University of Southern California Dissertations and Theses

Audience Effect And Age Differences On Tasks Of Increasing Complexity

(USC Thesis Other)

Audience Effect And Age Differences On Tasks Of Increasing Complexity

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content INFORMATION TO USERS
This material was produced from a microfilm copy of the original document. While
the most advanced technological means to photograph and reproduce this document
have been used, the quality is heavily dependent upon the quality of the original
submitted.
The following explanation of techniques is provided to help you understand
markings or patterns which may appear on this reproduction.
1. The sign or "target" for pages apparently lacking from the document
photographed is "Missing Page(s)". If it was possible to obtain the missing
page(s) or section, they are spliced into the film along with adjacent pages.
This may have necessitated cutting thru an image and duplicating adjacent
pages to insure you complete continuity.
2. When an image on the film is obliterated with a large round black mark, it
is an indication that the photographer suspected that the copy may have
moved during exposure and thus cause a blurred image. You will find a
good image of the page in the adjacent frame.
3. When a map, drawing or chart, etc., was part of the material being
photographed the photographer followed a definite method in
"sectioning" the material. It is customary to begin photoing at the upper
left hand corner of a large sheet and to continue photoing from left to
right in equal sections with a small overlap. If necessary, sectioning is
continued again — beginning below the first row and continuing on until
complete.
4. The majority of users indicate that the textual content is of greatest value,
however, a somewhat higher quality reproduction could be made from
"photographs" if essential to the understanding of the dissertation. Silver
prints of "photographs" may be ordered a t additional charge by writing
the Order Department, giving the catalog number, title, author and
specific pages you wish reproduced.
5. PLEASE NOTE: Some pages may have indistinct print. Filmed as
received.
Xerox University Microfilms
300 North Zeeb Road
Ann Arbor, Michigan 46106
74- 17,345
GRIBBIN, Kathy Jo, 1947-
A UD IEN C E EFFE C T A N D A G E D IFFER EN C ES O N
T A S K S O F INCREASING CO M PLEXITY.
University of Southern California, Ph.D., 1974
Psychology, general
|
) )
1 University Microfilms, A X ERO X Com pany, A nn Arbor, Michigan
THIS DISSERTATION HAS BEEN MICROFILMED EXACTLY AS RECEIVED.
AUDIENCE EFFECT AND AGE DIFFERENCES
ON TASKS OF INCREASING COMPLEXITY
by
Kathy Jo Gribbin
A Dissertation Presented to the
FACULTY OF THE GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
DOCTOR OF PHILOSOPHY
Psychology
February 1974
UNIVERSITY O F SOUTHERN CALIFORNIA
THE GRADUATE SCHOOL
UNIVERSITY PARK
LOS ANGELES, CALI FORNIA 9 0 0 0 7
This dissertation, written by
..J^£h)LJp__Gribbin.....................................................
under the direction of h D i s s e r t a t i o n Com
mittee, and approved by all its members, has
been presented to and accepted by The Graduate
School, in partial fulfillment of requirements of
the degree of
D O C T O R OF P H IL O S O P H Y
Dean
D a te ..^ ^ ^ L t.^ ]l
1SERTATION COMMITTEE
Chairman
TABLE OF CONTENTS
Page
ACKNOWLEDGEMENTS.................................................................... ii
CHAPTER
I. INTRODUCTION............................................................ 1
H. LITERATURE R E V IE W ................................................. 4
Audience Effect
Arousal
Arousal and the Older Person
A Suitable Experimental Task
Age and Task Difficulty
Sex Differences and Age
Summary
Experimental Hypotheses
IH. M E T H O D ........................................ 25
Subjects
Tasks
Procedure
Design of the Experiment
Data Analysis
IV. R E S U L T S ......................................................................... 30
V. DISCUSSION . ................................................................. 41
General Findings
Post-Experiment Questioning
Implications for Future Research
VI. SUMMARY AND CONCLUSIONS.................... 52
BIBLIOGRAPHY .................................................................... 55
iii
LIST OF TABLES
Table Page
1. Task M ean s............................................................................. 31
2. Multivariate Analysis of Variance
Summary T a b l e ............................................................ 32
3. Univariate F Tests: Levels of
Significance.................................................... 33
4. Univariate F Tests: Data for Males
Levels of S ig n ific an c e ................................................ 39
iv
LIST OF FIGURES
Figure Page
1. Age x Condition Interaction:
Stroop Test Initial Reaction T im e ........................... 35
2. Sex x Condition Interaction:
Stroop Test Initial Reaction T im e ........................... 36
3. Sex x Condition Interaction:
Stroop Test Level 3 - Level 2 . . . . .. . . 37
4. Standard E rror of the Means:
Stroop Test RT 1 Means (p < .0 5 )........................... 39
v
CHAPTER I
INTRODUCTION
This research is concerned with whether or not the presence of others
has an effect on the performance of older individuals on tasks on increasing
complexity.
The performance of an individual in the presence of others has long been
of significant interest to social psychologists. A disconcerting finding is that
sometimes performance is improved by the presence of others, and other
times it is impaired. Zajonc (1966) has presented a theory, which has been
modified by Cottrell (1972), in an effort to explain most of the research in
this area. The theory suggests that an audience impairs the acquisition of
new responses, i.e ., learning, and facilitates the emission of well-learned
responses, i.e ., performance. In essence, both Zajonc and Cottrell believe
that the presence of others acts as a drive or arousal producing agent and
thereby affects performance. It is this aspect of the theory which concerns
the present investigation.
Although there has long been much research concerned with Audience
Effect (A.E.), older people have not been of interest. If one focuses on the
initial level of arousal ( i.e ., the baseline, or normal, arousal level of the
individual) as being of significant interest, the study of older individuals and
the A .E. becomes extremely interesting due to the fact that little is known
about the arousal mechanisms of older people or their effect on the older
person's performance. At present there are two opposing positions—older
people are under-aroused, or they are over-aroused. Considering these
positions in light of the inverted-U arousal curve, increasing arousal by the
A .E. would lead to different performance expectations.
1
2
Most investigators have operationalized the arousal level of older per
sons in term s of various physiological indicators. Studies utilizing condition
ing procedures where peripheral measures of autonomic nervous system ac
tivity, such as the galvanic skin response (GSR), and finger vasoconstriction
were used, have concluded that the activation levels of older persons were
lower than those of young persons (Thompson and Marsh, 1973).
In contrast are those studies suggesting that old people are over
aroused. Eisdorfer, et al. (1970) attempted to reduce the arousal level of
older persons by injecting propranolol which purportedly produces a de
creased level of arousal. The results confirmed their contention that the dec
rement in learning performance found in older adults may be the result of
heightened rather than depressed autonomic end organ arousal. Amphet
amines have also been used; however, Jeffrey (1971) concluded that older
people were under-aroused. The concept of A .E. allows us to manipulate
arousal in a more natural situation.
Botwinick (1967) discusses these opposing positions and explains the
apparent inconsistent results by the fact that different measures were used to
assess arousal states, and that the various learning procedures upon which
the arousal states were measured imposed several different kinds of demands
upon the subject. After discussing various learning tasks used in studies
measuring the arousal of older people, Botwinick (1967) proposed the possi
bility that the more meaningful stimuli and the more stressful procedure of
Eisdorfer, et_al_. 's (1970) free fatty acid mobilization study differentially
raised the arousal levels of the older people whereas the less meaningful
stimuli and less stressful procedure of the GSR study differentially lowered
the arousal levels of older people.
This suggests that the task may be a very important factor when study
ing older people. Eisdorfer (1968) has suggested that an increase in task
difficulty may contribute to an increase in the arousal of older people. Many
studies have demonstrated that as tasks become more difficult, we can expect
a more pronounced decline of the performance of older people. Conse
quently, in conjunction with a condition manipulating A. E ., we would expect
an even greater difference for older people.
Relatively few investigations studying the performance of older people
have been concerned with sex differences. In fact, most studies dealing with
arousal and age have used only male subjects. Since sex differences have
been found in some studies where specific abilities are measured (See
Chapter n), it would seem fruitful to investigate the sex differences on tasks
of various complexities while manipulating A .E.
In order to achieve a more comprehensive understanding of the rela
tionship among these variables, a systematic review of those topics which
seem to have relevance to the research problem is presented in Chapter n .
CHAPTER H
LITERATURE REVIEW
Audience Effect
In an effort to differentiate between the various forms an audience may
take, research is normally conducted in one of two experimental paradigms—
the audience paradigm and the coaction paradigm. In the audience paradigm,
the experimenter manipulates the presence of passive spectators as the in
dependent variable, while in the coaction paradigm, the independent variable
is the presence of others who work independently at the same time. The de
pendent variable is the performance on the task for both paradigms.
Although much research using both paradigms has been conducted since
1898 when Triplett first observed that bicycle racers performed differently
when racing against a competitor than when racing against the clock alone,
little more about the phenomenon was understood even half a century later as
evidenced by Allee's (Allee, e^aL , 1950) noncommittal description stating
that:
In general, social facilitation refers to any increment or
decrement in an individual's behavior resulting from the pres
ence of another organism.
Such a description raises several questions. What defines the "presence of
another organism?" The literature shows that performance may increase or
decrease when an individual is in the presence of others. Under what condi
tions do these occur ? Is there an explanation to resolve such conflicting
results ?
Zajonc (1965, 1966) has suggested a hypothesis, utilizing the coaction
and audience paradigms, which integrates the apparent discrepant results of
performance sometimes being enhanced and other times being impaired by
the presence of others. This is an important achievement because most pre-
4
vious explanations have not been able to account for the vast range of the lit
erature that is encompassed by his hypothesis. After reviewing the vast lit
erature on this topic, he found a subtle consistence. An audience impairs
the acquisition of new responses; i . e . , learning and facilitates the emission
of well-learned responses; i . e . , performance. Continuing his explanation,
he states that in the early stages of the learning process a subject gives a
greater number of wrong responses than right ones; the wrong responses are
stronger and predominant. Correct responses predominate once the indi
vidual has successfully learned the task. This observation led him to the
statement that, "Audience enhances the emission of the dominant response"
(Zajonc, 1966).
The psychological process he used to explain this effect is "moti
vation." According to Birch and Veroff (1966), increased motivation in
creases the general arousal level of the organism, activates it, and energizes
it. Spence and others (e.g., 1956) have shown that the dominant response
seen under increased motivation tends to be sharper, more intense, and
quicker. He does not, however, take into account initial levels of arousal,
or conditions of over-arousal which create added dimensions to be resolved.
For example, Broen and Storm's (1961) concept of "reaction potential ceiling"
(See also Malmo's (1958) inverted-U arousal curve) hypothesizes that if a
drive is strong enough to bring the dominant response to excitation ceiling,
any further increase in drive would serve to increase the probability of the
occurrence of competing responses, while decreasing the probability of the
occurrence of the dominant response. Although not specifically tested in the
A .E. paradigm, there is evidence supporting this effect (e.g ., Frankel,
1969). Such an effect may prove very important to Zajonc's hypothesis. If
the subject's initial level of arousal is high upon entering the experimental
situation, and if the presence of an audience increases it beyond the "ceil
ing," other confounding variables may be introduced. This aspect of his
formulation is definitely in need of further development.
Another aspect of the Zajonc hypothesis not completely resolved con
cerns the presence of an audience having the effect of increasing drive or
arousal. There is some physiological evidence that the presence of other in
creases arousal (See Cottrell, 1972, for a thorough review), although not all
reported investigations have been positive (Henchy and Glass, 1968). It is
apparent that this issue is not yet resolved.
Cottrell (1968) has suggested a modification of Zajonc's hypothesis
which helps to resolve the divergent findings of the effect of others on arousal.
Zajonc implies that the drive increasing property of the presence of others
is biologically innate; Cottrell suggests, instead, that the presence of others
is a learned source of drive. To begin with the organism is motivationally
neutral in the presence of others, but with the classical conditioning associ
ated with the experience of consistent association with events that are them
selves drive-producing, the mere presence of others gradually loses its
neutral quality and becomes sufficient to increase the individual's drive level.
Such a modification of Zajonc's hypothesis seems better able to take into
account individual differences, especially as related to their individual levels
of arousal. The presence of others in some types of tasks may foment an
evaluative aspect, the anticipation of which triggers an arousal mechanism.
Such an explanation is consistent with the experimental findings.
Before proceeding further, it seems advisable to discuss the term
"dominant response," an integral term in the formulation of Zajonc's hypoth
esis. Normally the dominant response refers to the response with the
strongest habit strength. Consider a hierarchy of competing responses. As
one first learns a task, incorrect responses have a higher habit strength
(and thus are dominant). Consequently, they have a greater probability of
being elicited. As the subject learns the task, the correct response tends to
increase in habit strength, thereby becoming the dominant response in the
hierarchy, and is emitted more frequently.
The original literature on A .E. and social facilitation does not define
and, of course, does not experimentally manipulate the dominant response.
Zajonc had to infer this when formulating his hypothesis. When testing his
hypothesis, however, the dominant response must be independently derived,
explicitly identified and experimentally manipulated in order to be reasonably
certain that the experimental conditions are indeed affecting the elicitation
of the dominant response; otherwise, it would seem that all we have is a
circular definition—i . e . , it is the dominant response because its elicitation
is increased by the presence of others. Much of the recent literature, how
ever, has experimentally manipulated the dominant response by establishing
habits of different strengths during an initial period of training in the exper
iment itself (Cottrell, et_al., 1968; Zajonc and Niewenhuyse, 1964; Zajonc
and Sales, 1966), by observing the subject's preference for different response
alternatives before manipulating the independent variable (Strain, 1967), or
by using population norms to identify the dominant responses to the task
stimuli (Cottrell, Rittle and Wack, 1967; Matlin and Zajonc, 1968). The
studies just referred to will not be discussed in more detail (See Cottrell,
(1972) for a detailed comparison and description); it is sufficient to say that
their results are in line with what one would expect from the Zajonc hypoth
esis. Cottrell, et_al_. (1968) will be discussed more fully, however, because
one of his groups did not manifest the expected results.
Cottrell, et al. (1968) attempted to evaluate the proposal of Zajonc
that the mere presence of other persons is responsible for A.E. A total of
45 university students performed a pseudo recognition task (the stimuli are
flashed taehistoscopically) where the previously hierarchically well-estab
lished verbal habits were placed in competition with each other. Fifteen
students performed the task alone, 15 performed before an audience of two
blindfolded students (which will be termed the mere presence condition) and
the rest performed the task before two stooges who were there supposedly
waiting to be in another experiment. The presence of an audience enhanced
the emission of dominant responses, but the mere presence of the others did
8
not. In interpreting these results, Cottrell (1968) suggested that the key to
these unexpected results is evaluation—in the audience paradigm, the pres
ence of others will increase the emission of the dominant responses only when
the spectators can evaluate the individual's performance. In the blindfold
condition, subjects understood that they could not be evaluated because the
spectators were unable to see to what the subject was responding.
That evaluation is a critical factor in affecting performance in the A.E.
paradigm has been demonstrated in several other studies. Paulus and
Murdoch (1970) used a unique design to test the effect of evaluation. Unlike
the previous studies which asked subjects to vocalize their responses, they
instructed their subjects to write their responses to a pseudo-recognition
task. They used a 2 x 2 design in which they manipulated both the audience
(or no audience) and evaluation (or no evaluation). The evaluation condition
was induced by informing subjects that their written answers to the experi
mental task would be evaluated at the end of the session. The results support
the contention of Cottrell that evaluation (or expectation of evaluation) is the
critical element influencing performance, and not the "m ere presence" of
others.
Henchy and Glass (1968) sim ilarly tested the effect of perceived evalua
tion. They used four conditions. In the "alone condition" subjects were given
no special instructions concerning their performance and there were no ob
servers present. Another group, the "alone-recorded" group also had no ob
servers present but the subject was told that later specialists would evaluate
films and tape recordings of his performance. Observers sat quietly and
watched each subject in the other conditions; however, in one group the sub
ject was told that these individuals were experts in human learning and per
ception and were planning to evaluate his performance ("expert condition"),
and in the other group ("no-expert condition"), the subject was instructed that
the observers were undergraduate students who wanted to watch a psycholog
ical experiment. The results were as expected. The greatest increase of
the dominant response was found in the Expert and Alone-recorded conditions;
however, there was also a statistically significant difference between the No
expert and Alone conditions. Consequently, it was shown that evaluation is
certainly the more critical factor, but that the presence of others does have
an effect.
Martens and Landers (1972) also support that it is evaluation apprehen
sion, rather than mere presence which leads to differences in performance in
the coaction paradigm. Another interesting finding in their study is that as
the number of coactors increased in a motor task, performance decreased.
In summary, such results support the statement by Jones and Gerard
(1967), that social facilitation effects are "mediated prim arily through self-
consciousness on the part of the subject, a concern with how well he appears
in the eyes of others" (p. 607).
The only study using older subjects in the A.E. paradigm found that
older subjects were as susceptible to the phenomenon as were young subjects.
Gribbin (1973) used an age peer as the audience and observed the perform
ance of young and old women on tasks of increasing complexity. She found
that the audience impaired performance for both young and old; however, the
effect was constant. That is, older subjects performed more slowly in both
conditions than young subjects, but their performance was not differentially
affected by the introduction of an audience.
Arousal
Experimental social psychologists (e.g ., Cottrell, 1968; Cottrell,
et_al., 1967; Shapiro and Crider, 1968; Zajonc, 1965, 1966) use the concept
of motivation or arousal when conducting research on the influence of social
psychological variables on performance. Normally they employ two broad
classes of methods to classify human individuals as to motivational status.
One is to use psychometric procedures such as anxiety scales, the Thematic
10
Apperception Test, questionnaires or inventories, self-rating procedures or
rating by others, in order to get at motivational factors; the other method is
to use arousal conditions such as observers present, threats of shock, em
barrassm ent, stress, failure, e tc ., to establish the motivated state. Some
times both of these procedures are combined; that is, the effects of arousal
are studied in people classified as to motivational states by some psycho
metric procedure, or a psychometric procedure is used to assess the effect
of the arousal operations.
Regardless of which procedure is used, the motivational effects typi
cally are evaluated by a behavioral index. Seldom are physiological mea
sures employed. Such a procedure is theoretically sound as long as all other
relevant variables have been controlled. For example, if it is predicted that
a given motive will facilitate a particular type of performance and people who
score high on a measure of that motive (or whose motive has been experi
mentally aroused) do perform better than low-scoring people (or those in
whom no arousal has been manipulated), one may argue that the measure
ment (or the arousal operation) is valid because the expected effects have
been obtained. Such a procedure, however, merely tells us that the manipu
lations have produced changes. They tell us little about what is actually going
on or how these motive states are operating.
Even when physiological measures of change are taken as an index of
changes in motivation or "arousal," this problem is not always solved be
cause there is no general agreement on the operation of this basic variable
or its relation to manifest behavioral changes. Grossman (1967) states that
"most Western investigators use the term 'orienting response,' 'alerting*
response, or 'arousal' response interchangeably to refer to the electro-phys
iological changes elicited by a novel stimulus," while Botwinick (1967) refers
to "the 'degree of arousal' as a motive state of the generalized, undifferen
tiated type." However, he is still referring to physiological processes.
11
Such discussions emphasize the weakness of descriptions of "activation"
or "arousal." It is often stated that a concept can have only one definition;
nevertheless, it must have both a formal and an operational definition which
are not necessarily congruent. It seems that these investigators have opera
tionalized arousal, but have never really bothered to define it. Botwinick's
discussion comes closer to a formal definition than do many, but it still
refers to the physiological processes which operationalize it and would be un
suitable as a formal definition to those with a more social psychological ori
entation. It is obvious that there is still the need for a formal definition of
this concept. At present it remains an ambiguous term used only for conve
nience to refer to some "activated" state of the organism, and discussion of
it can only be undertaken with respect to the particular operational definition
being used.
A problem which arises when attempting to integrate the various find
ings regarding arousal is that few of the physiological measures that are
taken have high intercorrelations among themselves. This may be because
the physiological indicators were measuring "arousal" on a variety of tasks,
thus requiring a subject to utilize different processes. Alternatively, there
may be no generalized state of arousal that can be described across arousing
situations and types of tasks. Definitely more research is needed in this
area before science will be able to understand what arousal means and how
it is to be measured.
Arousal and the Older Person
Although there has long been much research concerned with A .E .,
older subjects have not been of interest. If one focuses on the initial level
of arousal as being of significant interest, the study of older individuals and
the A .E. becomes extremely interesting due to the fact that little is under
stood about the arousal mechanisms of older people or their effect on the
older individual's performance. At present there are two opposing positions
—older people are under-aroused, or they are over-aroused.
12
Botwinick and Kornetsky (1960) conducted a study in which the galvanic
skin response (GSR) was measured. The GSR is considered to be an index of
sympathetic activity of the autonomic nervous system and, therefore, indi
cates a general level of activation. These researchers conditioned and ex
tinguished the GSR’s of both elderly and young subjects. They felt easier
conditioning of the respective autonomic nervous systems would suggest a
generally more activated system. The results of their study showed that
elderly subjects conditioned less readily and extinguished more readily than
the younger subjects. Consequently, Botwinick and Kornetsky (1960) con
cluded that the activation levels of older subjects were lower than those of
younger subjects.
Schmavonian, Yarmat and Cohen (1965), also using a conditioning study
found the finger vasoconstriction and GSR of old subjects to be significantly
less than that of young subjects; however, a measure of EEG beta activity
was found to be significantly greater in the old subjects. They interpreted
this greater beta activity to be conditioned inhibition and thus consistent with
their other measures; however, others are not totally in agreement with such
an interpretation (Beta activity is usually seen to increase from adulthood to
middle-age and then to decrease. Obrist and Busse (1965) suggest that "be
cause fast activity is most prevalent among intellectually well-preserved
subjects during early senescence, its presence in an elderly person’s EEG
can probably be regarded as a favorable sign.") and consequently, would
consider it more as suggesting activation.
Silverman, Cohen and Schmavonian (1959) measured nonspecific GSR’s
(i.e ., those spontaneously emitted under various types of stimulation). They
found that the number of GSR nonspecifics during tone stimulation was sig
nificantly lower for older men than for younger; however, when more mean
ingful stimulation was used, spoken phrases, the differences between the age
groups diminished and there was no significant difference. Discussing this
13
data, Schmavonian and Busse (1963) concluded that the young were more
responsive than the old; nevertheless, the meaningfulness of the stimulation
was an important factor.
An additional study supporting the contention that old people are under
aroused was conducted by Surwillo (1966). He measured heart rate, palmar
skin temperature, and palmar skin potential during performance of a vigi
lance task. He found that although the number of e rro rs was greater with
time for the old than for the young, both groups were sim ilar in the rate of
change for heart rate and skin potential. However, he found a difference
between the rate of change of skin tem perature. The old tended to increase
while the young decreased. In light of the fact that a decrease in skin tem
perature usually occurs when subjects are attending to a task and an increase
in skin temperature is more consistent with a sleep state, he also concluded
that the elderly were under-aroused.
Before proceeding to the research suggesting that older people are
over-aroused, it is useful to discuss Lacey, et al. 's (1963) findings on heart
rate changes. These investigators have found that heart rate tends to in
crease in tasks involving an internal direction of attention (e .g ., cognitive
tasks) and to decrease in tasks involving an external direction of attention.
It is hypothesized by J . I. Lacey (1967) that heart rate deceleration may
facilitate environmental intake and related behavioral functioning. Support
for this position can be found in studies reporting beat-by-beat deceleration
during fixed foreperiods in traditional reaction time tasks (Lacey and Lacey,
1965). Heart rate deceleration reaches its lowest point at or shortly before
presentation of the response signal, and is then followed by rapid accelera
tion over the next several heart beats. In addition, the rate of deceleration
is positively correlated with response speed. In other words, the greater
the rate of change, the faster the response.
Testing this hypothesis using both older and younger subjects, Morris
and Thompson (1969) investigated the effects of age on heart rate deceleration
14
and respiration in a reaction time task with three different fixed foreperiod
intervals. They found that the observed changes were statistically signifi
cant. However, the old subjects decelerated less than one beat-per-minute
in all conditions, while young subjects showed significantly more deceleration
(approximately 3.5 to 4.0 beats per minute). In addition, supporting Lacey's
hypothesis, there was a relationship between rate of change and response
speed in the young, but not in the old. Such results would tend to support the
hypothesis that older people show less autonomic nervous system reactivity
than do young subjects.
Studies suggesting that old people are over-aroused have mostly been
carried out by Eisdorfer and his colleagues at Duke University. Powell,
Eisdorfer, and Bogdonoff (1964) found that the free fatty acid level (FFA) of
older people was higher both before and after a learning task and tended to
remain elevated long after that of the younger subjects had returned to base
line. (FFA mobilization is considered to be an indicator of autonomic
arousal.) In a later study Eisdorfer, Nowlin and Wilkie (1970) carried these
results further. Their experimental design called for the collection of two
blood samples, taken 15 minutes apart to determine base line plasma FFA.
Then subjects were randomly assigned to receive either an intravenous solu
tion of 10.0 mg of a placebo, isotonic saline, or an equivalent amount of
intravenous propranolol hydrochloride which purportedly produces a "partial
blockade of autonomic beta-adrenergic receptor sites in peripheral end
organs which largely mitigates more physiological concomitants of central
nervous system arousal." In other words, the effect of propanolol is to re
duce the level of arousal. Subjects, 28 paid male volunteers, aged 60+, were
then given a serial rote learning task (described in detail in Eisdorfer, 1968).
Eisdorfer, et_al., (1970) concluded that the findings of their study are signif
icant in two respects. The results confirm their contention that learning in
elderly people can be improved by pharmacologic modification of the auto
nomic nervous system, that the decrement in learning performance found in
15
older adults may be the result of heightened rather than depressed auto
nomic end organ arousal.
In a recent study, Herr (1973) has lent additional support to this hypoth
esis. Using young and old rats, he injected epinephrine, propranolol HC1, or
peanut oil (control), and observed their performance on a conditioned avoid
ance task. He found that his older controls performed significantly better
than did his young controls, that propranolol impaired performance of sub
jects of both ages, and that epinephrine impaired performance of only the
older subjects. He interpreted his results as tending to support Eisdorfer's
hypothesis.
Using a vigilance task, Bicknell (1971) manipulated arousal levels of
old and young subjects with conditions of constant white noise (a nonarousing
stimulus) and variable white noise (an arousing stimulus). She found age
differences in that the old subjects made more e rro rs of commission; how
ever, she was unable to demonstrate that her arousal condition significantly
affected performance.
In one of the few behavioral manipulations of arousal, Ganzler (1964)
studied the effect of highly motivating instructions vs less highly motivating
instructions on the performance of old and young subjects. The different
instructions did not differentially affect old as opposed to young subjects. He
concluded that the relative decrements in the performances of the elderly
subjects were probably not attributable to age differences in motivation.
Such opposing findings make it difficult to draw conclusions about
"arousal" in older people. Much of this research led Schmavonian, et al.
(1965, p. 255) to suggest, "Apparently the aged organism reacts to a novel
and possibly stressful situation with total mobilization of its biochemical
defense by 'giving it all it's g o t,' whereas the young men react more econom
ically; i . e . , by responding to the immediate situation rather than maintained
vigil." This statement seems logical in light of the research on FFA mobi
lization and also Schmavonian, et al. 's (1965) data demonstrating that older
16
men excreted adrenalin at a higher rate than did the younger men, both before
and after the study was concluded.
Botwinick (1967) discusses these opposing positions and explains the
apparent inconsistent results by the fact that different measures were used
to assess arousal states, and that the various learning procedures upon which
the arousal states were measured imposed several different kinds of demands
upon the subject. After discussing various learning tasks used in studies
measuring the arousal of older subjects, Botwinick (1967) proposed the pos
sibility that the more meaningful stimuli and the more stressful procedure of
the FFA study differentially raised the arousal levels of the older people as
opposed to the possibility that the less meaningful stimuli and less stressful
procedure of the GSR study differentially lowered the arousal levels of older
people.
In an alternative explanation of these opposing conclusions concerning
the autonomic reactivity of elderly people, Thompson and Marsh (1972) sug
gest that traditional physiological measures may not be appropriate for the
elderly. For example changes in the skin may confound measurement of GSR
while its activity has changed in some other way.
The studies concerned with manipulating the arousal level of older peo
ple have done so by the administration of drugs (e .g ., Eisdorfer, et_al_.,
1970; Jeffrey, 1971) to achieve the desired effect. The A.E. paradigm may
have particular relevance for aging research because it allow us to manipu
late arousal in what is a much more natural situation. We have all been and
will continue to be in situations requiring us to function in the company of
other people. Understanding such effects may enable us to plan more ade
quately the environments most conducive to the optimum performance of
older people.
This review of aging and "arousal" literature underscores the lack of
knowledge about and demonstrates the need for further study of the relation
ship between age and arousal and its effect on performance. Because of the
17
inconsistencies described above by Botwinick, it may be desirable to measure
simultaneously a variety of indicators of autonomic arousal on tasks purported
to get at different functions. Such knowledge about differences in age groups
should enable us better to understand why the older person's performance de
clines under certain conditions and may suggest possible intervention tech
niques.
A Suitable Experimental Task
One difficulty encountered when using the classical A. E. paradigm with
older subjects in that most experiments utilize a learning task. Most re
search on learning demonstrates a performance decrement with increasing
age (e.g., Arenberg, 1965; Canestrari, 1963; Gilbert, 1941). It is unre
solved whether the decline is due to a relative inability to learn, or to non-
cognitive factors which change with age, leading to poorer performance
(Botwinick, 1967). At any rate, trying to establish learning hierarchies in
the aged and then examining A.E. differences in young and aged subjects is
questionable. For example, the learning hierarchy may never truly be es
tablished and the subject's increased use of the hierarchically established
dominant response may be simply because it was the only learned response
and not due to any differences in arousal. Consequently, some other task is
needed. Nonlearning tasks such as color preferences, vowel cancellation,
e tc., have been used in the A.E. paradigm; however, merely to repeat these
studies using older and younger subjects tells us little more about the process
that is occurring.
When formulating his theory, Zajonc described a hierarchy of com
peting responses (Cottrell, 1972), the most probable of which is the dominant
response. This learning process can be viewed as the modification of the
hierarchies of competing responses elicited by the task stimuli. At first, the
correct response has a low probability of being emitted. Another response
is dominant. Consequently, the subject's overt responses are usually wrong.
The response to be learned is strengthened as the learning progresses,
18
thereby becoming dominant, and, naturally, is emitted more and more fre
quently. In certain tasks, this can be likened to improving discrimination
abilities; i . e . , one learns to be a good discriminator. Therefore, when the
discrimination is very apparent (i.e ., dominant), it is easily and quickly
elicited. As the discrimination becomes more difficult, however, the subject
begins to "block" (Birren, personal communication) and the response takes
longer to be elicited. Such an approach extends the theoretical basis of
Zajonc's "dominant response." If one completes a series of tasks varying
from a simple discrimination to tasks of increasing difficulty, while being
subjected to conditions designed to manipulate arousal, we should begin to
get a better picture of what is actually going on than Zajonc's theory is able
to give us. This more subtle distinction may be particularly important when
testing older subjects because of the interaction between age and task diffi
culty.
Age and Task Difficulty
Several investigators concerned with aging research have been inter
ested in performance changes due to the interaction between age and task
difficulty (e .g ., Birren, 1955; Birren and Botwinick, 1951, 1955; Kay, 1954,
1955; Klein, 1972; Rabbitt, 1964). In an experiment conducted by Birren and
Botwinick (1951), it was found that on more difficult tasks, the response la
tency of older subjects increased significantly more than did that of younger
subjects. In a later study these investigators suggested that the difficulty of
perceiving the stimulus accounted for only part of the increased latency which
was observed with older subjects on the more difficult tasks (Birren and
Botwinick, 1955). In this study they systematically varied the difficulty of a
discrimination task and found an interaction effect. In other words, as the
task difficulty increased, they found the response times of both younger and
older subjects increased; however, the increase was larger for older sub
jects. This illustrates the rather general finding that older subjects seem to
have a greater disadvantage with more difficult tasks.
19
Eisdorfer, et_al. (1963) varied task difficulty by changing the rate of
presentation of stimulus words in an 8 item serial rote learning paradigm.
As the time limits increased from 4 to 6 to 8 seconds between stimulus pre
sentations there was a significant improvement in the performance of the
older subjects, but not in that of younger subjects. The older subject's per
formance seemed to improve as a result of an increased tendency to respond
with increasing stimulus exposure time. Further experimentation demon
strated that the failure of the older subjects to respond was not a result of
their not having enough time to respond because in self pacing conditions they
responded within the time limits imposed by the fast paced conditions.
Eisdorfer (1968) attributed this lack of response to an increase in arousal
induced by the task. Bather than being uninvolved with the task, the older
individual is quite involved and therefore experiencing a greater degree of
internal activation. He suggests that task anxiety is maximized under the
faster paced stimulus presentation intervals and minimized under the slower
paced conditions. Such a conclusion would suggest that there is an inter
action between task difficulty and the task anxiety induced by it.
In a subsequent experiment designed to test the hypothesis that task
difficulty at 4 seconds is responsible for the greater number of erro rs,
Eisdorfer (1968) attempted to reduce the relative task complexity by using a
select sample of subjects of higher than normal intelligence who would find
learning tasks easier. His data supports the contention that with relatively
more simple tasks, the performance of aged subjects improves.
In summary, such studies suggest that on more difficult tasks we can
expect a more pronounced decline of the performance of older subjects. In
conjunction with a condition manipulating A. E ., we would expect an even
greater decline for older subjects.
Sex Differences and Age
Researchers interested in the aging process have paid relatively little
attention to sex differences in behavior. This seems particularly surprising
20
considering the well known vast physiological changes which women undergo
as they age and which definitely have an effect on behavior. Perhaps, how
ever, this is one explanation for why most of the research on aging arousal
discussed above was conducted solely with male subjects who do not undergo
these vast physiological changes and should, therefore, show less variability.
Most of the literature on sex differences has been concerned with intel
ligence on a general level and seldom are sex differences found. When spe
cific abilities are measured, however, sex differences do emerge. Women
do better than men on tasks which are verbally loaded such as analogies,
disarranged sentences, and vocabulary, and also digit symbol substitution,
while men are superior in spatial and arithmetic ability (Anastasi, 1967;
Garai and Sheinfeld, 1968; Jones and Conrad, 1933; Maccoby, 1966;
McNemar, 1942; Mellone, 1944-45). Such differences have also been found
with increasing age as demonstrated by continued discrepancies in arithmetic
(Doppelt and Wallace, 1955), verbal scores (Feingold, 1950), and recall
scores (Bromley, 1958).
In a longitudinal study designed to test changes in several different spe
cific abilities, Blum et_al_. (1972) measured older men and women on
Vocabulary, Sim ilarities, Digits Forward, Digits Backward, Tapping, Block
Design and Digit Symbol Substitution. They found that the mean scores of
older women were higher than those of men on all tests except Digits Forward
and Digits Backward, but the differences were statistically significant for
only Tapping and Digit Symbol Substitution. They concluded that the trend in
this study was for men to show an earlier and more rapid decline than women
which may reflect the shorter life span of the men, since earlier studies have
shown a significant decrease in intelligence shortly before death (Jarvik and
Blum, 1971). Certainly such an interpretation remains to be tested and,
possibly in the next testing of the longitudinal study, it may be validated. The
fact that there were differences depending on sex is substantiated, however.
Only a few studies of age and sex differences in problem solving have
been conducted. Interestingly, they come to opposite conclusions. Young
(1971) found that up to age forty, men perform better than women on problem
solving tasks, the results of which are in agreement with studies using young
adults. In the fifties, however, men performed more efficiently than women on
the simpler tasks, but did not show this advantage when the complexity was
increased. By the sixties there was little difference between the perform
ances of men and women. These results are consistent with those of Blum,
et_al_. (In press) in that it was the performance of the men that declined,
that of the women remained relatively constant. In contrast to this, Nehrke
(1972) studied syllogistic reasoning ability of persons classified according to
age, sex and education background. He found a significant age difference, as
expected, but also found that men were superior to women in all age groups.
Both studies classified their age groups on the same basis; however, Nehrke
used a much larger sample. Education was used as an independent variable
in Nehrkefs study and was found to have a significant effect. Young found
that years of education correlated positively with the problem solving scores
for men but not for women. Neither study covaried years of education to
parcel out this effect. Consequently, it is difficult to compare them on this
basis. Each study used a different problem solving task which is most likely
the prim ary reason for the different findings. Young used the Heuristic
Evaluation Problem Program mer, while Nehrke used logical syllogisms. It
would seem that until more research is conducted in the problem solving area
using different types of tasks and correcting for some of the weaknesses dis
cussed, our understanding of this ability with relation to age and sex differ
ences will not increase.
The above studies were dealing with cognitive abilities at a fairly high
level of cognitive functioning. Bettner, Jarvik and Blum (1971) used the
Stroop Color-Word test, which has been used as a measure of cognitive de
cline, and found that performance times were consistently longer for males
22
than for females. The findings of this study are relatively important because
the present investigation also uses the Stroop Color-Word test as one of the
task manipulations. The Stroop test requires the subject to read words and
to name colors; in addition, it provides an index of the conflict resulting when
word reading interferes with color naming. Consequently, it can be seen as
a task which increases in complexity. Nevertheless, it does not involve the
same level of difficulty as does problem solving.
Few studies have been concerned with age and sex differences and auto
nomic reactivity. Schmavonian, Yarmat, and Cohen (1965) conducted a study
testing age and sex differences in cardiovascular reactivity and conditioning.
They studied young men and women and old men, but failed to include a group
of aged females. They found sex differences in performance, with the young
men conditioning better than the young women. The young women, in turn,
conditioned better than the old men. Due to the lack of data, old women
could not be compared. In a later study, Schmavonian, et^al. (1970) mea
sured the respiration, heart rate, pulse pressure wave (by means of a digital
plethysmogram), EEG, and catecholamine levels of old men (age range 61 to
73), old women (age range 68 to 76), young men (age range 18 to 25) and
young women (age range 17 to 21). These subjects were tested in a discrim
ination condition paradigm. They concluded that their study demonstrated
clear-cut age and sex differences in cardiovascular reactivity and condi
tioning. In their conclusion, however, they fail to include much discussion
of the responses of old women. Most of the sex differences were contributed
by the young females and the aged females were given only cursory attention.
Consequently, our understanding of the responses of older women was not
greatly aided by this study.
In the vigilance task discussed above, Bicknell (1971) also studied sex
differences. She found that women performed significantly better than men
on correct signal detections and also found a sex by age interaction; however,
she did not describe the source of this variance.
23
Summary
Eisdorfer (1968) and Eisdorfer, et_al. 's (1970) data on FFA mobiliza
tion and adrenalin excretion levels suggest that elderly subjects are more
aroused by the total experimental situation than are younger subjects. How
ever, the galvanic skin response conditioning and extinction data of Botwinick
and Kornetsky (1960), the vasoconstriction conditioning and extinction data of
Schmavonian, et_al_. (1965) and the heart rate data of Morris and Thompson
(1969) support the position that older people are less aroused than are the
young. Such inconsistent findings are not easily explained; nevertheless, it is
possible that these inconsistencies are due to the different measures used and
the different procedures involved in measuring the motive states. For ex
ample, such a factor as the meaningfulness of the stimulation has been found
to contribute differentially more to the motivational state of older people than
to that of younger people. Consequently, it is extremely important to be
aware of such factors when conducting research using different age groups.
Yet at the present time very little is known about what factors contribute dif
ferentially in different age groups.
In the social psychological literature, it has long been substantiated
that the presence of others either improves or impairs the performance of
individuals; however, in the vast literature on A .E ., older subjects have been
used in only one study. Since Zajonc and Cottrell consider increasing arousal
to be the factor influencing these performances, and since it has been sub
stantiated that the arousal level of older people is different from that of
younger people, it would seem that observing older and younger subjects in
the A .E. paradigm is an effective means to gain further understanding of the
effect such a factor plays in the functioning of older as opposed to younger
persons.
It would seem that the initial level of arousal is very important when
using such a paradigm as the A.E. in aging research. It may be, as implied
by Eisdorfer, et al. (1970) that with certain tasks, the arousal level of older
24
people is already near ceiling and impairs rather than improves perfor
mance. Also, some indices of arousal, for example, blood pressure, are
already higher in older people than in younger and in absolute term s may not
be able to increase as much as those of younger people. At present such
questions cannot be answered.
It is the intent of this research to lay the ground work to answer such
questions and to discover what happens, when arousal is manipulated, to the
performance of older as opposed to younger subjects on a series of tasks of
differing complexities, and whether or not there are sex differences under
these conditions.
Experimental Hypotheses
The hypotheses being considered in this research concern age and sex
differences in performance in the A .E. paradigm under various levels of task
complexity. The hypotheses are based on the data available in the literature
on A .E ., and on arousal and age.
Hypothesis I: The Audience condition will improve performance on
the simple tasks and will impair it on the more complex tasks.
Hypothesis II: The performance of older subjects in the Audience con
dition will improve significantly more on the simple tasks and will be signifi
cantly worse on the more difficult tasks than that of the young subjects.
It is also expected that due to the nature of the tasks chosen, older
women will do significantly better than older men as the tasks increase in
complexity, while there will be no differences in the young subjects.
CHAPTER III
METHOD
Subjects
Sixty-six subjects were recruited for the experiment. The subjects
were all volunteers. Thirty-one of the subjects were older. Of these 12
were men (52-67 years; mean age 61.7 years) and 19 were women (52-85
years; mean age 66.5 years). Of the thirty-five young subjects, 15 were
men (19-28 years; mean age 21.3 years) and 20 were women(18-23 years;
mean age 20.5). The young subjects were students from the University of
Southern California who were enrolled in courses concerned with aging. The
subjects in the older sample were participants of the 1973 Summer Institute
in Gerontology at the University of Southern California. The older sample
was definitely not representative of their age population as they had a much
higher level of education than is typical. * Subjects were not screened by an
intelligence test; however, with older subjects education level is a good indi
cator of intelligence. The mean education level of the young subjects was
15.1 years, while that of the older subjects was 16.1 years.
*
An advantage to the sample used in this study is the fact that both old
and young subjects were students with an interest in aging. This suggests a
sim ilarity not normally possible when comparing performance of different
age populations. The artificial homogeneity of the sample, however, does
create some disadvantages due to the high selectivity of the group. In partic
ular, the responses of the older group may have been atypical due to the ex
perience of this group. Since a high education level in older people is nor
mally associated with many other positive characteristics, such factors may
have obscured otherwise typical age differences.
25
26
Tasks
A reliable battery of tasks ranging from simple to those of increasing
complexity is needed in order to assess the possible interaction effect of task
difficulty and performance which has been reported in the literature. The
tasks chosen include a vowel cancellation task, the Digit Symbol Substitution
test from the WAIS, the Stroop Color-Word test, and a "creativity" test.
The vowel cancellation test required the subject to cross out all of the
vowels which had been randomly placed among consonants on an 8 1/2 x 11
sheet of paper. The subject was asked to work as quickly as he could.
E rrors of commission and omission were measured, as well as the time it
took to complete the task.
The Digit Symbol Substitution (DSS) test was administered as described
in the WAIS except that subjects were timed to completion of the test rather
than limited to 90 seconds.
The Stroop test was administered in three stages. The initial stage of
the task is to read the names of colors printed in black ink. Next the subject
was required to name the color of ink printed on the page. In the final part of
the task, the names of colors are printed in an ink that is a different color
from the one signified by the word. The subject had to name the color of ink
that the word is printed in as opposed to the word printed. For example, if
the word blue is printed in brown ink, the subject was required to say
"brown." Reaction times were measured on all stages and erro rs were mea
sured on the final stage of the task.
The "creativity" test is a very difficult task which emphasizes abstract
verbal abilities. The test includes 20 sets of three words. The subject is
asked to find a word that the three given words have in common. For ex
ample: paint, doll, and cat. The answer is "house"—"house paint," "doll
house," and "house cat." Instructions were repeated until the subject under
stood the nature of the task. He was then given ten minutes to complete as
many of the 20 sets as he could. The number correct and the number wrong
27
were measured.
The order of these tasks was counter-balanced in order to eliminate
any progressive e rro r that might occur.
All of these tests are fairly brief. In general, it took about 30 minutes
to administer the entire battery. Since elderly subjects may not always have
the motivation to remain in a vigorous testing situation, the tasks were
chosen to be as interesting as possible, while still testing the desired effects
of the manipulation.
Procedure
The division of the subjects into the experimental groups was done ran
domly. Half of the subjects in each age group were tested in the No Audience
condition and half were tested in the Audience condition. Males and females
were sim ilarly divided into experimental and control conditions.
All of the subjects were tested in a one-way vision room. Subjects in
the No Audience condition were able to see into the observer's room in order
to be assured that they did not feel that they were being observed. Subjects
in the Audience condition could not see into the other room.
No Audience Condition. —The subject entered the experimental room
and was asked to be seated. The subject was then told:
"This experiment is concerned with how people
perform a series of relatively easy tasks which are
quite different in nature from each other. Of course,
all information will be kept confidential. I am sorry,
but this is the only experimental room I could get. It
is a one-way vision room, but as you can see, no one
is on the other side, so please don't let it bother you."
(The lights on the other side of the m irror were on and the subject was able
to see that no one was in there.) The tasks were then administered in their
pre-designated order.
Audience Condition. —The subject entered the experimental room and
was asked to be seated. The subject was then told:
28
"This experiment is concerned with how people
perform a series of relatively easy tasks which are
quite different in nature from each other. This
room is a one-way vision room. Two professors
from our Human Learning Laboratory who are inter
ested in assessing the performance level of all sub
jects being run here at USC will be observing your
performance. They can see you, but you cannot see
them. When you have completed the tasks, they will
look at your responses and will evaluate your per
formance in comparison with others they have al
ready observed. Of course, all information will be
kept confidential."
The tasks were then administered in their pre-designated order.
At this point subjects were debriefed and were allowed to leave.
Design of the Experiment
The model for the design was a 2 x 2 x 2 (Audience manipulation x sex
of subject x age) multivariate analysis of variance. Therefore, there were 8
independent groups of subjects.
Since in the older sample the average age of the men was less than that
of the women, a pseudo-variable, the deviations of the individual's age from
the mean age of his age group, was covaried to eliminate any effect the age
differences within the group would contribute. This resulted in the testing of
three main effects and four interaction effects.
Years of education is also known to have an effect on performance of
older people on certain tasks. However, this sample was pre-selected so
that the older subjects had all attained a greater than average education for
their age population. Consequently, it was felt that the young and old samples
were equated on this basis. In fact, the education level of the older group was
slightly greater than that of the younger.
Among others, the multivariate analysis allows one to examine the data
with respect to the following questions: a) Are there performance differences
29
on the various tasks as a function of having an audience ? b) A main effect
for age is expected on some tasks, but does A .E. differentially affect per
formance on tasks of different complexity? c) Are there sex differences in
performance ?
Data Analysis
The Stroop test increases in difficulty as the task progresses. In order
to get a measure of the increasing task difficulty which is not confounded by
the initial reaction or movement time, the time taken on the initial stage of
the task was subtracted from the time taken on the second state of difficulty.
The time taken on the second level of complexity was then subtracted from
that taken on the most difficult level. These difference scores were then
analyzed as dependent variables in the multivariate analysis of variance.
This permits one to determine the increasing effect of complexity on a task
which essentially is a repeated m easures design. Due to this, task difficulty
could not be analyzed as an independent variable as it violates the assumption
of independence of erro rs.
Analysis of the other tests required no conversion of the prim ary data.
CHAPTER IV
RESULTS
The A.E. literature predicts performance differences to occur regard
less of the task; however, four different tasks were used in this study and
only the Stroop test, successfully differentiated performance under the audi
ence conditions (See Table 1). The DSS, Stroop, and the vowel cancellation
task demonstrated age differences; however, the "creativity task" showed no
significant differences on any of the independent variables.
The means of each group for all of the tasks are presented in Table 1.
As can be seen in Table 2, age differences in the speed of response were
found for the DSS, Stroop test, and the Vowel Cancellation task (p < .001 for
all). On the Stroop test, there were no age differences on the initial speed
of reaction time (See Table 3, p < . 134); however, the interaction of age by
task difficulty was significant at the increasing levels of difficulty, even
though the speed had been equated for initial response time (Table 3,
p < .001, p< .001 respectively). In addition, young subjects made signifi
cantly more errors of commission on the Vowel Cancellation task than did
old subjects (p < .001).
30
31
TABLE 1
Task Means
CONDITION
WAIS
Time in
seconds
Error
STROOP
Time in seconds:
1 2-1 3-2
Error
YOUNG 116 0.9 35 24 37 1.9
MALE
OLD 176 0.7 44 35 68 2.5
AUD
YOUNG 127 0.5 41 24 49 4.3
FEMALE
OLD 172 0.4 44 36 76 4.1
YOUNG 142 1.1 45 27 46 3.9
MALE
NO
OLD 171 0.5 42 41 102 3.3
AUD
YOUNG 132 0.0 39 28 47 5.2
FEMALE
OLD 171 0.2 39 36 71 2.8
CONDITION
CREATIVITY
# #
correct wrong
VOWEL CANCELLATION
Time # Errors
(sec.) ommission commission
YOUNG 4.4 2.7 184 5.4 0.6
MALE
OLD 3.3 3.8 250 11.8 0.7
AUD
YOUNG 5.3 3.7 201 8.2 1.6
FEMALE
OLD 6.4 5.2 256 14.1 0.3
YOUNG 7.1 1.4 204 7.0 2.0
MALE
NO
OLD 5.8 4.5 254 9.5 0.5
AUD
YOUNG 5.9 3.4 195 11.1 1.7
FEMALE
OLD 5.0 5.1 232 10.6 0.7
TABLE 2
MULTIVARIATE ANALYSIS OF VARIANCE SUMMARY TABLE
32
Source of
Variation:
F
df :
hyp
df :
e rr
p less
than
Digit Symbol Substitution
Condition: .364 2 56 .697
Sex: 2.867 2 56 .065
Age: 17.553 2 56 . 001*:
Age x Condition: 1.593 2 56 .212
Age x Sex: 1.318 2 56 .276
Condition x Sex: .536 2 56 .588
Age x Sex x Condition: 1.142 2 56 .327
Stroop Test
Condition: .504 4 54 .733
Sex: 2.139 4 54 .088
Age: 12.240 4 54 .001*:
Age x Condition: 2.796 4 54 .035*
Age x Sex: 1.922 4 54 .120
Condition x Sex: 3.635 4 54 .011*
Age x Sex x Condition:
Creativity Test
Condition: .737 2 56 .483
Sex: 2.010 2 56 .143
Age: 1.426 2 56 .249
Age x Condition: .303 2 56 .739
Age x Sex: .814 2 56 .448
Condition x Sex: 2.013 2 56 .143
Age x Sex x Condition: .616 2 56 .544
Vowel Cancellation Test
Condition: 1.028 3 55 .387
Sex: .376 3 55 .770
Age: 12.001 3 55 .001*:
Age x Condition: 1.313 3 55 .280
Age x Sex: .835 3 55 .481
Condition x Sex: .863 .466
Age x Sex x Condition: .872 3 55 .461
*p < .05 Multivariate F Test
**p < . 001 Multivariate F Test
33
TABLE 3
Univariate F Tests:
Levels of Significance
Variable
Effect
C S A AxC AxS CxS CxSxA
Digit Symbol Substitution
Reaction Time . 707 .340 .001** .079 .174 .471 .170
E rrors .497 .022 .588 .868 .552 .383 .380
Reaction Time (RT1)
Stroop Test
.710 .140 .134 .004* .201 .003* .060
Level 2 - RT 1 .465 .459 .001** .576 .291 .787 .904
Level 3 - Level 2 .448 .443 .001** .739 .023 .031* .381
E rrors .407 .040 .237 .219 .446 .212 .602
# Correct
Creativity Test
.227 .405 .581 .516 .236 .054 .451
# Wrong . 768 .140 .094 .540 .934 .887 .584
Reaction Time
Vowel Cancellation
.532 .619 .001** .262 .246 .188 .801
# E rrors Omission .700 .463 .075 .071 .225 .956 .981
# E rrors Commission .114 .758 .001** .408 .444 .413 .115
*p < .05 Multivariate F Test
**p< .001 Multivariate F Test
C = Condition: Audience vs No Audience
S = Sex of Subject
A = Age group: Young vs Old
34
Hypothesis I states that the Audience condition will improve perfor
mance on the simple tasks and will impair it on the more complex tasks.
Since there were no significant differences on the simplest and most complex
tasks, this hypothesis was not supported by the data.
The second hypothesis, that the performance of older subjects in the
Audience condition will improve significantly more on the simple tasks and
will be significantly worse on the more difficult tasks than that of the young
subjects was partially supported. As can be seen in Figure 1, older subjects
performed much worse in the Audience condition than in the No Audience con
dition, while younger subjects performed better in the Audience condition than
in the No Audience condition. This result occurred on the initial reaction
time of the Stroop test (See Table 3) and the interaction was significant
(p < . 002). Planned comparisons were performed and it was determined that
the differences between the conditions were significant (p< .01 for both young
and old groups) and, therefore, accounted for the interaction. There were
no significant differences in the Audience conditions on the simpler tasks;
i.e ., the vowel cancellation and DDS; consequently, the hypothesis was only
partially supported.
It was also expected that due to the nature of the tasks chosen, older
women would do significantly better than older men as the tasks increased in
complexity, while there would be no differences in the young subjects. This
was not found. Although the univariate analysis of the sex by age interaction
on the most difficult part of the Stroop test was significant (See Table 3,
p< .046), it cannot be accepted because the overall multivariate analysis was
not significant.
The significant interaction between sex of subject and Audience con
dition, as shown in Figures 2 and 3, was unexpected. This result occurred
on the initial reaction time of the Stroop test (See Table 3, p < . 003), and on
the most difficult level of complexity (p< .031). A post-hoc analysis
(Scheffe, 1959) showed that the males in the No Audience condition performed
35
37-
3 6 -
ot
FIGURE 1
Age x Condition Interaction:
Stroop Test Initial Reaction Time
45
Old
Young
44--
43
42--
41--
40--
39 --
38
\
\
\
\
\
X
X
Audience No Audience
36
FIGURE 2
Sex x Condition Interaction;
Stroop T est Initial Reaction Time
44 --
43
42 --
41 --
40 --
39
\
Males
Females
Audience No Audience
37
80
FIGURE 3
Sex x Condition Interaction:
Stroop Test Level 3 - Level 2
75
70 --
65
60 --
55
50 --
Audience No Audience
Males
Females
38
significantly more poorly than males in the Audience condition on both levels
of task difficulty (p < .05 for both). Females, however, performed signifi
cantly more poorly in the Audience than No Audience condition on the simple
level ft) < . 05), while there were no differences on the more difficult level.
This result is particularly surprising since older subjects, which, of
course, includes the older men, did worse in the Audience condition. How
then do males improve with an audience ? Figure 4 shows the standard e rro r
of the means on the initial reaction time of the Stroop test plotted separately
for each condition. It can be seen that the performance of the older subjects
and the young females was worse in the Audience condition, while that of the
young male subjects was far better in the Audience condition. In addition,
young males performed much worse than any other group in the No Audience
condition.
It was felt that it might be wise to analyze the data of the men on the
Stroop test separately in order to get a clearer picture of the performance of
the males. As can be seen in Table 4, this analysis described the same re
lationship that was found in the previous analysis. In addition, however, it
suggested a relationship between age and condition on the initial reactiom
time of the Stroop test (p < . 005). This relationship was not significant on the
multivariate test ft) < . 095), so no conclusions may be drawn. Nevertheless,
this analysis does aid in interpreting the curious result found above, that
males improved under the Audience condition, while older men and women
were impaired.
39
FIGURE 4
Standard E rror of the Means: Stroop Test
RT 1 Means (p<.05)
50..
48..
46..
44..
42..
40..
38..
36..
34..
32..
0
■ 1 J
Young Young Old Old
Men Women Men Women
' V
\
' V
' V
Young Young Old Old
Men Women Men Women
AUDIENCE NO AUDIENCE
40
TABLE 4
Univariate F Tests: Data for Males
Levels of Significance
Stroop
Effect
Variable
Condition Age C x A
Reaction Time (RT 1) .049* .089 .005°
Level 2 - RT 1 .485 .025** .941
Level 3 - Level 2 .080 .001** .462
E rrors .159 .809 .468
*p <. 05 Multivariate F Test
**p <.001 Multivariate F Test
°p <. 095 Multivariate F Test
CHAPTER V
DISCUSSION
General Findings
One of the most surprising results of this study is the finding that the
A .E. affected performance on only one task. Results of previous literature
and the consequent theoretical formulation would lead one to expect to find
the effect on a wide range of tasks. There was agreement as to the order of
complexity of the tasks by both the subjects and a panel of judges. Unexpect
edly, performance on both the simplest and most difficult of these tasks was
unaffected by the A. E. manipulation, the exception being the middle-ranged
task, the Stroop test, where the A.E. phenomenon was found.
This particular manipulation of the A .E ., a one-way vision room with
the condition manipulated by instructions, has not previously been reported
as a manipulation in the A .E. literature. However, the use of a tape re
corder and film with the condition manipulated by instructions has been re
ported as successful (Henchy and Glass, 1968). It was felt that a one-way
vision room should be an even more effective manipulation.
Since the Stroop test was previously found to be sensitive to an A. E.
(Gribbin, 1973) and since it was also sensitive to it in this study, it was felt
that the A .E. manipulation was effective. Why in this particular battery only
the middle ranged task was affected cannot be answered at this tim e. Never
theless, some tentative explanations may be put forth. Since the Stroop test
was the only task that required the subject to vocalize his response, it may
be that the subject did not really feel in the other tasks that he was being
observed and evaluated to an extent significant enough to establish a perfor
mance difference.
41
42
Another explanation is that the "creativity task" was so difficult that
subjects more or less withdrew from the task. In other words, the internal
anxiety or arousal produced by the task was so great that subjects were un
affected by the addition of an external manipulation. Subjects could be said
to have been "over-aroused" by the difficulty of the task, and any additional
arousal produced no noticeable effects.
The explanation for the opposite case, the vowel cancellation task and
the DSS, is difficult to explain in this framework, however. Since the task
was simple, it is doubtful that subjects felt any specific task anxiety. If
there is no arousal being produced by the task, an external manipulation de
signed to increase arousal should improve performance. It is possible that
subjects were so bored by the task that they felt no involvement in it and even
with increased arousal were not motivated to improve performance on the
task—perhaps the increased arousal of the audience manipulation was di
rected elsewhere. Since no other criteria were measured, however, it is
impossible to support any such interpretation.
It is particularly difficult to explain the lack of findings on the vowel
cancellation task because it had previously been found to be affected by A. E.
(Allport, 1924). However, in that case the design was a coaction paradigm,
where other people were performing the same task at the same time. Con
sequently, it may take this particular circumstance in order to find the effect
on a specific task. However, if this is the case, it definitely weakens the
present theoretical formulation of Zajonc (1966).
Such disparate findings would lead one to believe that the effect of the
presence of others on performance may be very task specific, or at least
limited to a certain range of tasks. In addition, the type of "audience" may
be a very important factor, not to mention the effect of the initial level of
arousal of the subject. These are aspects which the present theoretical
formulation seems unable to handle. It may be that the "arousal" explanation
43
of A .E. is an unsuitable approach to explain the literature on A .E. On the
other hand, "arousal" differences may be a very efficient explanation for
handling a certain range of tasks; however, it must be accepted that it is valid
only within that limitation. Alternatively, "arousal" may be a valid explana
tion of performance for the entire range of tasks, but the specific effect of
A .E. is confounded by the arousal contributed by various other sources. For
example, other factors such as the anxiety induced by the task, or the anxiety
induced by the experimental environment, or individual differences in initial
level of arousal, confound the findings and make it difficult to parcel out any
arousal due specifically to the A .E. manipulation. Such a discussion of the
theoretical explanation of A .E . is beyond the scope of the present investiga
tion. It is obvious that thorough thought and testing of specific problem areas
is necessary before any conclusions can be drawn. The interaction between
task anxiety and A .E. is certainly a major area for further investigation.
Until the contribution of these various factors is parceled out, use of A .E.
for the purpose of investigating arousal is futile.
Since the Stroop test has repeatedly been shown to be a task in which
performance changes occur in the A. E. paradigm, the present study inter
prets the significant findings on this task as being the result of an increase
in arousal brought on by the instructions that experts were observing and
evaluating performance. This is justified because the arousal manipulation
was effective in changing performance on the task (See Chapter II). Ad
mittedly, interpretation of the findings of the entire study is severely ham
pered by the failure to find differences on the other tasks.
On the Stroop test, older subjects performed much worse in the
Audience condition than in the No Audience condition, while the reverse was
true for young subjects. Interpreting this finding as an inverted-U relation
ship, it would seem that the older subjects were at a more optimum level of
arousal without an audience because their performance was sim ilar to that of
44
the young subjects under that condition, and superior to older subjects in the
Audience condition. However, the addition of an audience improved the per
formance of the young subjects suggesting that their arousal level was less
than optimum under the No Audience condition. Since there were no signifi
cant differences as the task increased in complexity, it appears that any
additional increase in arousal brought on by task anxiety was not sufficient to
change this initial difference.
It was surprising to find that the Audience condition improved the per
formance of males, while it impaired that of females. (See Gribbin (1973)
where performance of females was worse on the Stroop test in the Audience
condition.) Females seemed to be performing at a more optimum level of
arousal initially because the increase in arousal impaired their performance.
This difference occurred on both the initial reaction time of the Stroop and
on the most difficult level of the task. Since the most difficult level of the
Stroop had been equated for performance on the simpler task, it is possible
that this is a case where the anxiety induced by the task had an interaction
effect with the anxiety induced by the presence of others. This level of diffi
culty put the performance of women in both conditions at a more sim ilar
level, while the performance of the males in the presence of an audience
still continued to improve.
It cannot be overlooked that the only group whose performance obvi
ously improved in the audience condition, as evidenced by the standard e rro r
of the means, was the young males. That they performed more poorly than
any other group, including the older groups, in the No Audience condition is
not as clear cut because the standard e rro r of the means show considerable
overlap with the other groups. It does seem safe, however, to conclude
that the young males were the major contributors to the significant differ
ences in performance between the two conditions.
45
Such findings would seem to suggest that males, at least young ones,
have a greater variability in the amount of arousal that they are able to with
stand before it begins to impair performance than do females; however,
since we have no measure of the initial levels of arousal, such an interpre
tation is tenuous at best.
These sex differences in performance are, nevertheless, very in
teresting and m erit further investigation in order to parcel out the effect of
some of these interacting factors. Why did only young males improve in
performance ? Much of the previous literature which found improvement on
performance was describing male subjects. Is this the explanation? This
study employed a young female experimenter. What effect did this have on
the differential performance of the subjects ? Sex differences in performance
were not the basis for the theoretical explanation of A .E .; however, this
study suggests that they may play an extremely important part in explaining
some of the literature, and cannot be overlooked in subsequent research.
It is somewhat difficult to interpret the findings of this study in term s
of Eisdorfer's (1968) work for several reasons. His formulation was based
on research with male subjects; this study included females. Another diffi
culty encountered when strictly comparing this work with Eisdorfer's is
that he attempted to lower arousal (by drug injection) and interpreted initial
arousal level from this basis, while this study attempted to raise arousal
level (by a social psychological manipulation), and to interpret initial arousal
level in this manner. This research adds additional support to his findings,
however, in that the performance of older subjects improved in the less
arousing situation suggesting that they were, indeed, over-aroused as he
postulates. This was particularly true for the older females. The results
of the performance of the young female subjects do not fit into his formula
tion. Certainly the unexpected findings of the female subjects establishes a
need for further study in this area.
46
Although it was expected that older women would do better than older
men on the DSS and the Stroop test, since previous researchers had found
sex differences on these tasks, this was not the case. The lack of significant
findings may be due to the highly select sample of older subjects which was
used in this study. Alternatively, since the older men were, on average,
younger than the older women, these results may have reflected that differ
ence even though the data were covaried to eliminate this effect. It is im
possible to draw any firm conclusions at this time.
It is interesting to note that although the age difference in speed of re
action time was significant on the DSS and the vowel cancellation task, which
would be expected from a vast amount of previous research, it was not found
on the initial time of the Stroop test. Since so many of the interaction effects
on the Stroop test were significant, the lack of findings on this part of the
task may be a spurious result confounded by the interaction effects of the
other factors. For example, the poor performance of the young males in the
No Audience condition probably contributed to this lack of an expected finding.
It may be, however, that this particular type of reaction time, where sub
jects are asked to read out loud as quickly as possible, may be an entirely
different type of speeded reaction time and may not be affected as other
speeded motor responses. Since this was a highly educated sample of older
people, reading and verbalizations would presumably be well learned and
well practiced responses and may not show as great age differences. The
data from this study is not designed to test such a conclusion, however, and
it remains a question.
A finding which has been demonstrated in numerous other studies was
again found in this study. Younger subjects made more errors of commis
sion on the vowel cancellation task. It may be that young subjects are re
sponding so quickly on the speeded part of the task that they are "response
prone" and tend to need less certainty before responding, consequently
47
making more e rro rs. Another way of looking at the same phenomenon is that
older people seem to need more certainty before responding and are con
sequently less quick in their responses. These differences in the number of
e rro rs were not found on the other tasks. Whether younger subjects found
the vowel cancellation task so simple that they hurried through it as fast as
possible and tended to make more e rro rs, while the other tasks were difficult
enough that they slowed down and tended to decrease the number of e rro rs, or
whether older subjects tended to increase their number of erro rs relative to
the type of task cannot really be ascertained since the task requirements are
different. Regardless of the number of erro rs made, however, younger
subjects are faster than older subjects, a finding which has, in general, be
come an established fact of aging.
Post-Experiment Questioning
Often some of the most interesting questions arise when discussing the
experiment with subjects after they have completed the tasks. Subjects in
the Audience condition were asked how they felt about the fact that someone
was observing them while they performed the task. Most young subjects
said that they were aware throughout the course of the experiment that there
was someone on the other side of the m irror. A few said that they did not
see how the observer could observe much of their performance since many
of the tasks required only written responses. These subjects felt that their
overt behavior was probably going to be correlated with their written re
sponses. Young subjects seemed, in general, to persist in their feeling
that those observing them were the critical elements, not that the experi
menter was also an important factor. This was in direct contrast to the
discussions of the older men. These subjects also were convinced that there
were observers in the other room; however, they said that it did not bother
them because they could not see them. The factor that they were aware of
and which they felt influenced their performance was that a young girl was
48
the experimenter and was observing their performance. Consequently, in
their subjective viewpoint, the experimenter was much more critical than
the "unknowns" on the other side of the m irror, and definitely influenced
their motivation to do well. The statistical analysis does not support their
contention because there was no overall interaction between age, condition,
and sex of subject. Since this particular variable was not manipulated, how
ever, no conclusive evidence can be put forth. Older women agreed that
they knew someone else was observing them, but that it did not really affect
their performance. They did not mention that the experimenter might have
an effect.
The idea that the experimenter might have an effect on the performance
of the subject is, of course, a very realistic notion; however, this particular
effect was not an issue in the current study. It is interesting to ponder what
the effects on performance would be if there were an older man or older
women conducting the experiment; however, that will have to remain for
future investigators to implement.
Subjects were also asked to rank the tasks by order of difficulty.
Their agreed upon order was discussed above.
There were four tasks used in this study. Only the "creativity" task
seemed to have a differential effect on the different age groups. Most older
subjects seemed to take especial concern to excuse what they felt was poor
performance. Some attributed it to old age; others said they never were
good at word games, etc. Young subjects also expressed concern with the
difficult of the task, but did not discuss it in these term s. Most young sub
jects seemed competitive in wanting to know how well other people performed
on the task. Two older subjects, one man and one woman, had to be en
couraged to continue trying to work the task. Each felt that it was more
than he or she wanted to handle at the moment. This did not occur with the
young subjects. It is interesting to note the subjective differences between
49
the age groups on this task since objective age differences were not statis
tically significant.
Subjects were all debriefed and told that there were really no observers
(even so, to this day, one older woman is convinced that there actually were
observers in the other room). The nature of the study was then discussed.
As would be expected, old subjects tended to spend a great deal of time in
questioning and in suggesting other ways the experiment could also be run.
Most seemed very interested in the implications for such a study. Young
subjects were generally interested in the hypotheses, and in what trend the
results seemed to suggest. Debriefing them took very little time in com
parison to the older subjects.
Implications for Future Research
Many of the questions which were raised above are certainly worthy of
further investigation. There seems little doubt that the present theoretical
formulation of the A .E. is too simplistic to handle many of the findings of
the present investigation. The entire area of the effect of the presence of
others on performance is much more complex than first imagined. Does the
phenomenon exist only on certain tasks or is it a result of an interaction with
the complexity of the task or some other factor ?
How does the effect of the coaction paradigm differ from that of the
audience paradigm ? Are some tasks more susceptible to one paradigm than
to another ? Older people were found in this study to perform differently
from young people. Would they also be so affected in the coaction paradigm ?
Certainly a number of questions arose in the present investigation which are
unanswerable at the present time.
Another area of interest concerns the nature of the observer. No study
has been conducted where the observers were older people. Would this dif
ferentially affect young and old subjects ? Does the status or some other
characteristic of the older person determine the nature of the effect?
50
The effect that the experimenter has on the nature of the differences
observed has yet to be tested. Do older people respond differently depending
upon who is requiring the performance ? What about the sex of the experi
menter ? Does this play a significant role in determining performance ?
Perhaps one of the most interesting areas which the present investiga
tion brought to light has to do with the nature of sex differences in perfor
mance. The present study seems to offer evidence that women were at a
higher level of initial or baseline arousal (at least on the Stroop test as in
terpreted by the inverted-U explanation). It also appeared that older people
started out at a higher level of arousal. Future investigations of the A.E.
should attempt to establish what the initial level of arousal is before attempt
ing to manipulate it, and then inferring what the initial level was.
The important question raised by this investigation concerns the utility
of using the A .E. paradigm when investigating the performance of older peo
ple. Since many different social situations include the presence of other
people, A .E. has great potential importance. In situations where older
people are being retrained; e .g ., stroke victims, knowledge of the effect of
the presence of others may greatly increase the effectiveness of such pro
gram s.
The utility of using the A .E. paradigm when studying the effect of
arousal and age differences in performance, however, is questionable.
Until further research is conducted which establishes that A .E . does, in
deed, bring about a change in arousal levels, and until some of the other
arousal contributing factors are better understood, it does not seem worth
while to use it as an arousal manipulation for research with old people.
Research on arousal and age is an extensive area worthy of further
research. Data on the level of arousal of older women is sadly lacking, both
on physiological and behavioral bases. Certainly this area requires much
more extensive investigation than it has received to date.
51
One of the most critical questions, which is still unresolved, concerns
the concept of arousal as used by social psychologists in contrast to that
used by physiological psychologists. More research in this area seems
very important in order to gain greater understanding of the behavior not
only of older individuals, or of younger individuals, but of men as opposed
to women. All of these factors seemed to result in performance differences
in this study. Can we establish the physiological differences which would
predict these behavioral differences ?
There is little doubt that thorough thought and extensive research is
necessary before science can begin to answer these questions.
CHAPTER VI
SUMMARY AND CONCLUSIONS
The present study dealt with the influence o f Audience Effect (A.E.) on
the performance of young and old subjects as tasks increased in complexity. The
following hypotheses were tested:
1) The Audience condition will improve performance on simple tasks and
will impair performance on more complex tasks.
2) The performance of older subjects in the Audience condition will be
significantly better on simple tasks and will be significantly worse on complex
tasks compared with young subjects.
It was also expected that due to the nature of the tasks chosen, older
women would do significantly better than older men as the tasks increased in
complexity, while there would be no sex differences in the young subjects.
The tasks varied from the least difficult, a vowel cancellation test, to the
Digit Symbol Substitution from the WAIS, which was more difficult, to the Stroop
Color Naming Test, which was increasingly more complex, and finally to a
“creativity” test, which was extremely difficult. After completing the test
battery, subjects were asked to order the tasks from least to most difficult.
Their subjective impressions on the task difficulty coincided with this ordering.
Performance was measured by reaction time and errors for all tests, except the
“creativity” test, where a time limit of ten minutes was imposed, and the
number correct and the number wrong were measured.
Sixty-six subjects participated in the experiment. O f the thirty-one
older subjects, 12 were man (52-67 years of age) and 19 were women (52-85
years). Of the thirty-five young subjects, 15 were men (19-28 years), and
52
53
20 were women (18-23 years). Half of the subjects were given instructions
that they were being observed while they performed the tasks. The re
maining subjects were assured that no one, except the experimenter, was
observing them.
A multivariate analysis of variance was carried out, which covaried
the deviations of each subject from the mean age of his group. This analysis
revealed: a) only the Stroop test was successful in differentiating perfor
mance under the Audience Effect; b) older subjects were affected differently
by the A .E. than were young subjects; i . e . , older subjects performed worse
with an audience, while younger subjects performed better; c) men per
formed significantly better in the Audience condition on both the initial re
action* time and the most difficult level of complexity on the Stroop test
than did men in the No Audience condition, while women performed better in
the No Audience condition on the initial reaction time, while no difference
was found between the conditions on the most difficult level; d) older sub
jects are slower to begin with than are young subjects and as the task be
comes more complex, they become disproportionately slow; e) young sub
jects made more erro rs of commission on the vowel cancellation task than
did old subjects.
From the results of the literature review and the findings of this ex
periment, the following conclusions were drawn:
1) The present theoretical formulation of the Audience Effect literature
seems inadequate to explain the findings of this study. Such factors as the
individual's initial level of response, the anxiety contributed by being in an
experimental situation, and the anxiety induced by the task may confound the
effect of the presence of others.
, 2) Although the arousal manipulation was effective on only one of the
tasks, the results on this task give tentative support to the explanation that
performance decrements in old age may be the result of over arousal. This
54
finding was particularly supported by the results of the performance of older
women. Eisdorfer, et al. (1970), in a study using older male subjects,
suggests sympathetic nervous system activity is the determinant of this over
arousal state. Since this study had no objective measurement of the physiol
ogical activity and since it used a behavioral manipulation of arousal, it is
impossible to attribute the performance decrements to this specific physiol
ogical state; nevertheless, the results were consistent with such an inter
pretation. In addition, this study extends his interpretation to older women
as well as to older men, although further research measuring the physiol
ogical indicators of older women is certainly warranted.
3) The sex differences in performance between conditions suggests
that males and females may differ in initial levels of arousal, or alterna
tively, in ability to withstand increases in arousal without impairing perfor
mance. Until further investigations are conducted, preferably using physiol
ogical indicators, it would be prem ature to make a definitive statement re
garding sex differences.
It was also discussed that within certain limitations, the Audience
Effect paradigm has potential importance for use with older people. It may
be highly practical to study the phenomenon for specific purposes, such as
retraining. Its utility in studying arousal and age differences in performance,
however, does not seem worthwhile at this time.
BIBLIOGRAPHY
Allee, W. C ., Park, O ., Emerson, A. E ., Park, T ., Schmidt, K. P.
Principles of animal ecology. Philadelphia: Saunders, 1950.
Anastasi, A. Individual differences. New York: John Wiley and Son,
1967.
Arenberg, D. Anticipation interval and age differences in verbal learn
ing. Journal of Abnormal Psychology, 1965, 70, 419-425.
Bettner, L. G ., Jarvik, L. F ., and Blum, J . E. Stroop color-word test,
non-psychotic organic brain syndrome, and chromosome loss in aged
twin. Journal of Gerontology, 1971, 26(4), 458-469.
Bicknell, A. T. Aging, arousal and vigilance. Unpublished doctoral
dissertation, Texas Tech University, 1970.
Birch, J. D ., and Veroff, J. Motivation: a study of action. Belmont,
California: Wadsworth, 1966.
Birren, J. E. Age changes in speed of simple responses and perception
and their significance for complex behavior. In Old age in the modern
world. London: E. and S. Livingstone, 1955.
Birren, J. E ., and Botwinick, J. Rate of addition as a function of diffi
culty and age. Psychometrika, 1951, 16, 219-232.
Birren, J. E ., and Botwinick, J. Speed as a function of perceptual
difficulty and age. Journal of Gerontology, 1955, 19, 433-436.
Blum, J. E ., Fosshage, J. L ., and Jarvik, L. F. Intellectual changes
and sex differences in octogenarians: a twenty year longitudinal
study of aging. Developmental Psychology, 1972, j[> 178-187.
Botwinick, J. Cognitive processes in maturity and old age. New York:
Springer Publishing Company, In c., 1967.
Botwinick, J . , and Kornetsky, C. Age differences in the acquisition and
extinction of the GSR. Journal of Gerontology, 1960, 15, 83-84.
55
56
Broen, W. E ., J r ., and Storm, L. H. A reaction potential ceiling and
response decrements in complex situations. Psychological Review,
1961, 68, 405-415.
Bromley, D. B. Some effects of age on short-term learning and rem em
bering. Journal of Gerontology, 1958, 13, 398-406.
Canestrari, R. E ., J r. Paced and self-paced learning in young and
elderly adults. Journal of Gerontology, 1963, 18, 165-168.
Cottrell, N. B. Performance in the presence of other human beings:
Mere presence, audience, and affiliation effects. In E. C. Simmel,
R. A. Hoppe, and G. A. Milton (Eds.), Social facilitation and imi
tative behavior. Boston: Allyn and Bacon, 1968, 91-110.
Cottrell, N. B. Social facilitation. In C. G. McClintock (Ed.), Experi
mental social psychology. New York: Holt, Rinehart and Winston,
1972, pp. 185-236.
Cottrell, N. B ., Rittle, R. H ., and Wack, D. L. The presence of an
audience and list type (competitional or noncompetitional) as joint
determinants of performance in paired-associates learning*
Journal of Personality, 1967, 35, 425-434.
Cottrell, N. B ., Wack, D. L ., Sekerak, G. J . , and Rittle, R. H. Social
facilitation of dominant responses by the presence of an audience and
the mere presence of others. Journal of Personality and Social Psy
chology, 1968, 9, 245-250.
Doppelt, J. E ., and Wallace, W. L. Standardization of the Wechsler
Adult Intelligence Scale for older persons. Journal of Abnormal and
Social Psychology, 1955, 51, 312-330.
Eisdorfer, C. Arousal and performance: Experiments in verbal learning
and a tentative theory. In G. A. Talland (Ed.), Human aging and
behavior. New York: Academic P ress, 1968, 189-216.
Eisdorfer, C ., Axelrod, S ., and Wilkie, F. L. Stimulus exposure time
as a factor in serial learning in an aged sample. Journal of Abnormal
and Social Psychology, 1963, 67, 594-600.
Eisdorfer, C ., Nowlin, J . , and Wilkie F. Improvement of learning in the
aged by modification of autonomic nervous system activity. Science,
1970, 170, 1327-1329.
57
Feingold, L. A psychometric study of senescent twins. Unpublished
doctoral dissertation, Columbia University, 1950.
Frankel, A. S. Effects of arousal on hierarchically organized responses.
Journal of Experimental Psychology, 1969, 82(12), 385-389.
Ganzler, H. Motivation as a factor in the psyhological deficit of aging.
Journal of Gerontology, 1964, 19, 425-429.
Garai, J. E ., and Scheinfeld, A. Sex differences in mental traits.
Genetic Psychology Monographs, May, 1968.
Gilbert, Jeanne G. Memory loss in senescence. Journal of Abnormal and
Social Psychology, 1941, 36, 73-86.
Gribbin, K. J. Age differences in arousal and the audience effect. Un
published m aster's thesis, University of Southern California, 1973.
Grossman, S. P. A textbook of physiological psychology. New York:
John Wiley and Sons, Inc., 1967.
Henchy, T ., and Glass, D. C. Evaluation apprehension and the social
facilitation of dominant responses. Journal of Personality and Social
Psychology, 1968, 10, 446-454.
H err, J. J. Differential effects of epinephrine and propranolol on shuttle
box avoidance learning in rats of different ages. Unpublished doc
toral dissertation, University of Southern California, 1973.
Jarvik, L. F ., and Blum, J. E. Cognitive declines as predictors of
mortality in discordant twin pairs: A twenty-year longitudinal study
of aging. In E. Palmore and F. C. Jeffers (Eds.), Prediction of
life span. Lexington, M ass.: D. C. Heath, 1971.
Jeffrey, D. W. The effect of dextroamphetamine on arousal and on per
formance in young and old men. Paper presented at the twenty-fourth
annual meeting of the Gerontological Society, Houston, October, 1971.
Jones, H. E ., and Conrad, H. S. The growth and decline of intelligence:
A study of a homogeneous group between the ages of 10 and 60.
Genetic Psychology Monographs, 1933, 13, 223-298.
58
Jones, B. E ., and Gerard, H. B, Foundations of social psychology. New
York: Wiley, 1967.
Kay, H. The effects of position in a display upon problem solving.
Quarterly Journal of Experimental Psychology, 1954, 6,155-169.
Kay, H. Some experiments on adult learning. In J. E. Birren (Ed.),
Old age in the modern world. London: E. and S. Livingstone, 1955.
Klein, R. L. Age, sex, and task difficulty as predictors of social con
formity. Journal of Gerontology, 1972, 27(2), 229-236.
Lacey, B. C ., and Lacey, J. I. Cardiovascular and respiratory corre
lates of reaction time. Appended preprint No. 3, Progress Report,
Grant MH-00623, June, 1965.
Lacey, J . I. Somatic response patterning and stress: Some revisions of
activation theory. In M. H. Appley andR. Trumbull (Eds.), Psycho
logical stress: Issues and Research. New York: Appleton Century
Crofts, 1967, pp. 14-42.
Lacey, J. I ., Kagan, J . , Lacey, B. C ., and Moss, H. A. The visceral
level: situational determinants and behavioral correlates of auto
nomic response patterns. In P. J. Knapp (Ed.), Expression of the
emotions in man. New York: International Universities P ress,
1963, pp. 161-196.
Maccoby, E. Sex differences in intellectual functioning. In E. Maccoby
(Eds.), The development of sex differences. Stanford California:
Stanford University P ress, 1966, pp. 25-55.
Malmo, R. B. Measurement of drive: an unsolved problem in psychology.
In M. R. Jones (Ed.), Nebraska symposium on motivation. Lincoln:
University of Nebraska P ress, 1958, pp. 220-265.
Martens, R ., and Landers, D. M. Evaluation potential as a determinant
of coaction effects. Journal of Experimental Social Psychology, 1972,
8(4), 347-359.
Matlin, Margaret W ., and Zajonc, R. B. Social facilitation of word
associations. Journal of Personality and Social Psychology, 1968,
10(4), 455-460.
59
McNemar, Q. The revision of the Stanford-Binet Scale; An analysis of
the standardization data. Boston: Houghton-Mifflin, 1942.
Mellone, M. A. A factorial study of picture tests for young children.
British Journal of Psychology, Sept.-May, 1944-1945, 35, 9-16.
M orris, J. D ., and Thompson, L. W. Heart rate changes in a reaction
time experiment with young and aged subjects. Journal of Geron
tology, 1969, 24, 269-275.
Nehrke, M. F. Age, sex, and educational differences in syllogistic
reasoning. Journal of Gerontology, 1972, 27(4), 446-470.
Obrist, W. D ., and Busse, E. W. The electroencephalogram in old age.
In W. P. Wilson (Ed.), Applications of electroencephalography in
psychiatry: A symposium. Durham: Dulse University P ress, 1965,
pp. 158-185.
Paulus, P. B ., and Murdock, P. Anticipated evaluation and audience
presence in the enhancement of dominant responses. Journal of
Experimental Social Psychology, 1971, 7(3), 280-291.
Powell, A. H ., Eisdorfer, C ., and Bogdonoff, M. D. Physiologic
response patterns observed in a learning task. Archives of General
Psychiatry, 1964, 10, 192-195.
Babbitt, P.M .A . Perceptual discrimination and choice of response. In
J. E. Birren, The psychology of aging. Englewood Cliffs, New
Jersey: Prentiee-Hall, 1964.
Scheffe, H. The analysis of variance. New York: Wiley, 1959.
Schmavonian, B. M ., and Busse, E. W. Psychophysiological techniques
in the study of the aged. In R. H. Williams, C. Tibbitts, and W.
Donahue, (Eds.), Processes of aging. New York: Atherton P ress,
1963, pp. 160-183.
Schmavonian, B. M ., Miller, L. H ., and Cohen, S. I. Differences among
age and sex groups with respect to cardiovascular conditioning and re
activity. Journal of Gerontology, 1970, 25(2), 87-94.
60
Schmavonian, B. M ., Yarmat, A. J ., and Cohen, S. I. Relationships
between autonomic nervous system and central nervous system in
age differences in behavior. In A. T. Welford and J. E. Birren
(Eds.), Behavior, aging and the nervous system. Springfield,
Illinois: Charles C. Thomas, 1965, pp. 235-258.
Shapiro, D ., and Crider, A. Psychophysiological approaches in social
psychology. In G. Lindzey and E. Aronson (Eds.), The handbook of
social psychology, Volume three. Menlo Park, California:
Addison-Wesley Publishing Company, 1968, pp. 1-49.
Silverman, A. J . , Cohen. S. I., and Schmavonian, B. M. Investigation
of psychophysiologic relationships with skin resistance m easures.
Journal of Psychosomatic Research, 1959, 4, 65-87.
Spence, K. W. Behavior theory and conditioning. New Haven: Yale
University P ress, 1956.
Strain, G. S. Social facilitation of oddity preference by children. Paper
presented at the meeting of the Midwestern Psychological Association,
Chicago, May, 1967.
Surwillo, W. W. The relation of autonomic activity to age differences in
vigilance. Journal of Gerontology, 1966, _21, 257-260.
Thompson, L. W ., and Marsh, G. Psychophysiological studies of aging.
In C. Eisdorfer and M. P. Lawton (Eds.), The Psychology of Adult
Development and Aging. Washington, D. C .: American Psychological
Association, 1973, pp. 112-148.
Triplett, N. The dynomogenic factors in pacemaking and competition.
American Journal of Psychology, 1898, 9, 507-533.
Young, Marguerite L. Age and sex differences in problem solving.
Journal of Gerontology, 1971, 26(3), 330-336.
Zajonc, R. B. Social facilitation. Science, 1965, 149, 269-274.
Zajonc, R. B. Social psychology: An experimental approach. Belmont,
California: Wadsworth, 1966.
61
Zajonc, R. B ., and Niewenhuyse, B. Relationship between word fre
quency and recognition: Perceptal process or response bias ?
Journal of Experimental Psychology, 1964, 67, 276-285.
Zajonc, R. B ., and Sales, S. M. Social facilitation of dominant and
subordinate responses. Journal of Experimental and Social Psychol
ogy, 1966, 2, 160-169.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Changes In Memory As A Function Of Age

PDF

Psychomotor Performance And Change In Cardiac Rate In Subjects Behaviorally Predisposed To Coronary Heart Disease

PDF

Age, Sex, And Task Difficulty As Predictors Of Social Conformity: A Search For General Tendencies Of Conformity Behavior

PDF

Differential effects of epinephrine and propranolol on shuttle box avoidance learning in rats of different ages

PDF

Age Differences In Primary And Secondary Memory Processes

PDF

Influences Of Interoperative Experience And Age On Recovery Of Visual Function Following Two-Stage Lesions Of The Striate Cortex

PDF

Age differences in arousal and the audience effect

PDF

Some Behavioral Consequences Of Career Success: A Synthesis Of Reward Andbalance Approaches

PDF

Biofeedback Control Of The Eeg Alpha Rhythm And Its Effect On Reaction Time In The Young And Old

PDF

Dying And Death Role-Expectation: A Comparative Analysis

PDF

Age Differences In Serial Reaction Time As A Function Of Stimulus Complexity Under Conditions Of Noise And Muscular Tension

PDF

Electrophysiological Properties Of Reaction Time And Aging

PDF

Signal Processing Time And Aging: Age Differences In Backward Dichoptic Masking

PDF

Dissonance and Self-Perception Analysis of "Forced Compliance": When Two Theories Make Competing Predictions

PDF

Reticulocyte Protein Synthesis By Heterologous Tissue Components: Implications For The Programming Of Development And Aging

PDF

Verbal And Motor Mediation In Normal Children Of Two Ages And Retarded Individuals

PDF

The Effects Of Justice, Balance, And Hostility On Mirth

PDF

Performance Evaluation Based On Multidimensional Job Behaviors

PDF

Sex Differences In The Organization Of Spatial Abilities In Older Men And Women

PDF

City Manager, City Council Role Consensus And Its Effect On Municipal Performance

Asset Metadata

Creator Gribbin, Kathy Jo (author)

Core Title Audience Effect And Age Differences On Tasks Of Increasing Complexity

Degree Doctor of Philosophy

Degree Program Philosophy

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag OAI-PMH Harvest,psychology, general

Language English

Contributor Digitized by ProQuest (provenance)

Advisor Birren, James E. (committee chair), Bengtson, Vern L. (committee member), Walker, James Paul (committee member)

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c18-685026

Unique identifier UC11356667

Identifier 7417345.pdf (filename),usctheses-c18-685026 (legacy record id)

Legacy Identifier 7417345.pdf

Dmrecord 685026

Document Type Dissertation

Rights Gribbin, Kathy Jo

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the au...

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus, Los Angeles, California 90089, USA