THE ROLE OF AWARENESS IN VERBAL LEARNING

Thais fer the Degree of Ph. D.
MICHBGAN STATE UNWERSEW
Rosaréa A. Eubamfla

1965

 

  
 
  

fl'HESlS

LIBRARY

Michigan Staw
University

 

This is to certify that the

thesis entitled

THE ROLE OF AWARENESS IN VERBAL LEARNING

presented by

Rosaria A. Bulgarella

has been accepted towards fulfillment
of the requirements for

_Eh_._D..._ degree in My

ea 9W

Major fliessor '

Date 1'38.

0-169

 

, r-—~--'

ABSTRACT
THE ROLE OF AWXRENESS IN VERBAL LEARNING

by Rosaria A. Bulgarella

The present study investigated the problem of verbal learning
and awareness. Previous studies have indicated that learning
frequently does not take place without awareness. However, these
studies are illustrative of a specific area of learning in which
the elicitors are weak and awareness is necessary for specification
of the response.

Two experiments using strong elicitors (reinforcers) were
carried out. The first was a classical learning situation in which
gs were classified as aware or unaware on the basis of a post-
conditioning interview. The scores of the Aware and Unaware §s on
a recall test were compared. No significant differences were found
between these two groups. However, both the Aware and Unaware §s
differed significantly from the Control Group indicating that
learning took place without awareness.

The second experiment was an Operant learning situation in
which the‘gs were able to relieve the monotony of a task by giving
a particular response. The §s were then given a rest interval
containing a recorded Joke. The extent of the §s' awareness was

determined by a specific, lengthy postconditioning interview. The

Rosario I. Bulgarells

date from the Aware §s' were eliminated from.the analysis. The
findings indicated that learning did occur without ewereness since
one group did show a significant rise in the correct responses end
differed significantly from the Control Group.

These results were predicted and interpreted in the framework
of elicitation theory.

ii

THE ROLE 0F.IHARENESS IN VERBAL LEARNING

A THESIS

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

NCTOR OF PHILOSOPHY

Department of Psychology

1965

TABLE OF CONTENTS

INTRODUCTION . . . . . . . . . . . . . . . . . . . . . . . . .
EXPERIMENT ONE . . . . . . . . . . . . . . . ... . . . . . . .
Gethod . . . . . . . . . . . . . . . . . . . e . . . . .
Results . . . . . . . . . . . . . . . . . . . . . . . .
Discussion . . . . . . . . . . . . . . . . . . . . . . .
EXPERIMENT TVO . . . . . . . . . . . . . . . . . . . . . . . .
Method . . . . . . . . . . . . . . . . . . . . . . . . .
Results . . . . . . . . . . . . . . . . . . . . . . . .
Discussion . . . . . . . . . . . . . . . . . . . . . . .
GENERALDISCUSSION.....................
SUMMARY . . . . . . . . . . . . . . . . . . . . . . . . . .
REFERENCES . . . . . . . . . . . . . . . . . . ... . . . . . .

”PMDICES O O O O O O O O O- O O O O O O O O O I O O» O O O O O

Page

16
16

27
30
30
35

47

50

53

ACKNOWLEDGEMENTS

The author wishes to express her appreciation
to m. M. R. Denny, chairman of her conittee, for
his theory, suggestions, and assistance in the
planning of this research. Appreciation is also
expressed to 11-. c. Henley, Dr. D. M. Johnson, and
an. R. E. MoMichael, who gave of their time and
served as members of her comittee.

In addition, Mr. Bill; Fenchr deserves thanks
for providing several of the jokes used in
kperinent Two.-

iv

LIST OF TABLES

TABLE Page

1. Frequency distribution of scores (number correct) on
the conditioning task for the eXperimental groups. . . . 76

2. Distribution of gs in the experimental groups on the
basis of postconditioning interview. . . . . . . . . . . 22

3. Frequency distribution of scores (number correct) on
the recall test for the Neutral-Neutral and Control
Group 0 O O O O O O O O O O O O O O O O O O O O O O O O O 77

4. Frequency distribution of scores (number correct) on
the recall test for aware and unaware §s in the
Neutral-SpeCifiC Group. a e a e e e e e e e e e e e e e 78

5. Frequency distribution of scores (number correct) on
the recall test for aware and unaware §s in the
Specific~heutral Group. . . . . . . . . . . . . . . . . 79

6. Frequency distribution of scores (number correct) on
the recall test for the experimental groups. . . . . . . 80

7. Number of correct recalls on the two halves of form B
bylfis in the Neutral-Neutral and Specific-Neutral
Groups. 0 C O O I O O O O O O O D C O O O O O O O O O O 81

8. Distribution of gs in the experimental groups on the
basis of the postrecall interview. . . . . . . . . . . . 25

9. Frequency distribution of scores (number correct) on
the recall test for aware and unaware §s in the Neutral-
Neutral Group based on the postrecall interview . .. . . 83

10. Frequency distribution of scores (number correct) on
the recall test for aware and unawareԤs in the
Specific-Neutral Group based on the postrecall interview. 84

11. Frequency distribution of scores (number correct) on

the recall test for aware and unaware §s in the Neutral-
Specific Group based on the posrecall interview. . . . . 85

vi

TABLE

12.

13.

14.

15.

16.

Number of aware §s in each group according to which
question indicated awareness. . . . . . ... . . . . .

Operant rates for the Control Group and unaware §s in
the expfirimental grOUPSe e e e e e e e e e e e e e 0

Summary of analysis of variance of operant rate and
first five blocks of trials. . . . . . . . . . . . .

Summary of analysis of variance of Operant rate and
firSt bIOCk Of trials. e e e e e e e v e o e e o a a

Summary of an analysis of variance on the first 30
conditioning trials. e e e e e e e e a e o e e e o 0

vii

Page

35

$91

38

[40

41

LIST OF FIGURES

FIGURE Page

1. Mean number of responses (75) as a function of
blocks of (30) trials for the aware §s. . . . . . . . . 37

2. mean number of responses (78) as a function of blocks
of (30) trials for the unaware §s. . . . . . . . . . . 39

3. Mean number of responses (75) as*a function of five

blocks of (6) trials (the first block of 301 trials)
fortheunaware§8eee'ee—e~eee~e-eeeeeeeeo. [52

viii

INTRODUCTION

The presumed role of awareness in verbal learning has changed
considerably throughout the last thirty years of exPerimentation.

The results of the studies first supported the S—R positions of
Thorndike and Hull in which the response is considered to be
mechanically (automatically) strengthened simply because it occurs

in association with a reinforcer which has satisfying (or drive-reduc-
ing) properties. However, with the development of more sensitive
techniques for the assessment of awareness, the later findings of

most experiments on verbal learning indicated that the response is not
automatically strengthened. It was generally found that only aware‘fis
showed an increment in performance, and this caused many investigators
to abandon the 8-H positions of Thorndike and Hull. Since there appeared
to be no better SAR alternative, the cognitive learning theory approach
was adapted as the one best suited to predict and explain these find-
ings. The investigators who follow the cognitive learning approach
suggest that awareness is a necessary condition for learning and that
unaware gs do not learn.

In the present study, an alternative S-R mechanistic approach,
elicitation theory as espoused by Danny and Adelman (1955), is used to
predict and interpret the results obtained. In the framework of
elicitation theory, awareness is necessary only in a particular area
of learning. Learning occurs automatically in situations as specified

by the theory in a subsequent section.

Historical égpggtg

One of the first experiments examining the problem of the relation-
ship between awareness and verbal learning was that devised by Thorndike
and Rock (1934). Learning without awareness was of some importance in
Thorndike's theory, particularly in the realm of learning emotions,
attitudes, conduct, and other similar forms of social behavior.
Consequently, an experiment on verbal learning, in addition to one on
motor learning, was carried out. In these studies, learning without
awareness was assumed to have occurred when the rewarded responses
gradually increased whereas learning with awareness was assumed when
these responses increased suddenly from a moderate percentage to a
high one.

In the verbal experiment, E read a series of words, one at a time,
to the §,who was instructed to respond by saying the first word or
words which came to his mind. If the §_gave a sequential response to
the word, §,called it ”right.” Hewever, if the‘§ gave a response
indicating word-meaning or interpretive connections, it was called
“wrong.” Thorndike and Rock found that there was a gradual rise in the
number of sequential associations given by §, Since there was gradual,
rather than rapid, improvement, they concluded that this was evidence

fer learning without awareness.

3
Irwin, Kauffman, Prior, and Weaver (1931.) questioned Thorndike

and Rock's assumption that a gradual increase in the percentage of
correct responses indicated that learning took place without the §
being aware of such learning. They replicated the experiment by
Thorndike and Rock, except that, as a control, some of the fie were
given the reason for _E_s announcements of 'right' or "wrong.“ It was
found that these control gs who were told of the correct principle,
and thus were aware, also showed a gradual increase in the number of
sequential associations. Consequently, it was concluded that Thorndike
and Bock had not given any evidence of learning without awareness.
Eastman and Jarrett (1952) also used the same procedure as
Thorndike and Rock. However, they used an additional criterion of
awareness, that of asking the § to state the principle which he thought
was correct. The fie were divided into an informed group and an un-
informed group. The first group was informed of the correct principle
in the beginning of the experiment, whereas the latter was not. At the
and of each block of twenty words, the uninformed fie were asked if they
could state the correct principle. An increase in the number of
sequential associations occurred in both groups, and the a: ority of
uninformed fie were. able to verbalise the principle correctly. The fie
who could verbalise showed gradual but significant improvement even
before verbali sation. However ,. the fie who were not able to verbalise
the correct principle did not show a significant anount of learning.

Consequently, this experiment did not present convincing evidence of
learning without awareness.

4
Philbrick and Postman (1955) investigated the role of awareness

when the correct principle was quite simple and easily applicable,
rather than complex as in the previous studies. It was expected that,
since the principle was more simple, learning without awareness would
be more likely to occur and, with verbalisation, a rapid improvement
would take place. The fie were presented with a series of stimulus
words to each of which they were to respond by giving a number. The
number was called correct if it was equal to the number of letters in
the word minus one. The expectation that this situation would enhance
learning without awareness was fulfilled in that the group which could
not verbalise the correct principle did show a significant amount of
learning, although performance was consistently poorer than that of the
fie who eventually were successful.

t study (DiVesta and Blake, 1959) utilised the same procedure as
Philbrick and Postman. It was found that, regardless of whether fie
are set for or are discouraged from looking for a principle, learning
takes place- However, a set to look for a principle facilitated both
awareness and learning.

Greenapoon (1955) determined the effectiveness of two stimuli,
other then 'right" and "wrong' as in the above studies. Either I'nm-
hmm" or ‘huh-uh“ was given by 5 after the plural nouns for some fie or
after any word not a plural noun for other fie. The fi's task was to
continue saying words until 3, told him to stop. Four groups, one for
each condition, were used. (lolly data from fie who were determined to

be unaware by a brief, general interview were used. It was found that

5
the contingent stimulus, ”huh—uh, " had a different effect on the two

groups which received it. The group which received the stimulus after
non-plural responses increased their non-plural responses, whereas

the group which received I'huh-uh" for plm'al responses decreased their
plural responses. The contingent stimulus, "ms-ham, " on the other
hand, had the effect of increasing both the plural and non-plural
responses.

A study similar to the one above was carried out by Sidowski
(1951.). The task was the same as Greenspoon's in that the fie were
asked to any words one at a time, but the reinforcing stimulus was a
light blink rather than a verbal stimulus by 3. lbs results indicated
that the light was an effective reinforcer. It was found that the
unaware fie did increase the number of plural words when these were
followed by the flash of light.

Dulany (1961), also using Greenspocn's procedure, found that
although the fie my not have been aware of the correct contingency
between the response and the reinforcer, they may have held a correlated
hypothesis. “his led them to increase their responses in such a way
that it appeared they had the correct hypothesis. Some fie showed
learning because they held a correlated hypothesis (saying words in a
series, e.g., dianonds, rubies, and pearls), rather than the correct
hypothesis (saying plural words). Delany concluded that unaware fie
who showed learning may have had correlated hypotheses of which they

were very much aware.

Yerplank (1955) attempted to condition the content of conversation

6

by reinforcement. the verbal behavior chosen for reinforcement was
the statements of opinion, and the reinforcer consisted of agreement
with, or the paraphrasing of, the opinion. The meriment was carried
out by a number of Es who, during an unstructured half-hour conversa-
tion with an unsuspecting fi, first determined the fi's Operant level.
§ then followed the fi's opinions by paraphrasing or agreement. It was
found that statements of epinion did increase with agreement or para-
phrasing. Btinction was more difficult since some fis became angry or
left to 'study' or gave some other such excuse. It was assumed that,
since the fie did not indicate that the conversation was an unusual
one, they were unaware that they had been taking, part in an experiment.

Cohen, Relish, 1hurston, and Cohen (1951.) used a method co-only
known as the Taffel procedure which consisted of giving fie 80 stimulus
cards with a verb and the pronouns, "I, we, you, he, she, they,‘ on
each. The fie were asked to make up sentences using one of the pro-
nouns and the verb. If the fi used either pronoun of the first person,
I or we, the g said 'good" in a flat unemctional voice. It was found
that an increment occurred in the pronouns followed by the .E. saying
'good. " Questioning of the fie did not reveal awareness of the relation-
ship between their responses and the E's behavior. mm (1955), using
the some method, obtained similar results.

Levin (1961) has examined the problem of whether learning without
awareness is an artifact of insensitive interviewing. His method as
similar to Taffel‘s sentence construction task. To test for awareness,

both a brief, general interview (BI), such as in the previous studies,

7
and an extended, specific interview (El), developed by Levin, were
used. Using the BI questions alone, only three out of the 60 ge.

were classified as being aware. Using the additional EI questions,

16 more gp were classified as being aware.

when only the BI was used to detect awareness, there was
evidence for learning without awareness. However, when the E1 was
used in addition to the BI, the fie who were feund to be unaware did
not, as a group, exhibit learning. It is interesting to note that
when the fie were aware of being reinforced for only one of the two
pronouns, the fie showed learning for that pronoun alone.

Levin's results threw serious doubts upon those studies which
reported conditioning without awareness when only a.brief, general
interview was used. Following Levin's study, experiments on awareness
and verbal learning (Spielberger, Levin, and Shepard, 1962; Spielberger
and DeNike, 1962; Spielberger, Berger, and Howard, 1963; DeNike and
Spielberger, 1963; DeNike, 1964; and Paul, Eriksen, and Humphreys,
1962) utilised the extended interview to assess awareness. Hhen fie
were determined to be unaware by the El, no significant learning
occurred.

The experiment by Spielberger, Levin, and Shepard (1962) examined,
in addition to awareness, the attitude of the fie toward the reinforcer,
Since the unaware‘fis did not exhibit learning, only the aware fie were
used for the analysis of attitude upon learning. The fie were asked to

choose among three alternatives on an increasing scale indicating how

much they wanted the E to say "good.“ The task was similar to that of

8
Taffel (1955). The results indicated that the fin who desired the
reinforcer very much showed greater learning than those who wanted
it only somewhat or did not care one way or another about it.

Spielberger and DeNike (1962) replicated Greenspoon's (1955)
study with a few refinements. Spielberger and DeNike matched
their fie more carefully as to operant rate than did Greenspoon.
Using only data from unaware fie, they did not find a significant
difference between these experimental fie and the control group.
They concluded that Greenspoon’s positive results were artifacts
of an insensitive, brief postconditioning interview and unequally
matched operant rates for the experimental and control groups.

In another study, utilizing Teffel's sentence construction task,
Spielberger, Berger, and Howard (1963) also found no learning by
unaware fie. They examined, in addition, such variables as need
fer social approval (as measured by the Marlowe-Crowns Social
Desirability Scale) and desire to receive the reinforcer. Again, a
relationship was found between desire to receive the reinforcer,
"good,“ and amount of learning for the aware fie; data from unaware
gs were not used for this analysis since they showed no improvement.
No relationship was observed between need for social approval and
learning or desire to receive the reinforcer.

The experiments by DeNike and Spielberger (1963) and DeNike
(1964) add to the impressive list of studies which showed no
evidence for learning without awareness. The first of these

experiments considered, among other variables, the role of mediat-

9

ing cognitive states (awareness) as measured by a rather detailed
interview. The fis were divided into three groups on the basis of
whether they held the correct hypothesis, other hypotheses which
could have augmented fis giving the response in question, or no
hypotheses which affected the giving of the response. Only the‘fis
having the correct hypothesis showed a significant increase in
performance.

DeNike (1964) attempted to determine whether the point at which
awareness occurs and the point at which performance gains take place
are the same. In addition to a detailed postconditioning interview,
the fig were interrupted during the task (after each trial block of
25 words) and were asked to write down their thoughts about the
experiment. These notes by the fis were later used to determine when
awareness first took place. The results indicated that performance
increments and awareness occurred on the same trial block.

Paul, Eriksen, and Humphreys (1962) devised a completely novel
method of studying learning and awareness. Though this experiment
does not deal with verbal behavior, it is of interest because of
the use of a strong elicitor. The{fis were required to sit through
an experimental session in a chamber maintained at 105° 1?. They
were deliberately misled as to the purpose of the study and were given
I pseudo task. During this time, a puff of cool air contingent about
particular physical movements, e.g., face or mouth movements, was
given whenever fi gave the appropriate response. Only the aware fie

showed an increase in response although almost all of the fis

10
indicated they wanted very much to receive the cool.air.

However, the picture for learning without awareness is not as
bleak as may appear from the above account. There have been isolated
cases in which learning appears to hate occurred without awareness,
and these suggest that this latter should not be considered settled.
These positive.results seen to be found using strong aversive stimuli,
as the elicitors (Eriksen and Kuethe, 1955; Turner and Solomon, 1962),
and using responses over which the fi has no apparent control
(Hefferline,, Keenan, and Earford, 1959; sorted-lino. and Keenan, 1963;
Drrett, 1962).

Eriksen and.Kuethe:(1955) used shock as the elicitor. £;list of
words was presented to the fig several tines, and they were instructed
to give the first association (word) which came to.aind.fer each of
the words.. The‘fis-were misled as to the purpose of the study and
were shocked.ilaediate1y after their first response to certain words.
Each time they repeated the same association for which they were
shocked the first tine, they were shocked again. Awareness was
assessed by an intensive postconditioning interview. Both the high
and the low awareness groups showed a decline in the repetition of
shocked associations as compared to the nonshocked responses. There
I were no significant differences in the amount and.rate of learning
between the two groups.

However, the high awareness group differed significantly from
the low awareness group in amount of time taken to give associations

to the critical and nencritical words. For the high awareness group,

11

following the first shocked trial, reaction time increased for the
critical words, whereas for the low awareness group, there was no
difference in reaction time for the critical and noncritical words.
The increase in reaction time for the aware group corresponded with
their report of deliberate suppression of the shocked responses
while trying to think of new responses. The authors concluded that
this difference in reaction time was an independent check upon the
validity of classifying the fin into the respective groups.

Turner and Solomon (1962) carried out a rather complex stub
(eight experimental groups) of avoidance learning in which both
the type of instruction and the amount of reflexiveness of the
escape responses (respondents versus operants) was varied. nae fie
were either given minimal instructions that a tone and shock would
be presented or adequate instructions which also included the
knowledge that they would do something to terminate or avoid the
shock. The fie were interviewed at the end of the first session and
at the beginning and end of the other sessions. Considering the
adequately instructed fie, all groups learned to avoid and escape
shock and were able to verbalise the correct avoidance response.
In the minimally instructed operant groups, one group did not even
show the escape response and was unable to verbelise whereas another
group, similar to the shuttle box situation, was able to verbalise
and learned to avoid the shock. However, in the minimally instructed
respondent groups, the is were unable to verbalise either the correct
escape or avoidance response and yet showed consistent escape from
shock.

12

The remaining studies have in common the fact that the‘fis did
not appear to be in a position to deliberately control the responses.
The series hy Hefferline et al. used a covert response which was so
Imall that the‘fis were unaware that they did anything. In Hefferline,
Keenan, and Hartford's study (1959), fig could either escape or avoid
an aversively loud noise superimposed on music by giving a small,
invisible thumb-twitch. These minute responses were able to be seen
hy‘fi through electromyographic amplification. ‘fis who were unaware
that they could avoid or escape showed increases in the thumb-twitch
from Operant level. Some of the fie who were instructed that they
could avoid the aversive noise by a tiny twitch of their thumb were
unable to deliberately produce such a small twitch as the unaware‘fis.

The second study, (Hefferline and.Keenan, 1963), also used the
small thumb-twitch as the response to be learned. fi was seated in
a reclining chair and told that g was measuring ability to relax.
After some time, for the establishment of operant rate, numbers
appeared on a box before the fi, indicating the number of nickels fi
earned. Each time 3 gave a thumb-twitch of a particular amplitude,
. the number on the box would advance one. During the conditioning
period, all fie increased the number of thumb-twitches over operant
level. than questioned at the end of the session, no fi was able to
tell how he earned the money. All as reported intense annoyance
during the extinction period, indicating that earning money is a
strong elicitor.

Other evidence comes from studies such as Barrett's (1962) in

13

the area of medicine. A patient suffering from neuromuscular tics
participated in an experiment designed to help reduce the tics. Music,
which was pleasing to the g; was interrupted whenever a tic occurred.
Barrett found that these tic-produced interruptions were effective in
reducing the tics. Deliberate attempts by the fi’to reduce tics were
not as effective as the interruptions of music.

Another interesting study (Lindsley, Hobika, and Etsten, 1961), in

the medical field, examined what occurred to a previously conditioned

avoidance response when the fi,was anesthetized for surgery. fis who had
learned an avoidance response to a loud tone prior to surgery, were
found to give the response while recovering from.anesthesia even though
they could not answer questions or give motor responses to verbal
commands.

This last group of studies presents some difficulty for those who
. hold that awareness is a necessary condition for learning. The findings
of these studies indicate that learning without awareness is possible,
particularly under circumstances in which the elicitors are strong.

The present study examines the role of awareness in verbal learning as

conceived within the elicitation framework.

Elicitation Ihgggz

Elicitation theory'(Ibnny and Adelman, 1955; Ebony and Adelman, 1956)
is presented here as an alternative S—R mechanistic position which
suggests the hypotheses of the present study. It is a contiguity theory

which states as an acquisition hypothesis that “the stimulus complex‘fi

14
which closely precedes in time any response elicited by any stimulus

(Se) acquires the property to elicit this response" (Denny and
Adelman, 1955, p. 290). unlike the drive reduction theory of Hull or
Thorndike, the reinforcer serves as a form of bribery for the elicita-
tion of the response to be learned rather than as a drive reducer. In
other words, the response is strengthened merely because it is produced
rather than because it is associated with a satisfier or reward.
Learning occurs when a response is consistently elicited during a
series of trials or period of time in a particular situation. Examples
of consistent (strong) elicitors are the U03 in a classical condition-
ing paradigm and food, water, or shock in operant learning situations.

If the elicitor is a strong, consistent one and naturally elicits
the response, the stimulus-response situation is automatically
strengthened. In this case, the stimulus itself provides the
appropriate response, and the fi,need not be aware in order to make
the correct response. Awareness may occur particularly when the fi,is
responding at a high level, but it is not a necessary condition for
learning. In studies of animal learning, where awareness defined as
verbalization cannot take place, the elicitors are necessarily quite
strong, such as food which consistently elicits approach responses or
shock which elicits escape responses.

0n the other hand, the stimuli, such as 'mmm-hmm' and 'good,‘
used in verbal learning studies are not strong, consistent elicitors.
They do not produce, or elicit, any particular response consistently.

For this reason, the positive results obtained in the early experiments

15
were unexpected. Elicitation theory predicated the opposite results

from the drive reductionists who maintained that these stimuli could
automatically strengthen responses since they were satisfying (showh
ing approval) to the fig Elicitation purports that since these stimuli
do not directly produce the response, the fi.must become aware before
learning can take place (i.e., the fi’himself must provide the response
in accordance with some hypothesis he holds). These weak elicitors
are first used by fi,as feedback that he has made a particular response.
In other words these stimuli function as knowledge of results that the
fi has made an appropriate response. The _E_'s saying "mmm—hmm" or ”right"
merely specifies the response that fi_should be making according to his
hypothesis as to what the experiment is about. Consequently, in the
area of learning with weak elicitors in which the stimulus situation
does not provide the correct response, awareness is necessary for -
learning. Awareness is defined here, similar to previous studies, as w
the ability to verbalise (i.e., give the correct hypothesis) about the
experiment. 6

In the present study, the stimuli used are presumed to be strong
elicitors for most gs. It is divided into two experiments. The first
of these is a classical conditioning paradigm in which, of course, the
elicitor is a strong and consistent one. The second experiment is an
instrumental learning situation also using a strong elicitor. The
major hypotheses to be tested arse. 1) Verbal learning takes place in
a classical conditioning paradigm without awareness. 2) verbal learn-
ing without awareness occurs in an instrumental learning situation if

the elicitor is strong.

EXPERIMENT ONE

Method

Subjects

The gs were 246 students enrolled in the introductory course in
psychology at Michigan State University. Three experimental groups
consisted of 82, 51, and 61 §s respectively, and a control group
consisted of 52 §s. The data were obtained in five classroom situa—
tions prior to termination of the teaching hour. They were collected
in two sessions, a conditioning session and a recall session, with a
24 hour intersession interval. Consequently, the §s worked individually

but as part of a group.

Materials Used

 

The conditioning procedure was a multiple-choice task in which a
phrase, consisting of an adjective and a noun, and five alternative
verbs were given (see Appendix A and B). The §pcompleted the phrase
by choosing one of the five verbs. The nouns were names of occupational
groups and functioned as CSs even though they followed the adjectives
which served as the 0088. Each of these nouns was repeated three times
with a different adjective each time, and since there were 24.noune, or
occupational groups, the task consisted of 72 phrases to be completed.
The adjectives, accompanying each noun, were such that one verb (the DOB)
was more likely to be chosen from.the alternatives given. For example,
the awkward barber, the careless barber, and the unskilled barber were
placed throughout the task. These all had in common, among the

16

1’1
alternatives, the verb 'nicks.‘ The other alternatives, though
plausible, did not fit the U08 as well as 'nicks" (UCR). Therefore,
most gs chose "nickedt The 003 quite consistently elicited the
expected 068.1

Two different forms of the conditioning task were used. Form A
(see Appendix A) had the 24 nouns randomly placed three times through-
out the task. For Form B (see Appendix B), the 21. nouns were divided
into. two groups of 12 nouns, or 36 items, constituting the first half
and the second half respectively, of Form B. These were: divided so
that the number of incorrect responses was equal for each half (i.e.,
equal conditionability for the two halves). The eliciting value
(number correct) for each noun was determined by a pilot study.

The control task (see Appendix 0) consisted of 72 items, each
being composed of two verbs, one of which the g was to circle as the
Pfeferred. This task included the 2!. verbs from the conditioning task,
each repeated three times as in the conditioning task. Thus, the
control fie saw the critical verbs, or response items, as often as the
merimental fie, though they did not see them paired with the meetive
and occumation (“control for familiarity or recency of exposure).

The postconditioning interview (see Appendix 1)) consisted of four

 

1 he adequacy of the conditioning task was determined by pilot
studies. The adjectives, nouns, and verbs were adjusted until a vast
majority of the §s chose the verb to be conditioned to the occupation.
In a pilot study of 76 fie, the appropriate verbs were selected 21 to 72
times out of a possible 72. The median score was 66. Only 15 fit ”01'“

below 60 and, of these, only four scored below 52.

18
questions which allowed g to determine objectively whether fi actually
knew that he was in a learning situation. Another question asked the
a to explain his choice in order to determine whether § was merely
guessing..

The recall test (see Appendix 3) consisted of 50 occupations, 24
of‘which were the same as in the conditioning task. The §‘s task was
to add a verb-to these nouns.

A;postrecall.interview (see Appendix 19 consisting of two questions
was also given to sole of the fig after the recall test. This was used
to determine whether the fie realised at that point that they had been
in a learning situation on the previous day.

Manama-as

There were three main experimental groups, one given neutral
instructions at the beginning of both the conditioning and the recall
session (Neutral-Neutral Group), a second group which received instruoh
tions specifying the task at the beginning of the conditioning session
and neutral.instructions at the beginning of the recall session
(Specific-Neutral Group), and a third group which received neutral
instructions at the beginning of the conditioning session and instruc-
tions sppcifying the responses on the recall.session (Neutral-Specific
Group). V

The neutral conditioning instructions given at the beginning of
the first session were as follows: 'Select one alternative which best
fits the beginning phrase.' Specific conditioning instructions used in

19

place of the neutral instructions at the beginning of the first session
were as follows:

You will note that in this task an occupation is

given along with a descriptive adjective. This

combination of an adjective and an occupation will

strongly suggest one verb among the choices given

as the best fitting one. It is your task to choose

the best fitting one. Note also that the occupa-

tion is repeated and the same verb. may be chosen.

be neutral instructions given at the beginning of the recall

session were, |'Couplete the following by adding the first verb that
comes to mind.“ The specific instructions were as follows:

Testerday you had to choose one verb from several

which seemed to best fit the occupation. Today

some of the same occupations are repeated. See

if you can recall the verbs you chose yesterday.

Complete the following occupations by adding the

appropriate verbs.

The control group was merely given the following instructions on

the. first session: "Circle the letter in front of the alternative which
you prefer.‘ Instructions on the: second session were the neutral ones

given above for the recall session.

Procedure

The fin participated in a classroom situation with the cooperation
of the instructors who taught the classes. The §s did not receive
advance notice that an experiment was to take place. The Q merely
showed up during the class period and was given a segment of the normal
class hour. The E then asked the fis' cooperation and handed out the
materials to the class.-

20

5:93 W. The control group received the task and
instructions as shown in Appendix C. has §s in the Neutral-
Neutral Group, which was made up of three small classes, received
either Form A or Form B" of the conditioning task with neutral
instructions. The Specific-Neutral Group and the Neutral-Specific
Group were run at the same time, since the class obtained was quite
large and no other class meeting within a 21. hour interval was
available. All instructions were written on the first page, so
the tasks for the two. groups were interlaced and handed out in
such a way that part of the class received specific instructions
and part received neutral instructions. The §s in the Specific-
Neutral Group received either Ibrm A or Form B. be is in the
[antral-Specific moup received only Form A.

Along with one form of the conditioning task which included
the instructions, all experimental fie received an IBM five-choice
answer sheet and the postconditioning interview sheet (Appendix D).

m me All groups received the test for recall of
Verbs (Appendix E). Pa-t of the Neutral-Neutral (h'oup and all of
the Specific-Neutral Group and Neutral-Specific Group received,
in addition, the postrecall interview of two questions (Appendix F).

RESULTS

ConditioningITask

The data for the three experimental groups on the conditioning
task (number of correct responses) are presented in Table 1, Appendix
G. The mean scores for the Neutral-Neutral, the Neutral-Specific, and
Specific-Neutral Groups were 62.0, 65.2, and 64.2 respectively. A 2?
test was used to test the differences between these experimental
groups. The Neutral-Neutral and Specific-Neutral Groups did not
differ significantly from each other (3’: 1.562, p ) .05), and the
Neutral-Specific and Specific-Neutral Groups also did not differ (£_=
.6329, p‘)».05). The differences between the Neutral-Neutral and
Neutral-Specific Groups, however, was significant at the .05 level
(1:, = 2.078).

Table 1 also presents the data for the Aware and Unaware subgroups
within the experimental groups. These subgroups within the Neutral-
Neutral Group and Specific-Neutral Group did not differ significantly
from each other (3’: .369, and 3,: 1.771 respectively). The Aware
and Unaware subgroupswithin the Neutral-Specific Group showed a
significant difference (3 = 2.087, p( .05).

Postconditioning Interview
Table 2 below presents a distribution of the experimental groups
divided into Unaware and Aware §s on the basis of the postconditioning

interview.

21

22

Mic 2. Distribution of §s in the experimental groups
on the basis of postconditioning interview.

 

 

 

 

Group Aware Unaware Total
Neutral-N antral 36 1.6 82
Neutral—Specific 2/. 27 51
Specific-Neutral 27 34 61

Total 87 107 194

 

The divisions were almost equal with slightly more of the fie
being categorised as unaware. The gs were: determined to be aware
if they chose two of the correct multiple-choices or chose one
correct alternative and gave a reasonable answer as to why they chose
the alternative. If the g chose neither of the correct alternatives,
the g was placed into the unware category. In cases where it was
difficult to determine whether the 5 was aware or not, the § m
randomly placed into either of the categories. However, in most
cases, the nature of the interview allowed a clear division.

A I? test was used to determine whether these were approximately
the same preporticn of aware. and unaware fie in each of the experi-
mental groups. The groups did not differ significantly in this
"I?!“ (12 = .1381, £11: 2,. p > .90).

23

Recall Test
Table 3 in Appendix G presents the recall data for the

Neutral-Neutral Group, divided into Aware and Unaware gs, plus
the data for the Control Group. The medians were 2.0 correct

recalls for the Aware §s, 1.5 for the Unaware gp, and 0 for the
Control Group. The Mann-Hhitney Q_test was used to determine
whether the differencesbetween the groups were significant. As
predicted, the difference between the Aware and Unaware §s within
the Neutral-Neutral Group was not significant (2 ='1.02, p = .3078).
The differencesbetween the Control Group and the Aware and Unaware
subgroups of the Neutral-Neutral Group were both highly significant
(3 = 4.90, p <..00003, and z = 4.14, p (..OOOOB), indicating
learning occurred without awareness.

Tables 4 and 5 of Appendix G present the recall data for the
Neutral-Specific and the Specificheutral Groups again divided into

Aware and Unaware subgroups. In the Neutral-Specific Group the
medians were 19.5 for the Aware §s and 18.0 for the Unaware gs. In
the Specific-Neutraerroup, the medians were 7.0 for the Aware §s
and 4.0 for the Unaware §s. Neither of these differences between
the Aware and unaware §s was significant (3 = 1.443. P = .1498,
and z = 1.043, p = .2984).

Table 6, Appendix G. presents the recall data for the three
intact experimental groups (i.e., irrespective of awareness). The
median number of verbs recalled were 2, 19, and 5 respectively for

the Neutral-Neutral Group, the Neutral-Specific Group, and the

24
Specific-Neutral Group. When the differences were tested by means of
the g’test, each of the distributions was found to be significantly

different from the others (between Neutralpfieutral and Neutral-
Specific, s = 8.63, pm< .00003; between Neutral-Neutral and Specific—
Neutral, 2 = 4.19, p <_.00003; between Neutral-Specific and Specific-‘

Neutral, 2 = 7.07, p < .00003)-

Lemme;

As mentioned previously in the method section, Form B of the
conditioning task was made up of two halves of equal conditionability.1
The number of correct recalls for the nouns ia.the first half of Fbrm B
was compared to the number of correct recalls for the nouns in the
second half. It was assumed that if awareness developed during the
conditioning session, and thereby facilitated learning, there would be
a greater number of correct recalls on the second half than on the
first half.

Table 7, Appendix G, presents the number of correct recalls on
the first half and second half of Form B for the fie in the Neutral-
Neutral and Specific-Neutral Groups. The trtest for matched groups

showed no significant difference between §s'performance on the first

and second halves (3,: .453, p >».60), indicating that §s did not do
better on the second half.

 

1 The nouns of equal conditionability on the conditioning
task were also found to be of equal recallability on the recall test
in that gm who were given Form A showed no difference in recall to the
nouns of the two halves of Farm B. 0f 40‘§s with Fbrm A, 18.§s did
the same on those nouns from each half of B, 11 §s did better on the
second half, and 11 §s did worse on the second half.

25

Postrecall Interview

A final division of the experimental groups into Unaware and
Aware fie was carried out on the basis of the postrecall interview
(see Table 8 below). Only the data from initially Unaware _S_s were

Table 8. Distribution of §s in the experimental groups
on the basis of the postrecall interview.

 

 

 

 

Group; Aware unaware Total
Neutral-Neutral 24 23 47*
Neutral-Specific 29 18 471".
Specific-Neutral 4D 21 61

Total 93 62 155

 

* A class of 35 §s did not receive the postrecall in-
terview and were not included above.

** Fbur §s did not fill out the postrecall interview

and were not included above.
recategorized. Consequently, the Aware subgroups consisted of two
types of aware fist 1) those who knew that they were in a learning
situation during the conditioning task in the first session and, 2)
those who only realized, during the recall test in the second
session, that they had been in a learning situation on the previous
day.

Tables 9, 10, and 11 of Appendix G show the number of correct
recalls for the Aware and Unaware §s in each of the experimental

groups. The medians for the Aware and Unaware §s in the Neutral-

Neutral Group:were 3 and 1 correct responses respectively (Table 9).

26

The medians for the Aware and Unaware §s in the Specific-Neutral

Groups were 7 and 3 respectively (Table 10). The differences

in recall between the Aware and UnawareԤs in the Neutral-Neutral
Group and the Specific-Neutral Group were significant (2 = 3.15,

p = .0016, and s = 2.14, p = .0324). However, the difference in

recall between the Aware and Unaware §s for the Neutral-Specific

Group (Table 11), the medians of which were 19 and 18 correct

responses, was not significant (3 = .562, p = .5754).

DISCUSSION

The most important finding of this study appears to be the lack
of significant differences in recall scores between the Aware and
Unaware §s in all experimental groups. This finding strongly
supports the major hypothesis that classical conditioning may take
place without awareness. The crucial factor seems to be the use
of strong elicitors. The negative results obtained by Spielberger,
DeNike, Levin, and others who follow'the cognitive learning theory
approach seem representative of a particular area in which the
elicitors are weak.

It appears that, when the response is naturally elicited, fig
learn even though they are not aware that they are in a learning
situation. Many Unaware fie, in this study, indicated that they
believed their prejudices, or Opinions of occupations, were being
measured. Others believed that the conditioning task was some sort
of personality test. Yet, these §s who had no clear idea that they
were in a learning situation appeared to learn as well as the Aware fie.

Pbrm B afforded a completely objective check as to whether
awareness, independent of its estimation during the postconditioning
interview, played a role in conditioning the response. If awareness
is irrelevant in classical conditioning, no improvement in perform-
ance ofl§s on the second half of Form.B should have taken place. No
significant improvement was found, indicating that awareness, objectively
defined, did not facilitate learning.

27

28

The significant differences among the experimental groups on
the recall test may be considered performance differences resulting
from the sets induced by the different instructions given each group.
However, as discussed above, the instructions did not differentially
affect the performance of the Aware and Unaware subgroups within
each of the experimental groups. This finding is consistent with
the prediction that set, or awareness, should make a large differ-
ence in operant learning situations where the role of the weak
elicitor must be interpreted as knowledge of results, but not in
classical conditioning where the apprOpriate response is clearly
defined.

The postrecall interview can best be considered a measure of
whether the §s were using an unlimited pool of responses, or
associations, for the recall task (Unaware fie) or whether they were
using the pool supplied 24 hours earlier (Aware fie). If the §s
were aware during recall that they were being tested for responses
previously learned, they would have, most likely, given the learned
response. 0n the other hand, for the §s who were not aware that
they were being tested, competing responses had more of an
Opportunity to interfere with learned responses. (For example, it
would be much easier to give the competing response, "the barber 923g,"
rather than the learned response, IIthe barber gigkg.')

The significant differences between the Aware and Unaware §s in
the Neutral-Neutral and SpecificeNeutral Groups probably cannot be

interpreted as the Aware §s showing more learning as a result of

29

awareness, since the awareness assessed by the postrecall interview
did not occur during the conditioning session.

It appears that the Aware §s who connected the recall test
with the conditioning task suppressed competing responses whereas
the Unaware §s did not. Evidence for this view comes from the lack
of a significant difference between the Aware and Unaware §s in the
Neutral-Specific Group who were told to give the same responses on
the recall test as they did the day before. In this case, it did
not matter whether the §s were aware of the learning situation,
since specifying the response on the recall test eliminated stronger
competing responses. The general finding of the superiority of the
Neutral-Specific Group on the recall test can be attributed to the

suppression of these competing responses.

EXPERD’IENT TWO

Method

Subjects
The fie were 72 fie enrolled in the introductory course in

psychology at Michigan State University. They volunteered to serve
for credit. They were each run individually during sessions of
approximately one hour in duration. Four §s were eliminated
because of apparatus failure. There was: a total of 17 fie in each

ef three experimental groups and in the control group.

mean on man

The experiment was carried out in a mll room equipped with
a desk and two chairs, one for g at the desk and one for g near one
end of the desk facing the g. A screen prevented the s from seeing
what the 5 was doing at the desk. A stopwatch was prominently
displayed though not used as an accurate time piece. be major
piece of equipment was a tape recorder ( a Uollensak 1580). The tape
contained 85 short jokes or one-line gags by some of the current
popular comedians, as well as a few amateurs, each separated by five
seconds of canned laughter (see Appendix H, for examples of jokes).
The laughter facilitated the presentation of one joke at a time.
The tape recorder was started and stepped by a foot pedal which was
concealed from the 5's view. However, when the foot pedal was operated,

there was a distinct slit.

31'

Procedure

The § was seated near the desk by the tape recorder. A sheet
of paper and a pencil were placed on the desk before the §. The
_E_ first engaged the f, in casual conversation to establish rapport.
When the § was comfortably seated, the following instructions
were given:

The purpose of this experiment is to measure the
effect of disruptive stimuli on one’s ability to

judge time.

Your task is to say or give a number from care
to 1000 every 3 seconds. However, you cannot
repeat the same number consecutively, count, or
consistently give multiples of any number,
especially 10. (An example of each was given.)

Before the eXperiment actually begins, we will
have a practice session. I will nod.my head
every three seconds to pace you, until you are
able to pace yourself. Continue to say numbers
every three seconds until I say stop. Hesse
remember: he not count, do not repeat the same
number, and do not give multiples of any number,
Are there any questions? (Questions were

gener answered by repeating the appropriate
phrases.

At this point, p displayed the stopwatch and pretended to pace
_8_, making sure that he had understood the directions. If g appeared
to be considerably off in his timing, § suggested that he speed up
or slow down-whichever was the appropriate direction. However, 5
kept a record only of the numbers which § had given. The practice
session included either two, three, or four blocks of 30 trials (i.e.,
a total of 60, 90, or 120 trials). This served the dual purpose of

32.
determining the §¥s operant rate and increasing the §‘s boredom and
fatigue with the task.1 The §s were randomly assigned to either of
three groups having 60 trials (low boredom), 90 trials (medium
boredom), or 120 trials (high boredom) of practice.
Immediately after the practice session, p continued with.the
following instructions:

Now that you have had some practice, you will be
given disruptive stimuli from the tape recorder

to interrupt your giving of numbers and perception
of time. These interruptions will be quite irreg-r
ular. You are to do the same thing you did during
the practice session, following the same rules.
Hewever, when the tape recorder turns on, stcpv
giving numbers until it turns off. Continue your
task immediately after each interruption until I
say, 'stop.‘ Are there any'questions? All right,

begin.

These instructions began the experimental session in which the
tape recorder was turned on and the g was allowed to listen to one
joke each time he gave a number containing the digit, "'7.‘t This digit
was chosen, since it is the only two-syllable digit (besides O) and
thusawas more likely to be distinguishable from the others. The jokes
varied in length, but most of them.were about ten seconds long.

After the firat block of 30 trials, the g asked the g to stop and
read the following instructions:

we are also very interested in your subjective
reactions and thoughts as you participate in this
experiment. So, any thus you have any kind of
thought about the experiment, please stop and
write your thought down on the paper before you.
Number each time you stop. Continue the task again
after you record your thought or reaction until I
say stop. Are there any questions?

 

1 This task was determined to be very dull and fatiguing in a pilot
study, both from gs verbal reports and from.thel§s' reluctance to smoid an
aversively loud stimulus even when they were able to do so. A.fewԤs
increased the response which led to the aversive stimulus and reported
they did so to break the monotony of the task.

33
The experiment continued until" g gave a total of 360 numbers,

including the cperant rate, or until the § had received 85 jokes.
This was the total number of jokes available and, consequently, the
experiment was terminated out of necessity after the supply of jokes
was exhausted.

The task was interrupted in three ways. The ,3 interrupted the
task by giving a number which included the digit, ‘7." 'me 3 also
was able to interrupt the task by saying he had a thought to write
down. Lastly, the g interrupted the task at the end of each block
of 30 trials and asked the § to write his thoughts down. This last
interruption took place if the _8_ did not ask to stop (lining the last
block of 30 trials.

After the task, the postconditioning interview (see Appendix it).
was given to determine whether the § had become aware. The interview
is a modified version of the one used by Paul, Erikscn, and Bullphreys,
(1962). If the s indicated he had become aware, g checked to make
certain that 3 had written down the point at which he became amre.
In the postconditioning interview, questions 1, 5, 7, 8, and 9 were
used to assess awareness. mar the interview, § was asked if he
had prior knowledge about the experiment. (No g indicated he had.)
p then asked §'s cooperation in maintaining secrecy and gave a brief
explanation of the purpose of the experiment.

The procedure for the control group was the same as the 90 trial
practice (medium boredom) group). The instructions did not differ.
The only difference was in the way in which the jokes were given.

34

he control group randomly received six to seven jokes for each
block of 30 trials during the experimental session after cperant
rate was determined. The number of jokes given was based on the

operant rates of the experimental groups.

RESULTS

Th3 Assessment 2; Awareness

0f the 51 experimentalԤs, 14 were determined to have become
aware during the conditioning session. The breakdown of the number
ofԤs, who first expressed awareness in response to questions 1, 5, 7, 8,
or 9 of the postconditioning interview interview (Appendix I), is shown
in Table 12.

Table 12. Number of Aware §s in each group according
to which question first indicated awareness.

 

 

 

 

Question
Group
1 5 7 8 9
Low Boredom. 2 1 1
Medium Boredom 2 1
High Beredom. 2 5

 

There was little difficulty in assessing awareness since most of
these §s gave unprompted evidence of awareness to questions 1 and 5.
Two additional §s gave the correct response (numbers containing the digit,
'7,') to question 8. However, when asked to pick one digit out of the
number, the §s did not pick '7' and conceded it could have been some
other digit. They were not certain as to which was the correct response.
Therefore, these'fis were classified as unaware. Seven of the 14.Aware §s
were from the High Boredom Group.

An additional factor minimized the difficulty of determining which

§s were aware. With one exception, all Aware §s showed a decrease in

35

36

giving the digit, '7,“ as evidenced by a sharp drOp in the correct
response in the block of trials after awareness.1

The performance for the Aware _S_s are shown in Figure 1. Both
the Medium and High Boredom Groups showed a downward trend as more
fie became stars. The low Boredom Group showed an increase toward
the and because one fi gave nothing but”??? after awareness (coopera-
tive rather than contrary). The slight upward trend at the end of
the Medium Boredom Group, appears to support the fis‘ verbalizations
that they could not suppress the response even though they tried.

The dates from these Aware §s were eliminated from the main

analysis.

3.2 (la—Man Brass

The operant rate of each group was based on the two blocks of
trials (60) of the pro-experimental practice. session. In order that
the operant rates could be compared with each block of the condition-
ing trials, the mean of the two blocks was used. The mean operant
rates of the Control, low Boredom, Medium Boredom, and High Boredom
Groups were 7.06, 5.73, 6.68, and 6.65 numbers containing the digit,
‘7," respectively. An analysis of variance (Simple-Randomized Design,
Lindquist, .1953) indicated no significant differences among the four
groups in mean Operant rate (F = .7916). The raw data are included
in Table. 13, Appendix J.

 

1' than questioned, :38 gave a variety of reasons for the decline,
but most said they tried to avoid the response because they felt E was
trying to get them to increase it. Five of the 11. gs reported great
difficulty avoiding the digit, "'7.a Even though they tried, some of the
7's slipped out anyway.

A

3558. on he £03m
or o w b e m

acccccm homo I
somehow Sago: I
Eonoaom 33 4| l IQ

 

. r enema

 

0r

,:

N».

.9.

.2

(8L) SESNOdSS’H JO HEHNIIN mm

38

Berformance g; unaware fig

The performance curves of the three experimental groups and the
control group are presented in Figure 2. Because of the drapping out
of §s who had exhausted the supply of jokes after the fifth rewarded
block of trials, only the operant rates and the first five blocks are
shown in Figure 2. These data were subjected to an analysis of
variance (Type I Design, Lindquist, 1953). The results of this
analysis are presented below in Table 14. The significant main effects

Table 11.. Summary of Analysis of Variance of Operant
Rate and First Five Blocks of Trials.

 

 

Error

 

line ' ' secures df MS Term. F
a. Between Groups. 3 151.293 (b) 4.163.
b.. Between §s in Same Group 50 36.341.

c. Between Trials 5 30.308 (6) 6.178**
d. Interactiont'Trials x Groups 15 7.833 (c) 1.601
s. Pooled §s 1 Trials 250 4.892

 

Total 323

9 Significant beyond .05 level of confidence
** Significant beyond .01 level of confidence

 

of Groups (1? = 4.163, with 3 and 50, p. (.05) and of Trials (1‘ = 6.178,
With 5 and 250 9;; p <f.01) indicate; that the total number ofq75” given
by each group differed significantly and that the number of such re-
sponses differed significantly across trials. The interaction effect,
Groups x Trials, did not quite reach .05 significance level (P = 1.601,
with 5 and 250 9;, p >.05).

.N 0.5m?"

m.H<Hm.H. on he unoogm

m e m N F acetone
\7
\q/
\\ // \QI'lI|4“\\|‘
\\ ’14\\\..\\\
4
8'!"
\\ 10/ \o
\ /
°\ // \
(I‘ll! ‘II-°\

 

'\T

fiovohom swam I
Homeuom E2302 OIII...
Hopeuom 33 4| 1 lo.
H9350 0.1 1 IO

(9 L) sasnoassu so enemas nvaw

40
The Operant rate and only the first rewarded block of trials
were also subjected to a separate analysis of variance which is
summarized in Table 15.

Table 15. Summary of Analysis of Variance of Operant
Rate and First Block of Trials.

 

 

 

 

Error

Line Source df MS Term F
a.. Between Groups 3 33.64 (b) 2.988”
b. Between §s in Same Groups 50 11.26
c. Between Trials 1 99.19 (e) 21.521**
d. Interaction; Trials 1 Groups 3 15.14 (a) 3.284?
e. Pooled.§@ x Trials 50 4.61

Total 107

 

* Significant beyond .05 level of confidence
** Significant beyond .01 level of confidence

Again both the main effects of Groups and Trials were significant
(F = 2.988, with 3 and 50 gig, p <.05, and F = 21.521, with 1 and so 11;,
pi< .01). However, here the interaction effect of Groups x Trials was
found to be significant (F = 3.284, with 3 and 50 511;, p 4 .05), indicating
that some groups showed a significantly larger increase from operant rate
to first block of trials than did other groups.

Individual comparisons were made between the number of“73”given in
the operant rate and the first block of trials for each group. The trtest
for matched groups showed a significant difference between the Operant rate

and the first block of trials for the Medium Boredom.Group>only (t_= 5.929,

41
p <(.01). (For the Control, the Low Boredom, and High Boredom Groups,
the respective results are t,='1.574, p > .10; t,='.1780, p )’.803 and
3,: 1.724, p >..1o.) A trtest for independent groups was also computed
between the Medium Boredom and the Control Group on the first block of
trials. This difference was significant (3,: 2.209, p‘< .05), indicat—
ing the expected learning effect.

Since there was such a large disparity between Operant rate of
emitting I'7,."and the rate in the first block of 30 trials, this first
block was broken down into five blocks of six trials each. The
performance curves for all groups over these trials are shown in
Figure 3. These data were subjected to an analysis of variance

(Type I, Lindquist, 1953). The results of the analysis are summarized
below in Table 16.

Table 16. Summary of an Analysis of variance on the
First 30 Conditioning Trials.

 

 

 

 

, Error
Line Source d‘f MS Term F
a. Between Groups 3 8.857 (b) 4.335**
b. Between gs in Same Groups 50 2.043
0. Between Trials 4 .4700 (e) .6564
d. Interaction::Trials x Groups 12 .4375 (e) .6110
e. Pooled.§p x Trials 200 .716

Total _l_ 269

 

** Significant beyond the .01 level of confidence

3453. 0 ho mmooqm

m Q n N P
I‘llh’ \Ila

I‘lll‘ t, A 1“...
0/ \‘O ''''' $ ''''' A

 

aoponom swam I
souenom same: I
Eovmhom 30H 4| I.- I4
acupeoo OI I '0

 

(BL) SQSNOM J0 manna MN

43
Only the main effect of Groups was significant (F = 4.335,

with 3 and 50 if, P < .01). An examination of the curves shows
a slight increase on the second block of six trials. However,

the last few blocks show a decrease.

DISCUSSION

To evaluate the results properly, perhaps it would be best to
examine the task in detail. The long practice session was used to
augment the amount of boredom, or fatigue, with the task. (The
greater the number of trials in the practice session, the greater
the amount of boredom.) Therefore, there existed a partial elicit-
ing state, within the organism, so that the organism tended to make
a response which relieved his boredom, namely to approach the joke
or rest interval, as concieved in elicitation theory. However, with
each response that reduced boredom, as the background elicitor, the
rest session and the joke lost their high eliciting power. In other
words, the strength of the elicitor fluctuated with the building up
and dissipation of boredom.

Presumably, for this reason, the Medium and High Boredom Groups
showed a jump from the operant rate to the first block of trials.
They began to respond initially at a high rate showing that the
elicitor was quite strong at that point. However, with each response
that was made, the elicitor lost its effectiveness. The lack of
change in performance, after the initial jump, indicates that the
boredom was only enough to maintain the response performance.

Confounded with the above was the‘fifs ability to interrupt the
task whenever he had a “thought about the experiment." Some §s

appeared to take much longer than necessary to write their thoughts

44

45
down, and one § interrupted the conditioning task as much as 40

times. These self-regulated interruptions by the §s also could
hale kept them from increasing their giving of the digit”...

Yet, in spite of these factors, the Medium Boredom Group
showed a significant increase in the response over the Control
Group. The Low Boredom Group; did not show any evidence ofnlearning.
The curve of the High Boredom Group showed an upward tendency from
the operant rate, though this increase was not significant either.
However, almost half of these §s became aware, as might be expected
if awareness follows learning, and their data were removed from the
analysis. The curves for these §s showed that they gave a high
number of "73:“ before awareness. As a whole, the results of this
study support the hypothesis that operant learning will occur with-
out awareness.

The experiment by Paul, Eriksen, and Humphrcys (1962) seems? to
be the only study using a strong elicitor which presents negaive
results. However, the assessment of awareness was made at the end
of the experiment only. No attempt was made to see whether §s
improved before awareness, as in this study, where it appears that
the Aware §s showed inprovement before awareness. his indicates
that awareness follows learning rather than learning follows awareness.

In addition, 7 out of 16 §s classified as aware in the study
by Paul, et al., were classified as such by their answer to
question 8 only of the postconditioning interview (similar to
question 8, Appendix I, in this study). Since this is a guess

46
question, it is possible that these §s may not here been aware

during the conditioning task but gave the correct answer because
they had been responding at a high rate and were quite near the

point of awareness. Therefore, it is not felt that the study by
Paul, at sl, presents a serious contradiction to the hypothesis

and findings of this study.

ENERAL DISCUSSION

The findings of these two eXperiments strongly suggest that
learning may occur without awareness, particularly with the use of
strong, consistent elicitors. The negative findings of previous
studies appear to result from the use of weak elicitors. These do
not naturally provide the response to be learned and necessitate
mediating states (awareness) so that the § himself provides the
response in accordance with a hypothesis which he holds. Meat of
the laboratory studies of verbal learning involve- only weak elicitors,
and the cognitive learning theory approach is adequate for the
interpretation of these studies.

However, the cognitive learning theory approach would be hard
pressed to explain the findings of the present study. These results
are consistent with elicitation theory which is a 8-H contiguity
position. Elicitation theory is offered as an alternative to the
Hullian type S—R theory which fails to explain past negative results
in verbal learning since it holds that the stimulus-response associations
are automatically strengthened because of the satisfying effect of the
reinforcer. Elicitation theory, on the other hand, emphasizes the
mechanistic strengthening of the response only when the stimulus
situation itself provides the response naturally. Consequently, elicitation
theory appears more adequate in predicting and interpreting the role of

awareness in verbal learning.

47

SUMMER!

A review of the literature on learning and awareness indicated
that unaware §s typically do not learn. These negative results
were found with weak elicitors.

The purpose of this study was to examine verbal learning and
awareness when the elicitors are strong. It consisted of two
experiments, one being a classical conditioning situation and the
other, an operant learning situation with strong elicitors. the
major hypotheses are as follows: 1) Awareness is not necessary fer
learning in classical conditioning.. 2) Awareness is not necessary
for learning in an cperant learning situation if the elicitors
are strong.‘

A.unique conditioning task was devised for the classical
conditioning situation. The §s were classified as aware and unaware
on the basis of a postconditioning interview. the scores of the
Aware and unaware §s en a recall test were compared. as significant
differences were found between these two groups. Hewever, both the
Aware and unaware s; differed significantly from the Control Group
indicating that learning took place without awareness.

The operant learning situation was one in which the £9 could
relieve the monotony of a task by giving a particular response.

The 5p were given a rest interve1.containing a prerecorded joke.
the task consisted of giving a number every 3 seconds, and the
response to be elicited was the digit, '7." The extent of the‘fis'

LI

49

awareness was determined hy a specific, lengthy postconditioning
interview; The data of the Aware 5s were eliminated from the
analysis.” The findings indicated that learning did occur without
awareness since one group showed a significant rise in the
correct response and differed significantly from.the Central Group.

Both of the hypotheses were supported. It was found that
learning does occur without awareness in classical conditioning

and in operant learning with strong elicitors.

References

Barrett, B. H. Reduction in multiple tics by free operant

conditionin methods. ,1. 9; Nervoug m Mental Diego”
1962, 135, $87-189.

Cohen, D. D., Relish, H. I.., Thurston, J. R. , and Cohen, E.
Experimental manipulation of verbal behavior. g. g2.
Psychol., 1954. 47, 106-110.

DeNike, D. L. The temporal relationship: between awareness “’
and performance in verbal conditioning. l. 22. Pa chol,

DeNike, D. L., and Spielberger, C. D. Induced mediating states

in verbal conditioning. ,1. Verb_g_l MEI—111.2 and, 122m
BahaViore, 1963, 1, 339.3450

Denny, M. R., and Adelman, H. M. Elicitation Theory: I. An
analysis of two typical learning situations. Psychol. £31.,
1955, 62, 290—296.

’7

Danny, M. R., and Adelmsn, H. M. Elicitation theory: II. The
formal theory and its application to instrumental escape and
avoidance conditioning. Unpublished theoretical paper.
MiOhigan State University. 1956.

DiVestra, F. J ., and Blake, E. The effects of instructional

sets om learning and transfer. Amer. ,1. szchol. 1959
72,, 57-67. , ’

Dulany, D. E. Hypothesis and habits in verbal “cperant
conditioning." l. abnorn. oc. Pg zcholu 1961, 63, 251-263.

Eriksen, G. U., and Kuethe, J. L. Avoidance conditioning without V’
awaeness. J. abnorm. so . 231212.91.» 1956, 63, 203-209.

Greenspoon, J. The reinforcing effect of two spoken sounds on

the frequency of two responses. her. 1. $31101. 1955
68, 409-416. ' '

Hefferline, R. F., and Keenan, B. Amplitude-induction gradient
of a small-scale (covert) operant. J. m. gal. 91 behavior. ,
1963, 6, 307-315.

Hefferline R. F.. Keenan 3., and Harford, R. A. Escape and
uoidanée conditioning,in human subjects without their
observation of the response. mange,” 1959, 130, 1338-1339.

Irwin, F. 3., Kaufman, K., Prior, 6., and weaver, H. E. On V
”learning without awareness of what is being learned.”
2].... 6 0‘ me, 193/» 17, 823-827e

Levin, 8. M. The effects of awareness on verbal conditioning. V
is me PBEChOIe, 1961, 61, 67-75.

Lidsley,. O. 3., Hobeke, J. H., and Etsten, D. E. Operant
behavior during anesthesia recovery: a continuous and
objective method. Anesthesiology., 1961, 22, 937-946.

Linden-int. 3- F. mmwawnm
gag edggtion. Boston: Houghton-Mifflin 60.,“ 1953.

Phul’, G. I», Eriksen, C. U.,. and Humphreys, L. G. Use of
temperature stress with cool stir reinforcement for human
cperant conditioning. J. 953. Psychol., 1962, 64,. 329-335.

Philbrick, E. B., and Postman, L. A further analysis of “learning u.
without ameness.“ Amer. ,1. B‘szchol., 1955, 68, 417-424.

Ecstman, L., and Jarrett, H. An experimental analysis of
|'learning without awareness." Amer. J. Fsychol., 1952, V
65. M255. """'

Sidowski, J. B. Influence of awareness of reinforcement on
verbal conditioning. J. 252. Echol” 1951., 1.8, 355-360.

Spielberger, C. D.,. Berger, A. , and Howard, K. Conditioning
of verbal behavior as a function of awareness, need for social
approval, and motivation to receive reinforcement. g. abnorm. ‘

£22.» Lszcholu 1963,. 67, 24T‘246e

Spielberger,. O. D., and DeNike, D. L... Operant conditioning of
plural nouns: a failure to replicate the Greenspoon effect.
global. _R_ep., 1962,. 11,. 355-366.

Spielberger, C. D’..,;Ievin, 8.. $1., and Shepard, M. The effects of
awareness and attitude toward the reinforcement on the cperant .x‘
conditioning of verbal behavior. ,1. Per!" 1962,, 30, 106-121.

I/

Taffel, G. Anxiety and the conditioning of verbal behavior.
do Mm 292- mg," 1955, 51,; 496-501.

59.

Thorndike, E. L., and Rock, R. T. Learning without awareness,,
of what is being learned or intent to learn it. J. gap.

MO, 1934, 17, 1-19.

Thmner, L. H., and Sblomon, R. L. Human traumatic avoidance
learning: theory and experiments on the operant-respondent
distinction and failure to learn. szchol. Mbnog;., 1962,
76, 1.32..

Verplank, H. 3. The control of the content of conversation:
Reinforcement of statements of opinions. g. abnorm. Egg.
ngchol., 1955, 51, 668-676.

Appendices

Lemma

Conditioning Task—Porn A'.

1‘.

2.

3.

1..

5.

9.

10.

11.

12.

13.

14.

15.

55'

The thirsty carpenter (1) shovels (2) drinks* (3) leaves
(4) rescues (5) drills.

The careless barber (1) knows (2) sets (3) shaves
(4) manipulates (5) giggg.

The unluckly pilot (1) steers (2) listens (3) assists
(4) delivers (5) oragheg.

The unthinking dentist (1) understands (2) hurt; (3) operates
(1.) files (5) drills.

The hard-working laborer (1) £33223 (2) sleeps (3) sings
(A) hints (5) believes.

The absent-minded professor (1) buns (2) votes (3) writes
(A) punishes (5) forgetg.

The poor farmer (1) gtgggglgg (2) dances (3) gossips (A) 3dkes
(5) prospers.

The clumsy typist (1) bathes (2) suffers (3) 221.1
(A) underestimates (5) learns.

The patient fisherman (1) waits (2) rushes (3) collects
(A) repairs (5) navigates.

The well-liked bartender (1) gripes (2) cocks (3) Judges
(4) ligigag (5) swears.

The cowardly soldier (1) listens (2) gaggenders (3) doubts
(4) rescues (5) defends.

The quiet librarian (1) gambles (2) whigperg (3) staggers
(4) snickers (5) blushes.

The untipped cabdriver (1) honks (2) waves (3) moves
(4) gaggglgg (5) agrees.

The over-weight movie star (1) fights (2) seduces (3) diet;
(4) hates (5) annoys.

The miserly shepkeeper (1) ggzgg (2) worries (3) loses
(1.) pulls (5) curses-

 

* The underlined verbs are the correct responses (UCH).

16.

1'7.
18.
19.

20.
21.

22.

23.

25..

26.

27.

31..

32.

55

The unhappy housewife (1) dusts (2) instructs (3) cocks
(1.) loves (5) £533.

The flattered secretary (1) Leafs (2) writes (3) fights
(4) 21.12.; (5) starves.

The cunning banker (1) embessle! (2) counts (3) laughs
(1.) travels (5) obeys.

The cocky athlete (1) help. (2) hires (3) hides (A) giggle.
(5) fights.

The tired maid (1) rest! (2) shouts (3) entertains (A) flirts.

The cheating gambler (1) preposes (2) repairs (3) manages
(1.) £1.91 (5 educates.

The gay waitress (1) orders (2) giggle; (3) writes (A) slinks
(5) discourages.

The ill-Immored teacher (1) Jokes (2) constructs (3) trips
(1.) reads (5) geoldg.

The bad-tempered umpire (1) trenbles (2) zellg (3) waves
(A) leafs (5) smiles.

The stalled cabdriver (1) grins (3) resists (3) synpathises
(1.) drives (5) $321323.

The thrifty shopkeeper (1) buys (2) hinders (3) exercises
(4) lies (5) gel”,

The typical female movie star (1) paddles (2) diet;
(3) Ph110phises (1.) dances (5) understands.

. The hoarse librarian (1) m (2) questions (3) giggles

(A) rages (5) arts.

The agreeable bartender (1) annoys (2) listen! (3) criticises
(1.) gossips (5) overcharges.

_ The calls fisherman (1') deceives (2) sinks (3) mi, t;

(1.) wrestles (5) swims.

$151; $2? pilot (1) guides (2) rests (3) dances (1.) crageg

The tipsy carpenter (1) trades (2) was (3) drinks (4) digs
(5) PoundIe

 

33.

35.

36.

37.

38..

39..

1.0..

1.1.

1.2..

1.3..

1.5.

46'.

47.

5:1

The «hard barber (1) understand. (2) nick; (3) succeeds
(1.)" lassages (5) steals.

The bad dentist (1) hurt; (2) talks (3) bums (A) comforts
('5) anesthetises. ‘

The preoccupied professor (‘1) criticises (2) forget;
(a) advances (1.) chats (5) laughs.

The nervous typist (1) yields (2) m (3) profits (1.) hedges
(5) practices.

The weary farmer (1) remembers (2) W (3) watches
(1.) whistles (5) grins.

The hot laborer (1) speaks: (2) trusts (3) 2229.2! (4) neddles
(5) pushes.

The frightened soldier (1) blackneils (2) challenges
(3) investigates (4) bargins‘ (5) surrenderg.

The frustrated housewife (1) scrubs (2) remembers (3) smiles
(1.) gen, ('5) washes.

The aggressive athlete (1) chuckles (2) specialises (3) sighs
(‘3 m (.5) ”Vela

The happy secretary (1) edgy; (2) telephones (3) argues
(4) types (5) annoys.

The desperate banker (1.) adds (2) enbessles (3) smiles:
(1.) investigates (5) Jokes.

The contested umpire (V1) 1.11.12. (2) pretends (3) stands
(4) signals (5) snares.

The lucky gambler (1) walks (2) shaves (3) fails (1.) errs

The. strict teacher (1) gcoldg (2) prints (3) overlooks
(1.) soothes (5) swears.

Th . waitr .
(523131?” ”axialwaits (2) rescues (3) giggles,

Th exha tednaid (1) mm 2 Egg
('4) sack: (5) gossips. es ( ) re (3) acorns

49.

50.

51 ..

52.

53.

54.

55.

56.

5'7.

58.-

59.

63.

64.

$5
The bushing librarian (1') guesses (2) assists (3) shim
(1.) steals (5) 800!!-

n. sympathetic bartender (1) contradicts (2) gossips
(3) reasons (4) hives (5) ligeng.

The hopeful fisherman (1) protects (2) gait; (3) rests
(4) drives (5) cooperates.

The hopeless soldier (1) gender; (2) suspects (3) cares
(4) smiles (5) courts.

The tough athlete (1') marries (2) cries (3) £391: (A) dances
(5) synpathises.

The angry umpire (1 ) pitches (2) catches (3) yell: (4) mints
(5) 81.11130

11:. experienced gambler (1) gig; (2) forgets (3) falters
(4) wheedles (5)” watches.

The teen-age waitress (1) listens (2) snores (3) Profits
(1.) simian (5) snubs.

The lacy mid (1) shops (2) scrubs (3). rest. ('4) perspires
(5) breathes.

The neurotic teacher (1') rescues (2) advises (3) diets
(A) field; (5) comforts.

The sleepy pilot (1) crgghel (2) flirts (3) tries (4) units
(5) tibia

The unkind dentist (1) examines (2) fills (3) 911.. (1.) gm
(5) wonders.

The inexperienced typist (1) rides (2) boasts (3) hurries
(4) 931 (5) resigns.

(111.). 31131:; 1:22;: (1) protests (2) E2932! (3) questions

n: er-burd d f ‘
(’5 gharges (gejmazer (1) sings (2) visits (3) m

The thoughtless: professor (1) talks (2) searches (3) understands
(4) mam (5) «plains.

59'

65. The unscrupulous banker (1) runs (2) calculates (3) embezzleg
(4) waves (5) thinks.

66. The hostile housewife (1) ..u. (2) cherishes (3) seas.
(i) forgets (5) irons.

67. The friendly secretary (1) files (2) wonders (3) sits
(4) diehonors (5) smileg.

68.’ The unskilled barber (1) nick! (2) hurts (3) smokes (4) runs

(5) shampooee.
69. The greedy shopkeeper (1) toils (2) sells (3) save; (4) pays
struggles.-

70. The figure-conscious movie star (1) diet; (2) promises
(3) sings (4) cries (5) disagrees..

71. The typical.New Turk cabdriver (1) g;ggblgg,(2) stops
(3) approves (4) pushes (5) sobs..

Appendix B

Conditioning Tadk-Fbrm B

1.

3.

5.

7.

8.

9.

10.

11.

12.

13.

14.

15.

54

The well-liked bartender (1) gripes (2) cocks (3) Judges
(4) ligtggg? (5) aware.

The careless barber (1) knows (2) eats (3) shares
(4) manipulate. (5) nickg.

The flattered secretary (1) leafs (2) writes (3) fights
(4) IEiLSI (5) starves.

The quiet librarian (1) gambles (2) ghiggggg (3) staggers
(i) snickers (5) blushes.

The absent-minded professor (1) bums (2) votes (3) writes
(A) punishes (5) forgetg.

The unthinking dentist (1) understands (2) hurt! (3) operates
(4) £11.. (5) drills.

The unhappy housewife (1) dusts (2) instructs (3) cocks
(4) loves (5) uses.

The tired.maid (1) regtg (2) shouts (3) entertains (4) dusts
(5) flirts.

The over-wei ht movie star (1) fights (2) seduces (3) diet.
(4) hates (5 annoys.

The cowardly soldier (1) listens (2) Egggenderg (3) doubts
4) rescues (5) defends.

The unlucky pilot (1) steers (2) listens (3) assists
(4) delievers (5) cgggheg.

The cunning banker (1) embeszleg (2) counts (3) laughs
(4) travels (5) obeys.

The agreeable bartender (1) annoys (2) ligtggg (3) criticises
(4) gossips (5) overcharges.

The exhausted maid (1) fumbles (2) regtl (3) .eern.
(4) smokes (5) gossips.

The awkward barber (1) understands (2) nick; (3) succeeds
(4) masses (5) “Odie

 

* The underlined verbs are the correct responses (UGR).

16.

17.

18..

19..

Z).

21.

22..

23.

25.

27.

28.

29.

30.

31. n

62

The desperate: banker (1) adds ('2) embezzles (3) miles
(4) investigates (5) Jokes.

The happy secretary (1) gig! ('2) telephones (3) argues
(4) types (5) annoys.

The confused pilot (1) guides (2) rests (3) dances
(4) 222211216) pushes.

The frightened soldier ('1) blackmails (2) challenges
(3) investigates (4) barng (5) surrenderg.

The hoarse librarian (1) flags“ (2) questions (3) Biggles
(4) 1'8ng (5) errs.

The typical female movie star ('1) paddles (2) diet;
(3) philosophizes (1.) returns (5) understands.

The frustrated housewife (’1) scrubs (2) remembers (‘3) smiles
(4) 22s.; ('5) washes.

The bad dentist (:1) hurt; (2) talks (3) bums (4) comforts
(5) anesthetises.

The preoccupied professor ('1) criticizes (2) forget;
(3) advances (4) chats (5)1aughs.

The sympathetic bartender (’1) contradicts (2) gossips
(3) reasons (4) drives (5) listggg.

The unscrupulous banker (1) runs (2) calculates (3) embessleg
(4) waves (5) think...

The unskilled barber (1) nick! ('2) hurts (3) smokes (4) runs
(5) shamppoes.

The sleepy pilot (1) crggheg (2) flirts (3) tries (4) waits
(5) ribs.

The friendly secretary (1) files (2) wonders ('3) sits
(1.) dishonors (5) £2.1-

The hopeless soldier (1) wonder: (2) suspects (3) cares
(1.) smiles (5) courts.

The figure-conscious movie star (1) diet; ('2) promises
(3) since (A) cries (5) dieters...

The unkind dentist (1) examines (2) fills (3) smiles (4) hurts
(5) wonders.

...-....

33.

34.

35.

41.

42.

43.

45.

46.

47.

48.

a.
The lazy maid (1) shops (2) scrubs (3) regtg (A) perspires
(5) breathes.

The bushing librarian (1) guesses (2) assists (3) ghigpggg
(4) steals (5) goofs.

The hostile housewife (1) sews (2) cherishes (3) nags
(A) fergets (5) irons.

. The thoughtless professor (1) talks (2) searches (3) under-

stands (4) forggtg (5) explains.

The patient fisherman (1) wait. (2) rushes (3) collects
(A) repairs (5) navigates.

The hardpworking laborer (1) 32232; (2) sleeps (3) sings
(4) hints (5) believes.

The untipped cabdriver (1) honks.(2) waves (3) moves
(4) EEEEELQQ (5) agrees.

The clumsy typist (1) bathes (2) suffers (3) gag!
(4) underestimates (5) learns.

The poor farmer (1) gtgggglgg (2) dances (3) gossips (4) Jokes
(5) prospers.

The badptempered umpire (1) trembles (2) 1911; (3) waves
(4) loafs (5) smiles.

The ill-humored teacher (1) iokes (2) constructs (3) trips
(4) reads (5) goolds.

The cheating gambler (1) proposes (2) repairs (3) manages
(4) giag,(5) educates.

The thirsty carpenter (1) shovels (2) drink; (3) leaves
(4) rescues (5) drills.

The miserly shopkeeper (1) ggggg (2) worries (3) loses
(A) nulls (5) wees.

The cocky athlete (1) helps (2) hires (3) hides (4) giggle.
(5) Mg.-

The gay waitress (1) orders (2) gigglgg (3) writes
(A) smirks (5) discourages.

49.

50.

510

52.

53.

54.

55.

56.

57.

58.

59.

61.

62.

63.

61...

a
The calm fisherman (1) deoeives (2) sinks (3) §§;§£_
(‘4) wrestles (5) swims.

The lucky gambler (1) walks (2) shares (3) fails (4) errs
('5) 11.92.

The hot laborer (1) speaks (2) trusts (3) £22532
(4) meddles (5) pushes.

The nervous waitress (1) waits (2) rescues (3) giggles
(A) yells (5) socialises.

The stalled cabdriver (T, grins (2) resists (3) HNNPathizes
(4) drives (5) M.

The aggressive athlete (1) chuckles (2) specialises (3) sighs
(1.} right; (5) moves.

The thrifty shopkeeper (1) we (2) hinders (3) exercises
(4) lies (*5) 12293..

no nervous typist (1) yields (2) 92.: (3) profits (1.) hedges
(5) practices.

The tipsy carpenter (1) trades (2) saves (3) drinks (4) digs
(5) pound'.

The strict teacher (1) scold; (2) prints (3) ovsrlooks
(A) soothes (5) swears.

The contested umpire (1) zelll (2) pretends (3) stands
(4) signals (5) snares.

The wear: farmer (1) remembers (2) struggles (3) watches
(4) whistles (5) grins..

The hopeful.fishorman (1) protests (2) ggi§g_(3) rests
(A) dives (5) cooperates..

The teen-age waitress (T) listens-(2) shores (3) profits
(4) $8815.93 (5) snubs.

The shirtless laborer (1) protests (2) £2222! (3) questions
(4) fails (5) snacks.

The tough athlete (1) marries (2) cries (3) fight;
(A) dances (5) sympathises.

65'.

66.

67.

68.

69..

70.

71‘.

72.

'65

The typical New York cabdriver (1) M (2) stops
(3) wwoves (4) push“ (5) sobs.

The greedy: shopkeeper ('1) toils (2) calls (3) meg
(A) pays (,5) struggles.

The dry. carpenter (1) drink; ('2) beers (3). fixes (A) crys
(5, ”U8.

The angry umpire ('1) pitches (’2) catches (3) zellg
(4)point. (5) grins»

The experienced gambler (1) 3131(2) forgets (3) falters
(1.) wheedles ('5) watches.

The inexperienced typist (1) rides (2) boasts (3) hurries
(4) 22: (F5) resigns»

The neurotic behavior teacher (’1) rescues ('2) advises (3) diets
(1.) .coldg (“5) comforts.

The over-burdened farmer (1) sings (‘2) visits (3) mm
(1.) charges (5) jumps.

Appendix G

Control Task

. 63
W: circle the letter in front of the alternative which
you prefer.
1. (a) shovels (b) drinks 26. (a) buys (b) saves
2. (a) knows (b) nicks 2'7. (a) diets (b) understands
3. (a) assists (b) crashes 28. (a) whispers (b) questions

4. (a) hurts (b) files 29. (a) annoys (b) listens
5.. (a) sweats (b) sleeps 30. (a) waits (b) swims
6. (a) writes (13) forgets 31. (a) guides (b) crashes

7.. (a) struggles (b) PIMP”! 32.. (’a) drinks (b) pounds

8. (a) errs (b) learns 33. (a) nicks (b) steals
9.. (a) waits (b) collects 34. (a) hurts (b) hums
10. (a) listens (b) swears 35. (a) criticises (b) forgets

1‘1 ,,, (a) menders (b) doubts 36. (a) errs (b) practices
12.. (a) whispers (b) blushes 37. (a) struggles (b) whistles

13.. (a) waves (1)) grumbles 38. (a) sweats (b) saddles
11... (a) diets (b) annoys 39.. (a) bargins Cb) surrenders
15. (a) saves (b) worries 40. (s) scrubs (b) nags

16.. (a). instructs (b) range 1.1. (a) sighs (b) fights
17. on leafs (b) smiles 1.2.. (a) smiles (b) argues
18. (a) embessles (13) counts 1.3. (a) embassles (b) Jokes
19. (a) hires (b). tights 4/.- (1.) yells (b). pretends
20. C.) rests (b) flirts 1.5. (.3 fails (b) wins

21.. (a) repairs (b) wins 46. (a) scolds (b) overlooks
22. (9.) orders (b). giggles 1.7.. (a) rescues (b) giggles
23. (a) reads (1)) scolds 1.8. (a) rests (b) smokes

24. (a) trembles (b) yells 1.9. (a) guesses (b) whispers
25. (a) drives (b) grumbles 50. (.3 contradicts (b) listens

51.
52.
53.
54.
55.
56.
57.
58.
59..
60.
61 .

63. t

65.

66..
t (a) dishonors (b) miles

68..
69.
70..

71.-

72.

(a) waits (b) cooperates
(a) surrenders (b) cares
(a) marries (b) fights
(a) yells (b) points

(a) wins (b) wheedles
("d giggles (b) snubs
(a) rests (b) breathes
(a) advises (b) scolds
(a) crashes (b) fibs

(.1 examines (b) hurts
(a) errs (b) resigns

(a) protests (b) sweets
(’a)’ struggles (b) Juaps
(a) emches (,b) forges
(a) runs ('b) embezsles
(a) cherishes (b) nags

(a) nicks (b) runs

.(a) toils (b) saves

(a) diets (b) promises,
(a) grumbles (b) approves
(a) drinks (b) fixes

Appendix D

Postconditioning Interview

1.

2.

3.

mi

The reason for this experiment is (circle one of the following):

8.

b.

c.‘

d.

To find what students"reactions are to certain occupations

To see whether students are consistent in their responses
to certain occupations over a short period of time

To teach students a particular response to a certain
occupation

Tb compare the favorableness or unfavorableness of one
occupation to the others

None of the above.

Explains why you chose the alternative you did in question 1,
if possible (use other side tocxif necessary):

This task involves (circle one of the following):

a.
b.
0.
d.’

Judgment
Interests
Prejudice
Learning
Personality

Indicate your second choice to question 1 (on this sheet):

 

* This response was used to assess awareness

Appendix E

Recall Test

1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.

20..

21.

23.
24-
25.

The
The
The
The
The
The
The
The
The

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

politican 26.
teacherI_____J!§flgyg1______ 27.
psychologist 28.
barber nicks 29.
bank-r M— 30.
statistician 31.
baker 32.
advertiser 33.‘
pilot craghes 34.
farmer gtggggleg n, 35.
plumber 36.
typist errs 37.
accountant a. 38.
cook 39.
athlete fifigtg 40.
policeman 41.
bartender listens 42.
nurse 43o
soldier ggggendggg 44.
movie star diet; 45.
musician 46.
dentist hurts 47.
truckdriver 48.
laborer ggggtg 49.
lawyer 50.

 

‘

 

The

storeclerk A.

Librarian ..222!22££_.
engineer
professor Mia.

 

Thomd__.1:29_ts____

The

The

mechanic
painter

sociologist
waitress __gigglgg____
singer-x

fi-hm .mn—
designer

 

 

 

 

housewife ___gggg_____
gardener

proofreader
cabdriver w

fireman

carpenter grinkg
milkman

 

 

 

post
“was" ...sezss____
secretary AL.

gambler wing
electrician

 

 

umpire. yell;

* The correct verbs (CR) are written in.

Appendix F

Postrecall Interview

74
ON 3 Answer each question as fully as possible.

1.. that do you think the experiment was about?

2. During the task, did you get the idea that you were supposed
to choose the verbs in any particular way? If yes, in which

way or ways?

Appendix G
Tables of Data). From Rperimental Group

76

Table 1. Frequency Distribution of Scores (Number Correct)
on the Conditioning Task for the Experimental Groups.

 

 

Neutral-Neutral Neutral-Specific Specific-Neutral
Score Group Group Group

 

Aware unaware Aware Unaware Aware Unaware

4‘04

\h
0‘
COCO-4004O-sOOOMOdUdOOONNdU-fi-‘NWNUU-fi

oooooooooooooooooo-s-so—so-aaowwwbmmw
DOA-40400000004dOOdO-AwONOb-IomhhoxO-e

 

g 0-.COCO-54AO-b-lOO-b—bO-bOlS-tKQNmb—Ibmmwoow

1
3
A
3
1
4
1
1
4
O
O
2
2
0
O
O
O
O
O
O
1
O
O
0
O
0
0
0
0
O
O
0
O

27'

3 dOOOOOOOOOOdOOO-lOOO-b-l—BN-I-b-bwmow

Total 36

N
‘x
\e
5

Mean 62.7 61.5 67.3 63.3 66.3 62.6

 

GROUP Total 82 51 61

 

77

Table 3. bequency Distribution of Scores (Number Correct)
on the Recall Test for the Neutral-Neutral and Control Groups.

 

 

 

 

 

 

Score Neutral-Neutral Group Control Ch‘oup
Aware Unaware

24 O O O
23 O 0 0
22 0 O O
21 O O 0
20 1 O 0
19 O 0 O
18 0 0 0
17 O 0 O
16 0 0 0
15 0 0 O
14 0 1 O
13 0 0 0
12 1 1 0
11 1 O O
10 o 0 o
9 1 1 O
8 0 2 0
7 1 2 o
6 2 1 0
5 2 2 o
A O 0 0
3 6 I. o
2 7 9 2
1 7 10 1o
0 7 13 40
Total 36 46 52

 

78

Table 4. Frequency Distribution of Scores (Number Correct)
on the Recall Test for Aware and Unawareifis in the
Neutral-Specific Group.

 

 

 

Score Aware Unaware
24 2 1
23 2 2
22 3 2
20 3 5
19 3 2
18 z. 3
17 1 1
16 2 2
15 O 1
14 0 2
13 1 1
12 0 1
11 O 0
10 o 1

9 1 O
8 o o
7 0 O
6 O 1
5 o o
4. 0 0
3 0 0
2 0 1
1 0 o
0 0 o
Tbtal 24 27

median 19.5 18

 

‘79

Table 5. Frequency Distribution of Source (Number Correct) on
the Recall Test for Aware and Unaware §s in the Specific-
Neutral Group.

 

 

 

 

Score Aware Unaware
24 0 1
23 O O
22 1 O
21 0 2
20 1 0
19 1 O
18 0 0
17 O 1
16 O 1
15 O 0
14. O 2
13 5 0
12 1 o
11 1 1
1o 2 o

9 O 0
8 O 2
7 2 0
6 1 2
5 3 3
4 1 3
3 1 4
2 2 4
1 1 5
O L 3
Total 27 34

Median 7 4

80

Table 6. Requency Distribution of Scores (Number correct)
on the Recall Test for the Experimental Groups.

 

 

 

 

Score N antral-Neutral Neutral-Specific Specifi c—Neutrel
Group Group Group

21, o 3 1
23 0 I. 0
22 0 5 1
21 O 3 2
20 1 8 1
19 0 5 1
18 0 7 O
17 O 2 1
16 0 4 1
15 0 1 0
14 1 2 2
13 O 2 5
12 2, 1 1
1 1 1 0 2
10 0 1 2
9 2 1 O
8 2 0 2
7 3 0 2
6 3 1 3
5 4 O 6
4 O 0 4
3 10 0 5
2 16 1 6
1 17 O 6
O 20 O 7
Total 82 51 61
Median 2 19 5

 

81

Table 7. Number of Correct Recalls to the two halves of Form B
by §s in the Neutral—Neutral and Specific-Neutral Groups.

 

Subject First Half Second Half Sign
1 0 0
2 1 1
3 0 1 +1
4 1 0 -1
5 3 6 +3
6 0 2 +2
7 1 1
8 0 0
9 0 0
1o 0 1 +1
11 1 o -1
12 o 1 +1
13 1 1
1!. 1 5 +1.
15 0 O
16 0 1 +1
17 1 1
18 2 1 -1
19 0 o
20 O 0
21 1 2 +1
22 0 2 +2
23 0 1 +1
24 1 1
25 9 5 -1.
26 U 1 +1
2'7 2 3 +1
28 0 2 +2
29 8 6 -2
30 11 10 -1
31 2 1 -1
32 0 0
33 2 1 .1
34 1 1
35 2 4 +2
36 12 9 -3
3'7 O 1 +1
38 3 2 -1
39 I. 2 —2
40 0 1 +1
41 0 o
42 2 1 .1
43 6 5 -1

82.
(Table 7 continued)

 

 

Subject First Half Second Half Sign
a o 1 +1
45 O 0
46 1 2 +1
47 7 4 -3
48 o 1 +1
49 0 1 +1
50 1 1
51 5 2 -3
52 1 2 +1
53 0 1 +1
54 0 1 +1
55 0 0
56 7 5 -2
57 1 2 4+1
58 1 1
59 2 3 +1
60 o o
61 4 2 -2
62 1 4 +3
63 5 1 -4
64 5 6 +1
65 o o
66 5 7 +2
67 o 1 +1
68 5 5
69 2 2
70 3 2 -1

 

a:

g
.a
O
\O
N
e
O

83

Table ‘9. Hequency Distribution of Scores (Number Correct)
on the Recall Test for Aware and Unaware §s in the Neutral-
Neutral Group based on the Postrecall Interview.

 

 

 

 

 

Score Aware Unaware
24 O 0
23 0 0
22 O O
21 O 0
20 O 0
1 9 0 O
18 O 0
17 O 0
1 6 O O
15 O 0
14 1 0
1 3 0 O
12 o 1
11 1 o
10 O O

9 1 1
8 O O
7 2 1
6 2 0
5 1 O
4 0 0
3 6 o
$ 4 5

4 5
0 3 9

 

84»

Table 10. Erequency Distribution of Scores (Number Correct)
on the Recall Test for Aware and Unaware _S_s in the Specific-
Neutral Group based on the Postrecall Interview.

 

 

 

 

Score Aware Unaware
24 1 O
23 0 O
22 1 0
21 1 1
20 1 0
19 1 O
18 O O
17 1 0
16 1 o
15 0 O
14 1 1
13 5 0
12 1 o
11 2 O
10 2 o

9 0 O
8 1 1
7 2 O
6 1 2
2 4+ 2
3

3 3 2
2 3 3
1 2 4
O 5 2
Total 40 21

85

Table 11. Frequency Distribution of Scores (Number Correct)
on the Recall Test for Aware and Unaware §s in the Neutral-
Specific Group based on the Postrecall Interview.e

 

 

§

Unaware

 

 

21. 2 1
23 2 1
22 3 2
21 2 1
20 5 3
19 3 0
18 I. 3
17 1 1
16 3 1
15 O 1
14 O 2
13 1 1
12 1 O
11 O 0
10 0 1
9 1 0
8 O O
7 O 0
6 1 O
5 O O
4 O O
3 0 0
2 0 0
1 O 0
0 0 0
Total 29 18
Median 19 1g

 

Appendix H

Ekamples of Jokes in Experiment Two

Rossi:
Allen:
Rossi:
Allen:

Rossi:
Allen:
Rossi:
Allen:

Rossi:
Allen:
Rossi:
Allen:

Rossi:
Allen:
Rossi:
Allen:

87

“What qualifications do you need to be a spy?“

”You've got to be a liar, a sneak, a thief, a cheat—-“
'How did you train for this?"

'I was a used car salesman.“

“Are you married?“

”Yeah, my wife's a princess."
"A princess?”

"She's built like a phone."

"What would a spy do if he were attacked by a killer?“
“Some spies run, and some spies scream."

"And what are you?“

‘I'm a screaming runner."

''who's the world's greatest spy?“
“Herbie Margolis.I
“Herbie Margolis!
”You see!"

I never heard of him."

A telephone rings

First voice:

“hello”

Second voice: "What's the easiest way to catch a squirrel?"

First voice:

“Climb a tree and act like a nut.”

A telephone rings

First voice:

"hello"

Second voice: "Do you like frying?“

First voice:

'No, it hurts my arms.“

Appendix I
Postconditioning Interview of hperiment Two

1’.-

2.

3.

5.

6.

7..

8.

9.

10.

89

Would you tell me, in your own words, exactly what we are
studying?

For any reason, did you look forward to the. recorder turning on?
Relatively how much?

Here you able to concentrate on giving numbers the entire period?
How well’ do you believe you did?

Do. you feel you had any control over when the recorder was
turned on? How? then?

Would you mind telling me specifically what you thought about
when you were giving the numbers?

Did anything you said during the experiment have any influence
upon when the recorder was turned on? that?

If I were, to tell you that your saying a certain number
determined when the recorder was turned on, which number would
you guess it would be?

Did you feel the recorder was turned on after number 7? If
yes, when did you notice this?

Hould you describe the fellings' you had about the following
things on a five point scale from none at all to very much.
a. Boredm with giving numbers
b. Eagerness to hear recording

c. Disappointment when recorder turned off

Appendix J
Operant Rate for the Control and EXperimentsl Groups

91

Table 13. Operant Rates fer the Control Group and
Unaware‘fis in the Experimental Groups.

 

 

Control Low Boredom. medium Boredom High Boredom

 

 

[0&5 9e5 600 8‘0
10.0 4.5 4.5 8-0
9e5 705 7‘0 8'0
9.0 8.0 5.0 7.5
10.5 4.5 7.0 5.0
9.0 2.5 4.5 4.0
8.0 2.5 8.0 4.5
11.0 5.5 7.5 7.0
5.0 6.5 4.0 6.5
5.0 4.5 5.0 8.0
4.5 3.5 10.5
3.0 4.0 5.0
5.5 7.0 8.5
7.0 11.0
3.5
9.0
1.0
120.0 74.5 93.5 66.5
17 13 14. 10

7.06 5.73 6.68 6.65