For better or worse?

MISSING IMAGE

Material Information

Title:
For better or worse? Social relationships and physiological reactivity
Alternate title:
Social relationships and physiological reactivity
Physical Description:
ix, 117 leaves : ill. ; 29 cm.
Language:
English
Creator:
York, David James, 1963-
Publication Date:

Subjects

Subjects / Keywords:
Social Facilitation -- psychology   ( mesh )
Physiology   ( mesh )
Interpersonal Relations -- psychology   ( mesh )
Startle Reaction -- psychology   ( mesh )
Startle Reaction -- physiology   ( mesh )
Genre:
bibliography   ( marcgt )
non-fiction   ( marcgt )

Notes

Thesis:
Thesis (Ph.D.)--University of Florida, 1993.
Bibliography:
Bibliography: leaves 75-80.
Statement of Responsibility:
by David James York.
General Note:
Typescript.
General Note:
Vita.

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
oclc - 79179967
ocm79179967
System ID:
AA00011176:00001

Full Text














FOR BETTER OR WORSE? SOCIAL RELATIONSHIPS AND
PHYSIOLOGICAL REACTIVITY













BY
DAVID JAMES YORK



















A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF
THE UNIVERSITY OF FLORIDA
IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR
THE DEGREE OF DOCTOR OF PHILOSOPHY
UNIVERSITY OF FLORIDA
1993









ACKNOWLEDGMENTS

I thank Anthony Greene, my advisor. He has contributed

significantly to my development as a scientist-practitioner

by collaborating on projects of mutual interest, providing

many material resources with which to conduct this project,

and guiding me with optimum levels of constructive

criticism.

I also thank Peter Lang, Margaret Bradley, and Bruce

Cuthbert for their extensive contributions to my development

as a psychophysiological researcher. Despite my shift in

professional orientation, I continue to respect and admire

their rigorous scientific expertise.

I additionally thank Hugh Davis, Keith Berg, Rus Bauer,

and James Algina, all of whom took the time to become

involved in this project, and with whom I have been

especially fortunate to have studied with and learned from.

I thank Samel Celebi for his tutorial in PASCAL and

generous guidance with his hardware troubleshooting skills,

Christine Freeman for performing above and beyond the call

of duty as my research assistant, and Renee Melancon for her

amazing job soliciting and scheduling subjects for this

project.

Finally, I thank my parents, Frank and Susan, who have

provided me with the exceptional support that has allowed me

to achieve aspirations such as this, and my brothers, Greg,

Chris, and Carl for their support and friendship.









TABLE OF CONTENTS


Page

ACKNOWLEDGMENTS........................................ ii

LIST OF TABLES .................................... vii

LIST OF FIGURES............................... .... viii

ABSTRACT.................................... .......... ix

CHAPTERS

1 INTRODUCTION................................... 1

Theories of Social Facilitation................ 2

The Mere Presence Hypothesis............... 2

Non-Drive Theory Alternatives............ 5

Drive Theory Alternatives................. 6

Learning Modulates Social Presence Effects...... 12

The Psychophysiology of Social Facilitation..... 15

Arousal is Multifactorial................. 16

Early Psychophysiological Studies........... 18

The Reactivity Hypothesis.................. 19

Testing the Reactivity Hypothesis........... 22

The Startle Blink as a Measure of

Motivational State......................... 24

Experimental Problem............................ 27

Experimental Hypotheses........................ 28

Preexisting Anxiety Levels................. 28

Verbal Report of Emotion, Safety & Trust... 28

Baseline Levels of Physiological Activity.. 29

Phasic Heart Rate Responding............... 29


iii









Phasic Skin Conductance Responding......... 29

Blink Responding to Startle Probes.......... 30

Anagram Performance........................ 30

The Interaction of Preexisting Anxiety

Levels and Social Presence

Conditions ......... ................. 31

2 METHODS.......................................... 32

Subjects................................. ...... 32

Apparatus........................................ 32

Heart Rate ................................. 32

Skin Conductance........................... 32

Startle Blink............................ 33

Stimuli ....................................... 33

Procedure......................................... 34

Data Reduction and Analysis...................... 38

Preexisting Anxiety Levels................. 38

Verbal Report of Emotion, Safety and Trust. 38

Baseline Levels of Physiological Activity.. 39

Phasic Heart Rate Responding............... 40

Phasic Skin Conductance Responding......... 40

Blink Responding to Startle Probes......... 42

Anagram Performance........................ 43

The Interaction of Preexisting Anxiety

Levels and Social Presence Conditions. 44

Manipulation Check.......................... 45









3 RESULTS...... ................. .............. 46

Preexisting Anxiety Levels...................... 46

Verbal Report of Emotion, Safety, and Trust..... 46

Baseline Levels of Physiological Activity....... 48

Phasic Heart Rate Responding.................... 48

Overall Responding Across Time.............. 48

Differences Among Conditions............... 50

Phasic Skin Conductance Responding............... 52

Overall Responding Across Time............ 52

Differences Among Conditions............... 54

Blink Responding to Startle Probes............... 56

Anagram Performance.............. ............... 57

The Interaction of Preexisting Anxiety Levels

and Social Presence Conditions............. 59

Manipulation Check.............................. 60

4 DISCUSSION............................................ 61

The Presence of a Friend Increases

Appetitive Motivation...................... 61

The Company of Strangers is Not

Always Aversive........................... 66

A Revised Reactivity Model....................... 68

Implications for the Mere Presence and

Distraction-Conflict Models................. 69

Implications for the Social Support Construct... 71

Physiological Variables as Indices

of Motivation............................ 72









Summary and Conclusion.......................... 72

REFERENCES.... ...................................... 75

APPENDICES

A ANOVA & ANCOVA TABLES FOR VERBAL RATINGS......... 81

B ANOVA TABLES FOR BASELINE PHYSIOLOGICAL

ACTIVITY ................................... 84

C ANCOVA TABLES FOR PHASIC HEART RATE RESPONDING.. 87

D ANCOVA TABLES FOR PHASIC SKIN CONDUCTANCE

RESPONDING.. .............................. 91

E ANCOVA TABLES FOR BLINK RESPONDING.............. 101

F ANAGRAM SOLUTIONS, MEAN OVERALL ACCURACY,

AND RANKINGS ............................... 102

G ANOVA TABLES FOR ANAGRAM ACCURACY .............. 103

H PROCEDURE FOR TESTING THE INTERACTION OF

PREEXISTING ANXIETY LEVELS & SOCIAL

PRESENCE CONDITIONS.......................... 105

I INTERVALS BETWEEN ORIENTING & STARTLE STIMULI... 111

J INSTRUCTIONS AND EXPERIMENTAL PROTOCOL.......... 112

BIOGRAPHICAL SKETCH.................................. 117









LIST OF TABLES


Page



TABLE 1 Mean Preexisting Anxiety Levels by
Condition ............................. 47

TABLE 2 Correlations Between Anxiety Levels
and Verbal Ratings.................... 47

TABLE 3 Mean Verbal Ratings by Condition............ 47

TABLE 4 Skin Conductance Response Frequency
and Amplitude to First Orienting
Tone by Condition..................... 55

TABLE 5 Correlations Between Anxiety and
Physiological Reactivity............... 59

TABLE 6 Belief Ratings by Condition................ 60

TABLE 7 Verbal, Physiological, and Performance
Findings.............................. 62


vii









LIST OF FIGURES


Page


FIGURE 1

FIGURE 2


FIGURE 3

FIGURE 4

FIGURE 5


Overall Cardiac Responding Across Time....

Cardiac Responding to Orienting Tones
by Condition..........................

Overall SCR Frequency Across Time.........

Startle Eyeblink Magnitude by Condition...

Performance Accuracy on Anagram Task......


viii















Abstract of Dissertation Presented to the Graduate School of
the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy

FOR BETTER OR WORSE? SOCIAL RELATIONSHIPS AND
PHYSIOLOGICAL REACTIVITY

By

David James York

August, 1993


Chair: Hugh Davis, Ph.D.
Cochair: Anthony F. Greene, Ph.D.
Major Department: Clinical and Health Psychology

Social facilitation describes the enhancement or

impairment of an individual's performance by the simple

presence of other species mates. Recent psychophysiological

studies suggest that previous theories of social

facilitation may represent sufficient but unnecessary

explanations. In light of such findings, investigators have

recently proposed that social facilitation effects are

mediated by increases in physiological reactivity which

occur when the presence of one or more species mates is

perceived as a threat to any valued resource, whether it is

physical or social. This reactivity model also predicts

that physiological responsivity to stressors will be

diminished when the presence of other species mates is

perceived to enhance control over resources. Cardiac,









electrodermal, and blink responses were measured from

subjects who were either alone, in the presence of a friend,

or in the presence of a stranger while passively waiting to

begin an experiment which they believed might involve mildly

painful radiant heat stimulation. Subjects performed

ratings of emotion, safety and trust, were presented with

two series of acoustic probes designed to elicit orienting

and startle responses, then solved a series of 25 anagrams

which ranged from easy to difficult and served ostensibly as

the experimental task. The presence of a friend produced a

distinct pattern of verbal, physiological and performance

responses that indicated the activation of an appetitive

motivational set. However, the presence of a stranger did

not activate an anxious motivational set. The stranger's

stimulus value may have been modified by the presence of

physical threat or by perceptions of the stranger as

cooperative. These findings are consistent with a modified

reactivity model in which it is hypothesized that the

presence of species mates associated with punishment or

frustration activates an anxious-inhibitory motivational

system, while the presence of species mates associated with

reward or escape from punishment activates an active-

appetitive motivational system.
















CHAPTER 1
INTRODUCTION

Social facilitation refers to the enhancement of an

individual's performance due to the presence of one or more

members of the same species. This effect was first noted in

1898 by Triplett, who observed that bicyclists were faster

when paced by another rider than when'riding alone. While

Triplett (1898) originally examined the effects of others

who were performing the same task as the subject (i.e.,

coacting), subsequent experiments have more often focused on

the effects of others as passive observers. Although many

investigations of the social facilitation effect were

performed during the early part of this century, researchers

had difficulty reliably replicating Triplett's (1898)

original effect. In fact, social presence was found to

impair performance as often as it facilitated it (e.g.

Dashiell, 1930; Pessin, 1933). Because these paradoxical

results could not be explained with the theoretical

constructs which existed at that time, studies of social

facilitation declined, and the topic remained obscure for

many years.









2

Theories of Social Facilitation

The Mere Presence Hypothesis

In 1965 Zajonc published a theoretical formulation that

effectively accounted for the previous discrepancies in

social facilitation research. He used the Hull-Spence drive

model (DxH=E), in which the strength of a response is a

product of both drive (i.e., general arousal) and the habit

strength of the response, to account for different

performance effects. Specifically, he proposed that the

"mere presence" of one or more members of the same species

(i.e., species mates or conspecifics) increases an

individual's level of general drive, resulting in

facilitation of those responses high in habit strength, but

inhibition of those responses low in habit strength.

Zajonc noted that for any task, "dominant" (i.e., high

habit strength) responses are typically behaviors which are

instinctual, well-learned, or simple, while "non-dominant"

(i.e., low habit strength) responses are typically behaviors

that are non-instinctual, poorly-learned or complex.

Furthermore, he hypothesized that while the effects of

increased drive would depend upon the dominance of the

behavior being examined, increased drive would consistently

impair new learning. However, Zajonc (1965) did not specify

the mechanism by which the mere presence of others elevates

drive. Later, Zajonc (1980) proposed that mere presence of

others results in uncertainty, which innately increases drive.








3

Zajonc and his colleagues have conducted an extensive

line of research which provides support for his theory. For

example, Zajonc and Nieuwenhuyse (1964) varied habit

strength by exposing subjects to verbal stimuli for

different numbers of times. After dominant and subordinate

responses were established, subjects performed a

tachistoscopic recognition test under either high or low

motivational conditions. Imbedded within this task were

trials on which no stimulus word was presented (i.e.,

pseudo-recognition trials). Subjects in the high motivation

condition showed significantly more dominant responses on

pseudo-recognition trials than subjects in the low

motivation condition, directly supporting Zajonc's

hypothesis that elevated general drive facilitates the

emission of dominant responses.

Zajonc and Sales (1966) utilized.the Zajonc-

Nieuwenhuyse paradigm to test the Zajonc's hypothesis that

the presence of a conspecific increases general arousal.

Drive was manipulated through audience conditions rather

than direct instructions. As expected, subjects working in

the presence of an observer gave a higher number of dominant

responses on pseudo-recognition trials than subjects working

alone, while the opposite effect was observed for

subordinate responses.

Matlin and Zajonc (1968) used a word association

paradigm to investigate social facilitation. They measured









4

the probability and latency of responses as subjects

performed a word association task both alone and in the

presence of an observer. They found that response latencies

were shorter when subjects were observed than when they were

not observed. Both the commonality and the uniqueness of

each association were used as indices of the probability of

dominant response emissions. When subjects were observed,

they produced more common responses and fewer unique

responses than when they worked alone. However, this effect

only occurred if the subject received the observation

condition prior to the alone condition. These findings were

interpreted as evidence that the presence of a species mate

increases general drive, resulting in the facilitation of

dominant responses.

Zajonc's success in accounting for both enhancement and

impairment of performance under conditions of social

presence restored new interest in the study of social

facilitation. As a result, research on social facilitation

has increased tremendously (Bond & Titus, 1983), and a

number of rival theories have been proposed. Several

non-drive theories have been put forth, but these have been

problematic in that they lack parsimony as well as empirical

support. The more viable alternatives have retained

Zajonc's concept of generalized drive, but differ in

specifying the mechanisms by which drive is increased.










Non-Drive Theory Alternatives

The first of the non-drive theories of social

facilitation is the objective self-awareness model

(Wicklund, 1975). In this model, the presence of others is

thought to elevate self-attention, resulting in the

perception of a discrepancy between actual performance and

ideal performance. In order to decrease this discrepancy,

the individual's motivation increases, resulting in improved

performance. Moore and Baron (1983) have pointed out

several problems with this theory. First, it does not

adequately account for impairment in performance on

difficult tasks. Second, it does not-account for findings

that dominant but incorrect or socially inappropriate

responses can be enhanced by the presence of others. Third,

a theory based upon self-awareness cannot easily account for

social facilitation in non-humans. Fourth, standard

self-awareness manipulations (the presence of a mirror, for

example) have not produced the same results as social

presence manipulations (Geen & Gange, 1977, Paulus, Annis &

Risner, 1978). Combined, these problems make the objective

self awareness model less than adequate.

The second non-drive theory is the attentional overload

model (Cohen, 1978). In this model, the presence of others

is thought to result in distraction of attentional resources

from the primary task, resulting in a condition of

attentional overload, which is compensated for by








6

restriction of attention to central task demands. This

attentional restriction is thought to enhance performance on

simple tasks, but impair performance on complex tasks that

require extensive attentional allocation to a number of

cues. This model is the attentional equivalent of the

distraction/conflict drive model described below, but

emphasizes perceptual instead of motor processes. The

primary criticism leveled against this model is that social

facilitation has been found to occur in the absence of

attentional overload (Matlin & Zajonc, 1968; Cacioppo, et

al., 1990). Thus, both non-drive models of social

facilitation clearly suffer from serious shortcomings.

Drive Theory Alternatives

There are two drive theory alternatives to Zajonc's

mere presence hypothesis. The first of these is the

distraction-conflict hypothesis (Baron, Moore and Sanders,

1978; Sanders et al., 1978). In this model, the presence of

an observer is thought to provoke social comparison

processes which compete with primary task demands for

attentional resources, resulting in cognitive overload. The

individual attempts to compensate with increased motivation,

which results in enhanced performance on simple tasks and

impaired performance on complex tasks. The distraction-

conflict model suffers the same problem as its pure

attention counterpart; performance facilitation has been








7

obtained without any sign of attentional distraction (Matlin

& Zajonc, 1968; Cacioppo et al., 1990).

The most viable alternative to the mere presence

hypothesis is the evaluation-apprehension model. Cottrell

(1968) agreed with Zajonc (1965) that social facilitation

involves increased drive, but argued that the increases in

drive depend upon learned anticipation of positive or

negative consequences which occur when subjects perceive

their performance to be evaluated. Weiss and Miller (1971)

agreed with Cottrell's learned drive hypothesis but, based

upon conditioning research, argued that learned drives could

be founded upon noxious primary drives but not upon

appetitive primary drives. Thus, they concluded, the

presence of others enhances drive only when negative

consequences are anticipated.

Several studies provided support for Cottrell's (1968)

original evaluation apprehension formulation. For example,

Cottrell, Wack, Sekerak and Rittle (1968) utilized the

Zajonc-Nieuwenhuyse paradigm to examine the facilitation of

dominant responses under alone, mere presence, and audience

conditions. The mere presence condition involved the

presence of another person who was ostensibly being prepared

for a sensory deprivation experiment by being blindfolded

and having his ears plugged, and was thus unable to evaluate

the subject. The audience condition produced a higher

number of dominant responses than the alone condition, but








8

the mere presence condition did not. Henchy and Glass

(1968) used the same pseudo-recognition paradigm, but

examined dominant responding across conditions in which the

subject was 1) alone, 2) alone with the belief that

performance was being recorded for later evaluation, 3) in

the presence of an observer perceived to be expert, or 4) in

the presence of an observer perceived to be a non-expert.

They found that subjects in the expert and alone-recorded

conditions demonstrated more dominant and fewer subordinate

responses than those in the alone or nonexpert conditions.

More recently, a number of investigators have examined

the hypothesis that an individual's performance is

facilitated only when positive outcomes are expected. While

this hypothesis is derived from Cottrell's (1968) evaluation

perspective, it is unique in proposing that facilitation of

performance results from anticipation of positive, but not

negative, consequences.

Geen (1977) had subjects perform a relatively difficult

anagram task in one of four conditions: 1) alone, 2)

observed, but told nothing regarding the observation (i.e.,

mere observation), 3) observed and told only that they were

being evaluated (i.e., evaluation only), 4) observed and

told that they were being evaluated in order to provide help

on a future task (i.e., assistance evaluation). Unobserved

subjects took less time than all other conditions to

complete the anagram task, while subjects in the assistance








9

evaluation condition required less time than subjects in

both the mere observation and the evaluation only

conditions, neither of which differed. Furthermore, when

subjects were divided into high and low scorers on the Test

Anxiety Scale (Sarason, 1980), high anxious subjects took

less time in the assistance evaluation condition than the

other two observation conditions, while low anxious subjects

showed no differences between conditions.

Subsequently, Geen (1979) measured the learning of high

response competition (i.e., difficult and nonintuitive)

paired associates after either success or failure feedback

about another related task. For subjects who were observed,

those given positive feedback needed fewer trials to attain

criterion and displayed fewer errors than subjects who were

given negative feedback.

In addition, Good (1973) examined the latency and

commonality of free associations to high and low response

competition words. Subjects were assigned to one of four

conditions. Each subject was told either that she had

performed well or poorly on a previous related task, and

that either she would or would not be evaluated during the

subsequent word association task. Although there were no

effects for high response competition words, the latency and

commonality of associations to low response competition

words were facilitated only for those subjects who

anticipated they would perform well while being evaluated.








10

Based upon the findings described above, Sanna and

Shotland (1990) proposed an expectancy-based explanation of

social facilitation in which they hypothesized that the

degree of facilitation or impairment engendered by an

audience depends upon the individual's anticipation of

positive or negative evaluation. In order to examine this

theory, Sanna and Shotland (1990) performed an experiment in

which subjects were assigned to one of six conditions. An

initial verbal recall task was performed and subjects were

given either positive feedback, negative feedback, or no

feedback. Subjects then performed two more verbal recall

tasks either in the presence a confederate whom they had

been told would be evaluating their performance or in the

absence of an evaluator. Immediately prior to the two

subsequent verbal recall tasks, subjects rated how well they

expected to perform as well as the degree to which they

expected their performance to be evaluated positively or

negatively.

Subjects who received positive feedback on the initial

memory task reported significantly higher mean expectancies

of performance and evaluation than those receiving either no

feedback or negative feedback, while those subjects who

received negative feedback reported significantly lower mean

expectancies of performance and evaluation than either of

the other two groups. Those subjects receiving no feedback

reported neutral mean expectancies of performance and








11

evaluation. In addition, the high expectancy subjects who

were evaluated exhibited significantly better verbal recall

than the no expectancy subjects who performed alone, while

the low expectancy subjects who were evaluated exhibited

significantly worse verbal recall than the no expectancy

subjects who performed alone. Furthermore, when subjects

who had received no feedback on the first task were divided

on the basis of their ratings into high and low expectancy

groups, high expectancy subjects who were evaluated

performed better than alone subjects, while low expectancy

subject who were evaluated performed worse than alone

subjects.

In summary, studies examining the effects of

anticipated positive, negative, or unspecified evaluation

upon performance have produced findings that cannot be

accounted for by the mere presence hypothesis. These

findings have forced Zajonc (1980) to argue that while it

may modify drive, evaluation apprehension is not the sole

mechanism of social facilitation. Indeed, both Markus

(1978) and Schmitt, Gilovich, Goore and Joseph (1986) have

demonstrated that the presence of others does facilitate

dominant responses even when the potential for evaluation is

neutralized, suggesting that overt evaluation is not a

necessary condition for social facilitation. However, they

agree that evaluation "is an important variable that can

indeed increase people's general arousal level and thus









12

further facilitate their dominant response tendencies"

(Schmitt et al., 1986, p. 246).

Learning Modulates Social Presence Effects

In addition to the results of those studies examining

evaluation-related hypotheses, findings from other paradigms

suggest that the effects of social presence depend upon the

contingencies associated with that social presence. For

example, Schachter (1959) has shown that in situations of

fear or threat, individuals chose the company of others over

being alone and report that the presence of others is

comforting. Wrightsman (1960) also found that subjects who

waited with others for what they thought was an aversive

injection reported significantly less fear than subjects who

waited alone. Interestingly, in both'of these studies

effects were strongest for subjects who were first or only

children. Schachter (1959) suggested that such children

differ from latter-borns in that they received more

consistent parental care and attention and therefore learned

to value social presence as comforting and supportive.

Latane and his colleagues (Latane & Glass 1968; Latane,

1969) have shown that, in a laboratory situation, indices of

fear such as freezing and defecation occur less often when

rats are placed in the presence of another rat, and that the

occurrence of these fear indices decreases further as the

rats are together over time. In addition, Davitz and Mason

(1955) have shown rats to display significantly less








13

conditioned fear when in the presence of another calm rat

than when alone.

Davidson and Kelly (1973) found that, in samples of

both psychiatric and non-psychiatric patients, subjects who

viewed a stress-inducing film in the presence of a nurse

showed lower electrodermal levels than subjects who viewed

the film alone. Similarly, Kissel (1965) measured skin

conductance level both before and during a frustrating task

performed in one of three conditions; alone, in the presence

of a stranger, or in the presence of a friend. The friend

condition resulted in lower skin conductance levels than

either the stranger or alone conditions, both of which were

not different from each other.

Other experimenters have manipulated the degree to

which subjects were familiar with an observer. Rajecki,

Kidd and Ivins (1976) examined the open-field responses of

chicks tested in the presence of a familiar companion, in

the presence of an unfamiliar stranger, or alone. An

interaction was found between social condition and the type

of behavior displayed by the chicks. Chicks in the familiar

companion condition showed predominantly pecking behavior.

Those in the alone condition showed predominantly

immobility, while those in the stranger condition showed

predominantly distress calls. The authors suggested that,

instead of producing changes in general drive along a

continuous single response hierarchy as would be predicted








14

by Zajonc's drive model, manipulation of social stimuli

resulted in several qualitatively different responses.

Shaver and Liebling (1976) investigated familiarity

effects in humans. Thirty males were exposed to one of

three conditions; alone, familiar observer, or unfamiliar

observer. Familiar observers were those observers who sat

with the subject during the instruction period of the

experiment, while unfamiliar observers were introduced to

the subject immediately preceding performance of the

experimental task. Subjects performed either a simple or a

complex tactile stylus maze. While there were no group

differences for those subjects performing the simple tactile

maze, subjects who were alone while working on complex mazes

produced significantly fewer errors than subjects in either

of the social presence conditions, who did not differ in

errors. However, despite the lack of statistical

significance, the pattern of mean errors across groups

suggested difference between the two social presence

conditions, with means of 61, 130 and 174 errors for the

alone, familiar observer and unfamiliar observer conditions

respectively. The authors concluded that "the results .

suggest that audience-induced drive is lowered if the

subject is given the opportunity to incorporate the fact of

the observer's presence into his plans for working on the

task" (p. 265). The low number of subjects in each cell

(n=5) and the fairly weak manipulation of familiarity








15

suggest that differences might be found with improved

methods.

Finally, Clark and Fouts (1973) found that subjects who

have had previous positive experiences with an observer

showed less arousal when later exposed to the presence of

that observer than subjects who did not have previous

positive experiences with the same observer. Specifically,

they found that response intensity on a simple motor task

was lower for children working in the presence of an adult

with whom they had a previously positive experience than it

was for children working in the presence of an adult with

whom they had either a negative or neutral encounter. When

the children performed the same task alone, no effect of

previous experience was found.

In summary, the studies reviewed above provide ample

evidence that perceptions regarding the contingencies

associated with the presence of a conspecific affect the

manner in which the conspecific is responded to.

The Psvchophvsiology of Social Facilitation

Up to this point we have limited our discussion in two

ways. First, we have focused primarily on behavioral data.

Second, the aforementioned examinations of social

facilitation have generally assumed the Hullian concept of

nonspecific drive to be synonymous with generalized

autonomic arousal. To gain further insight into the








16

mechanism of social facilitation let us examine recent

neuropsychophysiological advances regarding motivation.

Arousal is Multifactorial

As noted by Fowles (1980), the concept of arousal as a

unidimensional phenomena long held a prominent position in

psychology due to the combined influence of 1) its intuitive

simplicity, 2) Cannon's work on the general, unitary

response of the sympathetic nervous system, 3) Hull's

concept of generalized drive, and 4) early studies showing

the reticular activating system (RAS) to be the neurological

substrate for generalized arousal. However, Lacey (1967)

presented a large body of evidence against the general

arousal concept and argued that the data strongly indicated

the existence of a second, limbically located arousal system

which plays a larger role in behavioral activation than the

RAS. His critique was followed by others (e.g. Routtenberg,

1968), and current investigators agree that the concept of

unidimensional arousal "failed the empirical test rather

badly" (Fowles, 1980, p.88).

Jeffery Gray has recently developed a major theory

regarding the neurobiological substrates of motivation.

Gray (1970, 1981) has found substantial evidence that human

motivation involves three separate arousal systems. The

first system, which he calls the Behavioral Inhibition

System (BIS), has been found to operate in situations of

response-dependent punishment or frustrative nonreward








17

(i.e., the absence of response-dependent rewards). Because

it inhibits behavior in aversive situations and is

responsive to anti-anxiety drugs, the BIS appears to be the

substrate of anxiety.

In contrast to the BIS, Gray has proposed the existence

of a Behavioral Activating System (BAS) which initiates

reward-seeking, appetitive behavior as well as active escape

and avoidance from punishment. In turn, both the BIS and

the BAS are thought to have input into a general activation

system, which Gray believes to be associated with the RAS

and to increase the intensity and vigor of behavior. Thus,

Gray's model includes two behavior directing systems, one

which initiates responses to appetitive stimuli and one

which initiates responses to aversive stimuli, both of which

have input into a general arousal system which increases the

intensity of behavior regardless of its direction.

Fowles (1980) utilized Gray's three-arousal model to

organize a large number of otherwise unrelated psychophysio-

logical findings. Fowles (1980) documented that while a

well-known relationship exists between somatic activity and

heart rate acceleration (for a review see Obrist, 1976),

heart rate is sensitive to incentive conditions above and

beyond its coupling with somatic activity. He presented

substantial evidence that findings of HR responsivity to

incentive conditions as well as findings of increased HR

during active coping relative to passive coping can be








18

accounted for by the coupling of BAS activation to heart

rate acceleration.

Fowles (1980) also reviewed evidence that electrodermal

activity is most sensitive to situations involving

punishment and as such provides an index of BIS activation.

Fowles was careful to note that the extent of directional

fractionation will depend upon the degree to which an

aversive stimulus results in increased incentive (as in

active avoidance), or decreased incentive (as in extinction

or learned helplessness). If an aversive stimulus is

perceived as an increased incentive for active avoidance,

both electrodermal and cardiac responding will increase. If

the aversive stimulus results in decreased incentive,

electrodermal responding will increase but cardiac

responding will decrease, resulting in directional

fractionation. Given Fowles' (1980) elegant integration of

psychophysiological data with Gray's three-arousal theory,

what do the psychophysiological data tell us about the

mechanisms of social facilitation?

Early Psychophysiological Studies

In 1983, Moore and Baron reviewed the existing

psychophysiological data on social facilitation. Of the six

studies they reviewed, no social facilitation effects were

found for heart rate. Of the studies utilizing palmar

sweating as a measure of electrodermal activity, results

were equivocal, with five out of eight studies revealing a









19

social facilitation effect. Of the four studies which had

examined electrodermal levels, none showed social

facilitation effects. They noted, however, that the only

study in which phasic electrodermal responses were measured

(Geen, 1979) found strong effects for social presence which

varied as a function of the threat value of the observer.

Maximum skin conductance level and heart rate were also

measured, but these measures, as in other studies, were

unaffected.

It is noteworthy that almost all other studies on the

psychophysiology of social facilitation had utilized only

one physiological variable, either electrodermal or cardiac,

as an index of general autonomic arousal (i.e., drive). In

contrast, Geen (1979) measured both electrodermal responses

and heart rate, allowing his findings to be interpreted more

specifically as increased anxiety rather than merely

increased general arousal or drive.

The Reactivity Hypothesis

In more recent review of the psychophysiological

research on social facilitation, Cacioppo and Petty (1986)

noted that in all three of the studies reviewed by Moore and

Baron (1983) which found palmar sweating unaffected by

social presence, the observer had been concealed rather than

visible. They suggested that the manipulations in those

studies may have been weak. They also noted that skin

conductance level findings have traditionally been









20

insensitive to "conditions involving moderate to high threat

and more subject to psychologically irrelevant factors

such as hydration" (Fowles, 1980, p.95). They focused

instead upon Geen's (1977, 1979) findings that phasic

electrodermal responding indexes social presence conditions,

facilitation effects are modulated by the observer's threat

value, and heart rate and tonic skin conductance level are

both unaffected by observer conditions.

Cacioppo and Petty (1986) made several inferences from

these observations. First, measurement of electrodermal

responding appears to be a more accurate index of the

activation of threat-related arousal than the relatively

crude palmar sweat and skin conductance level methods.

Second, the bulk of the psychophysiological data on social

facilitation fit well into the Gray-Fowles three arousal

model and suggest that social facilitation procedures, which

typically involve observation of the subject by a stranger,

may be intrinsically threatening. Third, they proposed that

all social facilitation effects involve at least one basic

cognitive processing stage in which stimulus threat value is

assessed.

In light of positive findings for phasic electrodermal

reactivity and negative findings for tonic electrodermal

levels, Cacioppo and Petty (1986) went on to refute Zajonc's

hypothesis that the physiological mechanisms of social

facilitation operate through the adrenal cortex. Zajonc








21

(1965), while acknowledging that his hypothesis was

speculative and had not been directly tested, originally

proposed that increases in drive occurring during social

facilitation were mediated by the same mechanism of elevated

adrenal corticosteroid production that had previously been

found in conditions of crowding. Cacioppo and Petty (1986)

noted that later studies have shown that changes in cortico-

steroid levels are slow to develop and to dissipate, and are

associated more with increased tonic readiness to respond

than actual physiological reactivity. They suggested that

enhanced physiological reactivity is more likely to be

mediated by the adrenal medulla, which is sympathetically-

enervated and quickly releases catecholamines. They also

noted that catecholamine release is associated with

responsivity to psychological stressors, contributes to

elevated sympathetic activity and dissipates quickly.

In summary, Cacioppo and Petty (1986) successfully

integrated the psychophysiological data on social

facilitation into the Gray-Fowles three arousal model. They

suggested that social facilitation manipulations elicit a

pattern of heightened autonomic reactivity which has been

described by Fowles (1980) as indicative of anxiety

processes, rather than causing the elevated nonspecific

drive/general arousal predicted by Zajonc (1965), and they

proposed that social facilitation occurs when an

individual's perception of an increased possibility of









22

threat activates anxiety processes. This formulation, like

the evaluation hypotheses reviewed above, allows for

individual differences in threat perception based on

learning, but additionally suggests that individuals can

experience threat, and thus social facilitation, in the

absence of overt evaluation.

Testing the Reactivity Hypothesis

Cacioppo et al. (1990) tested the reactivity hypothesis

of social facilitation against the mere presence and

distraction-conflict hypotheses. They suggested that

findings of tonic changes in autonomic physiology would

support the mere presence model (which assumes that social

facilitation is mediated by general arousal), rather than

either the evaluation apprehension or distraction-conflict

model. The latter two models yield opposing predictions

regarding physiological reactivity. If subjects are in a

state of distraction during observation, as predicted by the

distraction-conflict model, reactivity to irrelevant

auditory stimuli would be diminished. If subjects are in a

state of threat assessment, as predicted by the reactivity

model, responsivity to irrelevant auditory stimuli would be

enhanced.

Thus, for both heart rate and skin conductance tonic

levels and phasic responses to auditory probes were measured

from 27 women who were passively seated while they thought

that the experimenter was calibrating'equipment. Half of









23

the subjects believed that they were being observed by the

experimenter through a one-way mirror, while the other half

believed that they were unobserved. During the waiting

period, subjects were exposed to a series of moderate

intensity slow-rise time auditory tones of 2 seconds in

duration. Physiological data was collected prior to the

presentation of each tone as well as for 30 seconds after

tone onset. These data were divided into different epochs

to allow for the assessment of baseline as well as phasic

and tonic physiological reactivity. No differences in

baseline physiology were found. Analysis of reactivity to

the first tone in the series indicated the directional

fractionation of HR and SCR characteristic of an orienting

response. Comparison of groups indicated greater skin

conductance responding for observed subjects than alone

subjects, but no significant differences between groups in

cardiac deceleration. Analysis of tonic reactivity to the

first tone showed that skin conductance level remained

constant for alone subjects, while observed subjects showed

a quick recovery from initial elevations in skin

conductance. An analysis of responding to the subsequent

tones in the series revealed 1) quick electrodermal

habituation for both groups despite initially greater skin

conductance levels for the observation group and 2) no group

differences in cardiac habituation.









24

These findings provided support for the reactivity

hypothesis and evidence against the mere presence and

distraction-conflict hypotheses. Because these findings

matched Fowles' (1980) predictions for heart rate and skin

conductance responding during threat, Cacioppo et al. (1990)

suggested that all of the aforementioned theories of social

facilitation (including mere presence, self presentation,

distraction-conflict, and evaluation apprehension) may

represent sufficient but unnecessary psychological

explanations which may be components of a general process

that occurs when the presence of conspecifics is perceived

as a threat to any resource valued by the individual, be it

physical or social (approval, control, etc.). Furthermore,

they speculated that the presence of others who enhance the

individual's sense of control over resources may diminish

physiological reactivity to stressors.

The Startle Blink as a Measure of Motivational State

The startle is a whole body reflex which is elicited by

the suddenness rather more than the intensity or duration of

stimuli. Because it is the most consistent component of the

startle reflex and can be reliably measured through surface

electromyographic recording of orbicularis oculi muscle

activity, the eyeblink component has become the preferred

measure of the startle reflex (for review see Anthony,

1985).








25
Psychophysiologists have traditionally used the startle

blink to measure information processing in humans (Anthony,

1985) and aversive conditioning in rats (Davis, 1986).

However, in recent years researchers have effectively

utilized the startle reflex to study emotion and motivation

in humans. In a pioneering experiment, Vrana, Spence and

Lang (1988) probed subjects with brief rise-time white noise

bursts while they watched affect-laden slides for six second

intervals. Based on ratings by an independent sample, the

slides were categorized as either positive, neutral, or

negative, and categories were balanced for mean arousal

value. Compared with blinks elicited during neutral slides,

blink magnitude was augmented during negative slides, but

inhibited during positive slides. Given the defensive

nature of the startle reflex, these findings suggested that

blink magnitude is modulated by its match with ongoing

emotional processes. Thus, negative emotions appeared to

synergistically augment blink responding while positive

emotions appeared to attenuate blink responding due to

mismatch.

However, other researchers using similar slide-viewing

paradigms (Anthony & Graham, 1985; Simons & Zelson; 1985)

found the magnitude of acoustically-elicited blinks to be

modulated by how interesting the slides were, suggesting the

alternative explanation that blink magnitude is actually

modulated by the degree to which attentional resources are








26

allocated to the auditory channel. In order to test these

competing attentional versus emotional explanations of blink

modulation, Bradley, Cuthbert and Lang (1991) reexamined

blink magnitude during the viewing of positive, neutral and

negative slides, but this time varied the sensory modality

in which the probes were presented. For each subject, half

of the startle probes were visual flashes and half were

acoustic white noise bursts. Blink magnitude was found to

vary in the same manner as in Vrana et al. (1988), with

augmented blinks for negative slides and inhibited blinks

for positive slides, regardless of which sensory modality

the probes were presented in. These findings indicate that

emotional stimuli such as visual slides do not modulate

blink magnitude by shifting attentional resources, but

rather produce an emotionally valent state which either

matches or mismatches the blink reflex.

Lang, Bradley and Cuthbert (1990) reviewed the above

findings as well as other results from their laboratory

which showed blink magnitude to be sensitive to aversive

conditioning in humans (Greenwald, Hamm, Bradley & Lang,

1988), and to hemispheric differences in emotional

processing (Bradley, Cuthbert & Lang, 1991). They

integrated these data with the large literature on emotional

organization, suggested that emotions are integrated action

dispositions which vary along a fundamental dimension (from

approach, affiliation and consummation to avoidance, escape











and defense), and proposed that such affective-motivational

states can be indexed through their match or mismatch with

probed reflexes.

It is notable that, despite their different research

domains and methodologies, Gray and Lang have reached

similar conclusions regarding the fundamental organization

of emotional-motivational behavior. While Lang has

suggested that emotion is organized on a continuum, and Gray

and Fowles have proposed separate but complementary systems,

these notions are not incompatible, and both aid in the

psychophysiological delineation of appetitive and aversive

motivational states.

Thus, the work of Lang and his colleagues provides

substantial evidence that the startle blink is an effective

index of emotional processing, and provides a theoretical

framework with which to interpret blink findings. If, as

hypothesized by Cacioppo and Petty (1986), social

facilitation involves the activation of unpleasant affect

due to perceived increases in threat, the startle reflex

should provide a sensitive index of such negative emotional

processes.

Experimental Problem

The findings of Cacioppo et al. (1990) suggested that

mere observation by a stranger results in a pattern of

heightened autonomic reactivity indicative of anxiety

processes. However, the converse prediction of the









28

reactivity hypothesis, that control-enhancing, supportive

social presence results in decreased autonomic reactivity,

remained untested. The present experiment attempted to

replicate the findings of Cacioppo et al. (1990), and to

expand upon them by comparing verbal report, overt

performance and physiological reactivity in threatening

social conditions to that in supportive social conditions.

Cardiac, electrodermal, and blink responses were

measured from subjects who were either alone, in the

presence of a friend, or in the presence of a stranger while

passively waiting to begin an experiment which they believed

might involve mildly painful radiant heat stimulation.

Subjects performed ratings of emotion, safety and trust,

were presented with two series of acoustic probes designed

to elicit orienting and startle responses, then solved a

series of 25 anagrams which ranged from easy to difficult

and served ostensibly as the experimental task.

Experimental Hypotheses

Preexisting Anxiety Levels

Mean state and trait anxiety levels measured

immediately prior to the start of the experiment were not

expected to differ between conditions.

Verbal Report of Emotion, Safety and Trust

Pleasantness, control, and safety ratings were expected

to be highest for the friend condition and lowest for the

stranger condition. Arousal ratings were expected to be








29

highest in the stranger condition and lowest in the friend

condition. Trust ratings were expected to be greater in the

friend condition than the stranger condition.

Baseline Levels of Physiological Activity

Baseline levels of heart rate, skin conductance, and

electromyographic (EMG) activity in the orbicularis oculi

region were not expected to differ between conditions.

Phasic Heart Rate Responding

The tones were expected to elicit the deceleratory

phasic cardiac response typically associated with orienting,

while the white noise bursts were expected to elicit the

acceleratory phasic cardiac response typically associated

with startle. However, heart rate was not expected to

differ between conditions, since subjects were not expected

to assume the appetitive motivational set to which cardiac

acceleration is related. Additionally, while heart rate was

expected to habituate over time, such habituation was not

expected to differ between conditions.

Phasic Skin Conductance Responding

Skin conductance responding to orienting tones was

expected to be of greatest magnitude and shortest onset

latency in the stranger condition and of smallest magnitude

and longest onset latency in the friend condition.

Additionally, initial differences between conditions in skin

conductance response (SCR) magnitude were expected to

dissipate rapidly. Given the ephemeral nature of the SCR








30

effects found by Cacioppo et al., (1990) SCR differences

were not expected to exist by the time that startle probes

were presented. However, if differences were found in SCRs

to the startle probes, they were expected to follow the same

pattern as for the orienting tones, with the stranger

condition exhibiting the greatest responses and the friend

condition exhibiting the smallest.

Blink Responding to Startle Probes

Blink responses were expected to'be of greatest

magnitude and shortest onset latency in the stranger

condition and to be of smallest magnitude and longest onset

latency in the friend condition. Initial differences in

blink magnitude and latency were expected to dissipate as

blink responding habituated over time.

Anagram Performance

Differences between conditions in response latency and

accuracy were expected to change across levels of anagram

difficulty. Subjects in the alone condition were expected

to demonstrate a decline in response accuracy as anagram

difficulty increased. Because they were expected to

experience greater threat, subjects in the stranger

condition were expected to respond to easy anagrams less

quickly and accurately than subjects in the alone condition,

and this impairment was expected to increase as difficulty

increased. In contrast, because they were expected to

experience greater safety, subjects in the friend condition








31

were expected to respond to easy anagrams more quickly and

accurately than subjects in the alone condition, and, while

their performance was expected to decline as anagram

difficulty increased, they were expected to respond to

difficult anagrams more quickly and accurately than subjects

in the alone condition.

The Interaction Between Preexisting Anxiety Levels and

Social Presence Conditions.

Based upon Geen's (1979) findings of a relationship

between test anxiety scores and treatment effects, high

anxious subjects were expected to be more affected by social

presence conditions than low anxious subjects.

Specifically, state and trait anxiety scores from the State

Trait Anxiety Inventory (STAI; Spielberger, Gorsuch, &

Lushene, 1970) were expected to account for a significant

portion of the variance in both skin conductance response

and startle blinks, and the nature of the relationship

between anxiety and physiological responsivity was expected

to change across conditions. In the friend condition, high

anxious subjects were expected to demonstrate smaller

physiological responses than low anxious subjects. In the

stranger condition, high anxious subjects were expected to

demonstrate greater physiological responses than low anxious

subjects.














CHAPTER 2
METHODS

Subjects

Subjects were 45 female introductory psychology

students from the University of Florida who volunteered to

participate in order to receive partial class credit.

Subjects were randomly assigned to one of three conditions.

Apparatus

Subjects sat in a comfortable reclining chair in a

softly lit 6' by 6' room that was partially sound shielded.

Stimulus presentation and data collection were controlled by

a Northgate AT computer located in an adjacent room.

Heart Rate

Lead I EKG was obtained with standard Ag-AgCl

electrodes filled with Beckman electrode paste. The EKG

signal was filtered and amplified through a Coulbourn S75-01

Hi Gain Bio-Amplifier, then processed through a Coulbourn

S21-06 Bipolar Comparator and a Coulbourn S52-12

Retriggerable One Shot, which converted each cardiac R-spike

into a 5 volt pulse. The interval between each consecutive

5 volt pulse was measured by a clock within the computer.

Skin Conductance

Skin conductance was measured with standard Ag-AgCl

electrodes attached to the hypothenar surface of the

32









33

nondominant hand. The skin was rubbed with distilled water

prior to electrode attachment and electrolyte was K-Y jelly.

The signal was processed through a Coulbourn S71-22 skin

conductance coupler with sensitivity set at 50 microvolts

per micromho.

Startle Blink

Muscle activity of the orbicularis oculi region was

recorded unilaterally from the left side of the face with

Beckman Ag-AgCl miniature electrodes attached at the

standard placements recommended by Tassinary, Cacioppo and

Geen (1989), and was amplified by a factor of 30,000 with a

Coulbourn S75-01 bioamplifier with filters set to allow

signals between 8 and 1000 Hz to pass. The signal was then

rectified and integrated through a Coulbourn S76-01

contour-following integrator with a time constant set at 200

msec.

Stimuli

Auditory stimuli consisted of separate series of six

orienting tones and six startle probes. Orienting tones

were slow rise-time, 2000 Hz, 65 db auditory tones of 2

seconds duration. Orienting tones were generated by a

Coulbourn S24-05 Voltage Controlled Oscillator, processed

through a Coulbourn S84-04 Shaped Rise/Fall Gate set at 80

milliseconds, amplified by a Coulbourn S79-02 Adjustable

Gain Amplifier, and gated by a Coulbourn S82-24 Audio

Mixer-amplifier. Startle probes were instantaneous








34
rise-time, 100 db bursts of white noise of 50 milliseconds

duration which were generated from a white noise cassette

tape played and amplified through an AIWA CA-30 cassette

player and gated through a Coulbourn S82-24 Audio

Mixer-Amplifier.

Both sets of acoustic stimuli were further amplified

through a Realistic Amplifier and presented through an

Infinity RS 1000 speaker positioned on a shelf immediately

behind the subject's head. In each series, stimuli were

pseudo- randomly presented within a 2.5 minute period, with

interstimulus intervals (ISIs) varying from 6 to 41 seconds

(see Appendix I for intervals). In order to maximize

statistical power, the series of orienting tones was

presented first, followed by the series of startle probes.

Anagrams consisted of 25 scrambled words which ranged

in length from three to seven letters. Anagrams were

randomly presented on a Sharp portable computer. See

Appendix F for Anagrams and their solutions.

Procedure

An undergraduate research assistant who was naive to

experimental hypotheses called student volunteers, assigned

them randomly to condition, and scheduled meeting times. If

a subject was assigned to the friend condition, she was

asked to have her closest friend accompany her to the

experiment.








35

Upon arriving at the laboratory, the subject entered

the experiment room. If the subject was in the friend

condition, she remained in the presence of her friend, and

they were allowed to talk freely until after instructions

had been administered and electrodes attached. If the

subject was in the alone condition, she remained alone

throughout the entire experiment. If the subject was in the

stranger condition, she remained alone until after

electrodes were attached, when she was exposed to the

presence of a female confederate.

After signing informed consent, the subject completed

the State Trait Anxiety Inventory (STAI) and a bogus

questionnaire on perceived problem-solving abilities.

Instructions were then administered. The experiment was

described as having two parts. The first part was described

as an examination of psychophysiological responses during

problem solving. If the subject was in the friend or

stranger conditions, she was told that the problem solving

tasks would be interpersonal in nature. The second part was

described as an examination of the effects of mild pain upon

problem solving which involved the possibility of receiving

mildly painful radiant heat stimulation. The radiant heat

stimulator was then introduced and turned on so that several

dials on the front panel lit up. If the subject was

unwilling to accept the possibility of mildly painful








36

radiant heat stimulation the experiment was discontinued.

Three subjects refused to continue the experiment at this

point.

Next, electrodes were attached and physiological

signals checked. If the subject was in the stranger

condition, a female confederate was then introduced as a

research assistant who would be participating with the

subject in the problem solving tasks. The confederate was

instructed to sit quietly and look directly at the subject

with a neutral facial expression without communicating

verbally or nonverbally.

Next, the experimenter stated that the physiological

recording devices needed to be calibrated before the

experiment could start, and that calibration would take

several minutes during which the subject might hear some

random noises that she could ignore. The subject was then

given instructions for the calibration period, which

included resting her left hand directly below the barrel of

the radiant heat stimulator until instructed otherwise. The

experimenter then left the room for 45 seconds during which

the subject completed seven-point Likert type ratings of

pleasantness, arousal, control, and safety. The

experimenter then reentered the room, collected the ratings,

notified the subject that calibration would begin, and left

the room again, at which time the subject closed her eyes

and attempted to relax as much as possible. Presentation of








37

the series of orienting tones and startle probes was

initiated 30 seconds after the experimenter left the room.

After both stimulus series were presented, the

experimenter returned to the room, turned on a Sharp

portable computer, placed it in the subject's lap, and

administered anagram task instructions. The experimenter

then left the room and the subject unscrambled each of 25

randomly presented anagrams as quickly and accurately as she

could. The subject typed with her right hand. She pressed

the return key a first time to initiate presentation of each

anagram and to start a clock inside the computer. As soon

the subject obtained a solution, she pressed the return key

a second time in order to stop the clock and thus register

her reaction time. She then typed her answer into the

computer and entered it by pressing the return key a third

time. This procedure was repeated for each of the 25

anagrams. The subject indicated that she had completed the

task by tapping on the wall of the experiment room.

The subject was then informed that she was not assigned

to receive mildly painful radiant heat stimulation, and that

the experiment was actually completed. Next, the subject

rated on a seven-point likert type scale the extent to which

she had believed that she might receive mildly painful

radiant heat simulation. Electrodes were then detached,

debriefing completed, credit for the experiment given, and

the subject allowed to leave.








38

Data Reduction and Analysis

Preexisting Anxiety Levels

Before the start of the experiment subjects completed

the State-Trait Anxiety Inventory. Separate scores for

state anxiety (STAI-S) and trait anxiety (STAI-T) were

subjected to a one-way Analysis of Variance (ANOVA) with

condition (Alone, Stranger, Friend) as the between subjects

factor.

Verbal Report of Emotion. Safety, and Trust

During the experiment all subjects rated their

experience of pleasantness, arousal, control, and safety on

seven-point Likert-type scales. Subjects in the friend and

stranger conditions also rated their trust of the other

person on a seven point Likert-type scale. Higher scores

denoted more of each descriptor. In order to assess the

degree to which verbal ratings were influenced by individual

differences in preexisting anxiety levels, Pearson

correlations were performed separately among each of the

ratings and state and trait anxiety scores.

In order to remove the significant variance in verbal

ratings attributable to preexisting individual differences

in anxiety, each rating was subjected to a one-way Analysis

of Covariance (ANCOVA) with STAI-S as the covariate. If a

significant main effect was found, pairwise comparisons were

performed using the Bonferroni correction against Type I

error.








39

Baseline Levels of Physiological Activity

To measure tonic baseline activity, heart rate and skin

conductance were sampled for six seconds once immediately

before the series of orienting tones was presented and once

immediately before the series of startle probes was

presented. A one-way Analysis of Variance (ANOVA) was

performed on mean tonic baseline activity for each dependent

variable at each time.

To measure tonic baseline orbicularis oculi activity,

integrated electromyographic (EMG) activity was sampled from

the orbicularis oculi region for six seconds once

immediately prior to presentation of the series of startle

probes.

Phasic heart Rate Responding

For both orienting and startle trials, sampling of the

intervals between cardiac R-spikes was initiated two seconds

before stimulus onset and discontinued six seconds after

stimulus onset. These R-R intervals were reduced off-line

into average beats per quarter second interval as per Graham

(1980). For each trial, change scores were calculated by

subtracting the average for the half-second interval

immediately preceding stimulus onset from each quarter-

second interval following stimulus onset. Because subjects

generally exhibited a biphasic (initial deceleration or

acceleration) response waveform which typically returned to

baseline by the fifth second after stimulus onset, average








40
change during the first four seconds after stimulus onset

was calculated as the measure of phasic cardiac response

rather than using traditional heart rate leg scores.

Mixed-model ANCOVAs were performed separately on phasic

cardiac responses to the orienting and startle stimuli.

Condition (alone, stranger, friend) was the between subjects

factor, and Time (trials one to six) was the within subject

factor. In each analysis the six-second tonic prestimulus

baseline was used as the covariate in order to reduce

extraneous variance attributable to tonic differences in

cardiac activity. If Condition and Time were found to

interact, one-way ANCOVAs were performed for each trial. If

a significant main effect was found on any trial, pairwise

comparisons of means were performed using the Bonferroni

correction against Type I error. When qualitative

differences in phasic cardiac responses were found either

across time or among conditions, post-hoc one-way t-tests

were performed to confirm that such responses differed from

baseline levels. Huynh-Feldt corrections for inflated

degrees of freedom are reported for all analyses performed

on heart rate, skin conductance, or orbicularis EMG which

included time as a within subject factor.

Phasic Skin Conductance Responding

Skin conductance data was sampled at a rate of 20 Hz

during orienting and startle trials. Sampling was initiated

two seconds before stimulus onset and discontinued six










seconds after stimulus onset. Skin conductance response

(SCR) was scored off-line. For each trial, SCR was

calculated by subtracting the average of the half-second

interval immediately preceding stimulus onset from the

greatest value occurring between one and six seconds after

stimulus onset. No response was scored for those trials on

which change from baseline did not exceed .05 micromhos.

When a response was scored, peak latency from stimulus onset

was also recorded.

SCR magnitude was calculated to include all trials on

which a response could have occurred, while SCR amplitude

was calculated to include only those trials on which a

response actually occurred. Because each of these indices

has its advantages and disadvantages, they have been the

subject of some debate. Prokasy and Kumpfer (1973) argued

for the use of amplitude after they noted that magnitude

confounds response size with response frequency, and cited a

number of studies in which frequency was not associated with

response size. In contrast, Venables and Christie (1980)

recognized a conceptual problem in the calculation of

amplitude. They noted that the determination of non-

responses ultimately involves the use of some convention-

based criterion which may arbitrarily exclude small but

actual responses. Dawson, Schell, and Filion (1990)

recognized both arguments. They also noted that the use of

amplitude often results in the problem of differing cell








42
sizes, and suggested that both indices be calculated and

contrasted for optimal examination of skin conductance data.

Two-way (condition x time) ANCOVAs were performed on

the frequency and magnitude of SCRs separately for orienting

and startle trials. In each analysis the six-second tonic

prestimulus baseline was used as the covariate in order to

reduce extraneous variance attributable to tonic differences

in electrodermal activity. Because so many subjects had at

least one trial on which they did exhibit a response of at

least .05 micromhos, a two-way (condition x time) ANCOVA

could not be performed upon SCR amplitude data.

In addition, one-way ANCOVAs were performed upon SCR

frequency, amplitude and latency for the first orienting

trial and upon amplitude and latency for each of the startle

trials. Each ANCOVA utilized the six-second tonic

prestimulus baseline as the covariate. If a significant

main effect was found on any trial, pairwise comparisons of

means were performed using the Bonferroni correction against

Type I error.

Blink Responding to Startle Probes

EMG activity of the orbicularis oculi region was

sampled at a rate of 1000 Hz during startle trials.

Sampling was initiated two seconds before stimulus onset and

discontinued six seconds after stimulus onset. Startle

blinks were scored off-line by a computer program which

calculated baseline as the average EMG during the 20








43
millisecond interval immediately subsequent to stimulus

onset, located the blink onset as the first detectable

change between 21 and 120 milliseconds of stimulus onset,

located the greatest EMG value occurring within 150

milliseconds after blink onset, then subtracted the baseline

from the greatest EMG value and recorded the resulting

magnitude as well as latency from stimulus onset.

Two-way (condition x time) ANCOVAs were performed

separately on mean blink magnitude and latency. In each

analysis the six-second tonic prestimulus baseline was used

as the covariate in order to reduce variance attributable to

tonic differences in orbicularis oculi EMG activity.

Anagram Performance

Each anagram response and its latency from stimulus

onset was recorded on-line. Response accuracy was scored

off-line. Responses were considered accurate if they could

be found in Webster's New Collegiate Dictionary. Under this

criterion, 21 anagrams yielded one valid response each, and

four anagrams yielded two valid responses each. Overall

mean accuracy was calculated for each anagram and was used

to rank the anagrams in order of difficulty. If two

anagrams had the same overall accuracy the shorter one was

ranked as easier. Ranked anagrams were then divided into 5

difficulty levels, and mean accuracy was calculated for each

difficulty level.








44

A two-way ANOVA was performed on anagram performance

accuracy, with condition as the between subject factor and

difficulty as the within subject factor. If condition was

found to interact with difficulty, one-way ANOVAs were

performed at each level of difficulty. If a significant

main effect was found at any level, Bonferroni-corrected

pairwise comparisons of means were performed.

The Interaction of Preexisting Anxiety Levels and Social

Presence Conditions.

Pearson correlations were calculated between each

physiological variable and STAI-S and STAI-T separately for

the friend and stranger conditions. If these correlations

appeared to differ in either direction or magnitude, the

responses of subjects in the friend and stranger conditions

were then subjected to two regression-style Analyses of

Covariance (ANCOVA) with anxiety level and condition as

independent variables.

The first, reduced model ANCOVA did not include an

interaction term, while the second, complete model ANCOVA

did include an interaction term. The sum of squared errors

of both the complete and reduced models were compared

through the calculation of an F statistic (Agresti & Finlay,

1986, pp. 447-457). A significant F-value indicated that

the complete model accounted for more variance than the

reduced model and that the effects of condition upon

response physiology changed as a function of anxiety levels.










Manipulation Check

In order to rule out the possibility that findings

might be attributable to differences in subjects' belief

that they were likely to receive mild pain, a manipulation

check was added to the protocol after'the data collection

phase of the study had already started. Immediately after

completing the experiment, approximately two-thirds of the

subjects rated how much they had believed that they were

likely to receive mildly painful heat stimulation. Ratings

were on a 7-point Likert type scale, with higher scores

denoting greater belief. These ratings were subjected to a

one-way ANOVA with condition as the between subjects factor.















CHAPTER 3
RESULTS

Preexisting Anxiety Levels

Overall mean state and trait anxiety scores were very

close to the means reported by Spielberger et al. (1983) for

female college students (M=38.76 and SD=11.95 for STAI-S;

M=40.40 and SD=10.15 for STAI-T), indicating that subjects'

mean overall preexisting anxiety levels were within the

normal range. As predicted, there were no differences among

conditions in preexisting state or trait anxiety levels.

These findings are illustrated in Table 1, and ANOVA tables

are contained in Appendix A.

Verbal Report of Emotion. Safety, and Trust

Verbal ratings were related to preexisting state and

trait anxiety levels. As shown in Table 2, pleasantness,

arousal and safety ratings were most highly related to state

anxiety while control ratings were most highly related to

trait anxiety.

Contrary to hypotheses, neither arousal, control, nor

safety ratings differed among conditions. In contrast,

ratings of pleasantness did differ among conditions

(F[2,41]= 5.78, p<.01). As predicted, subjects in the

friend condition reported greater pleasantness than those in

the alone (F[1,29]=11.24, p<.01) or stranger (F[1,29]=7.24,

46


_ ____












Table 1
Mean Preexisting Anxiety Levels by Condition


Alone Stranger Friend Overall E

STAI-T 36.47 37.50 40.53 38.18 .50

STAI-S 39.33 42.87 43.40 41.87 .50





Table 2
Correlations Between Anxiety Levels and Verbal Ratings



Pleasant Arousal Control Safety Trust

STAI-T -.13* .30* -.46** -.24 -.03

STAI-S -.55** .49** -.33* -.47** -.13


*=p<.05, **=p<.01



Table 3
Mean Verbal Ratings by Condition



Pleasantness Arousal Control Safety Trust

Alone 4.20 3.47 5.20 4.80

Stranger 4.00 3.47 5.20 4.87 4.13

Friend 5.00 3.80 4.87 5.20 6.67

Overall 4.40 3.58 5.09 4.96 5.40

Signif. p<.01 p=.78 p=.86 p=.36 p<.001








48
p<.02) conditions. However, pleasantness ratings did not

differ among the alone and stranger conditions. Also as

predicted, subjects in the friend condition reported greater

trust than those in the stranger condition (F[1,29]= 74.96,

p<.001). These findings are shown in Table 3, and ANCOVA

tables are contained in Appendix A.

Baseline Levels of Physiological Activity

As expected, no differences existed among conditions in

tonic cardiac or electrodermal levels either before

presentation of the orienting tone series or before

presentation of the startle probe series, and no differences

existed among conditions in tonic levels of EMG activity of

the orbicularis oculi region prior to presentation of the

startle probe series. ANOVA tables are contained in

Appendix B.

Phasic Heart rate Responding

Overall Responding Across Time

As predicted, the tones elicited the general pattern of

phasic heart rate deceleration typical of an orienting

response. As shown in Figure 1, overall cardiac responding

changed across repeated tone presentations (F[5,210]=3.21,

p<.01, e=.94), habituating from vigorous deceleration on the

first trial to no deceleration on the sixth trial.

Also as predicted, the white noise probes elicited the

general pattern of phasic heart rate acceleration found as a

component of both startle and defensive responses. Overall








49

cardiac acceleration differed significantly across trials

(F[5,210]=3.16, p<.01, E=.94), resulting in a U-shaped

pattern. The first leg of the U consisted of an

acceleratory phasic cardiac response to the first tone

(p<.05), followed by a deceleratory phasic cardiac response

to the third tone (p<.02). This pattern of short-lived

initial acceleration has been associated with the startle

response (Graham, 1979). The second leg of the U consisted

of an increase in overall cardiac acceleration across the

last three trials culminating in a marginally acceleratory

phasic cardiac response to the last tone (p<.055), a pattern

characteristic of an emergent defensive response (Graham,

1979). These findings are illustrated in Figure 1.


Heart Rate Change (bpm)
2





\ Orienting
-1 /Startle
/~ ~ ~ \ \/ (Startle


Trial 2 Tial 3 Trial 4 Trial 5 Trial 6
Time


Figure 1. Overall Cardiac Responding Across Time










Differences Among Conditions

Contrary to predictions, phasic heart rate responding

to the orienting tones differed among conditions.

Differences among conditions changed across time (condition

x time F(10,210)= 2.92, p<.003, e=.94), with differences

occurring only for the second tone. All conditions

exhibited heart rate deceleration to the first tone, a

finding that was expected due to the first tone's novelty.

While, as shown in Figure 2, heart rate deceleration tended

to increase as a function of the presence and friendliness

of others, it did not do so reliably (p >.14). Since mean

deceleration for the friend condition continued beyond four

seconds post tone onset, a second one-way ANCOVA was

performed on mean cardiac response for the entire six second

period. While this second analysis did account for a

greater portion of the variance in cardiac responding, the

effect continued to be unreliable (p>.09). Thus, these

findings indicate that orienting to the first tone did not

reliably differ among conditions.

While the novel first tone elicited cardiac

deceleration across all conditions, the second tone, which

lacked novelty, evoked qualitatively different heart rate

responses among conditions (F[2,41]= 4.64, p<.02). Although

subjects in the alone condition continued to demonstrate a

deceleratory mean heart rate response, subjects in the

stranger condition showed little mean change from baseline,













First Tone (novel stimulus)


Heart Rate change (bpm)


Second Tone (non-novel stimulus)


1 2 3 4 5 6
Time secss)


Figure 2
Cardiac Responding to Orienting Tones by Condition


-2

-4

-6

-8

-10-
0








52

and subjects in the friend condition exhibited an

acceleratory mean heart rate response. Thus, the friend and

stranger conditions did not differ from each other, but

resulted in respectively greater (F[1,29]=6.95, p<.02) and

marginally greater (F[1,29]= 3.65, p<.07) phasic cardiac

acceleration than the alone condition. In addition, post

hoc tests revealed that the deceleratory response exhibited

by the alone condition was significantly below baseline

levels (p<.03), while the acceleratory response exhibited by

the friend condition was not significantly greater than

baseline levels (p>.10).

There were no differences among conditions in phasic

heart rate responses to any of the subsequent orienting

tones, and there were no differences among conditions in

cardiac responding to the startle probes. ANCOVA tables for

all analyses of heart rate data are contained in Appendix C.

Phasic Skin Conductance Responding

Overall Responding Across Time

As predicted, skin conductance responding to the

acoustic tones habituated rapidly after the first tone,

exhibited by a significant decrease in overall SCR frequency

across time (F[5,210]= 13.49, p<.001, e=1.0), and a

corresponding decrease in overall SCR magnitude across time

(F[5,210)= 15.16, p<.001, E=.52). These findings indicate

that skin conductance responses to the first tone reflected

an orienting response to novel stimuli.








53

Skin conductance responding to the white noise probes

also habituated, demonstrated by a significant decrease in

overall SCR frequency across time (F[5,210]= 6.03, p<.001,

E=1.0), and a corresponding decrease in overall SCR

magnitude across time (F[5,210)= 2.95, p<.02, e=.65). The

relatively slow habituation of SCRs to the white noise

probes indicates that SCRs were elicited not by novelty but

by aversiveness. Overall skin conductance frequencies

across time for both orienting and startle trials are

illustrated in Figure 3.




SCR response frequency (%)



0.8


0.6 \I-
\ Orienting
+ Startle
0.4


Tial 2 Trial 3 Trial 4
Time


Figure 3. Overall SCR Frequency Across Time










Differences Among Conditions

The frequency and magnitude of skin conductance

responses to the tones did not differ among conditions or

interact with condition and time. Because skin conductance

responses to the tones habituated so rapidly, only skin

conductance responding to the first tone could be further

analyzed.

For the first tone, frequency and magnitude of skin

conductance responses did not differ among conditions. In

contrast, the amplitude of SCRs to the first tone did differ

among conditions (F[2,35]=4.56, p<.02), but not as

predicted. SCR amplitude did not differ among the stranger

and friend conditions, but both conditions resulted in

respectively smaller (F[1,21]=8.17, p<.02) and marginally

smaller (F[1,23]= 4.69, p<.05) SCR amplitude than the alone

condition. Finally, there were no differences among

conditions in latency of skin conductance response to the

first tone. The frequency and amplitude of skin conductance

responses to the first tone are illustrated in Table 4.

The frequency and magnitude of skin conductance

responses to the white noise tones did not differ among

conditions or interact with condition and time. In

addition, no differences among conditions in SCR amplitude

or latency were found on any trial. ANCOVA tables for all

skin conductance data analyses are contained in Appendix D.





















Table 4
Skin Conductance Response Amplitude and Frequency to the
First Orienting Tone by Condition




Alone Stranger Friend

Amplitude (gmhos) .98 .37 .73

Frequency (n) 10 12 14








56

Blink Responding to Startle Probes

Startle blink magnitude differed among conditions

(F[2,41]= 3.37, p<.05). As predicted, subjects in the

friend condition exhibited significantly smaller blinks than

subjects in the alone condition (F[1,29]=8.35, p<.01) and

marginally smaller blinks than subjects in the stranger

condition (F[1,29]=4.34, p<.05). Both overall blink

magnitude and differences in blink magnitude among

conditions did not change across time. These findings are

illustrated in Figure 4 and ANCOVA tables are contained in

Appendix E.



Eyeblink Magnitude (A/D Units)


Figure 4. Startle Eyeblink Magnitude by Condition


Alone Stranger Friend
Condition








57

Anagram Performance

As shown in Figure 5, overall response accuracy

decreased as anagram difficulty increased (F[4,168]= 104.22,

p<.001, e=.78). In addition, differences among conditions

changed across difficulty levels (condition x difficulty

F[8,168]= 3.24, p<.003, e=.78). As predicted, accuracy did

not differ among conditions when anagrams were easy (i.e.,

overall accuracy levels remained above 65 percent).

However, when anagrams became moderately difficult (i.e.,

overall accuracy levels dropped to between 56 and 66

percent), conditions differed (F[2,42]=6.49, p<.004), with

subjects in the friend condition demonstrating significantly

greater accuracy than subjects in the alone condition

(F[1,29]= 15.23, p<.002), and marginally greater accuracy

than those in the stranger condition (F[1,29]=5.23, p<.04).

However, when anagrams became very difficult (i.e., overall

accuracy levels dropped below 56 percent), accuracy no

longer differed among conditions. Anagram solutions as well

as rankings based upon mean overall accuracy are contained

in Appendix F, while ANOVAs are contained in Appendix G.



















Accuracy (% correct)



0.8


0.6 Alone
Stranger
0.4 -- Friend


0.2


V Easy Mod Easy Mild Diff Mod Diff Very Diff
Difficulty Level


Figure 5. Performance Accuracy on Anagram Task








59

The Interaction of Preexisting Anxiety Levels and Social

Presence Conditions

Contrary to predictions, neither state nor trait

anxiety scores were significantly associated with

responsivity in any of the three physiological systems

measured in the present study. While the correlations

illustrated in Table 5 suggested that the effects of social

presence condition upon response physiology may have been

modulated by preexisting anxiety levels, there were no

statistically significant interactions between preexisting

anxiety levels and condition upon response physiology. A

detailed description of the ANCOVA procedure used to test

this hypothesis is contained in Appendix G.



Table 5
Correlations Between Anxiety and Physiological Reactivity


Cardiac Change to SCR Amplitude to Overall Blink
Second Orienting First Orienting Amplitude
Tone Tone


STAI-T

Stranger .03 p=.91 .36 p=.21 -.08 p=.78
Friend -.03 p=.93 -.07 p=.82 .09 p=.76
Overall .06 p=.72 .13 p=.42 -.12 p=.45


STAI-S

Stranger .29 p=.30 .08 p=.79 .14 p=.63
Friend -.10 p=.71 -.01 p=.98 .01 p=.96
Overall .15 p=.34 .02 p=.88 .05 p=.72








60

Manipulation Check

As illustrated in Table 6, overall belief in the

probability of receiving mild pain was relatively high (5.68

on a 7-point scale), suggesting that the threat of pain was

effectively presented. Belief also did not differ among

conditions, ruling this out as a possible explanation for

other findings. ANOVA table is contained in Appendix A.






Table 6
Belief Ratings by Condition


Alone Stranger Friend Overall

Belief 5.56 5.58 6.00 5.68

N (9) (10) (9) (28)
















CHAPTER 4
DISCUSSION

The present study examined the effects of the simple

presence of friendly and strange species mates upon

individuals under the threat of mild pain. Rather than

merely reducing anxiety, the presence of a friend resulted

in a distinct pattern of verbal, physiological and

performance responses which appears to reflect activation of

an appetitive motivational set. In addition, the presence

of a stranger did not activate an anxious-inhibitory

motivational set. These findings are suggest a modified

version of the reactivity model in which the presence of

species mates associated with punishment or frustration

activates an anxious-inhibitory motivational set, while the

presence of species mates associated with reward or escape

from punishment activates a positive-appetitive motivational

set.

The Presence of a Friend Increases Appetitive Motivation

The hypothesis that the presence of a friend reduces

anxiety-related physiological reactivity was not confirmed.

Rather, the combined verbal, physiological, and performance

data provide convergent evidence that the presence of a

friend elicited an appetitive motivational set. A number of

61


















Table 7
Summary of Verbal, Physiological, and Performance Findings


Verbal Report

Pleasantness Ratings

Trust Ratings



Physiological Responding

Heart Rate

SCR Amplitude

Startle Blink



Overt Behavior

Anagram Accuracy


Alone = Stranger < Friend

Stranger < Friend





Alone < Stranger = Friend

Stranger = Friend < Alone

Friend < Stranger = Alone


Alone = Stranger < Friend








63

findings support this conclusion. First, verbal ratings of

pleasantness and trust were both greater in the friend

condition than in either the stranger or alone conditions.

Second, heart rate responding to a non-novel tone was

acceleratory for the friend condition relative to little

change from baseline for the stranger condition and

deceleration for the alone condition. While the response

exhibited by the friend condition was not statistically

greater than baseline cardiac levels, the acceleratory mean

response pattern was qualitatively distinct from those

exhibited by the other conditions.

In light of Fowles' (1980) demonstration that heart

rate acceleration is related to activation of an appetitive,

incentive-related Behavioral Activation System (BAS), these

findings suggest that the presence of a friendly species

mate is may activate an appetitive motivational set, even

when the threat of mild physical pain is present. In

addition, correlations between cardiac responding and

preexisting anxiety levels, while also not statistically

significant, differed in direction between the friend and

stranger condition in a manner consistent with the above

interpretation. Specifically, heart rate increased as

anxiety increased (r= .29) in the stranger condition, while

heart rate decreased as anxiety increased (r= -.10) in the

friend condition. These findings are consistent with the

conclusion that the cardiac acceleration exhibited in the








64

presence of a friendly conspecific is related to a positive

emotional set.

Third, the amplitude of skin conductance responses to a

novel tone were smaller for both the friend and stranger

conditions than the alone condition. Given Fowles' (1980)

findings that skin conductance responding is strongly linked

to activation of an anxiety-mediating Behavioral Inhibition

System (BIS), these findings indicate that under threat of

mild pain, subjects in the presence of either a friend or a

stranger were less anxious than subjects who were alone.

Fourth, mean startle blink magnitude was smaller for

the friend condition than either the alone or stranger

conditions. In light of considerable evidence that blink

magnitude is modulated by the degree to which an

individual's affective state matches the hedonic valence of

the startle reflex (see Lang et al., 1990), these findings

are consistent with the interpretation that the presence of

a friendly species mate evoked a positive emotional state

that mismatched the valence of the startle probe and thereby

inhibited its magnitude.

Fifth, subjects in the friend condition solved

moderately difficult anagrams more accurately than subjects

in the alone or stranger conditions. This finding is

inconsistent with previous social facilitation experiments

which have shown the presence of others to impair

performance on difficult tasks (see Bond & Titus, 1983), but








65

is consistent with studies showing the presence of others to

result in enhanced performance when such presence is

associated with positive outcomes (Good, 1973; Geen, 1977,

1979; Sanna & Shotland, 1990). Given that socially-induced

impairment of performance on difficult tasks typically

occurs when the species mates are strangers who may be

implicitly threatening (Cacioppo & Petty, 1986), these

contrasting findings suggest that performance is modulated

by the contingencies associated with a particular species

mate.

This interpretation is consistent with findings that

perceptions of self-efficacy are involved in the social

facilitation of performance. Briefly, Bandura's (1977)

theory of self-efficacy proposes that motivation is mediated

by two interrelated beliefs; efficacy expectancy describes

belief in one's capacity to perform a specific behavior,

while outcome expectancy describes belief that a given

behavior will result in a desired outcome (e.g., recognition

by others). Sanna (1992) examined the effects of socially-

mediated efficacy and outcome perceptions upon performance,

and found that under conditions of high outcome expectancy,

performance was improved if efficacy expectancy was high but

performance was impaired if efficacy expectancy was low.

The current findings are consistent with Sanna (1992).

Because outcome expectancies were equal in both the friend

and stranger conditions (i.e., performance was equally








66

likely to be recognized), these findings suggest that

efficacy expectancies were increased in the presence of a

friend but not in the presence of a stranger or in the

absence of another species mate.

The Company of Strangers is Not Always Aversive

The stranger condition did not differ from the alone

condition in ratings of pleasantness, blink magnitude, or

accuracy of performance on anagram tasks. These findings do

not support the hypothesis that the presence of a stranger

results in increased anxiety levels. In fact, findings that

the stranger condition elicited less skin conductance

amplitude to a novel tone than the alone condition suggest

that the presence of a stranger may have actually decreased

activation of the anxious-inhibitory motivational system.

While these results contradict previous findings that

subjects exposed to the presence of strange observers

exhibited greater anxiety (Geen, 1977; 1979, Cacioppo et

al., 1990), there are two possible explanations for such

discrepancies. First, in contrast to previous studies, the

present investigation involved the threat of mild physical

pain. Thus, it appears that the perception of a stranger as

either potentially punishing or potentially rewarding

depends on the experimental context. Specifically, the

findings of Cacioppo et al. (1990) suggest that strangers

who are present in the absence of other aversive stimuli are

perceived as potentially punishing, while the current








67

findings suggest that strangers who are present in the

context of other aversive stimuli are perceived as potential

signals of escape from punishment. This interpretation is

consistent with previous findings that in conditions of

threat, subjects experience the presence of others as

calming and comforting (Schachter, 1959; Wrightsman, 1960).

A second possible explanation for these findings

involves the instructions to subjects in the stranger

condition that they would "complete the first part of the

experiment involving problem-solving and interpersonal

interactions with my research assistant." It is possible

that subjects perceived these instructions as indicating

that they would likely cooperate with rather than be

evaluated by the other individual, and thus experienced less

anxiety than those in the alone condition.

Because friends are typically distinguished from

strangers on the basis of learning histories, the current

findings suggest that the responses elicited by the presence

of a species mate differ qualitatively as a function of the

contingencies associated with that species mate through

learning. These findings are consistent with those of

Rajecki et al. (1976), who found that chickens who were in

the presence of familiar species mates exhibited

qualitatively different behaviors than chickens who were in

the presence of strange species mates:








68

A Revised Reactivity Model

The present findings largely support the reactivity

model. The absence of differences among conditions in tonic

levels of physiological activity as well as the quick

dissipation of skin conductance and heart rate differences

after presentation of only two tones indicate that social

presence increases autonomic reactivity rather than tonic

levels of autonomic activity. These findings support the

hypothesis that social presence effects on the autonomic

nervous system are mediated by the adrenal medulla rather

than the adrenal cortex (Cacioppo & Petty, 1986).

In addition, findings of different patterns of verbal,

physiological, and performance responding between the friend

and stranger conditions are consistent with the hypothesis

that the effects of social presence depend on perceptions of

threat or safety. However, the hypothesis that the presence

of friendly species mates would result in decreased anxious

responding was not supported. Instead, the presence of

friendly conspecifics resulted in a pattern of responses

suggestive of increased appetitive activation.

It is therefore proposed that the reactivity model be

revised to integrate the qualitative motivational changes

predicted by the Gray-Fowles three arousal theory. In light

of the present findings, it is proposed that the presence of

species mates who are associated with punishment or

frustration will activate a motivational set associated with








69

anxiety and inhibition (i.e., BIS), while the presence of

species mates who are associated with reward or escape from

punishment will activate a motivational set associated with

positive affect and active coping (i.e., BAS).

Implications for the Mere Presence and

Distraction-Conflict Models

The current findings have implications for both the

mere presence and distraction-conflict theories of social

facilitation. Contrary to the distraction-conflict

hypothesis, there was little evidence that the presence of

others is attentionally distracting. While the finding that

both social presence conditions resulted in smaller SCR

amplitudes to the first tone than the alone condition might

initially be construed as supporting the distraction-

conflict hypothesis, this effect could not have been due to

distraction of attentional resources to another modality.

All subjects had their eyes closed throughout presentation

of the acoustic probes and thus, subjects in the social

presence conditions should have actually exhibited increased

attention in the auditory modality, since the subjects were

able to attend to species mates only through the auditory

channel. Furthermore, hypotheses regarding attentional

distraction cannot adequately account for differences in

performance and heart rate acceleration between the friend

and stranger conditions.








70

The current findings are also inconsistent with the

mere presence model. First, tonic levels of physiological

activity were unaffected by social presence conditions,

providing further evidence against Zajonc's original

hypothesis that the presence of species mates increases

drive or general arousal and supporting Cacioppo and Petty's

(1986) alternative hypothesis that social presence effects

phasic physiological reactivity. Second, findings that the

presence of friendly species mates elicited an incentive-

related motivational set and enhanced performance on

difficult tasks cannot be accounted for by the mere presence

model, which predicts that the presence of any species mate

will increase uncertainty, elevate general arousal, and

impair performance on difficult tasks.

Third, the present findings provide evidence against

Zajonc's (1980) refutation of previous findings that the

presence of species mates decreases anxiety under conditions

of stress. Zajonc (1980) argued that in stressful and

threatening situations the anxiolytic effect of the presence

of conspecifics was due to the directive influence of the

conspecific through such processes as imitation. However,

since subjects closed their eyes throughout the entire

period in which acoustic probes were presented, the observed

appetitive physiological response set.cannot be attributed

to imitation of or direction by the species mate.








71

Furthermore, while Zajonc (1980) interpreted Latane and

Cappell's (1972) findings that rats in a stressful

laboratory environment exhibited increased heart rate in the

presence of a species mate as evidence that the presence of

others increases arousal, the present data suggest the

entirely opposite conclusion. Given the known association

between heart rate acceleration and incentive motivation

(Fowles, 1980), as well as the current findings of increased

heart rate acceleration in the presence of friendly species

mates, the findings of Latane and Cappell (1972) can be seen

as activation an appetitive motivational system by the

presence of a conspecific.

Implications for the Social Support Construct

Since the publication of epidemiological studies in the

1970s which indicated that affiliation with primary groups

(i.e., families, etc.) has health protective effects, social

support has become a major area of investigation (Gottleib,

1983). The current findings suggest that the degree to

which the presence of others protects against stress may be

related to the extent to which such presence activates the

behavioral activation system described by Gray and Fowles.

While the present findings were obtained in a controlled

laboratory experiment and thus may not necessarily

generalize to other situations, they suggest a testable

hypothesis regarding a single mechanism of social support

and may allow elucidation of circumstances in which the








72
presence of others is beneficial versus harmful. This

learning theory-based hypothesis appears potentially

preferable to the myriad existing operationalizations of

social support which are often used interchangeably and

which reflect some confusion regarding the exact mechanism

of social support (Boyce, 1985).

Physiological Variables as Indices of Motivation

In the present study mean startle blink responses were

highly congruent with both verbal ratings of affect and

performance accuracy. In contrast, heart rate and skin

conductance, both of which reflected autonomic nervous

system (ANS) activity, were less consistent over time, more

sensitive to and influenced by acute changes such as

stimulus novelty, and were not entirely congruent with the

verbal and performance data. These findings demonstrate the

susceptibility of autonomic variables to specific contextual

demands and the relative independence of the startle blink

from such demands. These findings are consistent with Lang

et al. (1990), and attest to the sensitivity of the startle

blink to superordinate emotional dispositions, its

independence from tactical or contextual demands, and thus

its utility as a measure of affective-motivational state.

Summary and Conclusion

The present study examined the general hypothesis that

the effect of the presence of species mates upon motivation

varies as a function of the nature of the relationship held








73

with the conspecific. These findings-indicated that,

contrary to the predictions of the mere presence hypothesis

of social facilitation, the presence of species mates does

not universally produce elevated general arousal which

facilitates performance on simple tasks and impairs

performance on difficult tasks.

Instead, the presence of species mates resulted in

qualitatively different patterns of verbal ratings,

physiological reactivity, and performance depending upon the

nature of the relationship between the subject and the

species mate. Since social presence conditions differed

primarily in the extent to which the subject had a positive

learning history with the species mate, these findings

suggest that the effects of the presence of species mates

depend upon the contingencies associated with such species

mates through learning.

In addition, findings that the presence of friendly

species mates resulted in activation of a positive-

appetitive motivational set despite the concurrent threat of

mild pain have implications for the study of social support.

While these findings are preliminary and not immediately

generalizable, they suggest that the mechanism of social

support may involve activation of the appetitive

motivational system.

The current findings also suggest several possible

directions for future research. The hypothesis that








74
qualitative differences among conditions are due to the

learned association of species mates with different

contingencies would be further validated if similar results

were obtained after direct manipulation of such

contingencies. Another avenue for future research involves

applying the current findings to the study of social

support. This might be done by examining the inter-

relationship between activation of the appetitive system,

affiliation with primary groups, and physical health.

Thus, given the present unexpected, highly consistent

and somewhat provocative findings and the implications of

these findings for the areas of social facilitation and

social support, this study represents a significant

advancement of our understanding of one of the oldest

phenomena studied in experimental psychology, the effect of

the presence of others upon human behavior.









REFERENCES


Agresti, A., & Finlay, B. (1986). Statistical methods for
the social sciences. San Francisco: Dellen.

Anthony, B.J. (1985). In the blink of an eye: Implications
of reflex modification for information processing.
Advances in Psvchophvsiologv. (Vol. 1, pp. 167-218).
Greenwich, CT: JAI Press.

Anthony, B.J., & Graham, F.K. (1985). Blink reflex
modification by selection attention: Evidence for the
modulation of 'automatic' processing. Biological
Psychology, 21, 43-59.

Bandura, A. (1977). Self-efficacy: Toward a unifying
theory of behavioral change. Psychological Review, 84,
191-215.

Baron, R.S., Moore, D., & Sanders, G.S. (1978).
Distraction as a source of drive in social
facilitation. Journal of Personality and Social
Psychology, 36, 816-824.

Bond, C.F., & Titus, L.J. (1983). Social facilitation: A
meta-analysis of 241 studies. Psychological Bulletin,
94, 265-292.

Boyce, W.T. (1985). Social support, family relations, &
children. In S. Cohen, & S.L. Syme (Eds.), Social
support and Health (pp. 151-174). New York: Academic
Press.

Bradley, M.M, Cuthbert, B.N., & Lang, P.J. (1990). Startle
reflex modification: Emotion or Attention?
Psvchophysiologv, 27, 513-522.

Bradley, M.M, Cuthbert, B.N, & Lang, P.J. (1991). Startle
and Emotion: Lateral acoustic probes and the bilateral
blink. Psvchophvsiology, 28, 285-295.

Cacioppo, J.T., & Petty, R.E. (1986). Social Processes. In
M.G.H. Coles, E. Donchin, & S.W. Porges (Eds.),
Psychophysiolovg: Systems, processes and applications
(pp.646-679). New York: Guilford Press.

Cacioppo, J.T., Rourke, P.A., Marshall-Goodell, B.S.,
Tassinary, L.G., & Baron, L.G. (1990). Rudimentary
physiological effects of mere observation.
Psychophvsiology, 21, 177-186.










Clark and Fouts (1973). Effects of positive, neutral and
negative experiences with an audience on social
facilitation in children. Perceptual and Motor Skills,
37, 1008-1010.

Cohen, S. (1978). Environmental load and the allocation of
attention. In A. Baum & S. Valins (Eds.), Advances in
environmental research (pp. 1-29). Hillsdale, NJ:
Erlbaum.

Cottrell, N.B. (1968). Performance in the presence of other
human beings: Mere presence, audience, and affiliation
effects. In E.C. Simel, R.W. Hoppe, & G.A. Milton
(Eds.), Social facilitation and imitation behavior (pp.
91-110). Boston: Allyn & Bacon.

Cottrell, N.B., Wack, D.L., Sekerak, G.J., & Rittle, R.H.
(1968). Social facilitation of dominant responses by
the presence of an audience and the mere presence of
others. Journal of Personality and Social Psychology,
9, 245-250.

Dashiell, J.F. (1930). An experimental analysis of some
group effects. Journal of Abnormal and Social
Psychology, 25, 190-199.

Davidson, P.O., & Kelly, W.R. (1973). Social facilitation
and coping with stress. British Journal of Social and
Clinical Psychology, 12, 130-136.

Davis, M. (1986). Pharmacological and anatomical analysis
of fear conditioning using the fear-potentiated startle
paradigm. Behavioral Neuroscience, 100, 814-824.

Davitz, J.R. & Mason, D.J. (1955). Socially facilitated
reduction of a fear response in rats. Journal of
Comparative Psychology, 48, 149-151.

Dawson, M.E., Schell, A.M., & Filion, D.L. (1990). The
electrodermal system. In J.T. Cacioppo, & L.G.
Tassinary (Eds.), Principles of Psvchophvsiologv:
Physical.Social & Inferential Elements (pp. 295-324).
New York: Cambridge University Press.

Fowles, D.C. (1980). The three arousal model: Implications
of Gray's two-factor learning theory for heart rate,
electrodermal activity, and psychopathy.
Psvchophvsiology, 17, 87-104.










Geen, R.G. (1976). The role of the social environment in
the induction and reduction of anxiety. In C.D.
Spielberger & I.G. Sarason (Eds.), Stress and anxiety
(Vol. 3, pp. 105-126). Washington, DC: Hemisphere.

Geen, R.G. (1977). Affects of anticipation of positive and
negative outcomes on audience anxiety. Journal of
Consulting and Clinical Psychology, 45, 715-716.

Geen, R.G. (1979). Affects of being observed on learning
following success and failure experiences. Motivation
and Emotion, 3, 355-371.

Geen, R.G. and Gange, J.J. (1977). Drive theory of
social facilitation: Twelve years of theory and
research. Psychological Bulletin, 84, 1267-1288.

Good, K.J. (1973). Social facilitation: Effects of
performance anticipation, evaluation, and response
competition on free associations, Journal of
Personality and Social Psychology, 28, 270-275.

Gottlieb, B.H. (1983). Social support as a focus for
integrative research in psychology. American
Psychologist, 38, 278-287.

Graham, F.K. (1979). Distinguishing among orienting,
defense and startle reflexes. In H.D. Kimmel, E.H. Van
Olst, & J.F. Orlebeke (Eds.), The orienting reflex in
humans. Hillsdale, NJ: Erlbaum.

Graham, F.K. (1980). Representing cardiac activity in
relation to time. In I. Martin, & P.H. Venables
(Eds.), Techniaues in Psvchophvsiology. Chichester,
England: Wiley.

Gray, J.A. (1970). The psychophysiological basis of
introversion-extraversion. Behaviour Research and
Therapy, 8, 249-266.

Gray, J.A. (1981). Issues in the neuropsychology of
anxiety. In A.H. Tuma and J.D. Maser (Eds.), Anxiety
and the anxiety disorders (pp.5-25). Hillsdale, NJ:
Erlbaum.

Greenwald, M.K., Hamm, A.O., Bradley, M.M., & Lang, P.J.
(1988). The acoustic startle probe in the assessment
of classical aversive conditioning [Abstract].
Psychophysioloyv, 25, 451.










Henchy, T. & Glass, D.G (1968). Evaluation apprehension and
the social facilitation of dominant and subordinate
responses. Journal of Personality and Social
Psychology, 4, 446-454.

Kissel, S. (1965). Stress-reducing properties of social
stimuli. Journal of Personality and Social Psychology,
2, 378-384.

Lacey, J.I. (1967). Somatic response patterning and stress:
Some revisions of activation theory. In M. H. Appley &
R. Trumbull (Eds.), Psychological stress: Issues in
research (pp.14-42). New York: Appleton-Century-
Crofts.

Lang, P.J., Bradley, M.M., & Cuthbert, B.N. (1990).
Emotion, attention and the startle reflex.
Psychological Review, 3, 377-395.

Latane, B. (1969). Gregariousness and fear in laboratory
rats. Journal of Experimental Social Psvchology, 5,
61-69.

Latane, B. and Cappell, H. (1972). The effects of
togetherness on heart rate in rats. Psychonomic
Science, 29, 177-179.

Latane, B. & Glass, D.C. (1968). Social and nonsocial
attraction in rats. Journal of Personality and Social
Psychology, 2, 142-146.

Markus, H. (1978). The effect of mere presence on social
facilitation: An unobtrusive test. Journal of
Experimental Social Psychology, 14, 389-397.

Matlin, M.M., & Zajonc, R.B. (1968). Social facilitation of
word associations. Journal of Personality and Social
Psychology, 10, 435-460.

Moore, D.L., & Baron, R.S. (1983). Social facilitation:
A psychophysiological analysis. In J.T. Cacioppo &
R.E. Petty (Eds.), Social svychoDhvsiology: A
sourcebook (pp. 434-466). New York: Guilford Press.

Obrist, P.A. (1976). The cardiovascular-behavioral
interaction As it appears today. Psychophysiology,
13, 95-107.

Paulus, P.B., Annis, A.B., & Risner, H.T. (1978). An
analysis of the mirror-induced objective self-awareness
effect. Bulletin of the Psychonomic Society, 12, 8-10.










Pessin, J. (1933). The comparative effects of social and
mechanical stimulation on memorizing. American Journal
of Psychology, 45, 263-270.

Prokasy, W.F., & Kumpfer, K.L. (1973). Classical
Conditioning. In W.F. Prokasy, & D.C. Raskin (Eds.),
Electrodermal activity in psychological research (pp.
157-202). New York: Academic Press.

Rajecki, D.W., Kidd, R.F., & Ivins, B. (1976). Social
facilitation in chickens: A different level of
analysis. Journal of Experimental Social Psvchologyv,
12, 233-246.

Routtenberg, A. (1968). The two-arousal hypothesis:
Reticular formation and limbic system. Psychological
Review, 75, 51-80.

Sanders, G.S. (1981). Driven by distraction: An
integrative review of social facilitation theory and
research. Journal of Experimental Social Psvchologv,
17, 227-251.

Sanders, G.S., Baron, R.S., & Moore, D. L. (1978).
Distraction and social comparison as mediators of
social facilitation effects. Journal of Experimental
Social Psychology, 14, 291-303.

Sanna, L.J. (1992). Self-Efficacy Theory: Implications for
social facilitation and social loafing. Journal of
Personality and Social Psvchology, _2, 774-786.

Sanna, L.J. & Shotland, R.L. (1990). Valence of anticipated
evaluation and social facilitation. Journal of
Experimental Social Psychology, 26, 82-92.

Sarason, I.G. (1980). Test anxiety: Theory, research and
applications. Hillsdale, NJ: Erlbaum.

Schachter, S. (1959). The psychology of affiliation.
Stanford, CA: Stanford University Press.

Shaver, P., & Liebling, B.A. (1976). Explorations in the
drive theory of social facilitation. Journal of Social
Psychology, 99, 259-271.

Schmitt, B.H., Gilovich, T., Gorre, N., & Joseph, L. (1986).
Mere presence and social facilitation: One more time.
Journal of Experimental Social Psychology, 22, 242-248.










Simons, R.F., & Zelson, M.F. (1985). Engaging visual
stimuli and reflex blink modification.
Psvchophysiology, 22, 44-49.

Spielberger, C.D., Gorsuch, R.L., & Lushene, R.E. (1970).
Manual for the State-Trait Anxiety Inventory, Palo
Alto: Consulting Psychologists Press.

Tassinary, L.G., Cacioppo, J.T., & Geen, T.R. (1989). A
psychometric study of surface electrode placements for
facial electromyographic recording: The brow and cheek
muscle regions. Psychophysiolocy, 26, 1-16.

Triplett, N. (1898). The dynamogenic factors in pacemaking
and competition. American Journal of Psychology, 9,
507-533.

Venables, P.H., & Christie, M.J. (1980). Electrodermal
Activity. In I. Martin, & P.H. Venables (Eds.),
Techniques in Psvchophvsiology (pp. 3-57). New York:
Wiley & Sons.

Vrana, S.R., Spence, E.L., & Lang, P.J. (1988). The startle
probe response: A new measure of emotion? Journal of
Abnormal Psychology, 97, 487-491.

Weiss, R.F. & Miller, F.G. (1971). The drive theory of
social facilitation. Psychological Review, 78, 44-57.

Wicklund, R.A. (1975). Discrepancy reduction or attempted
distraction? A reply to Liebling, Seiler, and Shaver.
Journal of Experimental Social Psychology, 11, 78-81.

Wrightsman, L.S. (1960). Effects of waiting with others on
changes in level of felt anxiety. Journal of Abnormal
and Social Psychology, 61, 216-222.

Zajonc, R.B. (1965). Social facilitation. Science, 149,
269-274.

Zajonc, R.B. (1980). Compresence. In P.B. Paulis (Ed.),
Psvchology of group influence (pp.35-60). Hillsdale,
NJ: Erlbaum.

Zajonc, R.B., & Nieuwenhuse, B. (1964). Relationship
between word frequency and recognition: Perceptual
process or response bias? Journal of Experimental
Psychology, 67, 276-285.

Zajonc, R.B., & Sales, S.M. (1966). Social facilitation of
dominant and subordinate response. Journal of
Experimental Social Psychology, 2, 160-168.









APPENDIX A
ANOVA AND ANCOVA TABLES FOR VERBAL RATINGS




Table 8
ANCOVA of Valence Ratings


Source of Sum of D.F. Mean F 2-Tail
Variation Squares Square Prob.

Covariate 22.63 1 22.63 22.81 .000
Condition 11.48 2 5.74 5.78 .006

Explained 34.11 3 11.37 11.46 .000

Residual 40.69 41 .99
Total 74.80 44 1.70





Table 9
ANCOVA of Arousal Ratings

Source of Sum of D.F. Mean F 2-Tail
Variation Squares Square Prob.

Covariate 19.73 1 19.73 13.37 .001
Condition .74 2 .37 .25 .780

Explained 20.47 3 6.82 4.62 .007

Residual 60.51 41 1.48
Total 80.98 44 1.84















Table 10
ANCOVA of Control Ratings

Source of Sum of D.F. Mean F 2-Tail
Variation Squares Square Prob.

Covariate 11.64 1 11.64 4.90 .032
Condition .71 2 .36 .15 .861

Explained 12.35 3 4.12 1.73 .175

Residual 97.30 41 2.37
Total 109.64 44 2.49





Table 11
ANCOVA of Safety Ratings

Source of Sum of D.F. Mean F 2-Tail
Variation Squares Square Prob.

Covariate 20.24 1 20.24 12.17 .001
Condition 3.50 2 1.75 1.05 .358

Explained 23.74 3 7.91 4.76 .006

Residual 68.17 41 1.66
Total 91.91 44 2.09















Table 12
ANCOVA of Trust Ratings

Source of Sum of D.F. Mean F 2-Tail
Variation Squares Square Prob.

Covariate 1.16 1 1.16 1.80 .19
Condition 48.55 2 48.55 74.96 .000

Explained 49.71 3 24.86 38.38 .000

Residual 17.49 27 0.65
Total 67.20 29 2.42





Table 13
ANCOVA of Belief Ratings

Source of Sum of D.F. Mean F 2-Tail
Variation Squares Square Prob.

Covariate 1.16 1 1.16 .45 .510
Condition .90 2 .45 .17 .842

Explained 2.05 3 .68 .26 .850

Residual 62.06 24 2.59
Total 64.12 27 2.37









APPENDIX B
ANOVA TABLES FOR BASELINE PHYSIOLOGICAL ACTIVITY




Table 14
ANOVA of Tonic Heart Rate Activity Prior
to Orienting Tone Series

Source of Sum of D.F. Mean F 2-Tail
Variation Squares Square Prob.

Condition 120.92 2 60.46 .48 .620

Explained 120.92 2 60.46 .48 .620

Residual 5251.33 42 125.03
Total 5372.25 44 122.10





Table 15
ANOVA of Tonic Heart Rate Activity Prior
to Startle Probe Series

Source Sum of D.F. Mean F 2-Tail
Squares Square Prob.

Condition 148.24 2 74.12 .65 .528

Explained 148.24 2 74.12 .65 .528

Residual 4797.12 42 114.22
Total 4945.35 44 112.39















Table 16
ANOVA of Tonic Skin Conductance Activity Prior to
Orienting Tone Series

Source of Sum of D.F. Mean F 2-Tail
Variation Squares Square Prob.

Condition 1.46 2 .73 .16 .854

Explained 1.46 2 .73 .16 .854

Residual 193.32 42 4.60
Total 194.78 44 4.43





Table 17
ANOVA of Tonic Skin Conductance Activity Prior to
Startle Probe Series

Source Sum of D.F. Mean F 2-Tail
Squares Square Prob.

Condition 2.56 2 1.28 .29 .748

Explained 2.56 2 1.28 .29 .748

Residual 183.58 42 4.37

Total 186.13 44 4.23















Table 18
ANOVA of Tonic Orbicularis EMG Activity Prior to
Startle Probe Series

Source Sum of D.F. Mean F 2-Tail
Squares Square Prob.

Condition 12174.17 2 6087.09 .21 .812

Explained 12174.17 2 6087.09 .21 .812

Residual 1218099.96 42 29002.38

Total 1230274.13 44 27960.78









APPENDIX C
ANCOVA TABLES FOR PHASIC HEART RATE RESPONDING


Table 19
ANCOVA of Heart Rate Responding


to Orienting Tones


Source of Sum of D.F. Mean F 2-Tail Huynh
Variation Squares Square Prob. Feldt

Within Cells 1466.66 41 35.77
Regression 143.64 1 143.64 4.02 .052
Constant 72.75 1 72.75 2.03 .161
Condition .35 2 .17 .00 .995

Within Cells 5802.33 210 27.63
Time 442.90 5 88.58 3.21 .008 .94
CXT 806.60 10 80.66 2.92 .002





Table 20
ANCOVA of Heart Rate Responding to Startle Probes


Source of Sum of D.F. Mean F 2-Tail Huynh
Variation Squares Square Prob. Feldt

Within Cells 2491.69 41 60.77
Regression 114.42 1 114.42 1.88 .177
Constant 123.46 1 123.46 2.03 .162
Condition 99.59 2 49.79 .82 .448

Within Cells 3973.92 210 18.92
Time 298.97 5 59.79 3.16 .009 .94
CXT 230.87 10 23.09 1.22 .280















Table 21
ANCOVA of Heart Rate Responding


to First Orienting Tone


Source of Sum of D.F. Mean F 2-Tail
Variation Squares Square Prob.

Covariate 38.53 1 38.53 .87 .357
Condition 177.38 2 88.69 2.00 .149

Explained 215.92 3 71.97 1.62 .200

Residual 1822.36 41 44.45
Total 2038.27 44 46.32





Table 22
ANCOVA of Heart Rate Responding to Second Orienting Tone


Source of Sum of D.F. Mean F 2-Tail
Variation Squares Square Prob.

Covariate 5.49 1 5.49 .18 .672
Condition 280.64 2 140.32 4.64 .015

Explained 286.13 3 95.38 3.15 .035

Residual 1241.26 41 30.28
Total 1527.40 44 34.72















Table 23
ANCOVA of Heart Rate Responding to Third Orienting Tone


Source of Sum of D.F. Mean F 2-Tail
Variation Squares Square Prob.

Covariate 197.22 1 197.92 4.91 .032
Condition 135.49 2 67.75 1.68 .199

Explained 333.42 3 111.14 2.76 .055

Residual 1653.26 41 40.32
Total 1986.67 44 45.15





Table 24
ANCOVA of Heart Rate Responding to Fourth Orienting Tone


Source of Sum of D.F. Mean F 2-Tail
Variation Squares Square Prob.

Covariate 7.76 1 7.76 .32 .572
Condition 102.64 2 51.32 2.14 .130

Explained 110.40 3 36.80 1.54 .220

Residual 982.59 41 23.97
Total 1093.00 44 24.84















Table 25
ANCOVA of Heart Rate Responding

Source of Sum of D.F.
Variation Squares


to Fifth

Mean
Square


Orienting Tone

F 2-Tail
Prob.


Covariate .51 1 .51 .02 .877
Condition 53.61 2 26.81 1.26 .294

Explained 54.12 3 18.04 0.85 .475

Residual 871.76 41 21.26
Total 925.89 44 21.04





Table 26
ANCOVA of Heart Rate Responding to Sixth Orienting Tone


Source of Sum of D.F. Mean F 2-Tail
Variation Squares Square Prob.

Covariate 11.47 1 11.47 .76 .390
Condition 16.11 2 8.06 .53 .592

Explained 27.59 3 9.20 .61 .615

Residual 622.24 41 15.18
Total 649.83 44 14.77