical

conditioning studies, particularly with the cebus monkey

eyelid response. For decades eyelid conditioning studies

have been conducted with organisms ranging on the phylogenetic

scale from the frog to man; and these studies have, for the

most part, ignored off-trial responding. The results of

Stages 2, 3, and 4 clearly demonstrate that a casual obser-

vation of spontaneous responding prior to any stimulation

(such as the measure made in Stage 1) is not an adequate con-

trol since the spontaneous response rate may be altered by

stimulation.

The results from the studies conducted in human eyelid

conditioning using a measure of off-trial responding have

not provided a clear picture of the effects of the condi-

tioning or control paradigms on Off-trial RR. Deaux and

Deaux (1966) found a marked but temporary increase in Off-

trial RR upon increasing the UCS intensity during a series of

conditioning trials. They also found a significant and pro-

longed increase in Off-trial RR when Ss were shifted from a

59





60

CS-UCS paradigm to a CS/UCS paradigm. This result is in

contrast to that of Moore and Newman (1964) who reported no

effect on off-trial responding due to the changing of several

variables. Moore and Newman failed to find any difference

in Off-trial RR between Ss receiving high or low UCS inten-

sity, 50' or 100, reinforcement, and high or low CS and

inter-trial interval stimulus similarity. It is possible

that the latter result is due in part to the particular

nature of their experiment. The discrepancy between the

results of Moore and Newman and those of Deaux and Deaux

suggests a need for further exploration of the interaction

of CS and inter-trial interval stimulus similarity with other

aspects of the conditioning situation. Another explanation

of this discrepancy is possible. Since Moore and Newman

used overall mean response rate as their measure of condi-

tioning, it is possible that their failure to find an effect

of UCS intensity manipulations on Off-trial RR is due the

rapid habituation to the effects of UCS presentations.

MacDonald (1946) fourd that although SBR increased signifi-

cantly after only 5 UCS-alone presentations, after 50 trials

the SBR had returned to the initial level. In the Deaux

and Deaux (1966) study the effect on Off-trial RR produced

by a shift in UCS intensity was also only temporary.

The apparent differences in the effects on Off-trial

RR of the different treatments involved in the studies of

Deaux and Deaux (1966) and Moore and Newman (1964) emphasize

the need for large-scale experimentation in human eyelid





61

conditioning oriented toward determining the interaction

of Off-trial RR with the variables which have been tradi-

tionally manipulated. Such issues as the effect of the

ready signal on the unpaired-UCS-presentation phenomena (Kimble

& Ost, 1961) may find new interpretation in the light of

knowledge about the effect of unpaired presentations on

Off-trial RR. In addition to the increment in Off-trial RR

found by Deaux and Deaux (1966) when shifting to CS/UCS

presentations, the results of both Stages 2 and 3 in the

present study showed a significant dampening effect of the

UCS facilitation of the Off-trial RR by preceding the UCS

with the CS. Those concerned with the issue of drive level--

either anxiety level (Spence, 1964) or UCS intensity (Spence,

Haggard, & Ross, 1958)--on conditioning may need to pay

greater attention to the possible influence of Off-trial RR

on conditioning measures. The standard control procedure

of just rejecting those Ss who have an abnormally high SBR

(Taylor, 1951) doesn't seem to be adequate considering the

often found positive correlation between anxiety and SBR

(Doehring, 1957).

The strikingly clear demonstration of eyelid condi-

tioning in the rhesus monkey by Hilgard and Marquis (1936)

seems at first to be in contradiction to the results of this

study and the prior studies in this laboratory using cebus

and squirrel monkeys. The difference can not easily be ac-

counted for in terms of learning ability since, at least for

the cebus monkey, Koch (1935) has shown that the cebus monkey




62

performed in a comparable manner to the rhesus monkey on

several complicated problem-solving tasks. The difference

is possibly a methodological one. For the On-trial PR for

the six monkeys that received five days of classical con-

ditioning, the daily means were .281, .311, .310, .45L, and

.378. The increasing linear trend seen in these data is

significant, F (1, 25) = 6.2912. Upon correction for Off-

trial RR and comparison with the appropriate control groups,

this trend disappears and the acquisition data fail to pro-

vide direct evidence for conditioning. The linear trend

found in the uncorrected data in this study was much smaller

than the effect found by Hildard and Marquis (1936), however.

Future experimentation should take into consideration another

methodological difference between this study and theirs--

Hilgard and Marquis used a light for the CS instead of a

tone. The finding in Stages 2 and 3 that Off-trial RR was

affected less by the presence of the UCS when the CS pre-

cedes the UCS suggests that perhaps some sort of preparatory

response may be occurring in the presence of CS which

diminishes the noxiousness of the UCS. This attenuating

effect might be resulting from classical conditioning of some

response other than the eyelid reflex. Kimmel (1965) has

discussed the possibility of an adaptive aspect of classical

conditioning, and noted that UCR diminution is a frequent

phenomenon in classical conditioning (Kimmel & Pennypacker,

1962; Morrow, 1966). On the other hand, the acquisition of a

preparatory response could occur through instrumental condi-

tioning. A third alternative is available; it may be that





63

the CS already elicits some sort of preparatory response

prior to the conditioning situation. This latter issue can

probably be most easily resolved by measuring simultaneously

other general responses such as the EMG or EEG. Measures of

on-trial responding other than rate or latency which would

not tend to be so completely masked by a high Off-trial RR

might provide not only the answer to the existence of a

preparatory response but also may demonstrate true eyelid

conditioning.

In GSR conditioning where spontaneous responding is

given more attention than in eyeblink conditioning the most

generally used measure is one of CR amplitude. Pennypacker

(1964) has found CR magnitude, or area, to be superior to

amplitude, latency, and frequency measures for a set of

eyelid conditioning data on which he compared the different

measures. Particularly for organisms with relatively high

Off-trial RRs some such measures should receive strong

consideration. If future experimentation continues to

reveal significant effects on Off-trial RR of the treatments

used to study conditioning, then increasing emphasis will

have to be placed in general on response measures other

than those related to frequency.

The Ss used in this study showed little or no

habituation to the restraining chair for the duration of the

experiment. The animals were still highly stressed when

introduced into the experimental setting on the first day of

the experiment and remained that way throughout the experiment.




64

This highly emotional state is not the state under which

conditioning trials are generally administered. This problem

is particularly serious for eyelid conditioning because the

blink reflex reflects general arousal level and is also con-

committant to struggling. Hence, a generalization from the

results of this study to classical conditioning as a whole

might be a serious over-generalization. The negative results

of the present study might be construed, however, as support

for the existence of an optimal drive level above which

eyelid conditioning becomes less likely to occur.

Jouvet's (1960) comments on the importance of chronic

experimentation seem most applicable to cebus monkey eyelid

conditioning. If attempts at eyelid conditioning are to be

continued with the cebus monkey, techniques should be

developed which allow conditioning to be assessed in a situa-

tion to which the monkey has been more thoroughly habituated.

Either modification of the chair such that S can live in it

during the experiment or use of a free-ranging situation may

be involved. The latter possibility might be accomplished by

using implanted electrodes which could be used to monitor EMG

and, also, to deliver a shock UCS.

The 15 CS-only test trials at the end of Stage 2 indi-

cated that there was an increase in On-trial RR due to sensi-

tization. The effect, if real, must be quite short-lived since

there was no evidence for sensitization in the Stage 3 shift

data or in the Stage 4 extinction data. The data were treated

in 25-trial blocks in the negative instances and in a 15-trial

block in the positive instance which suggests rapid habituation.




65

Evidence against an inhibitory effect of the CS in

Stage 3 outweighs the small amount of evidence for it.

Furthermore, in the Stage 2 test trials there was no clear-

out evidence of a UCR to the tone in either the frequency or

the latency measures.

The analysis of the Off-trial RR used in this study

indicated that, in general, the CSxUCS interaction was in the

opposite direction to that which one would expect if condi-

tioning had occurred and had been reflected in Off-trial RR.

The finding that an increase in Off-trial RR did occur in the

CS-UCS condition with a certain temporal pattern poses certain

questions. A discussion of these questions may be parti-

cularly useful in anticipating problems which may accompany

future use of the Off-trial RR as a covariate.

An hypothesis stating that the similarity in the On-

trial RR and Off-trial RR for the CS-UCS treatment results from

both of them reflecting learning has certain intuitive appeal.

If this were the case, the covariate analysis would be can-

celling the very effect which it is supposed to extricate. A

counter hypothesis, the one justifying the use of the co-

variate analysis in this study, is more complex and perhaps

less familiar. Involved are the assumptions that:

1. In addition to whatever else it measures, On-trial

RR measures S's operant level of responding as does

Off-trial RR; and

2. Classical conditioning, if and when it occurs,

results only in an increase in On-trial RR.





66

The first assumption is not difficult to justify logically.

Surely for the NS and UCS conditions the two measures are

equivalent. The other stimulus conditions involving the CS

should not destroy that equivalence unless those stimulus

conditions are altering the response contingencies. These

changes would then reflect the effects of the stimulus condi-

tions more accurately than the absolute rate of responding.

The second assumption must be justified experimentally and

would require a rigorous analysis of the Off-trial RR.

The first hypothesis could be stated in terms of two

stages. Classical conditioning could be hypothesized to

occur first,thereby elevating the response rate during the

inter-stimulus interval. This increased responding im-

mediately prior to the UCS could then result, through operant

conditioning, in an increase in the operant level of respond-

ing. There are several aspects of the data which are not

in agreement with this explanation. The above development

seems to imply that On-trial RR should not only increase

first but also, to whatever extent On-trial RR is a sample

of the operant response level, show a greater increase than

Off-trial RR. The data show, instead, that On-trial RR and

Off-trial RR change together and the changes are of the same

magnitude. These aspects of the data seem to support the

alternative hypothesis involved with the use of the covariate.

With regard to that hypothesis if one argues that it is an

increase in the operant level that is being measured by

both On- and Off-trial RR, then it is difficult to explain





67

why the increase occurred only for the CS-UCS condition and

why this increase had an apparently consistent time course.

It is known, however, that preceding the UCS with the CS

reduces the facilitatory effect of the UCS on the operant

rate according to the results of Stage 2 and the Non-shift

portion of Stage 3. If one assumes that CS adaptation acted

to reduce the attenuating effect of the CS, then the consis-

tency of the increase in Off-trial RR and the fact that Ss who

had the CS-alone condition in Stage 2 didn't show a marked

increase on the third day of conditioning in Stage 3 can be

accounted for in nonassociative terms. In fact, one of the

3 Ss in the CS--CS-UCS condition showed a decrease on the

third day of conditioning. This argument is not, however, in

agreement with the finding that the CS/UCS Ss in Stage 2

demonstrated a large Off-trial RR increase on the third day

of conditioning. Such an argument would seem to imply that

the increase in Off-trial RR would be more gradual--occurring

over each of the first three days of conditioning.

Future studies will have to cope with these problems

for which there are no simple explanations. Studies con-

cerned primarily with SBR as affected by experimental variables

such as inter-trial interval, the ratio of chair adaptation

time to length of the conditioning session, or, perhaps, UCS

intensity may be necessary to determine the nature of changes

in operant level during the course of conditioning. This

study strongly suggests that future classical conditioning

studies will have to deal with spontaneous responding and





68

the problems that accompany it. Future studies must also

consider conditioning not in isolation but within a frame-

work such as that provided by the CS-UCS factorial analysis

used in the present study.









SUMMARY


It was hypothesized that in eyelid conditioning of the

cebus monkey spontaneous blink rate (SBR) was a variable of

considerable importance in determining on-trial response rate

(On-trial RR) and that it was in turn influenced by the stimuli

presented to S. The means used to assess conditioning was a

2x2 factorial design with CS and UCS presence and absence as

orthogonal factors. The Ss were 30 cebus monkeys (Cebus albi-

frons). Following SER measurement, groups of 6 Ss each were

given either NS, CS, UCS, CS-UCS, or CS/UCS treatments for

two days with On-trial and Off-trial RR measured. Half of the

SS in each group then received CS-UCS trials for three days

while the other half continued to receive the control treat-

ments. All Ss were then given three days of extinction.

The results of the experiment revealed a significant

contribution to the On-trial RR variance due to Off-trial RR.

The Off-trial RR was furthermore found to be significantly

affected by the treatments used to assess conditioning. A

reliable finding was that preceding the UCS with a CS de-

creased the facilitory effect of the UCS on Off-trial RR. No

significant interaction of CS and UCS presence on On-trial RR

was found evidencing conditioning or Pseudo-conditioning.

The results were discussed in the context of other eye-

lid conditioning studies which have measured SBR. Some impli-

cations for future experimentation were discussed.


































APPENDICES


































APPENDIX A


LINEAR MODEL FOR STAGE 2










APPENDIX A


LINEAR MODEL FOR STAGE 2


y = a0 + alxI + a2x2 + a3xlx2+ a4x3 + a5x4+ a6x5 +

2
a7x6 +ag + 7 a9x7 + al0xlx6 + allx2x6 + al2x1x2x6 +

2 2
al3x1x7 + al4x2x7 + al5xlx2x7 + al6xlx7 + al7x2x7 -

2 2
al8x1x2x7 + al9x6x7 + a20x6x7 + a21xlx6x7 + a2x2x26x7 +


2 2 2
a23x1x2x6x7 +a24xlx6x7 + a25x2x6x7 + a26x1x2x6x7 + e

X1 = -1, for CS Absent
+1, for CS Present

-1, for UCS Absent
2 +1, for UCS Present

S-1, for Replication 1 or 3
3 = +1, for Replication 2

-1, for Replication 1 or 2
x4 +1, for Replication 3

x5 = Actual value of the covariate

-1, for Day 4
6 +1, for Day 5

-1, for Block of Trials 1
7 = 0 for Block of Trials 2
+1, for Block of Trials 3
2
e is distributed normally with mean 0 and variance 2

ei, ej independent, 1 = j.


































APPENDIX B


LINEAR MODEL FOR STAGES 3 AND 4






APPENDIX B


LINEAR MODEL FOR STAGES 3 AND 4

y = a0+ alx+ a2x2 + a3x1x2 + a4x + a5x4 + a6x5 +
2 2
a7x6 + a8x7 + a9x + al0x8 + allx8 + al2x1x7 +
2 2 2
a13X2x7 + a7lxlx27 +- a5x1x + a16x2x7 + a17x1x2x7 +
2 2
al8x1x8 + al9x2x8 + a20xlx2x8 + a2xlx + a22x2x8 +
2 2 2 2 2
a23x1x2x8 + a24x7x8 + a25xx8 + a26x7x8 + a27x7x8 +

a28x1x7x8 + a20x2x7x8 + a30xlx2x7x8 + a31xlx7x8 +
2 2 2 2
a32x2x7x8 + a33xlx2x7x8 + a34xlx x8 + a35x2x7 +
a36x1x2xxg + aD7xlx x + a38x2x x + a39x1x2x"x2 +

a40xlx6 + a4lx2x6 + a42xlx2x6 + a43x6x7 + . +

a72x6x1x2x x + e

X = -1, for CS Absent
1 1, for CS Present

x -1, for UCS Absent
2 +1, for UCS Present

x= -1, for Reoli+cation 1 or 3
x3 +1, for Replication 2

-1, for Replication 1 or 2
"4 = +1, for Replication 3

x5 = Actual value of the covariate

x = -1, for same treatment in Stage 3 as in Stage 2
6 +1, for shift to CS-UCS in Stage 3

-1, for Day 1
x7 0, for Day 2
+1, for Day 3

-1, for Block of Trials 1
S= 0, for Block of Trials 2
+1, for Block of Trials 3


e is distributed normally with mean 0 and variance c2
ei, ej independent, 1 = j.































APPENDIX C


LINEAR MODEL FOR STAGES 2 and 3 COMBINED






APPENDIX C

LINEAR MODEL FOR STAGES 2 AND 3 COMBINED


y = a0+ alx1+ a2x2+ a3x1x2+ a4x3+ a5x4+ a6x5+ a7x6+
a8xg+ a9x4+ al0xg+ alxl7+ al2x+ al3x1l6+
22
a14x1xg+ a15xlx 7+ a16xx ;H a17x2x6+ al8x2x +
2 2
al9x2x7+ a20x2x7+ a21x1X2x6+ a22xlx2x6+
a23x1x2x7+ a24x1x2x2+ a25x6x7+ a26x6x2

a27x2x7+ + 8x a29x1x6x7 a30x1x6x-
2 2 2 2
a31x1x6x7+ a32x1x6x7+ a33x2x6x7+ a34x2x6x7+

2 2 2
a39xlx2x6x7+ a40x1x2x6x7+ e

x1 = -1, for CS Absent
+1, for CS Present

x2 = -1, for UCS Absent
+1, for UCS Present

3 = -1, for Replication 1 or 3
+1, for Replication 2

X4 = -1, for Replication 1 or 2
+1, for Replication 3

Actual Value of the Covariate
x5 = -2, for Day 4
-1, for Day 5

0, for Day 6
x6 = 1, for Day 7
2, for Day 8

-1, for Block of Trials 1
x7 = 0, for Block of Trials 2
+1, for Block of Trials 3

e is distributed normally with mean 0 and variance a2

ei, independent of ej, i = j











REFERENCES


Bernstein, A.L. Temporal factors in the formation of con-
ditioned eyelid reactions in human subjects. Journal
of General Psychology, 1934, 10, 173-197.

Bitterman, M. E. Lighting and visual efficiency: the present
status of research. Illumination Engineering, 1948,
43, 906-922.

Bitterman, M. E., & Solway, E. The relation between frequency
of blinking and effort expended in mental work.
Journal of Experimental Psychology, 1946, 36, 134-136.

Black, A.H. Cardiac conditioning in curarized dogs: the
relationship between heart rate and skeletal behaviour.
In W. F. Prokasy (Ed.), Classical conditioning: a
symposium. New York: Appleton-Century-Crofts, 1965.
Pp. 20-47.

Boneau, A. The interstimulus interval and the latency of
the conditioned eyelid response. Journal of Experi-
mental Psychology, 1958, 56, 464-472.

Brimer, C. J., & Kamin, L.J. Disinhibition, habituation,
sensitization, and the conditioned emotional response.
Journal of Comparative and Physiological Psychology,
1963, 56, 508-516.

Carlton, P.L., & Vogel, J.R. Habituation and conditioning.
Journal of Comparative and Physiological Psychology,
1967, 63, 348-351.

Cook, W. A. The role of UCS type and intensity in classical
conditioning of the eyelid reflex in the cebus monkey.
Unpublished master's thesis, University of Florida,
1966.

Deaux, E., & Deaux, Kay K. Effect of extinction trials on
spontaneous blinking. Psychonomic Science, 1967, 7,
73-74.

Doehring, D. G. The relation between manifest anxiety and
rate of eyeblink in a stress situation. Research Pro-
ject NM 13 10 99, Subtask 1, Report :To. 6, U. S. Naval
School of Aviation Medicine, Pensacola, Florida, 1957.






Doehring, D.G., & Ferster, C.B. Psychological and physi-
ological responses in a human operant conditioning
situation. Psychological Record, 1962, 12, 251-261.

Drew, G.C. Variations in reflex blink-rate during visual-
motor tasks. Quarterly Journal of Experimental
Psychology, 1951, 3, 73-88.

Franks, C.M. Ocular movements and spontaneous blink rate
as functions of personality. Perceptual and Motor
Skills, 1963, 16, 178.

Franks, C.M. The effects of alcohol upon fluctuation in
perspective, blink rate, and eye movements. Quarterly
Journal of the Study of Alcohol, 1964, 25, 56-67.

Gale, E.N., & Stern, J.A. Conditioning of the electro-
dermal orienting response. Psychophysiology, 1967,
3, 291-301.

Grant, D.A. The pseudo-conditioned eyelid response. Journal
of Experimental Psychology, 1943 (a), 32, 139-149.

Grant, D.A. Sensitization and association in eyelid
conditioning. Journal of Experimental Psychology,
1943 (b), 32, 201-212.

Grant, D.A. A sensitized eyelid reaction related to the
conditioned eyelid response. Journal of Experimental
Psychology, 1945, 35, 393-402.

Grant, D.A., & Adams, J.K. Alpha conditioning in the eyelid.
Journal of Experimental Psychology, 1944, 34, 136-142.

Grant, D.A., Hake, H.W., & Schneider, D.E. Effects of
repeated pre-testing with conditioned stimulus upon
extinction of the contitioned eyelid response.
American Journal of Psychology, 1948, 61, 243-246.

Grant, D.A., & Meyer, H.I. The formation of generalized
response sets during repeated electric shock stimula-
tion. Journal of Genetic Psychology, 1941, 24, 21-38.

Grant, D.A., & Norris, Eugenia B. Dark adaptation as a
factor in the sensitization of the beta-response of
the eyelid to light. Journal of Experimental Psychology,
1946, 36, 390-397.

Grant, D.A., & Norris, Eugenia B. Eyelid conditioning as
influenced by the presence of sensitized beta-res-
ponses. Journal of Experimental Psychology, 1947,
37, 423-433.






Grant, D.A., Norris, Eugenia B., & Boissard, Suzanne. Dark
adaptation and the pseudo-conditioned eyelid response.
Journal of Experimental Psychology, 1947, 37, 434-439.

Grether, W.F. Pseudo-conditioning without paired stimulation
encountered in attempted backward conditioning.
Journal of Comparative Psychology, 1937, 25, 91-96.

Grosberg, J.M. Pseudo-conditioning, drive reduction, and
the acquired fear drive hypothesis. Psychological
Record, 1962, 12, 299-307.

Hanche, W.J., & Grant, D.A. Onset versus termination of a
stimulus as the CS in eyelid conditioning. Journal
of Experimental Psychology, 1960, 59, 19-26.

Harlow, H.F. Pseudo-conditioned responses in the cat.
Psychological Bulletin, 1939, 36, 625.

Harris, J.D. Forward conditioning, backward conditioning,
pseudo-conditioning, and adaptation to the conditioned
stimulus. Journal of Experimental Psychology, 1941
(a), 28, 491-502.

Harris, J.D. An analysis of certain nonassociative factors
inherent in avoidance conditioning in the rat.
Psychological Bulletin, 1941 (b), 38, 572.

Hilgard, E.R. The latency of conditioned eyelid reactions:
reply to Dr. Cason. Journal of Experimental Psychology,
1934, 17, 899-908.

Hilgard, E.R., & Biel, W.C. Reflex sensitization and condi-
tioning of eyelid responses at intervals near simul-
taneity. Journal of General Psychology, 1937, 16,
223-234.

Hilgard, E.R., & Marquis, D.G. Conditioned eyelid responses
in monkeys, with a comparison of dog, monkey, and man.
Psychological Monographs, 1936, 47, 186-198.

Hilgard, E.R., & Marquis, D.G. Conditioning and learning.
New York: Appleton-Century-Crofts, 1940.

Hovland, C.I. Inhibition of reinforcement and phenomena
of experimental extinction. Proceedings of the
National Academy of Sciences, 1936, 22, 430-433.

Hull, C.L. Learning II: the factor of the conditioned reflex.
In C. Murchison (Ed.), A handbook of general experi-
mental psychology. Worc~ester: Clark Ulniv. Press,
W1934 Pp."3e2-455.






Jensen, D.D. Operationism and the question, "is this
behavior learned or innate?" Behaviour, 1961, 17,
1-8.

Jouvet, M. Les approaches neurologiques les processes d'ap-
prentisage. Biologie Medicale, 1960, 49, 282-360.

Kennard, D.W., & Glaser, G.H. An analysis of eyelid move-
ments. Journal of Nervous and Mental Disease, 1964,
139, 31-4.

Kimble, G.A. Conditioning as a function of the time between
conditioned and unconditioned stimuli. Journal of
Experimental Psychology, 1947, 37, 1-15.

Kimble, G.A., & Ost, J.W.P. Influence of unpaired UCS ore-
sentations on eyelid conditions. Psychology Reports,
1961, 9, 239-205.

Kimmel, H.D., Instrumental inhibitory factors in classical
conditioning. In W.F. Prokasy (Ed.), Classical
conditioning: a symposium. New York: ADpleton-
Century-Crofts, 1965. Pp. 148-171.

Kimmel, H.D., & Pennypacker, H.S. Conditioned diminution of
the unconditioned response as a function of the number
of reinforcements. Journal of Experimental Psychology,
1962, 65, 20-23.

Kimmel, H.D.. & Goldstein, A.J. Retention of habituation of
the GSR to visual and auditory stimulation. Journal
of Experimental Psychology, 1967, 73, 401-404.

Koch, A.M. The limits of learning ability in cebus monkeys.
Genetic Psychology Monographs, 1935, 17, 165-234.

Korn, S.J., & Welch, L. CRs fact and artifact. Journal
of Psychology, 1962, 53, 301-309.

Kotaki, Y., & Miyata, Y. Our seventeen years of research
on conditioned responses in man. Psychologia, 1958,
1, 158-165.

Levey, A.B., & Martin, I. Reflex sensitivity and the
conditioned eyelid response. Psychonomic Science,
1967, 8, 153-154.

Lipkin, S.G., & Moore, J.W. Eyelid trace conditioning,
CS intensity, CS-UCS interval, and a correction for
"spontaneous" blinking. Journal of Experimental
Psychology, 1966, 72, 216-220.

Luckiesh, M. Comments on criteria of ease of reading.
Journal of Experimental Psychology, 1946, 36, 180-182.







Luckiesh, M. Reading and the rate of blinking. Journal
of Experimental Psychology, 1947, 37, 266-288.

Luckiesh, M., & Moss, F.K. Reading as a visual task. New
York: MacMillan, 1942.

MacDonald, Annette. The effect of adaptation to the un-
conditioned stimulus upon formation of conditioned
avoidance responses. Journal of Experimental Psychology,
1946, 36, 1-12.

Martin, Irene. Blink rate and muscle tension. Journal of
Mental Science, 1958, 104, 123-132.

Martin, Irene. GSR conditioning and pseudoconditioning.
British Journal of Psychology, 1962, 53, 365-371.

Mattson, Michaela, & Moore, J.W. Intertrial responding
and CS intensity in classical eyelid conditioning.
Journal of Experimental Psychology, 1964, 68, 396-401.

McAllister, W.R. Adaptation of the original response to a
conditioned stimulus. Iowa Academy of Sciences, 1953
(a), 60, 534-539.

McAllister, W.R. Eyelid conditioning as a function of the
CS-UCS interval. Journal of Experimental Psychology,
1953 (b), 45, 417-422.

Mendenhall, W. Introduction to linear models in the design
and analysis of experiments. Belmont, Calif.:Wadsworth,
1968.

Meyer, D.R. On the interaction of simultaneous responses.
Psychological Bulletin, 1953, 50, 204-241.

Moore, J.W., & Newman, F.L. On trial and intertrial res-
ponding in eyelid conditioning as a function of US-
intensity, percentage reinforcement, and the similarity
between the CS and intertrial stimulus. Psychonomic
Science, 1964, 1, 213-214.

Morrow, M.C. Recovery of conditioned UCS diminution following
extinction. Journal of Experimental Psychology, 1966,
71, 884-888.

Mourant, R.R. Eyelid conditioning in monkeys as a function
of CS intensity. Unpublished master's thesis,
University of Florida, 1965.

Ost, J.W.P., & Lauer, D.W. Some investigations of classical
salivary conditioning in the dog. In W.F. Prokasy
(Ed.), Classical conditioning: a symposium. New York:
Appleton-Century-Crofts, 1965. Pp. 192-207.






Pavlov, I.P. Conditioned reflexes. London: Oxford University
Press, 1927. Reprinted by New York: Dover, 1960.

Pennypacker, H.S. Measurement of the conditioned eyelid re-
flex. Science, 1964, 144, 1248-1249.

Pennypacker, H.S., & Cook, W.A. Acquisition and extinction
of the conditioned eyelid response in the squirrel
monkey as functions of the CS-UCS interval. Psychologi-
cal Reports, 1967, 20, 1235-1243.

Pennypacker, H.S., King, F.A., Achenbach, K.E., & Roberts, L.
An apparatus and procedure for conditioning the eye-
blink reflex in the squirrel monkey. Journal of the
Experimental Analysis of Behavior, 1966, 9, 601-604.

Ponder, E., & Kennedy, W.P. On the act of blinking.
Quarterly Journal of Experimental Physiology, 1928,
18, 89-110.

Prokasy, W.F. Extinction and spontaneous recovery of
conditioned eyelid responses as a function of amount
of acquisition and extinction training. Journal of
Experimental Psychology, 1958, 56, 319-324.7

Prokasy, W.F., & Ebel, H.C. Three components of the classi-
cally conditioned GSR. Journal of Experimental
Psychology, 1967, 73, 247-256.

Prokasy, W.F., Ebel, H.C., & Thompson, D.D. Response shaping
at long interstimulus intervals in classical eyelid
conditioning. Journal of Experimental Psychology,
1963, 66, 138-1 1.

Prokasy, W.F., Hall, J.F., & Fawcett, J.T. Adaptation,
sensitization, forward and backward conditioning, and
pseudoconditioning of the GSR. Psychological Reports,
1962, 10, 103-106.

Rescorla, R.A. Pavlovian conditioning and its proper control
procedures. Psychological Review, 1967, 74, 71-80.

Sears, R.R. Effect of optic lobo ablation on the visuo-
motor behavior of goldfish. Journal of Comparative
Psychology, 1934, 17, 233-265.

Shipley, W.C. An apparent transfer of conditioning. Journal
of General Psychology, 1933, 8, 382-391.

Siegel, S. Nonparametric statistics for the behavioral
sciences. York, Pa.: ; .ple Press, 1950. Pp. 83-86.

Spence, K.W. Anxiety (drive) level and performance in eyelid
conditioning. Psychology Bulletin, 1964, 61, 129-139.






Spence, K.W., & Deaux, E. Conditioning (habit growth) in
the absence of CRs. Psychonomic Science, 1966, 6,
61-62.

Spence, K.W., Haggard, D.F., & Ross, L.E. UCS intensity and
the associative (habit) strength of the eyelid CR.
Journal of Experimental Psychology, 1958, 55, 404-411.

Stewart, M.A.,Stern, J.A., Winokur, G., & Fredman, S. An
analysis of GSR conditioning. Psychological Review,
1961, 68, 60-67.

Taylor, J.A. The relationship of anxiety to the conditioned
eyelid response. Journal of Experimental Psychology,
1951, 4i, 81-92.

Taylor, J.A. Level of conditioning and intensity of the
adaptation stimulus. Journal of Experimental
Psychology. 1956, 51, 127-130.

Weber, Henrietta, & Wendt, G.R. Conditioning of eyelid
closure with various conditions of reinforcement.
Journal of Experimental Psychology, 1942, 30, 114-124.

Wendt, G.R. Types of conditioned reaction. Psychological
Bulletin, 1933, 30, 563.

Wickens, D.D., & Wickens, Carol D. Some factors related
to pseudoconditioning. Journal of Experimental
Psychology, 1942, 31, 518-526.

Zeaman, D., & Smith, R.W. Review of some recent findings
in human cardiac conditioning. In W.F. Prokasy (Ed.),
Classical conditioning: a symposium. New York:
Appleton-Century-Crofts, 1954. Pp. 378-418.

Zimny, G.H., Stern, J.A., & Fjeld, S.P. Effects of CS and
UCS relationships on electrodermal response and heart
rate. Journal of Experimental Psychology, 1966, 72,
177-181.













BIOGRAPHICAL SKETCH


William Andrew Cook was born on November 19, 19h2,

at Tampa, Florida. He attended public schools in Florida

and graduated in June, 1960, from Sarasota High School. In

April, 1964, he received the degree of Bachelor of Science

from the University of Florida. He has been a research

assistant for Dr. H.S. Pennypacker in the Department of

Psychology while he has pursued his work toward his graduate

degrees. He received his Master of Science degree in

April, 1966, from the University of Florida. In :arch, 1967,

he was awarded a Post-doctoral Fellowship by the National

Institute of Health to study under Dr. William K. Estes at

Stanford University in California. He will activate this

fellowship in January, 1968.

William Andrew Cook is married to the former Lynn

Anne Rader and is the father of a son, William Montgomery

Cook. He is a member of Psi Chi.










This dissertation was prepared under the direction

of the candidate's supervisory committee and has been

approved by all members of that committee. It was submitted

to the Dean of the College of Arts and Sciences and to the

Graduate Council, and was approved as partial fulfillment of

the requirements for the degree of Doctor of Philosophy.



March, 1968 /


Dean, Coll e of Arts and
Sciences





Dean, Graduate School



Super sory Committee:




Sairman









/-/




University of Florida Home Page
© 2004 - 2010 University of Florida George A. Smathers Libraries.
All rights reserved.

Acceptable Use, Copyright, and Disclaimer Statement
Last updated October 10, 2010 - Version 2.9.7 - mvs